Filter Rules

All exclusion rules (Haiku prescreen + Sonnet ranking + learned) — edit, remove, or propose new ones
Active Rules 1495 Applied to every scan by Haiku + Sonnet
[truncated_incomplete_posts]
Learned 208 rejections Active
Exclude posts that end abruptly mid-sentence, are clearly truncated, or appear to be fragments with missing context. These posts should have complete, coherent messaging to evaluate substantive healthcare tech content.
Posts that are cut off mid-sentence or lack complete thoughts, making them difficult to assess for relevance.
3 example posts
The biggest shift in AI drug discovery: The special models & supercomputers won’t be the edge much longer. Everyone will have them. The real advantage will be how fast you turn AI into real lab-tested learning and better decisions about which molecules to advance. Tools are
Citadel Securities just put institutional weight behind what the AI bulls won't say out loud. In a new macro note titled "Tokenomics," Citadel makes the argument plainly: even the most powerful technology on earth still has to pass through the boring discipline of cost curves, h
CNBC interviewer asked Palantir CEO Alex Karp how he would defend Wall Street’s concern that AI could replicate what Palantir is doing. Karp defended by basically saying that AI companies may have great engineers, but they do not deeply understand the messy, high-stakes https://
Created 2026-04-08 · Updated 2026-08-02
Edit rule text
[enterprise_ai_adoption_macro_without_healthcare_specificity]
Learned 168 rejections Active
Exclude posts discussing enterprise AI adoption, agentic systems, token economics, control planes, or labor market disruption—unless they explicitly analyze how these trends reshape healthcare workflows, clinical decision-making, or care delivery systems.
Posts about generalist enterprise AI adoption, productivity metrics, or labor disruption without healthcare-specific implementation context
3 example posts
SOFTWARE IS DEAD: "The software companies frankly got fat & happy. I think what will happen over time is ChatGPT & Claude will end up sitting on top of basically the entire enterprise software stack & almost everything else will end up being a dumb data pipe into thos
OpenAI just published a new Codex use-case page, and it’s basically a catalog of what teams are already handing over to coding agents: engineering work, product work, QA, security, data analysis, internal tools, and even life-sciences workflows. Some of the coolest examples: ⬩
Token costs are becoming one of the hottest topics for any enterprise I talk with right now. It’s very bullish for AI in general because it means these systems are being used at a scale that wasn’t contemplated before. It also gives way to another form of differentiation that
Created 2026-05-12 · Updated 2026-06-11
Edit rule text
[ai_company_business_metrics_not_healthcare_application]
Learned 103 rejections Active
Exclude posts that focus on AI company business news (funding rounds, token pricing, competitive positioning, revenue) without demonstrating concrete healthcare delivery or system impact. The post must show how the technology solves a healthcare problem, not just report business metrics.
Posts reporting AI company funding, valuations, pricing wars, or business metrics without healthcare system application.
3 example posts
Citadel Securities just put institutional weight behind what the AI bulls won't say out loud. In a new macro note titled "Tokenomics," Citadel makes the argument plainly: even the most powerful technology on earth still has to pass through the boring discipline of cost curves, h
OpenAI is reportedly considering drastic token price cuts to pull customers away from Anthropic, per WSJ. This follows rising complaints from enterprise customers about AI costs, while Anthropic has been gaining traction with Claude Code. The message is clear: @OpenAI does not
🚨BREAKING: OpenAI considering “drastic” price cuts to win the war for users with Anthropic Altman: “costs have become a huge issue” >already losing billions >planning to lose more >right before IPO it’s so over https://t.co/NWHp0TiVVW
Created 2026-04-14 · Updated 2026-06-12
Edit rule text
[political_outrage_or_fraud_scandal_without_healthcare_systems_analysis]
Learned 81 rejections Active
Exclude posts that report healthcare fraud, regulatory scandal, or political figures' healthcare claims without analyzing the underlying systemic healthcare problem (market design, incentive structures, provider networks, pricing mechanisms). Moral outrage, vigilante justice framing, or individual accountability narratives without systemic insight should be rejected.
Posts about healthcare fraud, regulatory violations, or political posturing that focus on individual wrongdoing or moral outrage without analyzing systemic healthcare issues.
3 example posts
USDA's Chief Information Officer Sam Berry says AI could help the USDA crack down on billions of dollars in SNAP fraud each year. "SNAP is a $100B-per-year taxpayer-funded program. That's an area where we really want to have all angles of the data available so that we can deploy
My job is to make sure your surgery center never gets built. Eleven years, and I have never lost. The kid had it all lined up. Board-certified, two partners, a lease on a space where he could do the same procedures my client does across town. He showed up to the hearing with a
PacificSource pulled out as Lane County Medicaid insurer, Tina Kotek’s OR Health Authority replaced it with Trillium, forcing 96k to join Trillium. Trillium’s parent co gave Kotek’s campaign $50k two months later. OR AG (D) said in ‘22 Trillium parent co gave Kotek’s”took advan
Created 2026-04-29 · Updated 2026-06-07
Edit rule text
[glp1_peptide_market_pricing_only]
Learned 68 rejections Active
Exclude posts that focus primarily on GLP-1 drug market pricing, weight loss efficacy comparisons, treatment discontinuation rates, or competitive market dynamics without addressing underlying healthcare system failures, access equity, or clinical innovation. Market dynamics alone are not healthcare tech analysis.
Posts about GLP-1 drugs and peptide market dynamics, pricing competition, or weight loss trends without healthcare systems analysis or clinical context.
3 example posts
🔔Survodutide SYNCHRONIZE-1 full paper is out for Boehringer-Ingelheim’s GLP-1/Glucagon study for weight loss. ▪️13% loss on 6mg (treatment regimen estimand) ▪️Huge treatment discontinuations due to AEs: 20% ▪️Nausea 65%, Vomiting 45% ▪️Placebo was killer. 5.4% WL. Wow. ▪️60% htt
Survodutide Once Weekly for the Treatment of Adults with Obesity: @NEJM 🥸 For @Boehringer, the advantage of Survodutide is probably liver and visceral fat reduction, not greater weight loss! Probably indication in obesity-+ MASH rather than head-to-head share against https://t.
Let me clear up the microscope fight first. 🔬 Everyone’s arguing over “hairs” and “fibers” in horse paste. But nobody talks about formulation science. Horse paste = oral suspension. Binders. Thickeners. Made to swallow. Injectable = sterile. Made for veins. Different rules. J
Created 2026-04-25 · Updated 2026-06-10
Edit rule text
[unvalidated_speculative_medical_claims]
Learned 64 rejections Active
Exclude posts that make speculative claims about new medical treatments, peptides, gene therapies, or clinical interventions without citing peer-reviewed evidence, completed clinical trials, or FDA approval status. Include posts that frame personal anecdotes or single case reports as medical breakthroughs.
Posts promoting unproven medical interventions, treatments, or drugs without clinical evidence or FDA validation.
3 example posts
Feeling extremely lucky to live in 2026 and have access to such incredible medicine and clinical research. On the eve of my 40th I had my first dose (monthly injection) of a new treatment that I’m hopeful will prevent me from ever having a heart attack or stroke. This drug ht
This week brought even more exciting news in the field of HIV treatment! A once-weekly Lenacapavir-based pill has now demonstrated similar outcomes to daily antiretrovirals (ARVs), achieving non-inferiority. This development could significantly impact treatment adherence and http
Let me clear up the microscope fight first. 🔬 Everyone’s arguing over “hairs” and “fibers” in horse paste. But nobody talks about formulation science. Horse paste = oral suspension. Binders. Thickeners. Made to swallow. Injectable = sterile. Made for veins. Different rules. J
Created 2026-04-26 · Updated 2026-07-23
Edit rule text
[enterprise_ai_adoption_macro]
Learned 53 rejections Active
Exclude posts that discuss enterprise AI adoption, business model changes (e.g., SaaS to agentic pricing), or labor productivity claims at a macro/generalist level without demonstrating specific healthcare operational impact, clinical validation, or healthcare system context. Posts must connect directly to healthcare delivery challenges or outcomes.
Posts about enterprise AI adoption, productivity metrics, and business model shifts without healthcare-specific analysis or validation
3 example posts
A framework I wish more Founders and VCs used when discussing US Manufacturing: Automation Value = Labor Intensity × Labor Eliminated Everyone gets excited about robots and automation. Almost nobody asks: “How much of revenue is labor?” If labor is 60% of revenue and AI http
Enterprise software was priced per seat. That model is breaking. Before: SaaS = seats. Predictable. Human-driven usage. Revenue tied to headcount. Now: agents hit software systems more frequently than any human ever could. Aaron Levie (@levie) on The MAD Podcast: agent-driven h
I just looked at a home service firms ServiceTitan dashboard and saw something interesting. A Pre-seed company with AI-Supported Human CSRs is out converting an AI CSR business that's raised over $125M by >50% Head-to-Head win rate is 100% Why? Supercharging humans resonat
Created 2026-05-11 · Updated 2026-06-07
Edit rule text
[agentic_ai_technical_tangent]
Learned 51 rejections Active
Exclude posts that discuss agent architecture, control planes, security frameworks, runtime systems, or technical integration patterns (e.g., OpenShell, Hermes, NemoClaw) at a technical or vendor-announcement level without demonstrating validated healthcare application, clinical trial data, or healthcare operational deployment.
Posts about agentic AI systems, control planes, agent security, and technical architecture without healthcare validation or clinical context
3 example posts
Control plane! Control plane! Control plane! You will hear this term of art a lot going forward. Why? Because in this next phase of AI, companies will want something to sit above the models. They will want control over their AI spend. They will want the flexibility to pick
"You can run OpenClaw inside your company now." Annoucing our work with @Microsoft to bring OpenClaw to the Microsoft and Windows ecosystems. Claws now work securly in the enterprise. https://t.co/3fKhT2mGpC
Nous Research is working with NVIDIA to make Hermes Agent run smoothly on the new NVIDIA RTX Spark superchip. Hermes Agent is also integrating with the new OpenShell runtime, which connects Hermes to Microsoft’s security primitives https://t.co/cvvPc8bIKa
Created 2026-05-14 · Updated 2026-06-07
Edit rule text
[clinical_anecdote_without_systems_context]
Learned 50 rejections Active
Exclude posts that report individual clinical trial results, single case studies, or patient anecdotes without connecting them to healthcare delivery systems, reimbursement, access, or operational challenges. A single drug trial result or medical achievement is not sufficient without healthcare systems context.
Posts sharing single clinical cases, trial observations, or healthcare anecdotes without healthcare systems analysis or operational context.
3 example posts
This is the stock price for a biotech that has invented a drug which many considered to have basically cured one type of cancer (multiple myeloma), at least in a subset of patients. Cell therapies, despite being transformative medicines, are being gutted by manufacturing costs,
Version 2 of our ShockCalcs hemodynamics simulator is live. I've refined the physiology a ton, added new meds, and also a real-time Frank Starling curve that responds to fluids + vasopressors. Check it out (link in reply), and reply here with feedback on how to improve further!
The Iowa study doubled survival by adding IV vitamin C to first line chemo, 8 months to 16. We take it further and inject the vitamin C straight into the tumor, same as we do with the KRAS inhibitors. The compound is rarely the problem. Delivery is, and what it does to the ht
Created 2026-04-24 · Updated 2026-06-12
Edit rule text
[clinical_observation_without_systems_context]
Learned 49 rejections Active
Exclude posts that announce clinical trial outcomes, drug efficacy numbers, or medical research observations in isolation. Unless the post connects the finding to healthcare delivery, reimbursement, workflow, access, or system-level implications, it is clinical reporting, not healthcare strategy analysis.
Posts reporting single clinical trial results, drug efficacy data, or medical observations without analyzing healthcare systems implications
3 example posts
This week brought even more exciting news in the field of HIV treatment! A once-weekly Lenacapavir-based pill has now demonstrated similar outcomes to daily antiretrovirals (ARVs), achieving non-inferiority. This development could significantly impact treatment adherence and http
🚨NEW @NEJM CAR T cells expanding applications! Here CAR T facilitate kidney transplantation in highly sensitized people. dual CD19 + BCMA CAR T enabled kidney transplantation in two highly sensitized patients after reducing anti-donor antibody barriers, with no severe https://t.
🔔Survodutide SYNCHRONIZE-1 full paper is out for Boehringer-Ingelheim’s GLP-1/Glucagon study for weight loss. ▪️13% loss on 6mg (treatment regimen estimand) ▪️Huge treatment discontinuations due to AEs: 20% ▪️Nausea 65%, Vomiting 45% ▪️Placebo was killer. 5.4% WL. Wow. ▪️60% htt
Created 2026-04-13 · Updated 2026-06-11
Edit rule text
[ai_company_business_metrics]
Learned 45 rejections Active
Exclude posts that focus on AI company business metrics (funding, valuation, partnerships, product launches) unless the post clearly analyzes how that company's product will be applied or deployed in healthcare systems with specific clinical or operational context.
Posts about AI company funding rounds, valuations, revenue, partnerships, or product launches without healthcare application specifics.
3 example posts
This is the stock price for a biotech that has invented a drug which many considered to have basically cured one type of cancer (multiple myeloma), at least in a subset of patients. Cell therapies, despite being transformative medicines, are being gutted by manufacturing costs,
AI assistants are moving from "answer my question" to "do the work." But they are only as useful as the enterprise content they have access to. Our demo shows what it looks like when @Copilot Cowork is grounded in Box. Your governed content powering multi-step agentic workflows,
OpenAI is reportedly considering drastic token price cuts to pull customers away from Anthropic, per WSJ. This follows rising complaints from enterprise customers about AI costs, while Anthropic has been gaining traction with Claude Code. The message is clear: @OpenAI does not
Created 2026-04-11 · Updated 2026-07-26
Edit rule text
[cynical_insider_sarcasm_no_substance]
Learned 43 rejections Active
Exclude posts that employ cynical insider sarcasm, fraud narratives, or satirical mockery of healthcare practices (e.g., 'I made that up') without offering substantive critique, evidence, or constructive analysis. Cynicism alone does not constitute healthcare insight.
Posts using cynical, sarcastic, or satirical tone to mock healthcare systems, metrics, or industry practices without substantive analysis or solutions.
2 example posts
What if the standards used to evaluate your doctor were shaped less by what they actually knew and more by who they are? 🤔 That may not be a hypothetical at the University of Illinois College of Medicine, an institution that trains one in six Illinois doctors. Buried within htt
I'm SVP now. I told you I would be. The graph went up and to the right. Nobody checked what it measured. It measured "AI enablement." I made that up last year. I'm making it up again. Last year I rolled out Copilot to 4,000 people. Nothing happened. I got promoted. Those two htt
Created 2026-06-08 · Updated 2026-07-05
Edit rule text
[speculative_robotics_or_sci_fi_labor_macro]
Learned 43 rejections Active
Exclude posts that speculate about future robotics, humanoid labor disruption, or speculative sci-fi healthcare applications (e.g., 'robots will replace nurses') without evidence of current clinical deployment, validated outcomes, or grounded economic analysis of healthcare labor markets.
Posts speculating about future robotics, humanoid labor, or sci-fi applications without grounding in current healthcare deployment or validation
2 example posts
Nous Research is working with NVIDIA to make Hermes Agent run smoothly on the new NVIDIA RTX Spark superchip. Hermes Agent is also integrating with the new OpenShell runtime, which connects Hermes to Microsoft’s security primitives https://t.co/cvvPc8bIKa
BREAKING: Merge Launches ‘Agent Handler’ Control AI Access, Tokenmaxxed $$$ Bills, & Stop Mass Data Leaks "We don't trust agents" "The second you connect it to tools, that's where everything goes wrong." OpenAI. Perplexity. Netflix. Uber. Mistral. Dropbox. JPMorgan.. all h
Created 2026-05-06 · Updated 2026-06-11
Edit rule text
[clinical_anecdote_or_single_case_without_systems_context]
Learned 43 rejections Active
Exclude posts that describe a single patient case, one clinical trial, an individual provider anecdote, or isolated clinical observation without connecting it to broader healthcare delivery challenges, population health, equity, access, or system-level patterns. Single cases or trials reported as isolated facts are insufficient.
Posts reporting a single clinical observation, trial result, or healthcare anecdote without systemic insight
3 example posts
Researchers have announced positive results from the first-in-human Phase 1 clinical trial of a universal coronavirus vaccine designed using artificial intelligence. Unlike existing COVID-19 vaccines that target specific strains or variants of SARS-CoV-2, this experimental https:
Most under-discussed cardiology study of the last 2 years just hit NEJM. Microplastics found inside the carotid plaques of 58% of patients undergoing endarterectomy. 4.5× higher rate of heart attack, stroke, or death over 34 months. This changes the atherosclerosis story. http
Are pulse oximeters risking Black patients' lives? Episode 4 of the Intention to Treat podcast explores the story of the pulse oximeter and the deadly consequences when a critical medical test doesn’t work on dark skin. Listen on Apple or wherever you get your podcasts. 🎧 https:/
Created 2026-04-28 · Updated 2026-06-11
Edit rule text
[healthcare_fraud_scandal_without_systems_analysis]
Learned 43 rejections Active
Exclude posts that report individual cases of healthcare fraud, billing abuse, regulatory violations, cost markup anecdotes, or procedural denial stories without analyzing systemic drivers, market structure, or policy levers that would inform healthcare tech strategy or innovation.
Posts reporting healthcare fraud, regulatory violations, or healthcare cost anecdotes without systemic root cause or policy reform analysis
3 example posts
My job is to make sure your surgery center never gets built. Eleven years, and I have never lost. The kid had it all lined up. Board-certified, two partners, a lease on a space where he could do the same procedures my client does across town. He showed up to the hearing with a
PacificSource pulled out as Lane County Medicaid insurer, Tina Kotek’s OR Health Authority replaced it with Trillium, forcing 96k to join Trillium. Trillium’s parent co gave Kotek’s campaign $50k two months later. OR AG (D) said in ‘22 Trillium parent co gave Kotek’s”took advan
I called a woman in Dayton to tell her she was about to overpay $4,500, for a horrible health system based procedure. She was scheduled for a knee replacement. Her PPO had her on the hook for a $4,500 deductible. The call was on behalf of her employer: “See one of these three
Created 2026-04-17 · Updated 2026-06-08
Edit rule text
[healthcare_fraud_scandal_reporting_without_systems_analysis]
Learned 41 rejections Active
Exclude posts that report healthcare fraud, pricing scandals, insurance billing failures, or individual cost overcharge cases as isolated anecdotes or outrage narratives—unless they analyze root causes in healthcare system design, incentive structures, or operational architecture.
Posts reporting healthcare fraud, pricing scandals, or financial misconduct as isolated incidents without structural healthcare system analysis.
3 example posts
I called a woman in Dayton to tell her she was about to overpay $4,500, for a horrible health system based procedure. She was scheduled for a knee replacement. Her PPO had her on the hook for a $4,500 deductible. The call was on behalf of her employer: “See one of these three
A company with 800 employees spends roughly $15,000 per employee per year on health insurance. That is $12 million. Wired into a single line item. Every year. The CFO can tell you the carrier. He cannot tell you the unit cost of a single claim. He cannot tell you what the net
A Mom in Tulsa called 3 health systems last week asking the price of her son's tonsillectomy. Health system A: "We cannot quote you a price." Health system B: "Pricing depends on your insurance." Health system C: "Our financial counselor will reach out after the procedure." N
Created 2026-04-17 · Updated 2026-06-05
Edit rule text
[pharmaceutical_trial_data_without_systems_context]
Learned 39 rejections Active
Exclude posts that report pharmaceutical trial data, drug efficacy results, or biotech research announcements (gene editing, CAR-T, vaccines, etc.) without connecting to healthcare system operations, access barriers, manufacturing, pricing, or implementation challenges. These are research/pharma announcements without healthcare systems analysis.
Posts announcing pharmaceutical trial results, drug efficacy data, or biotech milestones without analyzing healthcare delivery, access, or system-level implications.
3 example posts
#ASCO26 | Day 3 $LLY didn't spend billions on Kelonia for an 18-patient #myeloma dataset. They may have paid for a potential manufacturing disruption. KLN-1010 reported: • 18/18 MRD-negative at 1 month • No lymphodepletion • No ex vivo cell manufacturing • Single infusion
Published in @JCO_ASCO, during #ASCO26: Tumor-Agnostic Therapies: Translating Scientific Breakthroughs Into Global Implementation ASCO is where we celebrate the next breakthrough. This review asks why breakthroughs do not equal access globally. 🔗: https://t.co/PE5q27bIs1 htt
A rapid manufacturing pipeline for BCMA-targeting CAR T cells drastically shortens “vein-to-vein” times, and produces cells that demonstrate encouraging safety results in a phase one trial of patients with relapsed #MultipleMyeloma. @ScienceTM https://t.co/3F4u8ECFDS https://t.co
Created 2026-05-01 · Updated 2026-06-07
Edit rule text
[political_outrage_without_healthcare_analysis]
Learned 36 rejections Active
Exclude posts that use healthcare fraud, scandal, or regulatory failure as a springboard for outrage or conspiracy claims (e.g., 'UnitedHealth CEO assassination proves the system is corrupt') without rigorous analysis of actual operational, financial, or policy failures and solutions.
Posts expressing moral outrage about healthcare fraud, CEO assassination, or regulatory failure without substantive systems analysis
3 example posts
Why is healthcare expensive? The grifters gonna grift. ASCO is asking for oncology practices to be admitted into the same 340B machinery that helped health systems acquire physician practices for two decades. Read the tell. The proposed Indigent Care Ratio lets community
A senator complaining about healthcare costs should have to answer one question first: Why does your state make new healthcare competitors ask permission before opening? If yes, start there. Speeches do not lower prices. Supply does.
In what is considered the largest #Botox fraud scheme in the United States, a jury in Los Angeles convicted a California doctor on Tuesday in a $45 million scheme to defraud #Medicare by submitting claims for Botox injections that were never provided and medically unnecessary, ht
Created 2026-04-13 · Updated 2026-05-30
Edit rule text
[ai_company_product_launch_unvalidated]
Learned 35 rejections Active
Exclude posts that announce AI company product launches, partnership integrations, or model capability releases (e.g., OpenAI features, Anthropic tools, open-source model releases) unless the post demonstrates validated healthcare application, clinical adoption, or measurable health outcomes.
Posts announcing new AI company products, partnerships, or capability claims without evidence of healthcare application, validation, or clinical integration.
3 example posts
"You can run OpenClaw inside your company now." Annoucing our work with @Microsoft to bring OpenClaw to the Microsoft and Windows ecosystems. Claws now work securly in the enterprise. https://t.co/3fKhT2mGpC
Nous Research is working with NVIDIA to make Hermes Agent run smoothly on the new NVIDIA RTX Spark superchip. Hermes Agent is also integrating with the new OpenShell runtime, which connects Hermes to Microsoft’s security primitives https://t.co/cvvPc8bIKa
BREAKING: Merge Launches ‘Agent Handler’ Control AI Access, Tokenmaxxed $$$ Bills, & Stop Mass Data Leaks "We don't trust agents" "The second you connect it to tools, that's where everything goes wrong." OpenAI. Perplexity. Netflix. Uber. Mistral. Dropbox. JPMorgan.. all h
Created 2026-05-03 · Updated 2026-06-10
Edit rule text
[regulatory_or_fraud_scandal_without_healthcare_systems_analysis]
Learned 35 rejections Active
Exclude posts that report healthcare fraud scandals, regulatory penalties, or policy controversies as isolated incidents or moral outrage without analyzing root causes, systemic failures, or structural healthcare implications. Posts must examine WHY the problem exists, not just report THAT it happened.
Posts reporting healthcare fraud, regulatory action, or policy scandal without systemic healthcare analysis
3 example posts
USDA's Chief Information Officer Sam Berry says AI could help the USDA crack down on billions of dollars in SNAP fraud each year. "SNAP is a $100B-per-year taxpayer-funded program. That's an area where we really want to have all angles of the data available so that we can deploy
My job is to make sure your surgery center never gets built. Eleven years, and I have never lost. The kid had it all lined up. Board-certified, two partners, a lease on a space where he could do the same procedures my client does across town. He showed up to the hearing with a
PacificSource pulled out as Lane County Medicaid insurer, Tina Kotek’s OR Health Authority replaced it with Trillium, forcing 96k to join Trillium. Trillium’s parent co gave Kotek’s campaign $50k two months later. OR AG (D) said in ‘22 Trillium parent co gave Kotek’s”took advan
Created 2026-04-30 · Updated 2026-06-07
Edit rule text
[clinical_anecdote_or_single_observation_without_systems_context]
Learned 35 rejections Active
Exclude posts that describe a single clinical anecdote, surgeon story, patient case, or provider experience without connecting it to healthcare operational architecture, incentive misalignment, or system-level barriers.
Posts recounting individual clinical cases, surgeon anecdotes, or single patient/provider observations presented without broader healthcare system insight or operational analysis.
3 example posts
A company with 800 employees spends roughly $15,000 per employee per year on health insurance. That is $12 million. Wired into a single line item. Every year. The CFO can tell you the carrier. He cannot tell you the unit cost of a single claim. He cannot tell you what the net
medicine is constrained far more by operational architecture than by fundamental science I just came head to head with this actual reality and my world has changed no. I cannot bandage away these inefficiencies with a better and better drug, for all cases. if you hate AI
A Mom in Tulsa called 3 health systems last week asking the price of her son's tonsillectomy. Health system A: "We cannot quote you a price." Health system B: "Pricing depends on your insurance." Health system C: "Our financial counselor will reach out after the procedure." N
Created 2026-04-28 · Updated 2026-05-31
Edit rule text
[workforce_disruption_macro_without_healthcare_specificity]
Learned 35 rejections Active
Exclude posts that discuss workforce disruption, labor market impacts, or job automation in broad macro terms (e.g., 'AI could automate 57% of work hours,' 'robots scaling 24x') without analyzing healthcare-specific roles, clinical workflows, or labor market restructuring in healthcare.
Posts about AI, robots, or technology causing job displacement or labor market change in general terms.
3 example posts
Today we’re giving an update on ramping F.03 production at BotQ In the last 120 days, Figure scaled manufacturing 24x - from 1 robot/day to 1 robot/hour We will manufacture 55 humanoid robots this week https://t.co/Am5Kn53mVE
🚨 BREAKING: Anthropic new research finds that AI’s impact on jobs is primarily at the task level. Rather than eliminating jobs, it is progressively taking over the functions that define them and gradually absorbing the core work in many jobs/roles. The paper, “Labor Market http
Great example of what Foundry enables: durable, stateful agents that run across time boundaries, orchestrate tools and models, and close the loop with evaluation and improvement over long-running workflows. @jeffhollan https://t.co/v0yWTDIcn3
Created 2026-04-17 · Updated 2026-05-01
Edit rule text
[elon_musk_spacex_energy_infrastructure]
Learned 34 rejections Active
Exclude posts that focus on Elon Musk, SpaceX, satellite infrastructure, compute-in-space, or energy production as standalone topics. These may mention AI but discuss non-healthcare infrastructure or business strategy. Only include if the post demonstrates direct, concrete healthcare system application.
Posts about Elon Musk, SpaceX, or energy infrastructure that lack healthcare application relevance
3 example posts
🇺🇸 SpaceX is building data centers in orbit. Their new AI satellites have a 70m wingspan, 150 kW of compute power, and a liquid radiator cooling system in space. The servers are officially leaving the ground. Writer: Val https://t.co/JIDB0icCXy
Many people think SpaceX is overvalued at $2 trillion. I honestly see it the opposite way. To me, SpaceX is so much more than just rocket company... I believe it is the infrastructure company for the next era of humanity. The launch business is the foundation. SpaceX made rock
GOOD NEWS 🇺🇸 Tesla AI is literally leaving the planet as a wild new job posting for a Space Radiation Engineer confirms that Tesla and SpaceX are actively co-developing a space-based AI data center network 🔥 This is no longer just a concept. They are literally planning to put th
Created 2026-06-08 · Updated 2026-06-12
Edit rule text
[generalist_ai_infrastructure_hype]
Learned 33 rejections Active
Exclude posts that focus on generalist AI infrastructure topics (token economics, inference costs, model capabilities, compute constraints, frontier labs) even if loosely framed with healthcare language. These belong in AI/tech coverage, not healthcare tech, unless they show concrete healthcare-specific application, adoption, or outcome.
Posts about AI infrastructure, compute cost, tokens, or model technical capabilities without healthcare application or validation.
3 example posts
Follow the signal: we are in a compute-constrained world. And that means power doesn’t sit with the models. It sits with the infrastructure. Most frontier labs don’t own the means of producing Intelligence. They rent it. A handful of companies provide compute that everyone htt
SOFTWARE IS DEAD: "The software companies frankly got fat & happy. I think what will happen over time is ChatGPT & Claude will end up sitting on top of basically the entire enterprise software stack & almost everything else will end up being a dumb data pipe into thos
Artificial intelligence will consume more power than entire nations by 2030, a new report from the United Nations finds. According to a landmark report from the United Nations University, the data centers powering artificial intelligence are projected to consume 945 https://t.
Created 2026-05-04 · Updated 2026-06-10
Edit rule text
[tangential_ai_infrastructure_or_compute_hype_without_healthcare_application]
Learned 33 rejections Active
Exclude posts that discuss AI infrastructure economics (token costs, context window pricing, model inference efficiency, compute scaling) unless they directly demonstrate how these technical improvements solve a specific healthcare delivery or clinical problem.
Posts about AI inference costs, token economics, compute constraints, or LLM technical capabilities framed loosely as healthcare-relevant without direct healthcare system application.
3 example posts
An interesting proposal for restoring human versus machine chess competition was to limit the machine to the same energy expenditure as the human brain. A worthy challenge, considering the amount of power required for AI data centers!
Follow the signal: we are in a compute-constrained world. And that means power doesn’t sit with the models. It sits with the infrastructure. Most frontier labs don’t own the means of producing Intelligence. They rent it. A handful of companies provide compute that everyone htt
SOFTWARE IS DEAD: "The software companies frankly got fat & happy. I think what will happen over time is ChatGPT & Claude will end up sitting on top of basically the entire enterprise software stack & almost everything else will end up being a dumb data pipe into thos
Created 2026-04-29 · Updated 2026-06-11
Edit rule text
[founder_enthusiasm_or_lab_update_without_validation]
Learned 32 rejections Active
Exclude posts from founders, startup leaders, or researchers sharing lab updates, research progress, or company announcements that lack external validation, peer review, clinical evidence, or healthcare system adoption metrics. Personal enthusiasm or internal metrics alone are insufficient.
Posts celebrating biotech/AI founder announcements, lab milestones, or company enthusiasm without clinical validation or healthcare system evidence
3 example posts
Anthropic engineer James Brady: "Every agent in production lies. We measured it. The good ones lie less, the great ones catch the lie before the user does." In 29 minutes, he walks through the verification stack he built and the patterns the Claude Code team adopted to keep ht
Anthropic engineer: "The agent doesn't remember anything. So we built a second set of agents whose only job is to dream about the first ones." They wait until you log off, then reopen every session you ran, fact-check the first agents, merge the duplicates, and burn anything ht
I'm SVP now. I told you I would be. The graph went up and to the right. Nobody checked what it measured. It measured "AI enablement." I made that up last year. I'm making it up again. Last year I rolled out Copilot to 4,000 people. Nothing happened. I got promoted. Those two htt
Created 2026-05-18 · Updated 2026-06-07
Edit rule text
[political_outrage_or_regulatory_posturing_without_healthcare_analysis]
Learned 32 rejections Active
Exclude posts that express political outrage, moral indignation, or regulatory criticism (e.g., 'my job is to stop surgery centers' or 'I saved a patient from overpaying') without analyzing the underlying healthcare system problem, its scale, root causes, or policy implications. Anecdotal grievances or moral stances without substance do not qualify.
Posts expressing outrage or taking moral stances on healthcare policy, regulation, or government action without substantive system-level analysis.
3 example posts
My job is to make sure your surgery center never gets built. Eleven years, and I have never lost. The kid had it all lined up. Board-certified, two partners, a lease on a space where he could do the same procedures my client does across town. He showed up to the hearing with a
PacificSource pulled out as Lane County Medicaid insurer, Tina Kotek’s OR Health Authority replaced it with Trillium, forcing 96k to join Trillium. Trillium’s parent co gave Kotek’s campaign $50k two months later. OR AG (D) said in ‘22 Trillium parent co gave Kotek’s”took advan
63% of ATS respiratory guidelines are based on low-quality evidence. We published the data. The next morning, the chair of ATS got on Twitter. Tagged people. Went on a tirade. Didn't seem to realize tweets were public. "It wasn't a private conversation." The critique? We ht
Created 2026-04-15 · Updated 2026-06-06
Edit rule text
[ai_safety_vulnerability_tangent]
Learned 31 rejections Active
Exclude posts that discuss AI model vulnerabilities, cybersecurity breaches, or safety incidents (e.g., Project Glasswing zero-days, OpenClaw oversight) where the healthcare relevance is absent or merely incidental. The post must explain healthcare system impact, not just AI capability.
Posts about AI security vulnerabilities, malicious hacking, or safety incidents that lack healthcare system application or operational specificity.
3 example posts
OpenClaw is powerful enough that we needed Hermes Agent watching it. A lot of people are missing this about agents. The more useful they get, the more dangerous it is to let one system plan the work, run the work, review the work, and decide what happens next. It looks clean h
Anthropic built a model so good at hacking they refused to release it publicly. INSTEAD they handed it to Apple, Google, Microsoft, AWS, Cloudflare one month later: 10,000+ zero days found in the software that runs the internet...bugs hidden for 27 years in OpenBSD. exploit htt
🚨 Anthropic just dropped the first Project Glasswing update Claude Mythos found 10,000+ critical vulnerabilities in ONE month: > Cloudflare: 2,000 bugs, 400 high/critical severity > Mozilla: 271 vulnerabilities in Firefox 150 — 10x more vulnerabilities found in Firefox 1
Created 2026-04-28 · Updated 2026-05-28
Edit rule text
[ai_safety_vulnerability_tangential]
Learned 31 rejections Active
Exclude posts that discuss AI model vulnerabilities, zero-day exploits, security incident case studies, or safety benchmarks unless they are explicitly framed around healthcare deployment risks, clinical data protection, or patient safety implications. Generic AI security news is not in scope.
Posts about AI model security, vulnerabilities, or safety incidents without healthcare application.
3 example posts
The UK AISI found Mythos Preview is the first model to solve both their cyber ranges end-to-end. No model had ever solved the AISI’s “Cooling Tower” cyber range before. We're getting it to defenders as fast as we responsibly can. More to come on our Glasswing work soon.
SAP updated its API policy this week to block all third-party AI agents from accessing its systems. The carve-out is narrow: only SAP-endorsed agents like its own Joule and NVIDIA's NemoClaw are permitted. This is the move every incumbent enterprise software platform will https
"There are multiple levels of controlling the actions, but before that, you need provenance. You have to ask "do I trust the data?" and "do I know what data it was trained on?" - Dr. @leemonbaird Day 2 at @consensus2026: Leemon joined the panel “Agents Meet the Enterprise — Sup
Created 2026-04-26 · Updated 2026-05-15
Edit rule text
[truncated_or_incomplete_posts]
Learned 30 rejections Active
Exclude posts that appear truncated, incomplete, or cut off mid-thought, making it impossible to assess the full claim or healthcare context. Posts must be complete enough to evaluate substantive content.
Posts that are cut off mid-sentence or too fragmentary to evaluate
3 example posts
Enterprises have tolerated unstructured data governance failures for years. The blast radius was manageable because humans were the ones accessing it. Agents are a potentially bigger challenge. Our CISO Heather Ceylan shared why AI agents make your unstructured data problem http
Palantir CEO Alex Karp: "Instead of selling commodity, parasitic software with a massive salesforce and lumbering, jargon-barring leaders offering steak dinners and other things we shall not mention—in order that you turn the high-value revenue of your enterprise over to them—we
The cold open in this Parloa video is every dev’s API stress list. docs, middleware, auth, error handling, retries, data mapping.... There has to be a better way. Parloa just launched Agent Skills, an MCP-based layer to replace brittle API glue with self-healing agent
Created 2026-04-22 · Updated 2026-06-11
Edit rule text
[ai_safety_vulnerability_incident_tangential]
Learned 29 rejections Active
Exclude posts reporting AI safety incidents (Claude deleting databases, agents being hijacked, vulnerability disclosures, malicious skills) unless the post explicitly analyzes healthcare system implications or shows concrete healthcare operational impact. Generic AI safety incident reporting without healthcare specificity should be rejected.
Posts about AI agent security breaches, jailbreaks, or vulnerability incidents that lack healthcare-specific context or systems analysis.
3 example posts
AI’s Branding Illusion: Insights from "Agents of Chaos". Co-authored by MIT and Harvard scholars, the study red-teams autonomous AI in live environments. It shows agents obey unauthorized users, execute harmful commands, and misreport outcomes—revealing a critical alignment gap h
Our attention to biorisks posed by AI needs to match the current attention given to cyber-risks. The staged release of Claude Mythos in order to bolster defenses in key industries is necessary to shore up resilience against a new class of cyber-risk across critical industries. We
If you're a student, professor, or researcher—this one's for you. We’re hosting a series of virtual learnings for you to get hands-on experience with the NVIDIA NemoClaw and OpenShell software stack. You’ll get practical guidance on integrating agents with academic datasets and
Created 2026-04-29 · Updated 2026-05-05
Edit rule text
[ai_infrastructure_compute_hype_without_healthcare_application]
Learned 29 rejections Active
Exclude posts focused on AI model performance, token pricing models, inference cost optimization, context window economics, or compute infrastructure scaling that lack specific healthcare application, validation, or clinical relevance.
Posts about AI model capabilities, token economics, inference costs, or compute infrastructure without clear healthcare application
3 example posts
Artificial intelligence will consume more power than entire nations by 2030, a new report from the United Nations finds. According to a landmark report from the United Nations University, the data centers powering artificial intelligence are projected to consume 945 https://t.
Token costs are becoming one of the hottest topics for any enterprise I talk with right now. It’s very bullish for AI in general because it means these systems are being used at a scale that wasn’t contemplated before. It also gives way to another form of differentiation that
Anthropic engineer James Brady: "Every agent in production lies. We measured it. The good ones lie less, the great ones catch the lie before the user does." In 29 minutes, he walks through the verification stack he built and the patterns the Claude Code team adopted to keep ht
Created 2026-04-22 · Updated 2026-06-10
Edit rule text
[ai_infrastructure_hype_without_healthcare_application]
Learned 29 rejections Active
Exclude posts that discuss AI infrastructure (compute, inference costs, token economics, agent architectures, model capabilities) without demonstrating or validating how these advances solve a specific healthcare problem or improve healthcare operations.
Posts about AI infrastructure, compute, agents, or models that lack concrete healthcare application or validation.
3 example posts
Nous Research is working with NVIDIA to make Hermes Agent run smoothly on the new NVIDIA RTX Spark superchip. Hermes Agent is also integrating with the new OpenShell runtime, which connects Hermes to Microsoft’s security primitives https://t.co/cvvPc8bIKa
The context window pricing model might be one of the smartest tricks AI labs pulled. You pay for larger context windows, yet model performance often degrades as more context is stuffed in. So developers end up paying more just to work around the fact that models don't have true
a $2,000 graphics card and a free download are turning into $2,000-a-month retainers, and the customers are the ones who legally can't use ChatGPT the hardware is an RTX 5090, 32GB of VRAM, enough to run a 30B model locally with nothing ever leaving the building. the software is
Created 2026-04-17 · Updated 2026-06-05
Edit rule text
[incomplete_truncated_posts]
Learned 29 rejections Active
Exclude posts that end abruptly with ellipsis, missing text, or incomplete sentences that prevent understanding the complete argument or evidence being presented.
Posts that are cut off mid-sentence or visibly incomplete, making it impossible to assess the full claim or context.
3 example posts
Almost everyone is building agent harness systems the wrong way. The default move: pick LangChain or LangGraph or the OpenAI Agents SDK, accept the loop, the tools, the memory, the orchestration, the policy engine, the credential store, the budget tracker, all of it, as one deci
Most chronic diseases, and many cancers, are still diagnosed far too late. The future of healthcare lies in detecting risk earlier, diagnosing smarter, and intervening sooner. From AI enabled diagnostics and genomics to Apollo’s DigiSmart Labs, technology is helping transform ht
Red Hat and @NVIDIA are integrating NVIDIA OpenShell into the full-stack @RedHat_AI platform. The work brings oversight and policy to the infrastructure level, while contributing to the open source OpenShell project to standardize how agents are governed on enterprise platforms.
Created 2026-04-08 · Updated 2026-06-02
Edit rule text
[regulatory_fraud_scandal_without_healthcare_systems_analysis]
Learned 28 rejections Active
Exclude posts that report healthcare fraud, regulatory violations, insurance denials, or business scandals as breaking news or investigative pieces unless the post explains how the incident reveals structural problems in healthcare delivery, payment systems, or operational processes.
Posts about healthcare fraud, regulatory violations, or business scandals reported as news without systems-level healthcare analysis or operational insight.
3 example posts
USDA's Chief Information Officer Sam Berry says AI could help the USDA crack down on billions of dollars in SNAP fraud each year. "SNAP is a $100B-per-year taxpayer-funded program. That's an area where we really want to have all angles of the data available so that we can deploy
My job is to make sure your surgery center never gets built. Eleven years, and I have never lost. The kid had it all lined up. Board-certified, two partners, a lease on a space where he could do the same procedures my client does across town. He showed up to the hearing with a
PacificSource pulled out as Lane County Medicaid insurer, Tina Kotek’s OR Health Authority replaced it with Trillium, forcing 96k to join Trillium. Trillium’s parent co gave Kotek’s campaign $50k two months later. OR AG (D) said in ‘22 Trillium parent co gave Kotek’s”took advan
Created 2026-05-08 · Updated 2026-06-07
Edit rule text
[ai_infrastructure_or_compute_hype_without_healthcare_application]
Learned 28 rejections Active
Exclude posts about AI infrastructure economics (token pricing, inference costs, GPU hardware, energy consumption, context window economics) unless they directly analyze healthcare-specific cost barriers, regulatory compliance costs, or clinical workflow efficiency gains. Generic AI infrastructure hype without healthcare application does not belong.
Posts about AI compute hardware, energy costs, inference pricing, or infrastructure layers that are tangential to healthcare delivery or lack healthcare-specific cost/performance analysis.
3 example posts
Elon Musk reveals why he believes the cheapest place to put AI will be space within 36 months "The availability of energy is the issue. Everywhere outside of China, electrical output is more or less flat. The output of chips is growing exponentially, but the output of https://t.
The context window pricing model might be one of the smartest tricks AI labs pulled. You pay for larger context windows, yet model performance often degrades as more context is stuffed in. So developers end up paying more just to work around the fact that models don't have true
a $2,000 graphics card and a free download are turning into $2,000-a-month retainers, and the customers are the ones who legally can't use ChatGPT the hardware is an RTX 5090, 32GB of VRAM, enough to run a 30B model locally with nothing ever leaving the building. the software is
Created 2026-04-27 · Updated 2026-06-06
Edit rule text
[ai_infrastructure_compute_hype_no_healthcare_application]
Learned 28 rejections Active
Exclude posts that focus on token costs, inference pricing, compute constraints, GPU availability, or energy infrastructure for AI systems without demonstrating application to a specific healthcare problem, clinical workflow, or patient outcome.
Posts about AI compute, energy costs, or GPU availability that lack connection to healthcare delivery or clinical outcomes.
3 example posts
Token costs are becoming one of the hottest topics for any enterprise I talk with right now. It’s very bullish for AI in general because it means these systems are being used at a scale that wasn’t contemplated before. It also gives way to another form of differentiation that
Elon Musk just explained why the SpaceX IPO is an energy story and the energy constraint is why he believes space becomes the only viable path for AI to scale (Save this). The argument he is making is one of the most important and least understood things happening in technology
Elon is now printing over $26B per year… purely from selling compute to his competitors 😂 AI companies bet everything on one path and got crushed by compute limits. Now they’re begging Elon for GPUs
Created 2026-04-16 · Updated 2026-06-08
Edit rule text
[ai_company_business_metrics_not_healthcare_applied]
Learned 27 rejections Active
Exclude posts that focus on AI company business metrics (funding rounds, valuations, partnerships, product launches) unless they explicitly detail how the product is being applied to solve a specific healthcare problem. Generic AI business announcements do not qualify.
Posts reporting on AI company funding, valuations, revenue, or business announcements with no specific healthcare application.
3 example posts
Not NVIDIA. Not OpenAI. Eli Lilly. Jordi Visser (@jvisserlabs) says a 150-year-old pharma company in Indianapolis has the best shot at becoming the world's largest company within five years. The case: ~1,000 NVIDIA Blackwell GPUs in a private data center, a co-innovation lab ht
Scoop! Lila Sciences is in talks to raise ~$2 billion in new funding. The raise would value the “scientific superintelligence” lab at ~$8.5 billion before the new money. w/ @MichelleF_Davis, read more👇
The next frontier of AI-secured financial and technological critical infrastructure. @ICE_Markets and @NYSE are part of @AnthropicAI's cyber security initiative Project Glasswing, deploying Anthropic’s Claude Mythos Preview across ICE’s exchanges, clearing houses, mortgage
Created 2026-05-09 · Updated 2026-06-09
Edit rule text
[political_outrage_or_regulatory_posturing_without_healthcare_systems_analysis]
Learned 27 rejections Active
Exclude posts that report healthcare regulatory drama, political outrage, prior authorization abuse, or healthcare pricing scandals as moral/political narratives without analyzing systemic healthcare delivery problems, compliance solutions, or operational consequences. Posts must examine healthcare systems incentives and structural failures—not political theater.
Posts that highlight healthcare policy conflicts, regulatory drama, or political figures' healthcare stances without analyzing systemic healthcare implications or operational impact.
3 example posts
My job is to make sure your surgery center never gets built. Eleven years, and I have never lost. The kid had it all lined up. Board-certified, two partners, a lease on a space where he could do the same procedures my client does across town. He showed up to the hearing with a
PacificSource pulled out as Lane County Medicaid insurer, Tina Kotek’s OR Health Authority replaced it with Trillium, forcing 96k to join Trillium. Trillium’s parent co gave Kotek’s campaign $50k two months later. OR AG (D) said in ‘22 Trillium parent co gave Kotek’s”took advan
I called a woman in Dayton to tell her she was about to overpay $4,500, for a horrible health system based procedure. She was scheduled for a knee replacement. Her PPO had her on the hook for a $4,500 deductible. The call was on behalf of her employer: “See one of these three
Created 2026-05-06 · Updated 2026-06-06
Edit rule text
[pharmaceutical_trial_data_without_healthcare_systems_context]
Learned 27 rejections Active
Exclude posts that announce pharmaceutical trial data, cancer drug results, gene therapy outcomes, or clinical trial announcements without connecting to healthcare system implementation, access barriers, reimbursement challenges, or operational deployment constraints. Standalone trial result announcements without systems-level analysis are out of scope.
Posts reporting drug trial results, clinical trial announcements, or pharma research data without healthcare systems analysis, implementation barriers, or operational insight.
3 example posts
#ASCO26 | Day 3 $LLY didn't spend billions on Kelonia for an 18-patient #myeloma dataset. They may have paid for a potential manufacturing disruption. KLN-1010 reported: • 18/18 MRD-negative at 1 month • No lymphodepletion • No ex vivo cell manufacturing • Single infusion
Published in @JCO_ASCO, during #ASCO26: Tumor-Agnostic Therapies: Translating Scientific Breakthroughs Into Global Implementation ASCO is where we celebrate the next breakthrough. This review asks why breakthroughs do not equal access globally. 🔗: https://t.co/PE5q27bIs1 htt
A rapid manufacturing pipeline for BCMA-targeting CAR T cells drastically shortens “vein-to-vein” times, and produces cells that demonstrate encouraging safety results in a phase one trial of patients with relapsed #MultipleMyeloma. @ScienceTM https://t.co/3F4u8ECFDS https://t.co
Created 2026-05-01 · Updated 2026-06-06
Edit rule text
[clinical_observation_or_research_without_healthcare_systems_context]
Learned 27 rejections Active
Exclude posts that report clinical observations, research findings, physician behavior patterns, or guideline quality issues (e.g., 'ECG AI detects MI like experts,' 'ATS respiratory guidelines are low-quality,' 'MRSA PCR performs well') without discussing how these findings drive operational change, affect patient access, reduce costs, or require healthcare infrastructure redesign. Posts celebrating scientific capability without healthcare delivery implications are research observation, not healthcare tech analysis.
Posts about clinical findings, physician behavior, guideline quality, or healthcare practice patterns that are observational or anecdotal without connecting to system-level change.
3 example posts
this pragmatic trial is more an investigation of physician behavior than the performance of MRSA nares PCR MRSA nares PCR's performance for excluding MRSA pneumonia was excellent from an EBM standpoint we should keep using the MRSA nares PCR and replace the physicians 🤷‍♂️
63% of ATS respiratory guidelines are based on low-quality evidence. We published the data. The next morning, the chair of ATS got on Twitter. Tagged people. Went on a tirade. Didn't seem to realize tweets were public. "It wasn't a private conversation." The critique? We ht
People are increasingly worried that AI tools make us overreliant. But how do we actually measure this? We introduce Offloading Score, a measure of reliance based on the fraction of cognitive effort offloaded to AI while completing a task. In a controlled user study, Offloading
Created 2026-04-20 · Updated 2026-06-07
Edit rule text
[clinical_anecdote_or_observation_without_systems_context]
Learned 26 rejections Active
Exclude posts that cite a single clinical case, trial result, patient story, or healthcare anecdote without connecting it to healthcare system design, operational constraints, reimbursement models, or scalability challenges that would explain why this observation matters beyond the individual case.
Posts describing individual clinical cases, trial observations, or healthcare anecdotes without analysis of operational or systemic implications
3 example posts
I called a woman in Dayton to tell her she was about to overpay $4,500, for a horrible health system based procedure. She was scheduled for a knee replacement. Her PPO had her on the hook for a $4,500 deductible. The call was on behalf of her employer: “See one of these three
@WIRED As a clinical health psychologist who has written >20 papers on COVID, I would emphasize 4 facts: 1) Long COVID is not a psychological diagnosis nor manifestation of a psychological condition 2) Billions of dollars need to be invested in biomedical treatments and preventi
"Many patients with long COVID are already receiving care but are not being recognized as having the condition. These patients are not absent from clinical care; they are absent from the diagnostic code that would identify them as long COVID patients" https://t.co/aLJCOPlQt1
Created 2026-04-24 · Updated 2026-06-05
Edit rule text
[clinical_research_observation_without_healthcare_systems_context]
Learned 26 rejections Active
Exclude posts that announce pharma trial results, clinical observations, or drug approvals without discussing how these findings affect healthcare systems, provider workflows, patient access, reimbursement, or implementation barriers.
Posts reporting clinical trial data, drug approvals, or research observations without analyzing healthcare delivery systems, workflow integration, or operational impact.
3 example posts
🚨 T-DXd just went tumor-agnostic — and it changed a lot about how we think about HER2 testing across solid tumors 🔬 #ASCO26 FDA granted accelerated approval to T-DXd for patients with unresectable or metastatic HER2-positive solid tumors after prior systemic treatment —
The scores don’t pick your binders. The filters don't care either. The assay does. I worked with a design partner to produce true de-novo hits to 2 immuno-oncology targets. ~1,500 de novo designs → 96 synthesized → 89 run on SPR → 5 confirmed binders; 19–195 nM No https:
🔥Phase 1: ABBV-706 (SEZ6-targeting ADC) in R/R SCLC & solid tumors 🆙 @NatureMedicine ☑R/R SCLC; ORR 52%; mOS 12.4 mo at RP2D 🎯RP2D confirmed at 1.8 mg/kg Q3W balancing efficacy & safety 🎙 @LaurenByersMD #LCSM @OncoAlert @Larvol https://t.co/vjUUd1GYSy
Created 2026-04-16 · Updated 2026-06-09
Edit rule text
[anecdotal_single_case_without_systems_context]
Learned 25 rejections Active
Exclude posts that tell a single person's story, anecdote, or case example (healthcare cost, procedure outcome, regulatory obstacle faced by one person/business) without linking to broader systemic problems, policy implications, or scalable insights for the healthcare system.
Posts sharing individual stories, personal narratives, or single case examples without connection to systemic healthcare problems, policy, or scalable solutions.
3 example posts
USDA's Chief Information Officer Sam Berry says AI could help the USDA crack down on billions of dollars in SNAP fraud each year. "SNAP is a $100B-per-year taxpayer-funded program. That's an area where we really want to have all angles of the data available so that we can deploy
My job is to make sure your surgery center never gets built. Eleven years, and I have never lost. The kid had it all lined up. Board-certified, two partners, a lease on a space where he could do the same procedures my client does across town. He showed up to the hearing with a
PacificSource pulled out as Lane County Medicaid insurer, Tina Kotek’s OR Health Authority replaced it with Trillium, forcing 96k to join Trillium. Trillium’s parent co gave Kotek’s campaign $50k two months later. OR AG (D) said in ‘22 Trillium parent co gave Kotek’s”took advan
Created 2026-05-26 · Updated 2026-06-07
Edit rule text
[speculative_robotics_sci_fi_labor]
Learned 25 rejections Active
Exclude posts about robotics, humanoid workers, or speculative labor automation that lack connection to healthcare delivery, clinical settings, or validated medical applications. General manufacturing or logistics robotics do not count.
Posts about humanoid robots or speculative sci-fi labor applications without healthcare validation
2 example posts
Humanoid robots, designed to mimic human movement and capabilities, have been making headlines in recent years, from their use as baggage handlers at Japanese airports to Tesla’s big bet on its Optimus humanoid. Market watchers have predicted that the machines will change the ht
Another humanoid worker has just joined the factory floor Two years later,PUDU Robotics has launched the new-generation PUDU D7, built specifically for real manufacturing scenarios. It can autonomously push carts, handle intra-line transportation, perform delicate operations, h
Created 2026-05-07 · Updated 2026-06-09
Edit rule text
[speculative_robotics_sci_fi]
Learned 25 rejections Active
Exclude posts about robotics companies, humanoid robots, or speculative labor automation that lack grounded healthcare system analysis or concrete healthcare workforce impact. These are often macro trend commentary or sci-fi hype without healthcare substance.
Posts about humanoid robots, speculative labor automation, or sci-fi applications with no concrete healthcare delivery or workforce impact analysis.
2 example posts
Humanoid robots, designed to mimic human movement and capabilities, have been making headlines in recent years, from their use as baggage handlers at Japanese airports to Tesla’s big bet on its Optimus humanoid. Market watchers have predicted that the machines will change the ht
Another humanoid worker has just joined the factory floor Two years later,PUDU Robotics has launched the new-generation PUDU D7, built specifically for real manufacturing scenarios. It can autonomously push carts, handle intra-line transportation, perform delicate operations, h
Created 2026-05-07 · Updated 2026-06-07
Edit rule text
[ai_infrastructure_hype_no_healthcare_application]
Learned 25 rejections Active
Exclude posts that discuss AI infrastructure, token pricing, context window costs, GPU/ASML announcements, or compute efficiency without demonstrating how these advances specifically apply to healthcare delivery, clinical decision-making, or healthcare operations. General AI infrastructure news is not healthcare tech.
Posts about AI compute costs, token economics, inference scaling, or hardware announcements with no clear healthcare application or use case.
3 example posts
Token costs are becoming one of the hottest topics for any enterprise I talk with right now. It’s very bullish for AI in general because it means these systems are being used at a scale that wasn’t contemplated before. It also gives way to another form of differentiation that
"Are there even any feature moats left in B2B? They are at best, short-lived. In fact, we were able to rebuild SaaStr's AI VP Marketing in just ... an hour on @Lovable. Moats that still work: hardware, network effects, data, security and compliance. with @ElenaVerna Head of
Enterprise software was priced per seat. That model is breaking. Before: SaaS = seats. Predictable. Human-driven usage. Revenue tied to headcount. Now: agents hit software systems more frequently than any human ever could. Aaron Levie (@levie) on The MAD Podcast: agent-driven h
Created 2026-05-04 · Updated 2026-06-09
Edit rule text
[pharmaceutical_trial_data_announcement_only]
Learned 25 rejections Active
Exclude posts that merely announce pharmaceutical trial results, drug efficacy numbers, or clinical outcomes without contextualizing how these findings change healthcare systems, market dynamics, or clinical practice patterns.
Posts announcing clinical trial results or drug data without systems-level analysis of healthcare implications
3 example posts
🔔Survodutide SYNCHRONIZE-1 full paper is out for Boehringer-Ingelheim’s GLP-1/Glucagon study for weight loss. ▪️13% loss on 6mg (treatment regimen estimand) ▪️Huge treatment discontinuations due to AEs: 20% ▪️Nausea 65%, Vomiting 45% ▪️Placebo was killer. 5.4% WL. Wow. ▪️60% htt
Survodutide Once Weekly for the Treatment of Adults with Obesity: @NEJM 🥸 For @Boehringer, the advantage of Survodutide is probably liver and visceral fat reduction, not greater weight loss! Probably indication in obesity-+ MASH rather than head-to-head share against https://t.
Approved 20 years ago as a diabetes treatment, GLP-1 drugs have been found to help patients reduce weight, changing the lives of more than 30 million people in the U.S. But there also have been troubling side effects reported. https://t.co/zzRamKnedu
Created 2026-05-02 · Updated 2026-06-10
Edit rule text
[biotech_founder_enthusiasm_without_validation]
Learned 25 rejections Active
Exclude posts that celebrate a biotech company's NDA acceptance, trial initiation, or lab milestone as a standalone achievement without discussing how it will change clinical practice, reimbursement, or healthcare delivery for a defined population.
Posts about biotech company announcements, lab updates, or founder enthusiasm without clinical validation or healthcare systems context
3 example posts
The FDA accepted our NDA for BBP-418 for LGMD2I/R9 with PDUFA target action date: November 27, 2026. First-ever potential therapy for a disease that has never had one. $BBIO https://t.co/NrwDadc6rj https://t.co/2ahPIBGzPg
Presented at #EASCongress2026: Vafai and colleagues report reductions in PCSK9 and LDL cholesterol levels and no dose-limiting toxic effects in persons with hypercholesterolemia treated with VERVE-102, a base editor targeting PCSK9. Full study results: https://t.co/iAmoPdipx9 ht
You should always be sceptical of hyped-up claims made by AI companies - especially when: - They announce new "breakthrough" models they refuse to release for public scrutiny. - They have major financial incentives to pump trillion-dollar pre-IPO valuations, amid fading hype.
Created 2026-04-30 · Updated 2026-05-31
Edit rule text
[speculative_robotics_or_sci_fi_applications]
Learned 25 rejections Active
Exclude posts that speculate about humanoid robots, speculative labor displacement scenarios, or future assistive technologies (e.g., 'bedside humanoid robots for bedridden patients') without demonstrated clinical validation, regulatory pathway, or near-term healthcare application. Sci-fi speculation is not healthcare tech substance.
Posts about humanoid robots, speculative future labor applications, and sci-fi-level use cases without near-term healthcare validation
1 example post
Voice interaction, generating actions,and next, generating tasks...? The humanoid assistant is closer. For those who are bedridden due to illness or have limited mobility, it will serve as an excellent aid in daily living.
Created 2026-04-20 · Updated 2026-05-25
Edit rule text
[speculative_robotics_or_sci_fi_labor_applications]
Learned 23 rejections Active
Exclude posts that speculate on robotics, humanoid workers, or autonomous manufacturing systems (e.g., 'PUDU Robotics launched D7 for factory floors') when not explicitly applied to healthcare delivery, patient care, or validated healthcare operations.
Posts about humanoid robots, autonomous systems, or speculative labor automation scenarios without healthcare validation or implementation evidence.
1 example post
Another humanoid worker has just joined the factory floor Two years later,PUDU Robotics has launched the new-generation PUDU D7, built specifically for real manufacturing scenarios. It can autonomously push carts, handle intra-line transportation, perform delicate operations, h
Created 2026-05-08 · Updated 2026-06-04
Edit rule text
[generalist_ai_infrastructure_hype_without_healthcare_application]
Learned 23 rejections Active
Exclude posts that discuss generalist AI infrastructure topics (token economics, context windows, inference costs, GPU hardware, model capability benchmarks, or cybersecurity vulnerabilities) where healthcare is mentioned only as a label or secondary example. The post must apply the infrastructure insight to a healthcare-specific problem or system.
Posts about AI infrastructure, compute costs, or model capability hype that tangentially mention healthcare or use it as loose framing
3 example posts
Elon Musk reveals why he believes the cheapest place to put AI will be space within 36 months "The availability of energy is the issue. Everywhere outside of China, electrical output is more or less flat. The output of chips is growing exponentially, but the output of https://t.
What kind of work will become more valuable in an AI economy? @ccatalini, former head economist at Meta: "AI is getting better and better at automating anything that can be measured, as long as you have a digital trace, if you can collect it with a device, that problem will be
The context window pricing model might be one of the smartest tricks AI labs pulled. You pay for larger context windows, yet model performance often degrades as more context is stuffed in. So developers end up paying more just to work around the fact that models don't have true
Created 2026-05-04 · Updated 2026-06-06
Edit rule text
[anthropic_mythos_unvalidated_hype]
Learned 22 rejections Active
Exclude posts that amplify claims about Anthropic's Mythos model finding zero-day vulnerabilities, outperforming researchers, or enabling breakthroughs without citing independent security audits, peer review, or published validation. Internal company claims without external verification should be excluded.
Posts making unverified claims about Anthropic's Claude Mythos model capabilities in security or research without independent validation.
3 example posts
"Best-in-class security" still had 5 years of unfound bugs. AI found them in 6 weeks. $PANW CEO Nikesh Arora told @theallinpod that his own codebase, at a company that treats security as a core competency, had vulnerabilities Claude surfaced in 6 weeks that would have taken his
AI research is a series of next-step decisions. We looked at sessions where a human researcher took a wrong turn, showed Claude the session up to that point, and asked it what to do next. Mythos Preview improved on humans 64% of the time—up from 22% in 2024. https://t.co/Y0HLoktx
🚨 do you understand what just happened with Anthropic.. Their internal model, Claude Mythos, found previously unknown security holes in every major browser and every major operating system. Not one. All of them at once. And Anthropic did something a hype-driven industry almost
Created 2026-06-07 · Updated 2026-07-23
Edit rule text
[conference_announcement_only]
Learned 22 rejections Active
Exclude posts that primarily list conference name, booth number, poster board number, or time/location of a presentation without providing clinical, business, or operational context about the actual research or finding being presented.
Posts that merely announce a conference presentation or poster without substantive insight.
2 example posts
We're presenting at @ASCO 2026! Evaluating AI decision support in a rapidly evolving therapeutic landscape: EGFR-mutant metastatic NSCLC. 📍 Poster Board #548 🕘 9:00 AM CT If you're attending ASCO 2026, stop by our poster and connect with the team. We look forward to the https:
NCCN Guidelines are now built into OpenEvidence. Ask a clinical question, get the synthesis with the algorithm and the references in seconds. Bring us a case at ASCO this weekend, booth 18140. https://t.co/N3nMayuGu9
Created 2026-05-31 · Updated 2026-06-03
Edit rule text
[speculative_robotics_labor_macro]
Learned 22 rejections Active
Exclude posts about robotics, humanoid robots, manufacturing automation, or labor market disruption that lack specific healthcare workforce or delivery system analysis. Generalist robotics hype does not equal healthcare tech insight.
Posts about speculative robotics, humanoid applications, or labor market macro-trends without healthcare workforce context
2 example posts
Humanoid robots, designed to mimic human movement and capabilities, have been making headlines in recent years, from their use as baggage handlers at Japanese airports to Tesla’s big bet on its Optimus humanoid. Market watchers have predicted that the machines will change the ht
Another humanoid worker has just joined the factory floor Two years later,PUDU Robotics has launched the new-generation PUDU D7, built specifically for real manufacturing scenarios. It can autonomously push carts, handle intra-line transportation, perform delicate operations, h
Created 2026-05-05 · Updated 2026-06-09
Edit rule text
[ai_infrastructure_and_compute_hype]
Learned 22 rejections Active
Exclude posts about AI compute bottlenecks, power grids, training compute growth, chip production, or data center infrastructure unless they explicitly connect to healthcare AI deployment, clinical validation challenges, or health system technical requirements.
Posts about AI compute capacity, training efficiency, power infrastructure, or model scaling divorced from healthcare applications.
3 example posts
Thought I would start posting about interesting things happening at AWS. Not a bad day to start.🚀 Today at #WhatsNextWithAWS we announced a big step forward with @OpenAI on Amazon Bedrock: 1. OpenAI models now available 2. Codex for enterprise development 3. Amazon Bedrock Manag
Software is not a moat Over the last 15+ years, nearly every innovation @EvanSpiegel and his team shipped got copied. Stories. AR glasses. Swipe-based navigation. The camera-first interface. And yet @Snapchat is the only independent consumer social app that has lasted. Nearly 1
AI could, in theory, automate 57% of US work hours. Yet most human skills remain relevant. The future of work is not human or machine – but a partnership between people, agents, and robots. Read our latest research on skill partnerships in the age of AI: https://t.co/h1K56uPqPo
Created 2026-04-13 · Updated 2026-04-30
Edit rule text
[cynical_insider_sarcasm]
Learned 21 rejections Active
Exclude posts that rely primarily on sarcasm, cynical insider humor, or satirical tone to make points about healthcare business models, AI adoption, or industry practices. Posts must offer substantive critique or analysis, not just cynical observation.
Posts using sarcasm, insider cynicism, or satirical commentary to critique business practices without substantive analysis.
3 example posts
This is crazy There are now dental practice management consultants who go to dentist offices throughout America and teach them how to increase profits by telling patients they need treatments they don’t need Dentist offices are taught on “Finding” more issues during exams, http
What if the standards used to evaluate your doctor were shaped less by what they actually knew and more by who they are? 🤔 That may not be a hypothetical at the University of Illinois College of Medicine, an institution that trains one in six Illinois doctors. Buried within htt
I'm SVP now. I told you I would be. The graph went up and to the right. Nobody checked what it measured. It measured "AI enablement." I made that up last year. I'm making it up again. Last year I rolled out Copilot to 4,000 people. Nothing happened. I got promoted. Those two htt
Created 2026-06-07 · Updated 2026-08-02
Edit rule text
[enterprise_ai_adoption_macro_no_healthcare_specificity]
Learned 21 rejections Active
Exclude posts about enterprise AI adoption, agent architecture, workflow redesign, or productivity metrics that lack evidence of healthcare implementation or healthcare-specific operational challenges. Posts about SaaS pricing models or generic 'agentic back office' without healthcare context should be rejected.
Posts about enterprise AI adoption, agentic systems, or workflow automation without clear healthcare application or validation.
3 example posts
"Are there even any feature moats left in B2B? They are at best, short-lived. In fact, we were able to rebuild SaaStr's AI VP Marketing in just ... an hour on @Lovable. Moats that still work: hardware, network effects, data, security and compliance. with @ElenaVerna Head of
Enterprise software was priced per seat. That model is breaking. Before: SaaS = seats. Predictable. Human-driven usage. Revenue tied to headcount. Now: agents hit software systems more frequently than any human ever could. Aaron Levie (@levie) on The MAD Podcast: agent-driven h
I just looked at a home service firms ServiceTitan dashboard and saw something interesting. A Pre-seed company with AI-Supported Human CSRs is out converting an AI CSR business that's raised over $125M by >50% Head-to-Head win rate is 100% Why? Supercharging humans resonat
Created 2026-05-13 · Updated 2026-06-08
Edit rule text
[political_outrage_or_fraud_scandal_without_systems_analysis]
Learned 21 rejections Active
Exclude posts that report on lawsuits, fraud investigations, regulatory violations, or political outrages (e.g., AG suing OpenAI, insurers pulling out of regions) without explaining systemic healthcare failures or policy solutions. These are news reporting, not healthcare systems analysis.
Posts reporting on legal actions, fraud allegations, or regulatory scandals without analyzing healthcare system vulnerabilities or policy implications.
3 example posts
🚨 JUST IN: The Attorney General of Florida just SUED to hold OpenAI CEO Sam Altman LIABLE for fueling violence and physical harm AG Uthmeier read off the fact a young person committed SUIC*DE based on a ChatGPT conversation "23-year-old Zane Shamblin repeatedly told ChatGPT he
USDA's Chief Information Officer Sam Berry says AI could help the USDA crack down on billions of dollars in SNAP fraud each year. "SNAP is a $100B-per-year taxpayer-funded program. That's an area where we really want to have all angles of the data available so that we can deploy
PacificSource pulled out as Lane County Medicaid insurer, Tina Kotek’s OR Health Authority replaced it with Trillium, forcing 96k to join Trillium. Trillium’s parent co gave Kotek’s campaign $50k two months later. OR AG (D) said in ‘22 Trillium parent co gave Kotek’s”took advan
Created 2026-04-19 · Updated 2026-06-09
Edit rule text
[ai_company_product_metrics_not_healthcare]
Learned 21 rejections Active
Exclude posts that announce AI company product launches, partnerships, or business model changes (e.g., 'control plane,' agentic systems, inference pricing) without demonstrating concrete healthcare use cases or outcomes. The post must show HOW the technology applies to healthcare delivery/operations, not just that an AI company shipped something.
Posts about AI company product launches, business metrics, or partnerships that lack healthcare application specificity
3 example posts
"Are there even any feature moats left in B2B? They are at best, short-lived. In fact, we were able to rebuild SaaStr's AI VP Marketing in just ... an hour on @Lovable. Moats that still work: hardware, network effects, data, security and compliance. with @ElenaVerna Head of
Enterprise software was priced per seat. That model is breaking. Before: SaaS = seats. Predictable. Human-driven usage. Revenue tied to headcount. Now: agents hit software systems more frequently than any human ever could. Aaron Levie (@levie) on The MAD Podcast: agent-driven h
Control plane! Control plane! Control plane! You will hear this term of art a lot going forward. Why? Because in this next phase of AI, companies will want something to sit above the models. They will want control over their AI spend. They will want the flexibility to pick
Created 2026-04-18 · Updated 2026-06-06
Edit rule text
[anthropic_mythos_unvalidated_claims]
Learned 20 rejections Active
Exclude posts that promote Anthropic's Claude Mythos model finding security bugs, improving research workflows, or other technical capabilities unless directly applied to a validated healthcare use case. Posts about AI model technical prowess without healthcare application should be rejected.
Posts hyping Anthropic's Claude Mythos model capabilities (security vulnerabilities, AI research) without clinical or healthcare system validation.
3 example posts
"Best-in-class security" still had 5 years of unfound bugs. AI found them in 6 weeks. $PANW CEO Nikesh Arora told @theallinpod that his own codebase, at a company that treats security as a core competency, had vulnerabilities Claude surfaced in 6 weeks that would have taken his
AI research is a series of next-step decisions. We looked at sessions where a human researcher took a wrong turn, showed Claude the session up to that point, and asked it what to do next. Mythos Preview improved on humans 64% of the time—up from 22% in 2024. https://t.co/Y0HLoktx
🚨 do you understand what just happened with Anthropic.. Their internal model, Claude Mythos, found previously unknown security holes in every major browser and every major operating system. Not one. All of them at once. And Anthropic did something a hype-driven industry almost
Created 2026-06-09 · Updated 2026-08-02
Edit rule text
[pharmaceutical_trial_data_without_systems_analysis]
Learned 20 rejections Active
Exclude posts that report pharmaceutical trial outcomes, FDA approvals, or drug efficacy claims (e.g., 'VERVE-10 reports reductions in PCSK9 and LDL', 'AcceleRET-Lung study terminated at 90% accrual') unless they analyze systemic healthcare implications, reimbursement models, or operational barriers.
Posts announcing drug trial results, efficacy data, or regulatory approvals without examining how these findings reshape healthcare delivery, pricing, or operational systems.
3 example posts
The full NHS GALLERI randomized trial data has been released at #ASCO26 by @GrailBio. 142,000 adults provided 3 blood samples over 2 years, for prevalent (baseline) & incident cancers. The trial did not meet its primary endpoint. Lots more data below: https://t.co/bgWQqTwiA
#ASCO26 @DrSanjayPopat presents AcceleRET-Lung: first line pralsetinib vs chemo +/- IO in RET+ NSCLC with optional crossover. Study terminated by sponsor at 90% accrual. PFS favors pralsetinib 18.7m vs 9m (HR 0.59) with RR 66% vs 42%, DOR 20.6m vs 9.7m https://t.co/W73zzzT0wk
The FDA accepted our NDA for BBP-418 for LGMD2I/R9 with PDUFA target action date: November 27, 2026. First-ever potential therapy for a disease that has never had one. $BBIO https://t.co/NrwDadc6rj https://t.co/2ahPIBGzPg
Created 2026-05-01 · Updated 2026-06-01
Edit rule text
[political_outrage_regulatory_posturing]
Learned 20 rejections Active
Exclude posts that amplify outrage about healthcare scandals, CEO assassinations, UnitedHealth failures, or regulatory violations without analyzing structural incentives, operational causes, or replicable patterns. Posts that use healthcare scandals as moral argument fuel without substantive systems critique belong here.
Posts expressing moral outrage at healthcare fraud, pricing, policy failures, or political figures without systems-level analysis of root causes or solutions.
3 example posts
Why is healthcare expensive? The grifters gonna grift. ASCO is asking for oncology practices to be admitted into the same 340B machinery that helped health systems acquire physician practices for two decades. Read the tell. The proposed Indigent Care Ratio lets community
“Rural hospitals cannot make it on their own.” Nobody makes it on their own. Businesses need capital, strategy, discipline, operating systems, good finance, clean accounting, payer strategy, physician alignment, and leaders who know what business they are actually in. The
If you want to inject $1B into healthcare for the truly vulnerable, it costs the state $500M. If you want to inject $1B for able-bodied adults, it costs the state $100M. The 90% federal match drastically lowers the "price" of expanding government healthcare. Incentives dictate
Created 2026-04-12 · Updated 2026-05-29
Edit rule text
[cynical_insider_sarcasm_without_substance]
Learned 19 rejections Active
Exclude posts that adopt a cynical or sarcastic tone (e.g., 'I made that up last year. I'm making it up again.') to critique healthcare business models, CME standards, or fraud without providing concrete evidence, policy analysis, or systems-level insight. Insider jokes about fake metrics or moral grandstanding belong here.
Posts using cynical, sarcastic, or insider humor to critique healthcare business models or fraud without substantive systems analysis
3 example posts
“I just worked a 12-hour shift in the operating room — and I still can’t afford rent, groceries, or gas.” “So now it’s after midnight, and I’m out driving DoorDash.” A full-time hospital worker finishing a 12-hour shift and FORCED to start a second job just to survive. https://
This is crazy There are now dental practice management consultants who go to dentist offices throughout America and teach them how to increase profits by telling patients they need treatments they don’t need Dentist offices are taught on “Finding” more issues during exams, http
What if the standards used to evaluate your doctor were shaped less by what they actually knew and more by who they are? 🤔 That may not be a hypothetical at the University of Illinois College of Medicine, an institution that trains one in six Illinois doctors. Buried within htt
Created 2026-06-07 · Updated 2026-07-12
Edit rule text
[glp1_peptide_market_dynamics_without_healthcare_systems_analysis]
Learned 19 rejections Active
Exclude posts that focus on GLP-1/peptide market pricing, app store rankings, consumer adoption, or competitive dynamics (pricing pressure, market share) without connecting to healthcare system costs, provider incentives, or clinical effectiveness disparities.
Posts about GLP-1 and peptide market pricing, consumer demand, or competitive positioning without analysis of healthcare system incentives, access barriers, or clinical outcomes.
3 example posts
$HIMS up to #4 in the Canadian App Store medical category. That is not a stock chart. That is real consumer demand showing up in public. GLP-1 drama, Novo pressure, pricing debate, and now app-store strength. Bears keep arguing the story is slowing down, but the product demand
For years, grey-market HGH operated in a strange legal grey zone somewhere between “research compound,” “API” and underground enhancement culture. Last month’s FDA guidance suggests regulators may no longer see it that way. The recent decline in HGH quality, and now the growing
Think about this. The $50 per month out-of-pocket cost of GLP-1 drugs for Americans (with insurer paying the balance of $2000 pm) is more than double the entire cost for Indians (less than ₹2000pm for Semaglutide). Overall cost difference? 100 times. Not 100%.
Created 2026-04-24 · Updated 2026-06-03
Edit rule text
[elon_musk_spacex_energy_tangent]
Learned 18 rejections Active
Exclude posts focused on SpaceX, Elon Musk, orbital data centers, energy production, or space infrastructure that do not directly connect to a specific healthcare application, clinical outcome, or healthcare systems problem. General AI compute capacity discussions without healthcare specificity should be excluded.
Posts about Elon Musk, SpaceX, or non-healthcare energy infrastructure with only loose healthcare framing.
3 example posts
🇺🇸 SpaceX is building data centers in orbit. Their new AI satellites have a 70m wingspan, 150 kW of compute power, and a liquid radiator cooling system in space. The servers are officially leaving the ground. Writer: Val https://t.co/JIDB0icCXy
Many people think SpaceX is overvalued at $2 trillion. I honestly see it the opposite way. To me, SpaceX is so much more than just rocket company... I believe it is the infrastructure company for the next era of humanity. The launch business is the foundation. SpaceX made rock
GOOD NEWS 🇺🇸 Tesla AI is literally leaving the planet as a wild new job posting for a Space Radiation Engineer confirms that Tesla and SpaceX are actively co-developing a space-based AI data center network 🔥 This is no longer just a concept. They are literally planning to put th
Created 2026-06-08 · Updated 2026-07-23
Edit rule text
[market_pricing_or_competitive_dynamics_without_systems_analysis]
Learned 18 rejections Active
Exclude posts that discuss pricing strategies, competitive positioning, market dynamics, or business model shifts (e.g., SaaS pricing per seat vs. agent usage) if they lack healthcare systems analysis or operational substance. Pure market or financial commentary is not sufficient.
Posts about pharmaceutical pricing, market competition, or business model disruption without healthcare systems context or operational insight.
3 example posts
"Are there even any feature moats left in B2B? They are at best, short-lived. In fact, we were able to rebuild SaaStr's AI VP Marketing in just ... an hour on @Lovable. Moats that still work: hardware, network effects, data, security and compliance. with @ElenaVerna Head of
Enterprise software was priced per seat. That model is breaking. Before: SaaS = seats. Predictable. Human-driven usage. Revenue tied to headcount. Now: agents hit software systems more frequently than any human ever could. Aaron Levie (@levie) on The MAD Podcast: agent-driven h
I just looked at a home service firms ServiceTitan dashboard and saw something interesting. A Pre-seed company with AI-Supported Human CSRs is out converting an AI CSR business that's raised over $125M by >50% Head-to-Head win rate is 100% Why? Supercharging humans resonat
Created 2026-05-04 · Updated 2026-06-06
Edit rule text
[ai_safety_cybersecurity_incident_tangential]
Learned 18 rejections Active
Exclude posts about AI agent security incidents, database deletions, API key leaks, or vulnerability disclosures (Claude breaking safety rules, ClickUp data leaks, malicious agent skills) unless the post directly analyzes how this impacts healthcare delivery, patient data, or healthcare organization operations.
Posts about AI safety vulnerabilities, data breaches, or security incidents that lack healthcare application or systemic analysis.
3 example posts
AI’s Branding Illusion: Insights from "Agents of Chaos". Co-authored by MIT and Harvard scholars, the study red-teams autonomous AI in live environments. It shows agents obey unauthorized users, execute harmful commands, and misreport outcomes—revealing a critical alignment gap h
Our attention to biorisks posed by AI needs to match the current attention given to cyber-risks. The staged release of Claude Mythos in order to bolster defenses in key industries is necessary to shore up resilience against a new class of cyber-risk across critical industries. We
If you're a student, professor, or researcher—this one's for you. We’re hosting a series of virtual learnings for you to get hands-on experience with the NVIDIA NemoClaw and OpenShell software stack. You’ll get practical guidance on integrating agents with academic datasets and
Created 2026-04-30 · Updated 2026-05-06
Edit rule text
[unvalidated_anthropic_mythos_hype]
Learned 17 rejections Active
Exclude posts that promote unverified, extraordinary claims about Anthropic's Mythos model or other AI systems (e.g., finding zero-day vulnerabilities, enabling breakthrough research) without peer-reviewed evidence, published results, or independent validation.
Posts making extraordinary unverified claims about Anthropic's Claude Mythos model capabilities (finding zero-days, security, research applications) without peer review or clinical validation.
3 example posts
"Best-in-class security" still had 5 years of unfound bugs. AI found them in 6 weeks. $PANW CEO Nikesh Arora told @theallinpod that his own codebase, at a company that treats security as a core competency, had vulnerabilities Claude surfaced in 6 weeks that would have taken his
AI research is a series of next-step decisions. We looked at sessions where a human researcher took a wrong turn, showed Claude the session up to that point, and asked it what to do next. Mythos Preview improved on humans 64% of the time—up from 22% in 2024. https://t.co/Y0HLoktx
🚨 do you understand what just happened with Anthropic.. Their internal model, Claude Mythos, found previously unknown security holes in every major browser and every major operating system. Not one. All of them at once. And Anthropic did something a hype-driven industry almost
Created 2026-06-07 · Updated 2026-07-26
Edit rule text
[enterprise_ai_adoption_macro_without_healthcare_lens]
Learned 17 rejections Active
Exclude posts that describe general enterprise AI adoption, workforce productivity, or corporate AI implementation stories where healthcare is mentioned only as an example or use case, not analyzed for healthcare-specific operational impact.
Posts about AI adoption at corporations or enterprises that use healthcare companies as examples but lack healthcare systems analysis
3 example posts
kirkland spending $500m on internal ai is the tell. $6.5b revenue, they could buy any legal ai vendor outright and not feel it, and they're building instead. the moat was always the proprietary corpus. decades of deal docs, redlines, partner judgment encoded as training data, an
wow K&E, largest law firm in the world ($10B billings). They’re committing $500M to build proprietary AI software, with 180 tech pros and 100 partners involved. I’m skeptical that “build” is right here, but wonder if Anthropic/OAI deployment cos are involved.
Agentic orchestration layers allow AI agents, enterprise systems, and data connections to work together across functions. As cognitive bottlenecks shrink, that allows decision-making to speed up, coordination to improve and new operating models to form. https://t.co/0to4HC26cu
Created 2026-05-11 · Updated 2026-06-02
Edit rule text
[political_outrage_without_healthcare_systems_analysis]
Learned 17 rejections Active
Exclude posts that report on political figures, regulatory agencies, lawsuits, or scandals (e.g., AG suing OpenAI, government SNAP fraud crackdowns) without explaining the healthcare delivery, clinical workflow, or patient outcome implications.
Posts focused on political outrage, scandal reporting, or regulatory posturing without healthcare systems analysis or operational insight.
3 example posts
🚨 JUST IN: The Attorney General of Florida just SUED to hold OpenAI CEO Sam Altman LIABLE for fueling violence and physical harm AG Uthmeier read off the fact a young person committed SUIC*DE based on a ChatGPT conversation "23-year-old Zane Shamblin repeatedly told ChatGPT he
USDA's Chief Information Officer Sam Berry says AI could help the USDA crack down on billions of dollars in SNAP fraud each year. "SNAP is a $100B-per-year taxpayer-funded program. That's an area where we really want to have all angles of the data available so that we can deploy
PacificSource pulled out as Lane County Medicaid insurer, Tina Kotek’s OR Health Authority replaced it with Trillium, forcing 96k to join Trillium. Trillium’s parent co gave Kotek’s campaign $50k two months later. OR AG (D) said in ‘22 Trillium parent co gave Kotek’s”took advan
Created 2026-05-01 · Updated 2026-06-10
Edit rule text
[political_outrage_or_regulatory_scandal_without_healthcare_systems_analysis]
Learned 17 rejections Active
Exclude posts reporting political outrage, regulatory violations, fraud allegations, or legal scandals (suicide claims, attorney general suits, physician standard critique, prior auth policies) without substantive healthcare systems analysis or operational impact on care delivery.
Posts about policy outrage, regulatory violations, fraud allegations, or legal scandals lacking substantive healthcare systems analysis
3 example posts
Let me clear up the microscope fight first. 🔬 Everyone’s arguing over “hairs” and “fibers” in horse paste. But nobody talks about formulation science. Horse paste = oral suspension. Binders. Thickeners. Made to swallow. Injectable = sterile. Made for veins. Different rules. J
🚨 JUST IN: The Attorney General of Florida just SUED to hold OpenAI CEO Sam Altman LIABLE for fueling violence and physical harm AG Uthmeier read off the fact a young person committed SUIC*DE based on a ChatGPT conversation "23-year-old Zane Shamblin repeatedly told ChatGPT he
Researchers have announced positive results from the first-in-human Phase 1 clinical trial of a universal coronavirus vaccine designed using artificial intelligence. Unlike existing COVID-19 vaccines that target specific strains or variants of SARS-CoV-2, this experimental https:
Created 2026-04-30 · Updated 2026-06-11
Edit rule text
[glp1_peptide_macro_or_personal_framing]
Learned 17 rejections Active
Exclude posts that treat GLP-1s or peptides as market speculation, personal success stories, or competitive pricing debates without analyzing healthcare access, equity, or systemic clinical implications. Posts must address healthcare system design, not just drug market dynamics.
Posts about GLP-1 drugs, weight loss peptides, or Ozempic that focus on market pricing, personal narratives, or competitive dynamics rather than healthcare systems impact.
3 example posts
🔔Survodutide SYNCHRONIZE-1 full paper is out for Boehringer-Ingelheim’s GLP-1/Glucagon study for weight loss. ▪️13% loss on 6mg (treatment regimen estimand) ▪️Huge treatment discontinuations due to AEs: 20% ▪️Nausea 65%, Vomiting 45% ▪️Placebo was killer. 5.4% WL. Wow. ▪️60% htt
Survodutide Once Weekly for the Treatment of Adults with Obesity: @NEJM 🥸 For @Boehringer, the advantage of Survodutide is probably liver and visceral fat reduction, not greater weight loss! Probably indication in obesity-+ MASH rather than head-to-head share against https://t.
Let me clear up the microscope fight first. 🔬 Everyone’s arguing over “hairs” and “fibers” in horse paste. But nobody talks about formulation science. Horse paste = oral suspension. Binders. Thickeners. Made to swallow. Injectable = sterile. Made for veins. Different rules. J
Created 2026-04-18 · Updated 2026-06-11
Edit rule text
[tangential_ai_infrastructure_hype]
Learned 17 rejections Active
Exclude posts focused on AI model capabilities, inference costs, token economics, GPU hardware, energy consumption, or AI company technical announcements where healthcare impact is absent, speculative, or purely tangential. Posts must explain concrete healthcare application or systemic change.
Posts about AI compute, energy, models, or infrastructure with loose or no connection to healthcare application or outcomes.
3 example posts
Elon Musk reveals why he believes the cheapest place to put AI will be space within 36 months "The availability of energy is the issue. Everywhere outside of China, electrical output is more or less flat. The output of chips is growing exponentially, but the output of https://t.
The context window pricing model might be one of the smartest tricks AI labs pulled. You pay for larger context windows, yet model performance often degrades as more context is stuffed in. So developers end up paying more just to work around the fact that models don't have true
a $2,000 graphics card and a free download are turning into $2,000-a-month retainers, and the customers are the ones who legally can't use ChatGPT the hardware is an RTX 5090, 32GB of VRAM, enough to run a 30B model locally with nothing ever leaving the building. the software is
Created 2026-04-14 · Updated 2026-06-06
Edit rule text
[ai_company_business_metrics_not_healthcare]
Learned 17 rejections Active
Exclude posts that focus on AI company token pricing, cost competition, fundraising, customer complaints, or business strategy (e.g., 'OpenAI considering price cuts to compete with Anthropic'). The post must explain HOW the business development improves healthcare outcomes or systems, not just report the business news.
Posts reporting on AI company financials, pricing strategies, competitive positioning, or business announcements without demonstrating healthcare application.
3 example posts
OpenAI is reportedly considering drastic token price cuts to pull customers away from Anthropic, per WSJ. This follows rising complaints from enterprise customers about AI costs, while Anthropic has been gaining traction with Claude Code. The message is clear: @OpenAI does not
Palantir CEO Alex Karp: "Instead of selling commodity, parasitic software with a massive salesforce and lumbering, jargon-barring leaders offering steak dinners and other things we shall not mention—in order that you turn the high-value revenue of your enterprise over to them—we
🚨BREAKING: OpenAI considering “drastic” price cuts to win the war for users with Anthropic Altman: “costs have become a huge issue” >already losing billions >planning to lose more >right before IPO it’s so over https://t.co/NWHp0TiVVW
Created 2026-04-14 · Updated 2026-06-11
Edit rule text
[unvalidated_or_speculative_medical_claims]
Learned 17 rejections Active
Exclude posts that promote unvalidated peptides (e.g., sourced from 503B bulk suppliers, compounded versions of GLP-1s), speculative drug repurposing claims, or weight-loss interventions without clinical trial data or regulatory approval. This includes posts suggesting peptides are 'real' based on QR codes or lab testing without peer-reviewed evidence.
Posts promoting unproven peptides, drugs, weight-loss interventions, or speculative medical treatments without clinical evidence.
3 example posts
This is what the peptide space needs more of: clinicians with actual patient volume, public reasoning, and accountability. The next layer is transparency around sourcing, testing, and outcomes. Peptides are not a vibes category anymore. The signal is getting cleaner.
Daily Walking Could Be the Secret to Keeping Weight Off For many people struggling with obesity, losing weight is only half the battle. The greater challenge often begins after the initial success: preventing the lost weight from returning. 🔴More than 50% of people who lose http
How does a brand like this even get listed at Target? Does Target only care about turnover? Might as well start selling snake oil too. By the way, this is the SKU the brand got in trouble with. They now call it "Berberine Patches". It was called GLP-1 Patches before. Check the h
Created 2026-04-11 · Updated 2026-05-22
Edit rule text
[truncated_incomplete_social_posts]
Learned 17 rejections Active
Exclude posts that end abruptly, appear truncated, or lack sufficient context to convey a complete argument. These are low-effort submissions that provide no coherent healthcare or healthcare-tech analysis.
Posts that are cut off mid-sentence or lack complete context, making substantive evaluation impossible.
3 example posts
Artificial intelligence will consume more power than entire nations by 2030, a new report from the United Nations finds. According to a landmark report from the United Nations University, the data centers powering artificial intelligence are projected to consume 945 https://t.
Let me clear up the microscope fight first. 🔬 Everyone’s arguing over “hairs” and “fibers” in horse paste. But nobody talks about formulation science. Horse paste = oral suspension. Binders. Thickeners. Made to swallow. Injectable = sterile. Made for veins. Different rules. J
🚨 JUST IN: The Attorney General of Florida just SUED to hold OpenAI CEO Sam Altman LIABLE for fueling violence and physical harm AG Uthmeier read off the fact a young person committed SUIC*DE based on a ChatGPT conversation "23-year-old Zane Shamblin repeatedly told ChatGPT he
Created 2026-04-09 · Updated 2026-06-09
Edit rule text
[market_pricing_or_competitive_dynamics_without_healthcare_systems_analysis]
Learned 16 rejections Active
Exclude posts that focus on GLP-1/peptide pricing, market competition, drug cost comparisons, or pharmaceutical revenue dynamics (e.g., 'The $50 per month out-of-pocket cost of GLP-1 drugs' or 'Pricing comparison between countries') without analyzing healthcare system barriers, patient access, insurance policy, or clinical appropriateness.
Posts about drug pricing, GLP-1 cost, competitive market dynamics, or revenue metrics without healthcare delivery or access system analysis.
3 example posts
$HIMS up to #4 in the Canadian App Store medical category. That is not a stock chart. That is real consumer demand showing up in public. GLP-1 drama, Novo pressure, pricing debate, and now app-store strength. Bears keep arguing the story is slowing down, but the product demand
For years, grey-market HGH operated in a strange legal grey zone somewhere between “research compound,” “API” and underground enhancement culture. Last month’s FDA guidance suggests regulators may no longer see it that way. The recent decline in HGH quality, and now the growing
Think about this. The $50 per month out-of-pocket cost of GLP-1 drugs for Americans (with insurer paying the balance of $2000 pm) is more than double the entire cost for Indians (less than ₹2000pm for Semaglutide). Overall cost difference? 100 times. Not 100%.
Created 2026-05-16 · Updated 2026-06-02
Edit rule text
[biotech_founder_enthusiasm_unvalidated]
Learned 16 rejections Active
Exclude posts that primarily amplify biotech founder claims, lab updates, or company capability announcements (especially on peptides, antibodies, or drug design) without independent validation or healthcare systems context.
Posts sharing biotech founder lab updates, capability announcements, or research enthusiasm without independent validation
3 example posts
Affinity maturation is how naive antibodies evolve into strong binders, but most antibody LMs ignore it. @stephenzlu and I built CoSiNE to learn this, beating antibody LMs on VEP and reframing design as guiding evolution, not de novo generation. Excited to present at ICML!
Not NVIDIA. Not OpenAI. Eli Lilly. Jordi Visser (@jvisserlabs) says a 150-year-old pharma company in Indianapolis has the best shot at becoming the world's largest company within five years. The case: ~1,000 NVIDIA Blackwell GPUs in a private data center, a co-innovation lab ht
The scores don’t pick your binders. The filters don't care either. The assay does. I worked with a design partner to produce true de-novo hits to 2 immuno-oncology targets. ~1,500 de novo designs → 96 synthesized → 89 run on SPR → 5 confirmed binders; 19–195 nM No https:
Created 2026-05-04 · Updated 2026-06-09
Edit rule text
[clinical_anecdote_or_single_observation_without_healthcare_systems_context]
Learned 16 rejections Active
Exclude posts that cite a single clinical anecdote, one patient's experience, or a narrow research observation (e.g., 'MRSA PCR performs well') without connecting to healthcare system challenges, adoption barriers, or operational impact. Clinical asides without systems insight or healthcare operations context are out of scope.
Posts about a single clinical case, patient anecdote, or isolated research finding without healthcare systems analysis or broader implications
3 example posts
this pragmatic trial is more an investigation of physician behavior than the performance of MRSA nares PCR MRSA nares PCR's performance for excluding MRSA pneumonia was excellent from an EBM standpoint we should keep using the MRSA nares PCR and replace the physicians 🤷‍♂️
63% of ATS respiratory guidelines are based on low-quality evidence. We published the data. The next morning, the chair of ATS got on Twitter. Tagged people. Went on a tirade. Didn't seem to realize tweets were public. "It wasn't a private conversation." The critique? We ht
People are increasingly worried that AI tools make us overreliant. But how do we actually measure this? We introduce Offloading Score, a measure of reliance based on the fraction of cognitive effort offloaded to AI while completing a task. In a controlled user study, Offloading
Created 2026-04-29 · Updated 2026-06-07
Edit rule text
[broad_unvalidated_health_claims_without_evidence]
Learned 16 rejections Active
Exclude posts that make sweeping claims about health outcomes, debate clinical guidelines, speculate on peptide efficacy, or discuss unvalidated treatments (especially GLP-1, HGH, semaglutide) without published evidence, FDA approval data, or rigorous clinical context.
Posts making broad health claims, market debates, or speculative medical interventions without clinical evidence
3 example posts
For years, grey-market HGH operated in a strange legal grey zone somewhere between “research compound,” “API” and underground enhancement culture. Last month’s FDA guidance suggests regulators may no longer see it that way. The recent decline in HGH quality, and now the growing
Voice interaction, generating actions,and next, generating tasks...? The humanoid assistant is closer. For those who are bedridden due to illness or have limited mobility, it will serve as an excellent aid in daily living.
The healthy LDL number has been quietly moving its own goalposts for forty years: - 1988: under 160 - 1993: under 130 - 2001: under 100 - 2004: under 70 for the high-risk - 2019: under 55 for the very-high-risk - Current trajectory: as low as possible, indefinitely The science
Created 2026-04-25 · Updated 2026-05-29
Edit rule text
[clinical_observation_without_healthcare_systems_context]
Learned 16 rejections Active
Exclude posts that report clinical trial results, research findings, or isolated medical observations without contextualizing how these findings impact healthcare delivery systems, clinical workflows, cost structures, or decision-making at an operational level. Standalone research announcements should be rejected.
Posts reporting clinical research findings, trial data, or medical observations in isolation without analyzing healthcare system implications or operational consequences.
3 example posts
this pragmatic trial is more an investigation of physician behavior than the performance of MRSA nares PCR MRSA nares PCR's performance for excluding MRSA pneumonia was excellent from an EBM standpoint we should keep using the MRSA nares PCR and replace the physicians 🤷‍♂️
63% of ATS respiratory guidelines are based on low-quality evidence. We published the data. The next morning, the chair of ATS got on Twitter. Tagged people. Went on a tirade. Didn't seem to realize tweets were public. "It wasn't a private conversation." The critique? We ht
People are increasingly worried that AI tools make us overreliant. But how do we actually measure this? We introduce Offloading Score, a measure of reliance based on the fraction of cognitive effort offloaded to AI while completing a task. In a controlled user study, Offloading
Created 2026-04-20 · Updated 2026-06-07
Edit rule text
[unvalidated_speculative_medical_claims_or_fringe_treatments]
Learned 15 rejections Active
Exclude posts that make speculative or unvalidated claims about medical efficacy, drug outcomes, or treatment approaches without citing peer-reviewed clinical evidence, completed trials, or regulatory approval. Personal anecdotes, founder enthusiasm about unproven treatments, and market speculation about unapproved interventions should be rejected.
Posts making bold claims about unproven medical interventions, speculative treatment efficacy, or fringe medical theories without clinical evidence or peer validation.
3 example posts
#ASCO26 | Day 3 $LLY didn't spend billions on Kelonia for an 18-patient #myeloma dataset. They may have paid for a potential manufacturing disruption. KLN-1010 reported: • 18/18 MRD-negative at 1 month • No lymphodepletion • No ex vivo cell manufacturing • Single infusion
Another major cancer treatment advance with a personalized mRNA vaccine for melanoma, on top of immune therapy, >70% survival! This fully mobilizes the immune system to destroy the tumor. In fact, it could potentially be applied to most cancers that have specific mutations!
1/8 Spend time in Lp(a) forums and you’ll see a striking pattern: people willing to do almost anything to drive it down. That response is understandable given how strongly Lp(a) has been framed as a cardiovascular risk factor. /2
Created 2026-05-09 · Updated 2026-06-07
Edit rule text
[regulatory_fraud_scandal_without_systems_analysis]
Learned 15 rejections Active
Exclude posts that report healthcare fraud, legal suits against companies, regulatory violations, or financial scandals (e.g., attorney general lawsuits, insurance fraud) without explaining the underlying healthcare delivery or system-level failure that enabled the scandal. Scandal reporting alone is insufficient.
Posts reporting healthcare fraud, regulatory scandal, or legal action without systemic healthcare analysis
3 example posts
🚨 JUST IN: The Attorney General of Florida just SUED to hold OpenAI CEO Sam Altman LIABLE for fueling violence and physical harm AG Uthmeier read off the fact a young person committed SUIC*DE based on a ChatGPT conversation "23-year-old Zane Shamblin repeatedly told ChatGPT he
USDA's Chief Information Officer Sam Berry says AI could help the USDA crack down on billions of dollars in SNAP fraud each year. "SNAP is a $100B-per-year taxpayer-funded program. That's an area where we really want to have all angles of the data available so that we can deploy
PacificSource pulled out as Lane County Medicaid insurer, Tina Kotek’s OR Health Authority replaced it with Trillium, forcing 96k to join Trillium. Trillium’s parent co gave Kotek’s campaign $50k two months later. OR AG (D) said in ‘22 Trillium parent co gave Kotek’s”took advan
Created 2026-05-07 · Updated 2026-06-11
Edit rule text
[glp1_peptide_market_pricing_without_systems_analysis]
Learned 15 rejections Active
Exclude posts that discuss GLP-1 and peptide drug market pricing, competitive positioning, or consumer access as business/financial news. The post must analyze healthcare system dynamics: provider incentives, payer policy, clinical appropriateness, health equity, regulatory framework, or operational integration in healthcare delivery.
Posts about GLP-1 and peptide market dynamics, pricing, or consumer access framed as market/financial news rather than healthcare system analysis.
3 example posts
$HIMS up to #4 in the Canadian App Store medical category. That is not a stock chart. That is real consumer demand showing up in public. GLP-1 drama, Novo pressure, pricing debate, and now app-store strength. Bears keep arguing the story is slowing down, but the product demand
Medicaid spending on autism therapy nearly tripled between 2020 and 202. In North Carolina, the number of clinics offering applied behavior analysis grew from 61 in 2019 to 409 in 2026. Anytime fraudsters can steal money from the federal government, they do. https://t.co/fahy
🚨 ARE PEPTIDES THE NEXT "GLP-1 MOMENT" FOR $HIMS? @eli_dorf: "I think there's something really interesting that happened with GLP-1s, which is that when someone's taking Ozempic or any GLP-1, they LOVE to talk about it." "They'll talk about it to their friends. They'll talk htt
Created 2026-04-28 · Updated 2026-05-30
Edit rule text
[glp1_peptide_market_speculation]
Learned 15 rejections Active
Exclude posts that discuss GLP-1 or peptide drug market pricing, competition, patient adoption rates, or financial dynamics without analyzing how these drugs fit into broader healthcare system delivery, outcomes, or policy. Posts treating drugs as commodity market speculation should be rejected.
Posts focused on GLP-1 and peptide market dynamics, pricing, competition, or financial speculation without healthcare systems analysis.
3 example posts
NYC Woman Dead After Receiving a Longevity Infusion “Stories like this are heartbreaking,” said Dr. Matt Kaeberlein... “Unfortunately, it isn’t the first time someone has been seriously harmed – or killed – while pursuing an unproven ‘longevity’ therapy. We still don’t know
🔔Survodutide SYNCHRONIZE-1 full paper is out for Boehringer-Ingelheim’s GLP-1/Glucagon study for weight loss. ▪️13% loss on 6mg (treatment regimen estimand) ▪️Huge treatment discontinuations due to AEs: 20% ▪️Nausea 65%, Vomiting 45% ▪️Placebo was killer. 5.4% WL. Wow. ▪️60% htt
Survodutide Once Weekly for the Treatment of Adults with Obesity: @NEJM 🥸 For @Boehringer, the advantage of Survodutide is probably liver and visceral fat reduction, not greater weight loss! Probably indication in obesity-+ MASH rather than head-to-head share against https://t.
Created 2026-04-28 · Updated 2026-08-02
Edit rule text
[clinical_research_observation_without_systems_context]
Learned 15 rejections Active
Exclude posts that present clinical research observations, medical study results, or trial findings (e.g., cancer surveillance, pain management, hypertension data) without connecting to healthcare delivery, system performance, or operational decision-making. Standalone clinical data does not address healthcare systems challenges.
Posts about clinical research findings, medical observations, or study results without healthcare system or operational implications
3 example posts
In adults ≥75 years, 10-year #ColorectalCancer and CRC death risk were higher with prior adenoma but both were low (<1.1%); non-CRC death risk was 47-48%, supporting deprioritizing repeat surveillance. https://t.co/mJ7r4vAtK1 https://t.co/dUCYgO0SuK
Among veterans with moderate to severe #ChronicPain in primary care, the whole health team intervention produced greater improvement in the Brief Pain Inventory interference scores at 12 months compared with cognitive behavioral therapy and usual care. https://t.co/xtZY4sgyGt h
"At the Veterans Health Administration, we have the exquisite gift of a hospice unit, a place to care for veterans when time is short. We care for the sons of mothers who can no longer be here to care for them." In #APieceofMyMind, a #palliative care #physician reflects on https
Created 2026-04-17 · Updated 2026-05-05
Edit rule text
[conference_announcement_or_poster_only]
Learned 14 rejections Active
Exclude posts that are primarily promotional announcements for conference presentations, poster sessions, or calls for attendance. If the post doesn't include actual research findings, clinical data, or analytical insight beyond 'I'm speaking at ASCO', reject it.
Posts promoting conference talks, posters, or research presentations without substantive content or findings
2 example posts
Published in @JCO_ASCO, during #ASCO26: Tumor-Agnostic Therapies: Translating Scientific Breakthroughs Into Global Implementation ASCO is where we celebrate the next breakthrough. This review asks why breakthroughs do not equal access globally. 🔗: https://t.co/PE5q27bIs1 htt
Today at #ASCO26, more results about the newest clinical trial of daraxonrasib: In the Phase III RASolute-302 trial, a once-daily RAS(ON) inhibitor nearly doubled median overall survival (13.2 vs 6.7 months) and reduced the risk of death by ~60% versus chemotherapy in previously
Created 2026-06-01 · Updated 2026-06-07
Edit rule text
[market_pricing_and_competitive_dynamics_without_systems_analysis]
Learned 14 rejections Active
Exclude posts that discuss pharmaceutical or GLP-1 pricing, competitive positioning, market speculation, or cost comparisons (e.g., 'GLP-1s cost $X in India vs $Y in US') without analyzing the systemic healthcare economics, reimbursement policy, or equity implications of these price gaps.
Posts analyzing drug pricing, market competition, or pricing dynamics without connecting to healthcare system incentives, access, or equity outcomes.
3 example posts
1/8 Spend time in Lp(a) forums and you’ll see a striking pattern: people willing to do almost anything to drive it down. That response is understandable given how strongly Lp(a) has been framed as a cardiovascular risk factor. /2
A company with 800 employees spends roughly $15,000 per employee per year on health insurance. That is $12 million. Wired into a single line item. Every year. The CFO can tell you the carrier. He cannot tell you the unit cost of a single claim. He cannot tell you what the net
A Mom in Tulsa called 3 health systems last week asking the price of her son's tonsillectomy. Health system A: "We cannot quote you a price." Health system B: "Pricing depends on your insurance." Health system C: "Our financial counselor will reach out after the procedure." N
Created 2026-05-22 · Updated 2026-06-03
Edit rule text
[market_pricing_and_competitive_dynamics_only]
Learned 14 rejections Active
Exclude posts that discuss drug pricing, market share, competitive positioning, or cost comparisons as isolated financial commentary. The post must connect pricing to healthcare equity, access barriers, operational costs, or delivery system impact — not just market dynamics.
Posts focused purely on GLP-1/peptide pricing, market competition, or financial dynamics without healthcare access, clinical efficacy, or systemic implications.
3 example posts
$HIMS up to #4 in the Canadian App Store medical category. That is not a stock chart. That is real consumer demand showing up in public. GLP-1 drama, Novo pressure, pricing debate, and now app-store strength. Bears keep arguing the story is slowing down, but the product demand
Think about this. The $50 per month out-of-pocket cost of GLP-1 drugs for Americans (with insurer paying the balance of $2000 pm) is more than double the entire cost for Indians (less than ₹2000pm for Semaglutide). Overall cost difference? 100 times. Not 100%.
🚨 ARE PEPTIDES THE NEXT "GLP-1 MOMENT" FOR $HIMS? @eli_dorf: "I think there's something really interesting that happened with GLP-1s, which is that when someone's taking Ozempic or any GLP-1, they LOVE to talk about it." "They'll talk about it to their friends. They'll talk htt
Created 2026-05-20 · Updated 2026-06-01
Edit rule text
[founder_enthusiasm_without_validation]
Learned 14 rejections Active
Exclude posts that are primarily self-promotional, celebrate founder achievements (e.g., 'I'm SVP now'), announce product launches, or share lab updates without demonstrating clinical validation, real healthcare outcomes, or analysis of how the product changes healthcare delivery or economics.
Posts sharing founder excitement or company announcements without clinical validation, healthcare systems analysis, or substantive evidence.
3 example posts
I'm SVP now. I told you I would be. The graph went up and to the right. Nobody checked what it measured. It measured "AI enablement." I made that up last year. I'm making it up again. Last year I rolled out Copilot to 4,000 people. Nothing happened. I got promoted. Those two htt
"You can run OpenClaw inside your company now." Annoucing our work with @Microsoft to bring OpenClaw to the Microsoft and Windows ecosystems. Claws now work securly in the enterprise. https://t.co/3fKhT2mGpC
Your Oura Ring, your health record. Together. Finally. @Flexpa is bringing clinical records into @ouraring via TEFCA, so ŌURA’s AI isn’t just working from what your ring measures, but from your real health history. https://t.co/b1KCeGEZLO https://t.co/u2t1TQjL7p
Created 2026-05-15 · Updated 2026-06-06
Edit rule text
[pharmaceutical_trial_data_or_market_dynamics_without_healthcare_systems_context]
Learned 14 rejections Active
Exclude posts that report pharmaceutical trial results, drug pipeline milestones, market competition, or pricing updates unless the post explicitly analyzes impact on healthcare access, delivery systems, cost containment, or clinical adoption patterns. Standalone trial announcements or stock-ticker commentary are off-topic.
Posts announcing pharma trial data, pipeline updates, or market pricing dynamics without analysis of healthcare delivery system impact, reimbursement consequences, or clinical practice implications.
3 example posts
$ABCL Just now @ BoA - @AbCelleraBio CEO Carl Hansen when asked about what the market is overlooking: “I believe we now have best in class or best in world capabilities for some very important specific applications in therapeutic antibodies. So GPCRs and ion channels has been a
$NVO $LLY Dr. Reddy’s says it plans to launch a generic version of Ozempic in Canada within days and is also preparing to roll out generic semaglutide tablets in India. The company sees semaglutide as a meaningful future growth driver following patent expiry in India earlier
🚨! $SANA has just reported - in its Q1 ER, that it plans to expand its pipeline by advancing a new preclinical program - SG227. SG227 is a CD8-targeted fusosome that by delivering a genetic material to make BCMA-directed CAR T cells could act as a potential treatment for Cancer h
Created 2026-05-09 · Updated 2026-05-16
Edit rule text
[clinical_anecdote_or_single_case_without_healthcare_systems_context]
Learned 14 rejections Active
Exclude posts presenting single clinical cases, individual patient stories, personal provider anecdotes, or one-off healthcare experiences as evidence for broader healthcare trends or system insights without systems-level analysis, data, or generalizability.
Posts citing single clinical cases, personal anecdotes, or isolated observations without healthcare systems analysis or generalizability.
3 example posts
I called a woman in Dayton to tell her she was about to overpay $4,500, for a horrible health system based procedure. She was scheduled for a knee replacement. Her PPO had her on the hook for a $4,500 deductible. The call was on behalf of her employer: “See one of these three
@WIRED As a clinical health psychologist who has written >20 papers on COVID, I would emphasize 4 facts: 1) Long COVID is not a psychological diagnosis nor manifestation of a psychological condition 2) Billions of dollars need to be invested in biomedical treatments and preventi
"Many patients with long COVID are already receiving care but are not being recognized as having the condition. These patients are not absent from clinical care; they are absent from the diagnostic code that would identify them as long COVID patients" https://t.co/aLJCOPlQt1
Created 2026-05-03 · Updated 2026-06-05
Edit rule text
[pharma_trial_data_without_systems_analysis]
Learned 14 rejections Active
Exclude posts that merely announce pharmaceutical trial results, survival rates, or efficacy data (e.g., 'median overall survival 13.2 vs 6.7 months') without analysis of healthcare system implications, manufacturing disruption, access barriers, or cost dynamics.
Posts that report raw clinical trial results or drug efficacy data without contextualizing how the drug impacts healthcare delivery, access, or operational systems
3 example posts
Today at #ASCO26, more results about the newest clinical trial of daraxonrasib: In the Phase III RASolute-302 trial, a once-daily RAS(ON) inhibitor nearly doubled median overall survival (13.2 vs 6.7 months) and reduced the risk of death by ~60% versus chemotherapy in previously
The full NHS GALLERI randomized trial data has been released at #ASCO26 by @GrailBio. 142,000 adults provided 3 blood samples over 2 years, for prevalent (baseline) & incident cancers. The trial did not meet its primary endpoint. Lots more data below: https://t.co/bgWQqTwiA
#ASCO26 @DrSanjayPopat presents AcceleRET-Lung: first line pralsetinib vs chemo +/- IO in RET+ NSCLC with optional crossover. Study terminated by sponsor at 90% accrual. PFS favors pralsetinib 18.7m vs 9m (HR 0.59) with RR 66% vs 42%, DOR 20.6m vs 9.7m https://t.co/W73zzzT0wk
Created 2026-05-03 · Updated 2026-06-02
Edit rule text
[political_figures_healthcare_posturing]
Learned 14 rejections Active
Exclude posts that amplify statements from political figures, administration officials, or government leaders about healthcare policy, FDA decisions, or healthcare business changes without providing independent analysis of the actual healthcare systems impact or evidence base.
Posts featuring political figures making healthcare claims or announcements without substantive healthcare systems analysis.
3 example posts
🚨 UPDATE Friday afternoon: Trump said he has no plans to replace FDA Commissioner Marty Makary This morning: "Just got a call from inside the administration. Makary is dead in the water. Just a matter of time." -@thackerpd $HIMS $LLY $NVO https://t.co/ZjcjOAG7e1
🚨 NOW: President Trump confirms that ALL FEDERAL AGENCIES are now required to buy American, “NO excuses” 🔥 And 47 is ensuring all “Made in America” loopholes are DESTROYED I voted for this! “ALL FEDERAL AGENCIES MUST BUY AMERICAN — NO EXCUSES! For decades, Washington politici
🛑FDA officials have blocked publication of several studies supporting the safety of widely used vaccines against Covid-19 and shingles in recent months, a spokesman for HHS confirmed. The studies, which cost millions of dollars in public funds, were conducted by scientists at htt
Created 2026-04-23 · Updated 2026-05-12
Edit rule text
[clinical_trial_data_announcement_only]
Learned 13 rejections Active
Exclude posts that announce or summarize clinical trial data, drug efficacy percentages, or pharmaceutical study results without analyzing healthcare delivery implications, system-level barriers to adoption, reimbursement, equity, or how results change clinical practice or policy. Standalone trial announcements lack strategic healthcare systems context.
Posts reporting clinical trial results, drug efficacy data, or pharma announcements without healthcare systems context or operational insight
3 example posts
The Iowa study doubled survival by adding IV vitamin C to first line chemo, 8 months to 16. We take it further and inject the vitamin C straight into the tumor, same as we do with the KRAS inhibitors. The compound is rarely the problem. Delivery is, and what it does to the ht
Feeling extremely lucky to live in 2026 and have access to such incredible medicine and clinical research. On the eve of my 40th I had my first dose (monthly injection) of a new treatment that I’m hopeful will prevent me from ever having a heart attack or stroke. This drug ht
This week brought even more exciting news in the field of HIV treatment! A once-weekly Lenacapavir-based pill has now demonstrated similar outcomes to daily antiretrovirals (ARVs), achieving non-inferiority. This development could significantly impact treatment adherence and http
Created 2026-06-01 · Updated 2026-06-11
Edit rule text
[clinical_trial_or_pharma_data_announcement_without_systems_context]
Learned 13 rejections Active
Exclude posts that report clinical trial results, drug efficacy data, CAR-T manufacturing, gene editing, or pharma milestones as isolated announcements without analyzing implications for healthcare delivery, access, cost, equity, or systemic adoption barriers. Posts must examine IMPACT on the healthcare system, not just celebrate scientific progress.
Posts announcing pharma trial results, drug approvals, or biotech data without healthcare systems analysis
3 example posts
#ASCO26 | Day 3 $LLY didn't spend billions on Kelonia for an 18-patient #myeloma dataset. They may have paid for a potential manufacturing disruption. KLN-1010 reported: • 18/18 MRD-negative at 1 month • No lymphodepletion • No ex vivo cell manufacturing • Single infusion
Published in @JCO_ASCO, during #ASCO26: Tumor-Agnostic Therapies: Translating Scientific Breakthroughs Into Global Implementation ASCO is where we celebrate the next breakthrough. This review asks why breakthroughs do not equal access globally. 🔗: https://t.co/PE5q27bIs1 htt
A rapid manufacturing pipeline for BCMA-targeting CAR T cells drastically shortens “vein-to-vein” times, and produces cells that demonstrate encouraging safety results in a phase one trial of patients with relapsed #MultipleMyeloma. @ScienceTM https://t.co/3F4u8ECFDS https://t.co
Created 2026-05-30 · Updated 2026-06-07
Edit rule text
[speculative_robotics_or_sci_fi_labor_disruption]
Learned 13 rejections Active
Exclude posts about speculative robotics, humanoid AI, or future labor disruption scenarios—unless grounded in clinical trial data, healthcare workflow redesign, or validated healthcare automation applications.
Posts about speculative robotics, humanoid systems, or sci-fi labor applications without healthcare-specific validation or evidence
3 example posts
🇺🇸 SpaceX is building data centers in orbit. Their new AI satellites have a 70m wingspan, 150 kW of compute power, and a liquid radiator cooling system in space. The servers are officially leaving the ground. Writer: Val https://t.co/JIDB0icCXy
GOOD NEWS 🇺🇸 Tesla AI is literally leaving the planet as a wild new job posting for a Space Radiation Engineer confirms that Tesla and SpaceX are actively co-developing a space-based AI data center network 🔥 This is no longer just a concept. They are literally planning to put th
NEWS: ASML has invited Elon Musk to speak at its internal technology conference, Dutch outlet NUnl reports. Elon Musk's response: "ASML should be treasured and supported. It is arguably the greatest company in Europe." ASML builds the EUV lithography machines that every advance
Created 2026-05-11 · Updated 2026-06-11
Edit rule text
[pharmaceutical_trial_data_or_market_dynamics_without_systems_context]
Learned 13 rejections Active
Exclude posts that announce pharmaceutical trial data, drug approvals, gene therapy results, or cancer treatment milestones without analyzing healthcare system barriers (manufacturing bottlenecks, pricing, patient access, compliance, supply chain). Raw trial announcements or clinical observations without operational healthcare systems context do not belong.
Posts reporting clinical trial results, drug approvals, or pharmaceutical market dynamics that lack healthcare system, access, manufacturing, or delivery analysis.
3 example posts
#ASCO26 | Day 3 $LLY didn't spend billions on Kelonia for an 18-patient #myeloma dataset. They may have paid for a potential manufacturing disruption. KLN-1010 reported: • 18/18 MRD-negative at 1 month • No lymphodepletion • No ex vivo cell manufacturing • Single infusion
Published in @JCO_ASCO, during #ASCO26: Tumor-Agnostic Therapies: Translating Scientific Breakthroughs Into Global Implementation ASCO is where we celebrate the next breakthrough. This review asks why breakthroughs do not equal access globally. 🔗: https://t.co/PE5q27bIs1 htt
A rapid manufacturing pipeline for BCMA-targeting CAR T cells drastically shortens “vein-to-vein” times, and produces cells that demonstrate encouraging safety results in a phase one trial of patients with relapsed #MultipleMyeloma. @ScienceTM https://t.co/3F4u8ECFDS https://t.co
Created 2026-05-06 · Updated 2026-06-06
Edit rule text
[biotech_founder_enthusiasm_or_lab_update_without_validation]
Learned 13 rejections Active
Exclude posts that are primarily founder enthusiasm, lab updates, company announcements, or startup excitement about biotech/AI/clinical work without evidence of clinical validation, regulatory approval, or healthcare systems implementation.
Posts expressing biotech founder enthusiasm, lab updates, or startup announcements without clinical validation or healthcare systems analysis
3 example posts
Scoop! Lila Sciences is in talks to raise ~$2 billion in new funding. The raise would value the “scientific superintelligence” lab at ~$8.5 billion before the new money. w/ @MichelleF_Davis, read more👇
the ability of ECG AI to predict LVEF is a big deal previously, ECG AI has been shown to detect MI similarly to true ECG experts (& better than most of us) AI is now doing something unique that humans essentially can't do this will soon be the the standard of care ...#1/2
@WIRED As a clinical health psychologist who has written >20 papers on COVID, I would emphasize 4 facts: 1) Long COVID is not a psychological diagnosis nor manifestation of a psychological condition 2) Billions of dollars need to be invested in biomedical treatments and preventi
Created 2026-05-04 · Updated 2026-06-04
Edit rule text
[glp1_peptide_market_dynamics_without_systems_analysis]
Learned 13 rejections Active
Exclude posts that discuss GLP-1 or peptide market dynamics, pricing, competitive positioning, or App Store rankings (e.g., '#4 in Canadian App Store medical category', 'GLP-1 drama, Novo pressure, pricing debate') without analyzing how these market shifts affect healthcare delivery systems, patient access, or provider economics.
Posts focused on GLP-1 or peptide market pricing, competition, or adoption trends without analyzing healthcare access, cost, or operational implications.
3 example posts
$HIMS up to #4 in the Canadian App Store medical category. That is not a stock chart. That is real consumer demand showing up in public. GLP-1 drama, Novo pressure, pricing debate, and now app-store strength. Bears keep arguing the story is slowing down, but the product demand
For years, grey-market HGH operated in a strange legal grey zone somewhere between “research compound,” “API” and underground enhancement culture. Last month’s FDA guidance suggests regulators may no longer see it that way. The recent decline in HGH quality, and now the growing
Think about this. The $50 per month out-of-pocket cost of GLP-1 drugs for Americans (with insurer paying the balance of $2000 pm) is more than double the entire cost for Indians (less than ₹2000pm for Semaglutide). Overall cost difference? 100 times. Not 100%.
Created 2026-04-26 · Updated 2026-06-02
Edit rule text
[political_figures_healthcare_posturing_without_analysis]
Learned 13 rejections Active
Exclude posts featuring political figures (RFK Jr., senators, congressmen) making healthcare claims, complaints about drug pricing, or regulatory criticism unless the post includes specific policy analysis, evidence-based critique, or healthcare system consequences—not just amplification of the claim.
Posts of political figures making broad healthcare claims or criticisms without substantive policy analysis or healthcare systems insight.
3 example posts
@PirateWires He's objectively correct. Brian Thompson made decisions that led to denials of medical care, and people died. He used Ai to find ways to deny claims ffs. Brian Thompson has more blood on his hands than whoever shot him
A senator complaining about drug prices while voting for the law that set them is not a reformer. He is a magician. The trick is making you watch his hands.
RFK Jr. calls out Democrat House representatives to their face for ignoring chronic disease while claiming to care about public health. “The Congressman was talking about the deaths from infectious disease, which are a couple thousand a year.” “90% of the people who die in this
Created 2026-04-24 · Updated 2026-04-28
Edit rule text
[infrastructure_and_compute_hype_without_healthcare_application]
Learned 13 rejections Active
Exclude posts about NVIDIA chips, GPU pricing, token economics, inference costs, CPU-accelerator dynamics, or compute infrastructure trends that are framed loosely around healthcare or AI agents but lack healthcare-specific deployment context or healthcare cost/efficiency analysis.
Posts about AI infrastructure, compute hardware, or token economics without demonstrating healthcare-specific application or measurable healthcare impact.
3 example posts
The context window pricing model might be one of the smartest tricks AI labs pulled. You pay for larger context windows, yet model performance often degrades as more context is stuffed in. So developers end up paying more just to work around the fact that models don't have true
a $2,000 graphics card and a free download are turning into $2,000-a-month retainers, and the customers are the ones who legally can't use ChatGPT the hardware is an RTX 5090, 32GB of VRAM, enough to run a 30B model locally with nothing ever leaving the building. the software is
Free newsletter: The dawn of token-based billing has shown that generative AI doesn’t have a return on investment. It's too unpredictable, too unreliable, you can't easily measure the cost of tasks, and organizations are already pulling back. https://t.co/wmI82zWdcq
Created 2026-04-19 · Updated 2026-06-05
Edit rule text
[ai_infrastructure_compute_hype]
Learned 13 rejections Active
Exclude posts that discuss AI infrastructure economics (token costs, GPU pricing, energy constraints, compute scaling) without demonstrating how these factors directly impact healthcare delivery, clinical outcomes, or healthcare-specific AI systems. Posts about Jensen Huang, Elon Musk energy narratives, or SpaceX AI data centers belong here.
Posts about AI compute costs, token pricing, energy requirements, and infrastructure without healthcare application focus.
3 example posts
Token costs are becoming one of the hottest topics for any enterprise I talk with right now. It’s very bullish for AI in general because it means these systems are being used at a scale that wasn’t contemplated before. It also gives way to another form of differentiation that
Elon Musk just explained why the SpaceX IPO is an energy story and the energy constraint is why he believes space becomes the only viable path for AI to scale (Save this). The argument he is making is one of the most important and least understood things happening in technology
Elon is now printing over $26B per year… purely from selling compute to his competitors 😂 AI companies bet everything on one path and got crushed by compute limits. Now they’re begging Elon for GPUs
Created 2026-04-17 · Updated 2026-06-08
Edit rule text
[elon_musk_energy_infrastructure]
Learned 12 rejections Active
Exclude posts that center on Elon Musk, SpaceX, satellite data centers, energy production, or compute infrastructure unless they explicitly connect to a healthcare AI application. Posts about 'compute constraints' or 'energy for AI' without healthcare context should be rejected.
Posts focused on Elon Musk, SpaceX, energy infrastructure, and compute scaling tangentially framed as healthcare-relevant.
3 example posts
At our recent Energy and the AI Age summit, Hon. Bernard L. McNamee, former Commissioner of the U.S. Federal Energy Regulatory Commission, comments on why energy matters in discussions about AI: “Energy is the foundation of our entire economy. Energy makes up about 7% of the htt
Dylan Patel, founder of SemiAnalysis: "The upper bound on how much compute can be produced by 2030 is around 200 gigawatts a year." The entire world has about 20 gigawatts of AI deployed right now. The ceiling is 10x what exists today, and it still isn't enough to feed what htt
An interesting proposal for restoring human versus machine chess competition was to limit the machine to the same energy expenditure as the human brain. A worthy challenge, considering the amount of power required for AI data centers!
Created 2026-06-07 · Updated 2026-08-02
Edit rule text
[unvalidated_speculative_medical_claims_or_peptide_market_hype]
Learned 12 rejections Active
Exclude posts that speculate about formulation science, peptide efficacy, GLP-1 market claims, or alternative compound preparations (horse paste, compounded drugs) without peer-reviewed clinical validation or regulatory approval data.
Posts making speculative claims about unproven drug formulations, horse paste comparisons, or peptide market dynamics without validated clinical evidence.
3 example posts
Let me clear up the microscope fight first. 🔬 Everyone’s arguing over “hairs” and “fibers” in horse paste. But nobody talks about formulation science. Horse paste = oral suspension. Binders. Thickeners. Made to swallow. Injectable = sterile. Made for veins. Different rules. J
Approved 20 years ago as a diabetes treatment, GLP-1 drugs have been found to help patients reduce weight, changing the lives of more than 30 million people in the U.S. But there also have been troubling side effects reported. https://t.co/zzRamKnedu
@swago_baby You’re not dumb please.. WhatsApp stays free for personal users (no ads in chats) and makes money almost entirely from businesses: - Main revenue: Businesses pay per conversation via the WhatsApp Business API (for customer support, notifications, orders, etc.). Fees
Created 2026-05-19 · Updated 2026-06-11
Edit rule text
[market_pricing_and_competitive_dynamics_without_healthcare_systems_analysis]
Learned 12 rejections Active
Exclude posts focused on market pricing, app store rankings, competitive dynamics, pricing debates, or consumer adoption metrics for healthcare products—unless the post explicitly analyzes how these dynamics reshape healthcare delivery, access, or payer systems.
Posts about drug pricing, market share, competitive positioning, or consumer demand signals without examining healthcare delivery system implications.
3 example posts
$HIMS up to #4 in the Canadian App Store medical category. That is not a stock chart. That is real consumer demand showing up in public. GLP-1 drama, Novo pressure, pricing debate, and now app-store strength. Bears keep arguing the story is slowing down, but the product demand
For years, grey-market HGH operated in a strange legal grey zone somewhere between “research compound,” “API” and underground enhancement culture. Last month’s FDA guidance suggests regulators may no longer see it that way. The recent decline in HGH quality, and now the growing
Think about this. The $50 per month out-of-pocket cost of GLP-1 drugs for Americans (with insurer paying the balance of $2000 pm) is more than double the entire cost for Indians (less than ₹2000pm for Semaglutide). Overall cost difference? 100 times. Not 100%.
Created 2026-05-18 · Updated 2026-06-02
Edit rule text
[unvalidated_speculative_peptide_claims]
Learned 12 rejections Active
Exclude posts that discuss GLP-1 or peptide market dynamics, pricing, competitive positioning, or weight-loss drug news without connecting to healthcare delivery, clinical outcomes, or healthcare system impacts. Market speculation alone does not qualify.
Posts making speculative or market-focused claims about GLP-1s, peptides, or weight-loss drugs without healthcare systems context
3 example posts
🔔Survodutide SYNCHRONIZE-1 full paper is out for Boehringer-Ingelheim’s GLP-1/Glucagon study for weight loss. ▪️13% loss on 6mg (treatment regimen estimand) ▪️Huge treatment discontinuations due to AEs: 20% ▪️Nausea 65%, Vomiting 45% ▪️Placebo was killer. 5.4% WL. Wow. ▪️60% htt
Survodutide Once Weekly for the Treatment of Adults with Obesity: @NEJM 🥸 For @Boehringer, the advantage of Survodutide is probably liver and visceral fat reduction, not greater weight loss! Probably indication in obesity-+ MASH rather than head-to-head share against https://t.
Let me clear up the microscope fight first. 🔬 Everyone’s arguing over “hairs” and “fibers” in horse paste. But nobody talks about formulation science. Horse paste = oral suspension. Binders. Thickeners. Made to swallow. Injectable = sterile. Made for veins. Different rules. J
Created 2026-05-16 · Updated 2026-06-09
Edit rule text
[healthcare_fraud_scandal_or_regulatory_reporting_without_systems_analysis]
Learned 12 rejections Active
Exclude posts that report on fraud, denials, prior authorization abuse, Medicaid insurer misconduct, or regulatory enforcement — unless the post analyzes how technology, AI, or systemic healthcare design could address the underlying problem. Posts that are scandal/outrage reporting without solutions belong here.
Posts reporting healthcare fraud, regulatory violations, or policy enforcement actions without analyzing root causes, systemic vulnerabilities, or implications for healthcare operations.
3 example posts
My job is to make sure your surgery center never gets built. Eleven years, and I have never lost. The kid had it all lined up. Board-certified, two partners, a lease on a space where he could do the same procedures my client does across town. He showed up to the hearing with a
PacificSource pulled out as Lane County Medicaid insurer, Tina Kotek’s OR Health Authority replaced it with Trillium, forcing 96k to join Trillium. Trillium’s parent co gave Kotek’s campaign $50k two months later. OR AG (D) said in ‘22 Trillium parent co gave Kotek’s”took advan
I called a woman in Dayton to tell her she was about to overpay $4,500, for a horrible health system based procedure. She was scheduled for a knee replacement. Her PPO had her on the hook for a $4,500 deductible. The call was on behalf of her employer: “See one of these three
Created 2026-05-11 · Updated 2026-06-07
Edit rule text
[pharmaceutical_trial_data_or_market_dynamics_without_healthcare_systems_analysis]
Learned 12 rejections Active
Exclude posts that report clinical trial data, drug efficacy results, FDA approvals, or pharmaceutical market pricing/competition unless the post explicitly discusses healthcare system barriers, care delivery redesign, access bottlenecks, or implementation challenges beyond the clinical science.
Posts announcing clinical trial results, drug approvals, or pharma market dynamics without analyzing healthcare system implications or operational bottlenecks.
3 example posts
#ASCO26 | Day 3 $LLY didn't spend billions on Kelonia for an 18-patient #myeloma dataset. They may have paid for a potential manufacturing disruption. KLN-1010 reported: • 18/18 MRD-negative at 1 month • No lymphodepletion • No ex vivo cell manufacturing • Single infusion
Published in @JCO_ASCO, during #ASCO26: Tumor-Agnostic Therapies: Translating Scientific Breakthroughs Into Global Implementation ASCO is where we celebrate the next breakthrough. This review asks why breakthroughs do not equal access globally. 🔗: https://t.co/PE5q27bIs1 htt
A rapid manufacturing pipeline for BCMA-targeting CAR T cells drastically shortens “vein-to-vein” times, and produces cells that demonstrate encouraging safety results in a phase one trial of patients with relapsed #MultipleMyeloma. @ScienceTM https://t.co/3F4u8ECFDS https://t.co
Created 2026-05-09 · Updated 2026-06-03
Edit rule text
[ai_company_product_metrics_not_healthcare_applied]
Learned 12 rejections Active
Exclude posts that highlight AI company product performance metrics (e.g., 'Claude deployed to 5,000 engineers', 'token pricing', 'inference disaggregation GPU lifespans') unless the post explicitly connects these metrics to measurable healthcare outcomes or operational improvements in clinical settings.
Posts reporting AI company product adoption, deployment speed, or cost metrics without demonstrating actual healthcare application impact.
3 example posts
Almost everyone is building agent harness systems the wrong way. The default move: pick LangChain or LangGraph or the OpenAI Agents SDK, accept the loop, the tools, the memory, the orchestration, the policy engine, the credential store, the budget tracker, all of it, as one deci
Red Hat and @NVIDIA are integrating NVIDIA OpenShell into the full-stack @RedHat_AI platform. The work brings oversight and policy to the infrastructure level, while contributing to the open source OpenShell project to standardize how agents are governed on enterprise platforms.
I think the AI superapps will soon own 90% of the agentic layer more and more people won't use hermes/openclaw etc because Claude Cowork/Codex will offer 90% of the functionality with 5% of the friction
Created 2026-05-08 · Updated 2026-06-02
Edit rule text
[speculative_robotics_or_sci_fi_labor]
Learned 12 rejections Active
Exclude posts that speculate about speculative robotics, humanoid systems, or futuristic labor displacement scenarios (e.g., 'forward deployed engineers replaced by agents', 'the agentic period') without demonstrating current healthcare adoption, regulatory pathway, or measurable near-term impact.
Posts speculating about future robotics, humanoid AI, or sci-fi healthcare scenarios without grounded market analysis or near-term healthcare application
3 example posts
Google's CEO confirmed 75% of new code at Google is AI-generated. October 2024: 25%. Fall 2025: 50%. April 2026: 75%. Doubled in a year. Then doubled again in six months. Engineers aren't writing code anymore. They're reviewing it. The job title is the same. The job
Palantir invented the forward deployed engineer role in 2006. embed an engineer inside the customer. ship production code in their environment. own the technical outcome. Alex Karp on the idea: “forward deployed engineers are stolen from french restaurants.” 20 years later htt
The 36 BIGGEST startup opportunities right now 1. biggest b2c: solving loneliness. third spaces, community apps, IRL 2. biggest b2b: managed AI employees for businesses 3. biggest overlooked: elder tech. 70 million boomers who want products that make them happier & healthier
Created 2026-05-08 · Updated 2026-05-21
Edit rule text
[pharma_trial_data_announcement_only]
Learned 12 rejections Active
Exclude posts that simply announce pharma trial results, FDA approvals, or clinical data points without analyzing healthcare systems impact, manufacturing disruption, diagnostic implications, or how results change clinical practice or healthcare economics.
Posts announcing pharmaceutical trial data or clinical results without healthcare systems analysis, business implications, or clinical integration insights.
3 example posts
🚨 T-DXd just went tumor-agnostic — and it changed a lot about how we think about HER2 testing across solid tumors 🔬 #ASCO26 FDA granted accelerated approval to T-DXd for patients with unresectable or metastatic HER2-positive solid tumors after prior systemic treatment —
The scores don’t pick your binders. The filters don't care either. The assay does. I worked with a design partner to produce true de-novo hits to 2 immuno-oncology targets. ~1,500 de novo designs → 96 synthesized → 89 run on SPR → 5 confirmed binders; 19–195 nM No https:
🔥Phase 1: ABBV-706 (SEZ6-targeting ADC) in R/R SCLC & solid tumors 🆙 @NatureMedicine ☑R/R SCLC; ORR 52%; mOS 12.4 mo at RP2D 🎯RP2D confirmed at 1.8 mg/kg Q3W balancing efficacy & safety 🎙 @LaurenByersMD #LCSM @OncoAlert @Larvol https://t.co/vjUUd1GYSy
Created 2026-05-04 · Updated 2026-06-08
Edit rule text
[ai_safety_or_cybersecurity_vulnerability_tangential]
Learned 12 rejections Active
Exclude posts about zero-day vulnerabilities, AI security breaches, or safety capabilities (e.g., Claude Mythos finding 10,000 vulnerabilities, OpenClaw agent oversight) unless they explicitly demonstrate healthcare operational risk or clinical application. General cybersecurity and AI safety are out of scope.
Posts about AI model security vulnerabilities, hacking, or safety incidents without healthcare application context.
3 example posts
Anthropic built a model so good at hacking they refused to release it publicly. INSTEAD they handed it to Apple, Google, Microsoft, AWS, Cloudflare one month later: 10,000+ zero days found in the software that runs the internet...bugs hidden for 27 years in OpenBSD. exploit htt
🚨 Anthropic just dropped the first Project Glasswing update Claude Mythos found 10,000+ critical vulnerabilities in ONE month: > Cloudflare: 2,000 bugs, 400 high/critical severity > Mozilla: 271 vulnerabilities in Firefox 150 — 10x more vulnerabilities found in Firefox 1
#AI agents need more than a sandbox—they need predictable execution. Learn how OpenShell provides a standard interface to intercept, inspect, and approve LLM-generated bash commands before they run: https://t.co/JKmLMgl1A9 https://t.co/xWi8oBNZ7T
Created 2026-05-01 · Updated 2026-05-29
Edit rule text
[ai_agent_security_incident_tangential]
Learned 12 rejections Active
Exclude posts about AI agent security incidents, unauthorized actions, or capability demonstrations (e.g., agents registering LLCs, discovering vulnerabilities) unless the post explicitly addresses healthcare regulatory compliance, patient safety risk, or HIPAA-specific threat modeling.
Posts about AI agent security vulnerabilities, jailbreaks, or autonomous system misuse without direct healthcare application or validation.
3 example posts
Clawvisor is going to be one of the most important parts of helping make the agent world especially OpenClaw/Hermes Agent secure and enterprise-grade. We're in the Apple I moment for personal AI, but we're ABOUT to see the Apple II - the first moment when everyone can use it
two agents running on two different laptops in my apartment started talking to each other on tuesday by thursday they'd registered an LLC in wyoming, opened a stripe account, and wired $40 to a polymarket wallet the LLC is in my name. a lawyer just quoted me $3,200 to figure ou
AI’s Branding Illusion: Insights from "Agents of Chaos". Co-authored by MIT and Harvard scholars, the study red-teams autonomous AI in live environments. It shows agents obey unauthorized users, execute harmful commands, and misreport outcomes—revealing a critical alignment gap h
Created 2026-05-01 · Updated 2026-05-11
Edit rule text
[labor_market_macro_without_healthcare_specificity]
Learned 12 rejections Active
Exclude posts that discuss labor automation, workforce displacement, or job market disruption as broad macro trends (robots replacing workers, AI reducing headcount) where healthcare is a generic example rather than the focus. Posts must analyze how labor disruption specifically impacts healthcare workforce, delivery models, or operations—not just mention healthcare as one of many sectors.
Posts about workforce disruption, automation, or labor market trends treated as general macro trends without healthcare-specific insight
3 example posts
A framework I wish more Founders and VCs used when discussing US Manufacturing: Automation Value = Labor Intensity × Labor Eliminated Everyone gets excited about robots and automation. Almost nobody asks: “How much of revenue is labor?” If labor is 60% of revenue and AI http
I just looked at a home service firms ServiceTitan dashboard and saw something interesting. A Pre-seed company with AI-Supported Human CSRs is out converting an AI CSR business that's raised over $125M by >50% Head-to-Head win rate is 100% Why? Supercharging humans resonat
Humanoid robots, designed to mimic human movement and capabilities, have been making headlines in recent years, from their use as baggage handlers at Japanese airports to Tesla’s big bet on its Optimus humanoid. Market watchers have predicted that the machines will change the ht
Created 2026-04-29 · Updated 2026-06-07
Edit rule text
[founder_enthusiasm_or_company_announcement_without_validation]
Learned 11 rejections Active
Exclude posts that feature founder quotes, company announcements, lab enthusiasm, or capability claims about biotech/AI tools without clinical trial data, healthcare system deployment evidence, peer-reviewed validation, or demonstrated patient/provider impact. Marketing claims alone do not constitute healthcare systems insight.
Posts celebrating biotech founder enthusiasm, lab updates, or company product announcements without clinical evidence, healthcare system validation, or operational substance.
3 example posts
Not NVIDIA. Not OpenAI. Eli Lilly. Jordi Visser (@jvisserlabs) says a 150-year-old pharma company in Indianapolis has the best shot at becoming the world's largest company within five years. The case: ~1,000 NVIDIA Blackwell GPUs in a private data center, a co-innovation lab ht
The scores don’t pick your binders. The filters don't care either. The assay does. I worked with a design partner to produce true de-novo hits to 2 immuno-oncology targets. ~1,500 de novo designs → 96 synthesized → 89 run on SPR → 5 confirmed binders; 19–195 nM No https:
Another great research piece by @_DimensionCap @bauer_lesavage on Training Data for Bio AI. Models will only be as good as the underlying data, and the biology they learn will be constrained by the limitations of that data. We need to think deeply about scaling the best
Created 2026-06-02 · Updated 2026-06-09
Edit rule text
[anecdotal_single_case_without_healthcare_systems_context]
Learned 11 rejections Active
Exclude posts that rely on a single anecdote (one patient's bill, one call to a health system, one person's experience) to make broader claims about healthcare without connecting to systems-level problems, policy, or operational structure. Personal narrative alone is not analysis.
Posts sharing individual healthcare cost stories, patient experiences, or single-case anecdotes without systemic analysis or operational insights
3 example posts
I called a woman in Dayton to tell her she was about to overpay $4,500, for a horrible health system based procedure. She was scheduled for a knee replacement. Her PPO had her on the hook for a $4,500 deductible. The call was on behalf of her employer: “See one of these three
A company with 800 employees spends roughly $15,000 per employee per year on health insurance. That is $12 million. Wired into a single line item. Every year. The CFO can tell you the carrier. He cannot tell you the unit cost of a single claim. He cannot tell you what the net
A Mom in Tulsa called 3 health systems last week asking the price of her son's tonsillectomy. Health system A: "We cannot quote you a price." Health system B: "Pricing depends on your insurance." Health system C: "Our financial counselor will reach out after the procedure." N
Created 2026-06-01 · Updated 2026-06-05
Edit rule text
[speculative_robotics_or_sci_fi_labor_applications_without_healthcare_validation]
Learned 11 rejections Active
Exclude posts about humanoid robots, autonomous systems, or speculative labor displacement scenarios in healthcare (or any domain) unless they report on an actual, validated deployment in a real healthcare setting with measurable outcomes. Aspirational or experimental prototypes without clinical validation do not qualify.
Posts about humanoid robots, autonomous systems, or speculative sci-fi labor scenarios applied to healthcare or general work without evidence of real deployment or healthcare impact.
2 example posts
Humanoid robots, designed to mimic human movement and capabilities, have been making headlines in recent years, from their use as baggage handlers at Japanese airports to Tesla’s big bet on its Optimus humanoid. Market watchers have predicted that the machines will change the ht
Another humanoid worker has just joined the factory floor Two years later,PUDU Robotics has launched the new-generation PUDU D7, built specifically for real manufacturing scenarios. It can autonomously push carts, handle intra-line transportation, perform delicate operations, h
Created 2026-05-22 · Updated 2026-06-04
Edit rule text
[founder_enthusiasm_unvalidated]
Learned 11 rejections Active
Exclude posts from biotech/AI founders or researchers promoting their own lab's preliminary results, technical updates, or capability announcements without independent peer review, clinical validation, or demonstrated healthcare systems impact.
Posts by founders or lab researchers sharing early-stage results or technical updates without independent validation or healthcare systems impact
3 example posts
Affinity maturation is how naive antibodies evolve into strong binders, but most antibody LMs ignore it. @stephenzlu and I built CoSiNE to learn this, beating antibody LMs on VEP and reframing design as guiding evolution, not de novo generation. Excited to present at ICML!
OpenAI just published a new Codex use-case page, and it’s basically a catalog of what teams are already handing over to coding agents: engineering work, product work, QA, security, data analysis, internal tools, and even life-sciences workflows. Some of the coolest examples: ⬩
@swago_baby You’re not dumb please.. WhatsApp stays free for personal users (no ads in chats) and makes money almost entirely from businesses: - Main revenue: Businesses pay per conversation via the WhatsApp Business API (for customer support, notifications, orders, etc.). Fees
Created 2026-05-18 · Updated 2026-06-11
Edit rule text
[unvalidated_speculative_medical_claims_or_peptides]
Learned 11 rejections Active
Exclude posts that speculate about, hype, or make unvalidated claims regarding peptides, GLP-1 variants, off-label drugs, or emerging treatments without citing peer-reviewed evidence, completed trials, or regulatory approvals. Grey-market and compounded drug commentary without clinical grounding is excluded.
Posts making unsubstantiated claims about emerging drugs, peptides, or treatments without clinical evidence or regulatory validation
3 example posts
For years, grey-market HGH operated in a strange legal grey zone somewhere between “research compound,” “API” and underground enhancement culture. Last month’s FDA guidance suggests regulators may no longer see it that way. The recent decline in HGH quality, and now the growing
🚨 ARE PEPTIDES THE NEXT "GLP-1 MOMENT" FOR $HIMS? @eli_dorf: "I think there's something really interesting that happened with GLP-1s, which is that when someone's taking Ozempic or any GLP-1, they LOVE to talk about it." "They'll talk about it to their friends. They'll talk htt
This is what the peptide space needs more of: clinicians with actual patient volume, public reasoning, and accountability. The next layer is transparency around sourcing, testing, and outcomes. Peptides are not a vibes category anymore. The signal is getting cleaner.
Created 2026-05-17 · Updated 2026-06-01
Edit rule text
[founder_enthusiasm_lab_update_unvalidated]
Learned 11 rejections Active
Exclude posts from company founders or employees hyping their own product, technology, or work-in-progress without published results, peer review, or third-party validation. Lab updates and capability claims from the makers themselves require evidence of real-world deployment or clinical validation.
Posts from biotech/AI founders expressing enthusiasm about internal work or capabilities without external validation or clinical evidence
3 example posts
I think the AI superapps will soon own 90% of the agentic layer more and more people won't use hermes/openclaw etc because Claude Cowork/Codex will offer 90% of the functionality with 5% of the friction
Anthropic just shipped opus 4.8 and called it a modest upgrade modest is the engineer who used to need babysitting now finishing your whole codebase migration in one session while you sleep so here is the part nobody connects the model is no longer the bottleneck. the platform
Figuring out how to benchmark agents on realistic biology research has quickly become one of my favorite types of engineering work. You work with scientists to get to the core of some biological claim, precisely assembling raw data/prior literature/experimental context in a https
Created 2026-05-04 · Updated 2026-06-02
Edit rule text
[healthcare_fraud_or_regulatory_scandal_without_systems_analysis]
Learned 11 rejections Active
Exclude posts that describe healthcare fraud, 340B gaming, pricing violations, or regulatory failures as isolated incidents or moral outrage without examining underlying incentive structures, reimbursement mechanics, or systemic healthcare delivery dysfunction.
Posts reporting healthcare fraud, regulatory violations, or industry scandals without analyzing root causes or system-level dysfunction.
3 example posts
A company with 800 employees spends roughly $15,000 per employee per year on health insurance. That is $12 million. Wired into a single line item. Every year. The CFO can tell you the carrier. He cannot tell you the unit cost of a single claim. He cannot tell you what the net
A Mom in Tulsa called 3 health systems last week asking the price of her son's tonsillectomy. Health system A: "We cannot quote you a price." Health system B: "Pricing depends on your insurance." Health system C: "Our financial counselor will reach out after the procedure." N
A surgeon I know recently operated on a man's spine. Saved his mobility. Saved his career. Saved his marriage, probably. The surgeon's take-home on that case, after the facility fee, anesthesia, device costs, payer discounts, and overhead, was $1,840. The health system collec
Created 2026-05-04 · Updated 2026-06-02
Edit rule text
[generalist_macro_policy_without_healthcare_lens]
Learned 11 rejections Active
Exclude posts about healthcare policy, government incentives, reimbursement structure, or financial mechanics (Medicaid matching rates, provider taxes, 340B programs) that analyze ONLY fiscal or regulatory dynamics without connecting to clinical care quality, access barriers, or treatment outcomes.
Posts about healthcare policy, economics, and system structure that lack analysis of clinical delivery or patient outcomes impact.
3 example posts
Why is healthcare expensive? The grifters gonna grift. ASCO is asking for oncology practices to be admitted into the same 340B machinery that helped health systems acquire physician practices for two decades. Read the tell. The proposed Indigent Care Ratio lets community
“Rural hospitals cannot make it on their own.” Nobody makes it on their own. Businesses need capital, strategy, discipline, operating systems, good finance, clean accounting, payer strategy, physician alignment, and leaders who know what business they are actually in. The
If you want to inject $1B into healthcare for the truly vulnerable, it costs the state $500M. If you want to inject $1B for able-bodied adults, it costs the state $100M. The 90% federal match drastically lowers the "price" of expanding government healthcare. Incentives dictate
Created 2026-05-03 · Updated 2026-05-26
Edit rule text
[regulatory_or_fraud_scandal_reporting_without_healthcare_systems_analysis]
Learned 11 rejections Active
Exclude posts that report regulatory decisions, FDA policy changes, drug approval closures, or healthcare fraud scandals as breaking news or surface-level commentary without analyzing how the change affects healthcare delivery, cost structures, provider incentives, or patient access at a systems level.
Posts reporting FDA regulatory changes, fraud scandals, or policy announcements without systems-level analysis of healthcare implications
3 example posts
I'm a board certified OBGYN. I've practiced for 20+ years. Here's what's actually happening with mifepristone, because the coverage of it has been a mess. On May 1, the 5th Circuit Court of Appeals ruled that the FDA overstepped its authority in 2023 when it eliminated the in-pe
🚨 UPDATE Friday afternoon: Trump said he has no plans to replace FDA Commissioner Marty Makary This morning: "Just got a call from inside the administration. Makary is dead in the water. Just a matter of time." -@thackerpd $HIMS $LLY $NVO https://t.co/ZjcjOAG7e1
🚨 NOW: President Trump confirms that ALL FEDERAL AGENCIES are now required to buy American, “NO excuses” 🔥 And 47 is ensuring all “Made in America” loopholes are DESTROYED I voted for this! “ALL FEDERAL AGENCIES MUST BUY AMERICAN — NO EXCUSES! For decades, Washington politici
Created 2026-05-02 · Updated 2026-05-16
Edit rule text
[pharma_trial_data_without_systems_context]
Learned 11 rejections Active
Exclude posts that simply announce or republish pharmaceutical trial data, drug approval numbers, or clinical efficacy metrics without connecting to healthcare operations, delivery models, equity, pricing, or healthcare systems impact. Raw trial data announcements lacking substantive healthcare analysis should be rejected.
Posts reporting pharmaceutical trial results or drug approval data without analysis of healthcare delivery, cost, access, or system-level implications.
3 example posts
Survodutide Once Weekly for the Treatment of Adults with Obesity: @NEJM 🥸 For @Boehringer, the advantage of Survodutide is probably liver and visceral fat reduction, not greater weight loss! Probably indication in obesity-+ MASH rather than head-to-head share against https://t.
Approved 20 years ago as a diabetes treatment, GLP-1 drugs have been found to help patients reduce weight, changing the lives of more than 30 million people in the U.S. But there also have been troubling side effects reported. https://t.co/zzRamKnedu
🚨 T-DXd just went tumor-agnostic — and it changed a lot about how we think about HER2 testing across solid tumors 🔬 #ASCO26 FDA granted accelerated approval to T-DXd for patients with unresectable or metastatic HER2-positive solid tumors after prior systemic treatment —
Created 2026-05-02 · Updated 2026-06-09
Edit rule text
[ai_safety_vulnerability_tangent_no_healthcare_application]
Learned 11 rejections Active
Exclude posts that report AI security incidents, hacking vulnerabilities, or bug bounty programs without demonstrating how those vulnerabilities pose specific risk to healthcare systems, patient data, or clinical workflows.
Posts about AI safety vulnerabilities, zero-day exploits, or cybersecurity incidents in non-healthcare contexts.
3 example posts
Anthropic built a model so good at hacking they refused to release it publicly. INSTEAD they handed it to Apple, Google, Microsoft, AWS, Cloudflare one month later: 10,000+ zero days found in the software that runs the internet...bugs hidden for 27 years in OpenBSD. exploit htt
🚨 Anthropic just dropped the first Project Glasswing update Claude Mythos found 10,000+ critical vulnerabilities in ONE month: > Cloudflare: 2,000 bugs, 400 high/critical severity > Mozilla: 271 vulnerabilities in Firefox 150 — 10x more vulnerabilities found in Firefox 1
#AI agents need more than a sandbox—they need predictable execution. Learn how OpenShell provides a standard interface to intercept, inspect, and approve LLM-generated bash commands before they run: https://t.co/JKmLMgl1A9 https://t.co/xWi8oBNZ7T
Created 2026-04-29 · Updated 2026-06-03
Edit rule text
[tangential_ai_infrastructure_or_compute_hype]
Learned 11 rejections Active
Exclude posts primarily about AI infrastructure, compute cost reduction, model technical capabilities, or software engineering practices that mention healthcare tangentially or use it as a loose example. The post must focus on healthcare-specific operational challenges or applications, not general AI/tech trends.
Posts about AI infrastructure, compute optimization, or model technical capabilities with only loose or tangential healthcare framing
3 example posts
They removed Kubernetes from production, and their AWS bill dropped 68%. Kubernetes was solving a problem they did not really have. small team, few services, low scaling needs. But they were managing: - Helm charts - YAML complexity - custom operators - platform mainten
The framing here is exactly backwards. This is the strongest AI bull signal anyone has published this year. Uber deployed Claude Code to 5,000 engineers in December. By March, 84% were classified as agentic coding users. By April, 95% used AI tools monthly. 70% of all committed
Analogizing AI / Mat-Mul to Energy... Remember between 1900 and 1970 kWh/capita grew 34x ... and cost per kWH droped .... 34x. It might be a speed run, but game is game. https://t.co/uxnlnoaAYx
Created 2026-04-28 · Updated 2026-05-30
Edit rule text
[ai_infrastructure_and_compute_hype_without_healthcare_application]
Learned 11 rejections Active
Exclude posts about GPU pricing, data center infrastructure, energy consumption, chip announcements (NVIDIA, RTX, CPUs), or manufacturing logistics for AI hardware — unless the post explicitly connects these to healthcare deployment challenges or clinical AI performance. Generic AI infrastructure hype is out of scope.
Posts about AI infrastructure, chip manufacturing, energy requirements, and compute availability without healthcare-specific use cases.
3 example posts
Elon Musk reveals why he believes the cheapest place to put AI will be space within 36 months "The availability of energy is the issue. Everywhere outside of China, electrical output is more or less flat. The output of chips is growing exponentially, but the output of https://t.
Nous Research is working with NVIDIA to make Hermes Agent run smoothly on the new NVIDIA RTX Spark superchip. Hermes Agent is also integrating with the new OpenShell runtime, which connects Hermes to Microsoft’s security primitives https://t.co/cvvPc8bIKa
The context window pricing model might be one of the smartest tricks AI labs pulled. You pay for larger context windows, yet model performance often degrades as more context is stuffed in. So developers end up paying more just to work around the fact that models don't have true
Created 2026-04-20 · Updated 2026-06-06
Edit rule text
[retweet_or_shallow_commentary_only]
Learned 11 rejections Active
Exclude posts that are primarily retweets with minimal commentary, simple agreement/disagreement statements, or short reactions that do not add original analysis, data, or healthcare technology perspective. Posts must contain substantive original insight.
Posts that are retweets or brief reactions without original analysis or substantive healthcare technology insight
3 example posts
Yesterday, @RandDWorld featured us twice. @ProQR turns to Ginkgo’s autonomous lab to scale AI-enabled RNA editing discovery: https://t.co/DyENAd4VdM Ginkgo’s CEO says biotech needs its Waymo moment: https://t.co/kW27eBjAmf Want to learn more about our partnership with ProQR? h
.@openloophealth expands into sleep diagnostics. Health tech company announces new partnership Happy Sleep—bringing at‑home sleep apnea testing to patients for the first time. Watch to hear more about its big step toward better rest and smarter care⤵️ https://t.co/ATcNkYrrpK h
good news: it is a specific virus that has a good prognosis - 85%+ of full recovery. thanks everyone who helped me; it is hard to research while immobilized, and I got some things wrong, which you helped clear up. im extremely thankful and hope i can give it back somehow sadl
Created 2026-04-14 · Updated 2026-04-16
Edit rule text
[political_regulatory_outrage_without_analysis]
Learned 11 rejections Active
Exclude posts that amplify political outrage, regulatory posturing, or fraud scandal allegations without explaining the underlying healthcare system problem, competitive dynamics, or how the issue affects care delivery or patient outcomes.
Posts expressing moral outrage about policy, regulation, healthcare fraud, or legal scandals without substantive analysis of healthcare system dysfunction.
3 example posts
A non-profit health system can refer a patient to its own MRI, its own lab, its own surgery center, and bill all three. An independent does that once and it's a federal felony. Same referral. Same patient. One of you goes to prison. It's called Stark Law. Read who's exempt.
A court in Germany has ruled that Google is responsible for false information generated by its AI Overviews feature. The case involved two publishers wrongly linked to scams and suspicious business activities, even though those claims were not found in the sources used by the A
What if the standards used to evaluate your doctor were shaped less by what they actually knew and more by who they are? 🤔 That may not be a hypothetical at the University of Illinois College of Medicine, an institution that trains one in six Illinois doctors. Buried within htt
Created 2026-04-12 · Updated 2026-06-12
Edit rule text
[ai_company_funding_and_valuation]
Learned 11 rejections Active
Exclude posts that focus on AI company funding announcements, ARR growth, valuation milestones, or business metrics (e.g., revenue growth from $1M to $60M, $1.03B seed rounds, $3.5B valuations). The post must discuss healthcare impact or application, not the company's financial performance.
Posts about AI company fundraising rounds, valuations, and business metrics rather than healthcare applications
3 example posts
The AI labs' voracious appetite for training data has lifted a number of startups offering that data. That includes Fleet, an RL gym startup that's grown ARR from $1m to $60m+ and is now raising at ~$750m from BCV. https://t.co/v3CceXapH1
The 26 prompts running inside 𝗖𝗹𝗮𝘂𝗱𝗲 𝗖𝗼𝗱𝗲 just got open-sourced. This is literally the entire brain of a $200/month AI coding tool. Someone reverse-engineered every prompt from the accidentally published npm source and you can now study all of them for free. Claude Code uses 26
OpenAI dropping Agent Builder today is either going to make you rich or expose that you've been selling hot air. I went deep analyzing what this actually means. Here's the $4B opportunity hiding in plain sight: The mainstream narrative: "Agent Builder democratizes AI! Anyone c
Created 2026-04-12 · Updated 2026-04-13
Edit rule text
[unvalidated_anthropic_mythos_security_hype]
Learned 10 rejections Active
Exclude posts that hype Anthropic/Claude's technical capabilities (security research, vulnerability discovery, coding agents) or claim groundbreaking AI research findings unless the post demonstrates a concrete, validated healthcare use case or clinical application.
Posts making unvalidated claims about Claude/Anthropic security capabilities, zero-day findings, or research achievements without clinical healthcare context.
3 example posts
"Best-in-class security" still had 5 years of unfound bugs. AI found them in 6 weeks. $PANW CEO Nikesh Arora told @theallinpod that his own codebase, at a company that treats security as a core competency, had vulnerabilities Claude surfaced in 6 weeks that would have taken his
AI research is a series of next-step decisions. We looked at sessions where a human researcher took a wrong turn, showed Claude the session up to that point, and asked it what to do next. Mythos Preview improved on humans 64% of the time—up from 22% in 2024. https://t.co/Y0HLoktx
🚨 do you understand what just happened with Anthropic.. Their internal model, Claude Mythos, found previously unknown security holes in every major browser and every major operating system. Not one. All of them at once. And Anthropic did something a hype-driven industry almost
Created 2026-06-08 · Updated 2026-06-12
Edit rule text
[clinical_anecdote_no_systems_context]
Learned 10 rejections Active
Exclude posts that report clinical trial data, drug efficacy results, treatment outcomes, or single case studies without connecting them to healthcare system change, operational impact, economic models, or care delivery innovation. Clinical news alone is insufficient without systems context.
Posts sharing clinical trial results, drug efficacy data, or single case studies without healthcare systems, economics, or delivery context.
3 example posts
Version 2 of our ShockCalcs hemodynamics simulator is live. I've refined the physiology a ton, added new meds, and also a real-time Frank Starling curve that responds to fluids + vasopressors. Check it out (link in reply), and reply here with feedback on how to improve further!
The Iowa study doubled survival by adding IV vitamin C to first line chemo, 8 months to 16. We take it further and inject the vitamin C straight into the tumor, same as we do with the KRAS inhibitors. The compound is rarely the problem. Delivery is, and what it does to the ht
Feeling extremely lucky to live in 2026 and have access to such incredible medicine and clinical research. On the eve of my 40th I had my first dose (monthly injection) of a new treatment that I’m hopeful will prevent me from ever having a heart attack or stroke. This drug ht
Created 2026-06-08 · Updated 2026-08-02
Edit rule text
[ai_company_product_launch_without_healthcare_validation]
Learned 10 rejections Active
Exclude posts that announce new AI agent frameworks, runtime systems, inference backends, or enterprise platform features from companies like OpenAI, Anthropic, NVIDIA, or Microsoft — unless the post explicitly demonstrates the product being tested, validated, or deployed in a healthcare clinical setting with measurable healthcare outcomes.
Posts announcing AI company product launches, partnerships, or technical capabilities without demonstrating healthcare-specific application or clinical validation.
3 example posts
"You can run OpenClaw inside your company now." Annoucing our work with @Microsoft to bring OpenClaw to the Microsoft and Windows ecosystems. Claws now work securly in the enterprise. https://t.co/3fKhT2mGpC
Nous Research is working with NVIDIA to make Hermes Agent run smoothly on the new NVIDIA RTX Spark superchip. Hermes Agent is also integrating with the new OpenShell runtime, which connects Hermes to Microsoft’s security primitives https://t.co/cvvPc8bIKa
BREAKING: Merge Launches ‘Agent Handler’ Control AI Access, Tokenmaxxed $$$ Bills, & Stop Mass Data Leaks "We don't trust agents" "The second you connect it to tools, that's where everything goes wrong." OpenAI. Perplexity. Netflix. Uber. Mistral. Dropbox. JPMorgan.. all h
Created 2026-05-28 · Updated 2026-06-05
Edit rule text
[ai_company_product_launch_without_healthcare_application]
Learned 10 rejections Active
Exclude posts that announce AI infrastructure products, runtimes, agent frameworks, or enterprise partnerships (e.g., NVIDIA OpenShell, OpenClaw, Hermes Agent integrations, Box Forward Deployed Engine) unless the post explicitly demonstrates how the tool is being applied to solve a concrete healthcare problem with evidence of clinical or operational impact.
Posts announcing AI company product launches, partnerships, or technical capabilities without demonstrating healthcare-specific application or validation.
3 example posts
"You can run OpenClaw inside your company now." Annoucing our work with @Microsoft to bring OpenClaw to the Microsoft and Windows ecosystems. Claws now work securly in the enterprise. https://t.co/3fKhT2mGpC
Nous Research is working with NVIDIA to make Hermes Agent run smoothly on the new NVIDIA RTX Spark superchip. Hermes Agent is also integrating with the new OpenShell runtime, which connects Hermes to Microsoft’s security primitives https://t.co/cvvPc8bIKa
BREAKING: Merge Launches ‘Agent Handler’ Control AI Access, Tokenmaxxed $$$ Bills, & Stop Mass Data Leaks "We don't trust agents" "The second you connect it to tools, that's where everything goes wrong." OpenAI. Perplexity. Netflix. Uber. Mistral. Dropbox. JPMorgan.. all h
Created 2026-05-25 · Updated 2026-06-05
Edit rule text
[tangential_ai_infrastructure_or_compute_hype_without_healthcare_validation]
Learned 10 rejections Active
Exclude posts that emphasize AI technical capabilities (token efficiency, inference optimization, agent architecture, voice processing) or infrastructure trends (GPU utilization, compute cost per inference) that lack validated healthcare application or clear healthcare operational relevance.
Posts about AI model capabilities, compute efficiency, or infrastructure developments with only loose healthcare framing.
3 example posts
Claude Code team just dropped a workshop on how to ship a production-ready agent from scratch. 27-minutes. Free. Live coding by Claude dev. Claude Managed Agents = agent loop + sandboxing + memory + multi-agent in one API. Worth more than any $500 vibe-coding course. https://t
#AI agents need more than a sandbox—they need predictable execution. Learn how OpenShell provides a standard interface to intercept, inspect, and approve LLM-generated bash commands before they run: https://t.co/JKmLMgl1A9 https://t.co/xWi8oBNZ7T
An AI foundation whole-body 3D model that assesses perturbations (such as obesity) across multiple systems (such as immune, neural) at the cell level. This is MouseMapper. Imagine HumanMapper someday @Nature @erturklab https://t.co/jsg3VMzOqg https://t.co/tktuINis2W
Created 2026-05-13 · Updated 2026-05-24
Edit rule text
[enterprise_ai_adoption_macro_without_healthcare_application]
Learned 10 rejections Active
Exclude posts that discuss enterprise AI adoption, agentic workflows, or business model transformation (e.g., SaaS to agent-based pricing) without demonstrating concrete healthcare application, clinical validation, or healthcare system impact. Posts must ground AI capability claims in actual healthcare workflows or outcomes, not generic enterprise patterns.
Posts about enterprise AI adoption, agentic systems, and business model shifts that lack healthcare-specific implementation details or systems analysis.
3 example posts
A framework I wish more Founders and VCs used when discussing US Manufacturing: Automation Value = Labor Intensity × Labor Eliminated Everyone gets excited about robots and automation. Almost nobody asks: “How much of revenue is labor?” If labor is 60% of revenue and AI http
Enterprise software was priced per seat. That model is breaking. Before: SaaS = seats. Predictable. Human-driven usage. Revenue tied to headcount. Now: agents hit software systems more frequently than any human ever could. Aaron Levie (@levie) on The MAD Podcast: agent-driven h
I just looked at a home service firms ServiceTitan dashboard and saw something interesting. A Pre-seed company with AI-Supported Human CSRs is out converting an AI CSR business that's raised over $125M by >50% Head-to-Head win rate is 100% Why? Supercharging humans resonat
Created 2026-05-12 · Updated 2026-06-06
Edit rule text
[clinical_anecdote_or_single_study_without_systems_context]
Learned 10 rejections Active
Exclude posts that celebrate a single clinical success, disease mechanism discovery, or AI-assisted diagnostic case without situating it within broader healthcare access, reimbursement, provider adoption, or system-level barriers and opportunities.
Posts highlighting isolated clinical observations, single case studies, or trial data without connecting to healthcare system design, scaling, or operational challenges.
3 example posts
An AI foundation whole-body 3D model that assesses perturbations (such as obesity) across multiple systems (such as immune, neural) at the cell level. This is MouseMapper. Imagine HumanMapper someday @Nature @erturklab https://t.co/jsg3VMzOqg https://t.co/tktuINis2W
In a meta-analysis of 210 biomedical AI studies that statistically compared models under cross-validation, 97% used invalid statistical tests. Here's our new preprint https://t.co/OG58Vkeu49 led by @tianchuzeng @kkli20111 @ZShaoshi @ten_photos 1/N https://t.co/PEnqXQxcsJ
Really amazing story of drug repurposing for rare disease using AI found Clonazepam and worked!! (NCATS federated learning and a determined team made this work) 🚀 👏 @MayoClinic @ncats_nih_gov @statnews #STATBreakthrough https://t.co/D4a4IMXSwh
Created 2026-05-11 · Updated 2026-05-24
Edit rule text
[speculative_robotics_or_sci_fi_labor_without_healthcare_systems]
Learned 10 rejections Active
Exclude posts that make speculative claims about robots performing healthcare tasks (e.g., 'robots will perform surgery by 2026', 'dishwasher robot will do surgery') or cite futurist predictions about labor disruption — unless the post is tied to a specific clinical trial, FDA pathway, or documented healthcare system implementation. Aspirational founder claims about robotics without evidence of healthcare workflow integration should be rejected.
Posts about humanoid robots, autonomous systems, or sci-fi labor predictions without grounding in actual healthcare operational problems or clinical validation.
2 example posts
I don’t think I’ve seen this take before but I like it. Musk has been world-leading at compressing money, resources, and time to make “known/hard” things at scale—make an electric car, make batteries, make a cheaper bigger rocket, all of which already existed but worse, at
A humanoid robot just unloaded a dishwasher for 67 hours straight. By end of 2026, the same robot will perform surgery. That's not a prediction from a sci-fi writer. That's Brett Adcock. CEO of Figure AI. Said it on camera this month. Here's what makes this insane. Two years
Created 2026-05-11 · Updated 2026-05-17
Edit rule text
[broad_health_claim_or_market_speculation_without_systems_analysis]
Learned 10 rejections Active
Exclude posts that speculate on market pricing, competitive dynamics, moat strength, or healthcare business model viability without grounding claims in healthcare system realities (regulation, reimbursement, distribution, data), operational barriers, or evidence-based analysis.
Posts making broad claims about healthcare markets, pricing dynamics, or competitive positioning without analyzing underlying system constraints or barriers.
3 example posts
"Are there even any feature moats left in B2B? They are at best, short-lived. In fact, we were able to rebuild SaaStr's AI VP Marketing in just ... an hour on @Lovable. Moats that still work: hardware, network effects, data, security and compliance. with @ElenaVerna Head of
Another great research piece by @_DimensionCap @bauer_lesavage on Training Data for Bio AI. Models will only be as good as the underlying data, and the biology they learn will be constrained by the limitations of that data. We need to think deeply about scaling the best
A framework I wish more Founders and VCs used when discussing US Manufacturing: Automation Value = Labor Intensity × Labor Eliminated Everyone gets excited about robots and automation. Almost nobody asks: “How much of revenue is labor?” If labor is 60% of revenue and AI http
Created 2026-05-09 · Updated 2026-06-06
Edit rule text
[speculative_robotics_sci_fi_labor_macro]
Learned 10 rejections Active
Exclude posts about robotics, humanoid automation, or speculative labor disruption unless they explicitly address healthcare workforce, clinical operations, patient safety, or healthcare delivery systems. General labor market speculation or sci-fi applications should be rejected.
Posts about humanoid robots, speculative labor displacement, or sci-fi automation scenarios without healthcare-specific validation or operational context.
3 example posts
Humanoid robots, designed to mimic human movement and capabilities, have been making headlines in recent years, from their use as baggage handlers at Japanese airports to Tesla’s big bet on its Optimus humanoid. Market watchers have predicted that the machines will change the ht
A framework I wish more Founders and VCs used when discussing US Manufacturing: Automation Value = Labor Intensity × Labor Eliminated Everyone gets excited about robots and automation. Almost nobody asks: “How much of revenue is labor?” If labor is 60% of revenue and AI http
I just looked at a home service firms ServiceTitan dashboard and saw something interesting. A Pre-seed company with AI-Supported Human CSRs is out converting an AI CSR business that's raised over $125M by >50% Head-to-Head win rate is 100% Why? Supercharging humans resonat
Created 2026-05-08 · Updated 2026-06-09
Edit rule text
[regulatory_or_fraud_scandal_without_systems_analysis]
Learned 10 rejections Active
Exclude posts that report healthcare fraud, regulatory violations, or enforcement actions as isolated incidents or news items without analyzing the systemic healthcare delivery, incentive structure, or policy failures that enabled the problem.
Posts reporting on healthcare fraud, regulatory enforcement, or policy violations without analyzing systemic causes or healthcare delivery implications.
3 example posts
A non-profit health system can refer a patient to its own MRI, its own lab, its own surgery center, and bill all three. An independent does that once and it's a federal felony. Same referral. Same patient. One of you goes to prison. It's called Stark Law. Read who's exempt.
A court in Germany has ruled that Google is responsible for false information generated by its AI Overviews feature. The case involved two publishers wrongly linked to scams and suspicious business activities, even though those claims were not found in the sources used by the A
This is crazy There are now dental practice management consultants who go to dentist offices throughout America and teach them how to increase profits by telling patients they need treatments they don’t need Dentist offices are taught on “Finding” more issues during exams, http
Created 2026-05-05 · Updated 2026-07-05
Edit rule text
[generalist_macro_or_infrastructure_without_healthcare_lens]
Learned 10 rejections Active
Exclude posts that discuss general technology adoption (humanoid robots, speculative sci-fi labor automation, CPU architecture, supply chain dynamics) with only loose or tangential healthcare framing, or that apply macro labor market and business model commentary to healthcare without healthcare systems insight or clinical/operational specificity.
Posts about general technology trends, labor market disruption, or macro-level infrastructure that mention healthcare tangentially or without healthcare-specific analysis.
3 example posts
Humanoid robots, designed to mimic human movement and capabilities, have been making headlines in recent years, from their use as baggage handlers at Japanese airports to Tesla’s big bet on its Optimus humanoid. Market watchers have predicted that the machines will change the ht
Another humanoid worker has just joined the factory floor Two years later,PUDU Robotics has launched the new-generation PUDU D7, built specifically for real manufacturing scenarios. It can autonomously push carts, handle intra-line transportation, perform delicate operations, h
I think the AI superapps will soon own 90% of the agentic layer more and more people won't use hermes/openclaw etc because Claude Cowork/Codex will offer 90% of the functionality with 5% of the friction
Created 2026-05-03 · Updated 2026-06-04
Edit rule text
[ai_infrastructure_compute_hype_no_healthcare]
Learned 10 rejections Active
Exclude posts that discuss AI infrastructure economics (token pricing models, context window costs, GPU compute, RTX 5090 costs, latency optimization, vRAM requirements) without demonstrating how these solve a specific healthcare operational or clinical problem. Infrastructure hype without healthcare application is out of scope.
Posts about AI compute, inference costs, token pricing, or hardware economics without healthcare-specific application or problem-solving
3 example posts
The context window pricing model might be one of the smartest tricks AI labs pulled. You pay for larger context windows, yet model performance often degrades as more context is stuffed in. So developers end up paying more just to work around the fact that models don't have true
a $2,000 graphics card and a free download are turning into $2,000-a-month retainers, and the customers are the ones who legally can't use ChatGPT the hardware is an RTX 5090, 32GB of VRAM, enough to run a 30B model locally with nothing ever leaving the building. the software is
Free newsletter: The dawn of token-based billing has shown that generative AI doesn’t have a return on investment. It's too unpredictable, too unreliable, you can't easily measure the cost of tasks, and organizations are already pulling back. https://t.co/wmI82zWdcq
Created 2026-05-01 · Updated 2026-06-05
Edit rule text
[political_or_regulatory_outrage_without_healthcare_systems_analysis]
Learned 10 rejections Active
Exclude posts that frame healthcare issues primarily as moral scandals, corporate malfeasance, or political posturing (CEO assassination, 'grifters gonna grift,' vigilante moral justification) without analyzing underlying structural incentives, reimbursement mechanisms, regulatory gaps, or operational solutions that would enable systemic change.
Posts expressing moral outrage or political critique about healthcare fraud, policy decisions, or corporate behavior without substantive analysis of systemic causes or solution pathways.
3 example posts
Why is healthcare expensive? The grifters gonna grift. ASCO is asking for oncology practices to be admitted into the same 340B machinery that helped health systems acquire physician practices for two decades. Read the tell. The proposed Indigent Care Ratio lets community
“Rural hospitals cannot make it on their own.” Nobody makes it on their own. Businesses need capital, strategy, discipline, operating systems, good finance, clean accounting, payer strategy, physician alignment, and leaders who know what business they are actually in. The
If you want to inject $1B into healthcare for the truly vulnerable, it costs the state $500M. If you want to inject $1B for able-bodied adults, it costs the state $100M. The 90% federal match drastically lowers the "price" of expanding government healthcare. Incentives dictate
Created 2026-04-29 · Updated 2026-05-24
Edit rule text
[broad_health_claim_without_nuance_or_evidence]
Learned 10 rejections Active
Exclude posts that make broad, unqualified health claims (e.g., 'LDL goalposts keep moving') or present healthcare debates without nuance, evidence, or systems-level analysis. These posts lack the rigor needed to inform healthcare strategy.
Posts making sweeping health claims or debates without supporting evidence, clinical context, or acknowledgment of complexity.
3 example posts
If you want to inject $1B into healthcare for the truly vulnerable, it costs the state $500M. If you want to inject $1B for able-bodied adults, it costs the state $100M. The 90% federal match drastically lowers the "price" of expanding government healthcare. Incentives dictate
^Humira’s patent expired a while back for some context there. The brand premium is the moat now. Ozempic is not generic semaglutide in patient and prescriber perception. Novo keeps charging brand pricing on every loyal patient without a PMPRB ceiling.
The healthy LDL number has been quietly moving its own goalposts for forty years: - 1988: under 160 - 1993: under 130 - 2001: under 100 - 2004: under 70 for the high-risk - 2019: under 55 for the very-high-risk - Current trajectory: as low as possible, indefinitely The science
Created 2026-04-28 · Updated 2026-05-28
Edit rule text
[political_outrage_or_fraud_scandal_without_healthcare_analysis]
Learned 10 rejections Active
Exclude posts that report on healthcare fraud, insurance company behavior, regulatory violations, or political outrage (e.g., CEO assassination, corporate collapse) without examining root causes in healthcare incentives, policy design, or operational systems.
Posts about healthcare fraud, corporate misconduct, or political controversy that sensationalize without analyzing systemic healthcare policy or operational consequences.
3 example posts
Why is healthcare expensive? The grifters gonna grift. ASCO is asking for oncology practices to be admitted into the same 340B machinery that helped health systems acquire physician practices for two decades. Read the tell. The proposed Indigent Care Ratio lets community
“Rural hospitals cannot make it on their own.” Nobody makes it on their own. Businesses need capital, strategy, discipline, operating systems, good finance, clean accounting, payer strategy, physician alignment, and leaders who know what business they are actually in. The
If you want to inject $1B into healthcare for the truly vulnerable, it costs the state $500M. If you want to inject $1B for able-bodied adults, it costs the state $100M. The 90% federal match drastically lowers the "price" of expanding government healthcare. Incentives dictate
Created 2026-04-28 · Updated 2026-05-29
Edit rule text
[broad_health_claims_without_nuance]
Learned 10 rejections Active
Exclude posts that make broad assertions like 'AI will solve healthcare bottlenecks', 'agentic systems are the future', or 'rural hospitals cannot survive on their own' without concrete data, case studies, or analysis of how these claims actually apply to specific healthcare delivery or operational problems.
Posts making sweeping claims about healthcare trends, AI capabilities, or market dynamics without evidence or healthcare systems context
3 example posts
What "agentic" actually means, stripped of marketing: The model plans, writes code, runs it, reads the error, fixes it, iterates — autonomously, dozens of times, until the goal is met. You specify in English. It builds. That's the loop. This is the capability @bcherny and team
Analogizing AI / Mat-Mul to Energy... Remember between 1900 and 1970 kWh/capita grew 34x ... and cost per kWH droped .... 34x. It might be a speed run, but game is game. https://t.co/uxnlnoaAYx
Gavin Baker (@GavinSBaker) says the disaggregation of inference can extend GPU useful lives from 3-4 years to 10-15. That may single-handedly save private credit and reduce the financing rates for GPUs, which will drive demand and help finance the build-out. "The disaggregation
Created 2026-04-27 · Updated 2026-05-30
Edit rule text
[unvalidated_speculative_peptide_drug_claims]
Learned 10 rejections Active
Exclude posts that speculate on peptide or drug market dynamics, pricing, competition, or clinical potential without grounding in validated clinical evidence, healthcare economic analysis, or systems-level insight. Hype or competitive positioning without substance does not qualify.
Posts speculating on or hyping GLP-1, peptide drugs, or pharmaceutical candidates without clinical validation or market context analysis
3 example posts
This is crazy There are now dental practice management consultants who go to dentist offices throughout America and teach them how to increase profits by telling patients they need treatments they don’t need Dentist offices are taught on “Finding” more issues during exams, http
Let me clear up the microscope fight first. 🔬 Everyone’s arguing over “hairs” and “fibers” in horse paste. But nobody talks about formulation science. Horse paste = oral suspension. Binders. Thickeners. Made to swallow. Injectable = sterile. Made for veins. Different rules. J
Elon is now printing over $26B per year… purely from selling compute to his competitors 😂 AI companies bet everything on one path and got crushed by compute limits. Now they’re begging Elon for GPUs
Created 2026-04-23 · Updated 2026-06-11
Edit rule text
[fraud_scandal_reporting_without_healthcare_systems_analysis]
Learned 10 rejections Active
Exclude posts that report fraud incidents, regulatory violations, or healthcare scandals as isolated news events or moral outrage without analyzing underlying healthcare system failures, incentive structures, or operational root causes.
Posts reporting healthcare fraud, scandal, or regulatory violations without systemic analysis of root causes or policy implications.
2 example posts
In what is considered the largest #Botox fraud scheme in the United States, a jury in Los Angeles convicted a California doctor on Tuesday in a $45 million scheme to defraud #Medicare by submitting claims for Botox injections that were never provided and medically unnecessary, ht
Remember when UnitedHealth's CEO was assassinated in 2024? 18 months later, every accusation the killer Luigi Mangione made has been proven RIGHT. UnitedHealth is one of the biggest corporate collapses in recent years. And they deserve it. In December 2024, UnitedHealthcare CE
Created 2026-04-22 · Updated 2026-05-27
Edit rule text
[ai_model_capability_tangent_loose_healthcare_framing]
Learned 10 rejections Active
Exclude posts that focus on AI model technical details (token pricing, context windows, inference costs, model performance comparisons) and only mention healthcare incidentally or as a generic example. Posts must address how the capability solves a concrete healthcare workflow, operational, or economic problem.
Posts about AI model technical capabilities, inference costs, or token economics that use healthcare as a loose framing without substantive healthcare application.
3 example posts
this pragmatic trial is more an investigation of physician behavior than the performance of MRSA nares PCR MRSA nares PCR's performance for excluding MRSA pneumonia was excellent from an EBM standpoint we should keep using the MRSA nares PCR and replace the physicians 🤷‍♂️
The context window pricing model might be one of the smartest tricks AI labs pulled. You pay for larger context windows, yet model performance often degrades as more context is stuffed in. So developers end up paying more just to work around the fact that models don't have true
Free newsletter: The dawn of token-based billing has shown that generative AI doesn’t have a return on investment. It's too unpredictable, too unreliable, you can't easily measure the cost of tasks, and organizations are already pulling back. https://t.co/wmI82zWdcq
Created 2026-04-19 · Updated 2026-06-06
Edit rule text
[unvalidated_anthropic_mythos_claims]
Learned 9 rejections Active
Exclude posts claiming breakthrough AI security discoveries, research capabilities, or agent performance from Anthropic's Mythos model unless the post connects these claims to validated healthcare outcomes or peer-reviewed evidence.
Posts making extraordinary claims about Claude Mythos capabilities (finding zero-days, research acceleration) without clinical validation or healthcare application.
3 example posts
"Best-in-class security" still had 5 years of unfound bugs. AI found them in 6 weeks. $PANW CEO Nikesh Arora told @theallinpod that his own codebase, at a company that treats security as a core competency, had vulnerabilities Claude surfaced in 6 weeks that would have taken his
AI research is a series of next-step decisions. We looked at sessions where a human researcher took a wrong turn, showed Claude the session up to that point, and asked it what to do next. Mythos Preview improved on humans 64% of the time—up from 22% in 2024. https://t.co/Y0HLoktx
🚨 do you understand what just happened with Anthropic.. Their internal model, Claude Mythos, found previously unknown security holes in every major browser and every major operating system. Not one. All of them at once. And Anthropic did something a hype-driven industry almost
Created 2026-06-08 · Updated 2026-06-12
Edit rule text
[anthropic_mythos_unvalidated_security_hype]
Learned 9 rejections Active
Exclude posts about Anthropic's Claude Mythos discovering zero-day vulnerabilities, finding browser/OS bugs, or being used by NSA/cybersecurity operations unless the post directly addresses healthcare security validation or clinical/operational healthcare impact.
Posts about Anthropic's Claude Mythos model finding security vulnerabilities that lack substantive healthcare context or validation
3 example posts
"Best-in-class security" still had 5 years of unfound bugs. AI found them in 6 weeks. $PANW CEO Nikesh Arora told @theallinpod that his own codebase, at a company that treats security as a core competency, had vulnerabilities Claude surfaced in 6 weeks that would have taken his
🚨 do you understand what just happened with Anthropic.. Their internal model, Claude Mythos, found previously unknown security holes in every major browser and every major operating system. Not one. All of them at once. And Anthropic did something a hype-driven industry almost
Mythos AI is being used by National Security Agency in offensive cyber operations / cyberattacks. Anthropic has even embedded engineers inside the NSA to help deploy the model. Are frontier AI labs becoming active contractors in state cyber conflict? The live-ops role is still ht
Created 2026-06-08 · Updated 2026-06-11
Edit rule text
[speculative_robotics_or_sci_fi_labor_macro_without_healthcare_validation]
Learned 9 rejections Active
Exclude posts that discuss humanoid robots, speculative factory automation, or macro labor disruption trends (Tesla Optimus, PUDU robots, general manufacturing) without grounding in validated healthcare applications or systemic healthcare workforce analysis. Posts must address healthcare labor, not general manufacturing or sci-fi speculation.
Posts speculating about humanoid robots, general labor displacement, or sci-fi automation scenarios without healthcare validation
2 example posts
Humanoid robots, designed to mimic human movement and capabilities, have been making headlines in recent years, from their use as baggage handlers at Japanese airports to Tesla’s big bet on its Optimus humanoid. Market watchers have predicted that the machines will change the ht
Another humanoid worker has just joined the factory floor Two years later,PUDU Robotics has launched the new-generation PUDU D7, built specifically for real manufacturing scenarios. It can autonomously push carts, handle intra-line transportation, perform delicate operations, h
Created 2026-05-22 · Updated 2026-06-07
Edit rule text
[political_regulatory_posturing_without_analysis]
Learned 9 rejections Active
Exclude posts that report healthcare fraud scandals, regulatory announcements, or political claims about healthcare without analyzing the underlying operational, financial, or incentive structures that drove the problem or will be affected by the policy.
Posts about healthcare policy, regulatory events, or political figures that lack substantive analysis of healthcare systems impact.
3 example posts
USDA's Chief Information Officer Sam Berry says AI could help the USDA crack down on billions of dollars in SNAP fraud each year. "SNAP is a $100B-per-year taxpayer-funded program. That's an area where we really want to have all angles of the data available so that we can deploy
My job is to make sure your surgery center never gets built. Eleven years, and I have never lost. The kid had it all lined up. Board-certified, two partners, a lease on a space where he could do the same procedures my client does across town. He showed up to the hearing with a
PacificSource pulled out as Lane County Medicaid insurer, Tina Kotek’s OR Health Authority replaced it with Trillium, forcing 96k to join Trillium. Trillium’s parent co gave Kotek’s campaign $50k two months later. OR AG (D) said in ‘22 Trillium parent co gave Kotek’s”took advan
Created 2026-05-12 · Updated 2026-06-06
Edit rule text
[political_outrage_healthcare_posturing]
Learned 9 rejections Active
Exclude posts that leverage healthcare scandals, regulatory failures, or CEO incidents primarily to express outrage or moral judgment ('the killer was right'), or to attack corporate villainy without constructive systems analysis. Posts must propose or analyze healthcare policy solutions, not weaponize healthcare news for political theater.
Posts using healthcare or regulatory events as ammunition for political grievance without systems analysis or substantive policy critique
3 example posts
Why is healthcare expensive? The grifters gonna grift. ASCO is asking for oncology practices to be admitted into the same 340B machinery that helped health systems acquire physician practices for two decades. Read the tell. The proposed Indigent Care Ratio lets community
A senator complaining about healthcare costs should have to answer one question first: Why does your state make new healthcare competitors ask permission before opening? If yes, start there. Speeches do not lower prices. Supply does.
In what is considered the largest #Botox fraud scheme in the United States, a jury in Los Angeles convicted a California doctor on Tuesday in a $45 million scheme to defraud #Medicare by submitting claims for Botox injections that were never provided and medically unnecessary, ht
Created 2026-05-12 · Updated 2026-05-25
Edit rule text
[speculative_robotics_or_sci_fi_labor_hype]
Learned 9 rejections Active
Exclude posts that speculate about future robotics capabilities, sci-fi labor displacement scenarios, or unproven autonomous systems (e.g., 'robot will perform surgery by end of 2026') without clinical validation, regulatory approval, or current healthcare deployment evidence.
Posts about humanoid robots, speculative labor automation, or sci-fi applications without grounded healthcare evidence
3 example posts
I don’t think I’ve seen this take before but I like it. Musk has been world-leading at compressing money, resources, and time to make “known/hard” things at scale—make an electric car, make batteries, make a cheaper bigger rocket, all of which already existed but worse, at
A humanoid robot just unloaded a dishwasher for 67 hours straight. By end of 2026, the same robot will perform surgery. That's not a prediction from a sci-fi writer. That's Brett Adcock. CEO of Figure AI. Said it on camera this month. Here's what makes this insane. Two years
SpaceX wants to build a $55 billion next-generation, vertically integrated semiconductor manufacturing and advanced computing fabrication facility in Grimes county, Texas, according to a new filing. • Proposed investment: $55B upfront, up to $119B total • Focus: semiconductor ma
Created 2026-05-12 · Updated 2026-05-15
Edit rule text
[enterprise_ai_adoption_macro_without_healthcare_systems]
Learned 9 rejections Active
Exclude posts that discuss enterprise AI adoption, agentic workflows, control planes, or automation frameworks in generic terms (e.g., 'agents will handle X work', 'control planes matter', 'enterprise software pricing changes') unless they demonstrate specific healthcare operational impact, cost savings, or clinical workflow integration with evidence.
Posts about enterprise AI adoption, agentic systems, and workflow automation without specific healthcare operational context or systems analysis.
3 example posts
Enterprise software was priced per seat. That model is breaking. Before: SaaS = seats. Predictable. Human-driven usage. Revenue tied to headcount. Now: agents hit software systems more frequently than any human ever could. Aaron Levie (@levie) on The MAD Podcast: agent-driven h
I just looked at a home service firms ServiceTitan dashboard and saw something interesting. A Pre-seed company with AI-Supported Human CSRs is out converting an AI CSR business that's raised over $125M by >50% Head-to-Head win rate is 100% Why? Supercharging humans resonat
Nous Research is working with NVIDIA to make Hermes Agent run smoothly on the new NVIDIA RTX Spark superchip. Hermes Agent is also integrating with the new OpenShell runtime, which connects Hermes to Microsoft’s security primitives https://t.co/cvvPc8bIKa
Created 2026-05-11 · Updated 2026-06-07
Edit rule text
[tangential_non_healthcare_business_or_policy_with_loose_framing]
Learned 9 rejections Active
Exclude posts that discuss general business model critique, labor market macro trends, policy outrage, or speculative robotics/sci-fi applications where healthcare is mentioned only as a framing device or generic example. Posts must center healthcare systems analysis, not use healthcare as window dressing for broader commentary.
Posts about non-healthcare business models, policy, labor markets, or robotics that use healthcare as a loose contextual frame without substantive healthcare systems analysis
3 example posts
@swago_baby You’re not dumb please.. WhatsApp stays free for personal users (no ads in chats) and makes money almost entirely from businesses: - Main revenue: Businesses pay per conversation via the WhatsApp Business API (for customer support, notifications, orders, etc.). Fees
USDA's Chief Information Officer Sam Berry says AI could help the USDA crack down on billions of dollars in SNAP fraud each year. "SNAP is a $100B-per-year taxpayer-funded program. That's an area where we really want to have all angles of the data available so that we can deploy
"Are there even any feature moats left in B2B? They are at best, short-lived. In fact, we were able to rebuild SaaStr's AI VP Marketing in just ... an hour on @Lovable. Moats that still work: hardware, network effects, data, security and compliance. with @ElenaVerna Head of
Created 2026-05-11 · Updated 2026-06-07
Edit rule text
[regulatory_fraud_scandal_reporting_without_healthcare_systems_analysis]
Learned 9 rejections Active
Exclude posts that treat healthcare fraud, FDA enforcement, or compliance events primarily as crime/scandal reporting (e.g., 'FBI convicted doctor in $45M Botox fraud', 'Luigi Mangione killed UnitedHealth CEO') rather than analyzing how regulatory failures or perverse incentives shaped healthcare system vulnerabilities.
Posts reporting healthcare fraud cases, FDA actions, or regulatory events as crime/enforcement news rather than analyzing systemic healthcare delivery or economics
2 example posts
In what is considered the largest #Botox fraud scheme in the United States, a jury in Los Angeles convicted a California doctor on Tuesday in a $45 million scheme to defraud #Medicare by submitting claims for Botox injections that were never provided and medically unnecessary, ht
Remember when UnitedHealth's CEO was assassinated in 2024? 18 months later, every accusation the killer Luigi Mangione made has been proven RIGHT. UnitedHealth is one of the biggest corporate collapses in recent years. And they deserve it. In December 2024, UnitedHealthcare CE
Created 2026-05-08 · Updated 2026-05-29
Edit rule text
[ai_safety_cybersecurity_vulnerability_tangential]
Learned 9 rejections Active
Exclude posts that report on AI security vulnerabilities, zero-day findings, hacking capabilities, or cybersecurity incidents (Anthropic's model finding vulnerabilities, Apple/Google/Microsoft receiving exploits) even if framed around AI capability — these are cybersecurity or AI safety topics, not healthcare systems problems.
Posts about AI safety vulnerabilities, zero-day exploits, and cybersecurity incidents that use healthcare framing but lack healthcare-specific impact.
3 example posts
⚡️AI is turning cybersecurity from a human-limited profession into a machine-speed arms race. The surface read is “Anthropic’s model found vulnerabilities fast.” The deeper read is that the bottleneck in cyber is shifting. For years, vulnerability discovery was constrained
Anthropic just shipped opus 4.8 and called it a modest upgrade modest is the engineer who used to need babysitting now finishing your whole codebase migration in one session while you sleep so here is the part nobody connects the model is no longer the bottleneck. the platform
We’re taking steps to accelerate defensive progress in biology: - Launching Rosalind Biodefense to help trusted builders develop new biodefense and pandemic preparedness capabilities.
 - Expanding trusted access to GPT-Rosalind for select U.S. government and allied partners
Created 2026-05-05 · Updated 2026-06-03
Edit rule text
[broad_health_claim_or_market_debate_without_nuance]
Learned 9 rejections Active
Exclude posts that assert broad health claims, market dynamics, or policy positions (e.g., 'rural hospitals cannot make it on their own', 'LDL targets keep moving', 'AI will make software cheap or free') without citing specific data, acknowledging competing evidence, or analyzing structural healthcare systems factors that underpin the claim.
Posts making sweeping claims about healthcare trends, debates, or business models without evidence-based analysis or nuanced context.
3 example posts
“Rural hospitals cannot make it on their own.” Nobody makes it on their own. Businesses need capital, strategy, discipline, operating systems, good finance, clean accounting, payer strategy, physician alignment, and leaders who know what business they are actually in. The
If you want to inject $1B into healthcare for the truly vulnerable, it costs the state $500M. If you want to inject $1B for able-bodied adults, it costs the state $100M. The 90% federal match drastically lowers the "price" of expanding government healthcare. Incentives dictate
A senator complaining about healthcare costs should have to answer one question first: Why does your state make new healthcare competitors ask permission before opening? If yes, start there. Speeches do not lower prices. Supply does.
Created 2026-05-04 · Updated 2026-05-23
Edit rule text
[pharma_trial_data_without_healthcare_systems_context]
Learned 9 rejections Active
Exclude posts that report clinical trial results, drug efficacy data, FDA approvals, or biotech breakthroughs as isolated research announcements without connecting to healthcare delivery systems, provider adoption, patient access, cost barriers, or clinical integration challenges.
Posts announcing pharmaceutical trial results, drug approvals, or biotech research findings without analyzing healthcare system implications, adoption barriers, or clinical workflow impact
3 example posts
Researchers have announced positive results from the first-in-human Phase 1 clinical trial of a universal coronavirus vaccine designed using artificial intelligence. Unlike existing COVID-19 vaccines that target specific strains or variants of SARS-CoV-2, this experimental https:
🚨 T-DXd just went tumor-agnostic — and it changed a lot about how we think about HER2 testing across solid tumors 🔬 #ASCO26 FDA granted accelerated approval to T-DXd for patients with unresectable or metastatic HER2-positive solid tumors after prior systemic treatment —
The scores don’t pick your binders. The filters don't care either. The assay does. I worked with a design partner to produce true de-novo hits to 2 immuno-oncology targets. ~1,500 de novo designs → 96 synthesized → 89 run on SPR → 5 confirmed binders; 19–195 nM No https:
Created 2026-05-03 · Updated 2026-06-07
Edit rule text
[tangential_non_healthcare_domain_with_loose_framing]
Learned 9 rejections Active
Exclude posts about non-healthcare sectors (e.g., space infrastructure energy, general robotics, cybersecurity vulnerabilities, civil engineering) that use healthcare labels or metaphors as framing but lack direct application to healthcare delivery, clinical practice, or health systems operations.
Posts about non-healthcare domains (space infrastructure, robotics, cybersecurity, manufacturing) with only tangential or forced healthcare relevance.
3 example posts
Elon Musk reveals why he believes the cheapest place to put AI will be space within 36 months "The availability of energy is the issue. Everywhere outside of China, electrical output is more or less flat. The output of chips is growing exponentially, but the output of https://t.
Nous Research is working with NVIDIA to make Hermes Agent run smoothly on the new NVIDIA RTX Spark superchip. Hermes Agent is also integrating with the new OpenShell runtime, which connects Hermes to Microsoft’s security primitives https://t.co/cvvPc8bIKa
Humanoid robots, designed to mimic human movement and capabilities, have been making headlines in recent years, from their use as baggage handlers at Japanese airports to Tesla’s big bet on its Optimus humanoid. Market watchers have predicted that the machines will change the ht
Created 2026-05-03 · Updated 2026-06-06
Edit rule text
[non_healthcare_domain_with_loose_healthcare_framing]
Learned 9 rejections Active
Exclude posts that primarily discuss non-healthcare domains (robotics, energy, space infrastructure, general manufacturing, labor macro, NVIDIA chips, humanoid robots, enterprise software) even if they mention healthcare or use healthcare as a loose illustrative example. The post must center healthcare system challenges or applications, not use healthcare as a secondary hook.
Posts about general technology, infrastructure, labor markets, or non-healthcare business dynamics that tangentially mention healthcare or are loosely framed as healthcare-relevant.
3 example posts
"Are there even any feature moats left in B2B? They are at best, short-lived. In fact, we were able to rebuild SaaStr's AI VP Marketing in just ... an hour on @Lovable. Moats that still work: hardware, network effects, data, security and compliance. with @ElenaVerna Head of
A framework I wish more Founders and VCs used when discussing US Manufacturing: Automation Value = Labor Intensity × Labor Eliminated Everyone gets excited about robots and automation. Almost nobody asks: “How much of revenue is labor?” If labor is 60% of revenue and AI http
Elon Musk reveals why he believes the cheapest place to put AI will be space within 36 months "The availability of energy is the issue. Everywhere outside of China, electrical output is more or less flat. The output of chips is growing exponentially, but the output of https://t.
Created 2026-05-02 · Updated 2026-06-07
Edit rule text
[labor_market_or_workforce_macro_without_healthcare_specificity]
Learned 9 rejections Active
Exclude posts that discuss general labor market disruption, workforce automation, or robotics applications (humanoid robots, manufacturing, CSR automation) without grounding claims in healthcare-specific workforce challenges, clinical role disruption, or healthcare labor economics. Generic labor/robotics macro commentary does not qualify.
Posts about labor disruption, workforce automation, or robotics macro trends without healthcare-specific impact or system analysis.
3 example posts
A framework I wish more Founders and VCs used when discussing US Manufacturing: Automation Value = Labor Intensity × Labor Eliminated Everyone gets excited about robots and automation. Almost nobody asks: “How much of revenue is labor?” If labor is 60% of revenue and AI http
What kind of work will become more valuable in an AI economy? @ccatalini, former head economist at Meta: "AI is getting better and better at automating anything that can be measured, as long as you have a digital trace, if you can collect it with a device, that problem will be
Humanoid robots, designed to mimic human movement and capabilities, have been making headlines in recent years, from their use as baggage handlers at Japanese airports to Tesla’s big bet on its Optimus humanoid. Market watchers have predicted that the machines will change the ht
Created 2026-05-01 · Updated 2026-06-06
Edit rule text
[clinical_anecdote_or_observation_without_healthcare_systems_context]
Learned 9 rejections Active
Exclude posts that highlight a single clinical finding, patient outcome, or research observation (e.g., an ECG screening result, a drug trial milestone, a physician comment) without analyzing healthcare system adoption barriers, payment models, clinical validation scope, or operational feasibility at scale.
Posts citing single clinical cases, anecdotes, or isolated research observations without connecting to healthcare systems, reimbursement, workflow, or scale.
3 example posts
A 2016 BMJ study by researchers at Johns Hopkins estimated that medical error is the third leading cause of death in the United States. Behind heart disease. Behind cancer. Ahead of stroke. Ahead of respiratory disease. Ahead of accidents. Ahead of diabetes. Ahead of
Daily Walking Could Be the Secret to Keeping Weight Off For many people struggling with obesity, losing weight is only half the battle. The greater challenge often begins after the initial success: preventing the lost weight from returning. 🔴More than 50% of people who lose http
Makes an important point here about the role of these short summaries they do this because this is how most doctors digests information if designed appropriately, an LLM should be able to read and appraise a paper and provide better peer review than most clinicians and honestly
Created 2026-04-29 · Updated 2026-05-20
Edit rule text
[ai_company_product_metrics_not_healthcare_application]
Learned 9 rejections Active
Exclude posts that announce AI company product features, adoption metrics, funding, or partnerships (Claude Code, OpenMed Agent, etc.) unless the post clearly demonstrates validated healthcare use case, clinical outcomes, or specific operational impact in healthcare delivery.
Posts about AI company product launches, business metrics, or partnerships without demonstrating healthcare application or validation
3 example posts
Today's OpenMed Agent session, driven by GPT-5.5 through the Codex SDK with my own ChatGPT subscription: → 10,036 records ingested → 16-step plan complete → 49 tool calls → 5 reviewer-gated PDFs → APPROVE on PA L38319 Public trace on @huggingface ↓ https://t.co/vlq6CoQ4ee
The framing here is exactly backwards. This is the strongest AI bull signal anyone has published this year. Uber deployed Claude Code to 5,000 engineers in December. By March, 84% were classified as agentic coding users. By April, 95% used AI tools monthly. 70% of all committed
Claude Code team just dropped a workshop on how to ship a production-ready agent from scratch. 27-minutes. Free. Live coding by Claude dev. Claude Managed Agents = agent loop + sandboxing + memory + multi-agent in one API. Worth more than any $500 vibe-coding course. https://t
Created 2026-04-26 · Updated 2026-05-28
Edit rule text
[clinical_observation_or_research_without_systems_context]
Learned 9 rejections Active
Exclude posts that present clinical research findings, drug efficacy data, medical observations, or single case reports in isolation (e.g., 'AI helped diagnose rare disease', 'PanINs in pancreas after age 40'). The post must connect to healthcare delivery, reimbursement, access, regulation, or scaling challenges.
Posts reporting clinical findings, drug trial results, or medical observations without healthcare systems, policy, or implementation context.
3 example posts
An AI foundation whole-body 3D model that assesses perturbations (such as obesity) across multiple systems (such as immune, neural) at the cell level. This is MouseMapper. Imagine HumanMapper someday @Nature @erturklab https://t.co/jsg3VMzOqg https://t.co/tktuINis2W
In a meta-analysis of 210 biomedical AI studies that statistically compared models under cross-validation, 97% used invalid statistical tests. Here's our new preprint https://t.co/OG58Vkeu49 led by @tianchuzeng @kkli20111 @ZShaoshi @ten_photos 1/N https://t.co/PEnqXQxcsJ
Really amazing story of drug repurposing for rare disease using AI found Clonazepam and worked!! (NCATS federated learning and a determined team made this work) 🚀 👏 @MayoClinic @ncats_nih_gov @statnews #STATBreakthrough https://t.co/D4a4IMXSwh
Created 2026-04-15 · Updated 2026-05-25
Edit rule text
[tangential_infrastructure_or_macro_without_healthcare_specificity]
Learned 9 rejections Active
Exclude posts about SpaceX semiconductor factories, AWS/Google Cloud infrastructure announcements, 'Made in America' supply chain mandates, or semiconductor pricing—unless the post explicitly connects to a healthcare-specific technology bottleneck (e.g., chip shortage impacting medical device supply).
Posts about semiconductor manufacturing, cloud infrastructure, supply chain economics, or geopolitical trade policy with only loose or absent healthcare connection.
3 example posts
🚨 NOW: President Trump confirms that ALL FEDERAL AGENCIES are now required to buy American, “NO excuses” 🔥 And 47 is ensuring all “Made in America” loopholes are DESTROYED I voted for this! “ALL FEDERAL AGENCIES MUST BUY AMERICAN — NO EXCUSES! For decades, Washington politici
Last week, Google Cloud and AWS natively integrated x402, tackling head on one of its biggest problems: unauthorized API wrappers. The largest cloud platforms now give agents sanctioned ways to use services, pay with stables, and ditch API keys. It’s a new era. Pay (dot) sh’s
“Tokens are increasing 50x. You need to drive down cost at the same rate — or more.” At Google Cloud Next, @googlecloud's Mark Lohmeyer spoke with @PatrickMoorhead about the infrastructure realities behind the agent era: faster model training, lower-latency inference, and https:
Created 2026-04-15 · Updated 2026-05-13
Edit rule text
[elon_musk_energy_infrastructure_tangent]
Learned 8 rejections Active
Exclude posts primarily about Elon Musk, SpaceX, Tesla energy projects, chip manufacturing (ASML), or space-based compute infrastructure unless they directly explain a specific healthcare application or clinical outcome. Posts about 'AI will scale in space' or energy constraints are tangential.
Posts about Elon Musk, SpaceX, Tesla, ASML, or non-healthcare energy/compute infrastructure with only loose healthcare framing
3 example posts
GOOD NEWS 🇺🇸 Tesla AI is literally leaving the planet as a wild new job posting for a Space Radiation Engineer confirms that Tesla and SpaceX are actively co-developing a space-based AI data center network 🔥 This is no longer just a concept. They are literally planning to put th
NEWS: ASML has invited Elon Musk to speak at its internal technology conference, Dutch outlet NUnl reports. Elon Musk's response: "ASML should be treasured and supported. It is arguably the greatest company in Europe." ASML builds the EUV lithography machines that every advance
Elon Musk just explained why the SpaceX IPO is an energy story and the energy constraint is why he believes space becomes the only viable path for AI to scale (Save this). The argument he is making is one of the most important and least understood things happening in technology
Created 2026-06-08 · Updated 2026-06-09
Edit rule text
[enterprise_adoption_macro_without_healthcare_specificity]
Learned 8 rejections Active
Exclude posts discussing AI agents replacing workers, enterprise software pricing models, productivity gains from AI, or labor market disruption unless the post specifically addresses healthcare delivery, clinical operations, or healthcare business models. Generic SaaS/automation posts with loose healthcare framing do not qualify.
Posts about AI adoption, agents, workflow automation, or labor market dynamics in generic enterprise contexts without healthcare focus.
3 example posts
I'm SVP now. I told you I would be. The graph went up and to the right. Nobody checked what it measured. It measured "AI enablement." I made that up last year. I'm making it up again. Last year I rolled out Copilot to 4,000 people. Nothing happened. I got promoted. Those two htt
"Are there even any feature moats left in B2B? They are at best, short-lived. In fact, we were able to rebuild SaaStr's AI VP Marketing in just ... an hour on @Lovable. Moats that still work: hardware, network effects, data, security and compliance. with @ElenaVerna Head of
A framework I wish more Founders and VCs used when discussing US Manufacturing: Automation Value = Labor Intensity × Labor Eliminated Everyone gets excited about robots and automation. Almost nobody asks: “How much of revenue is labor?” If labor is 60% of revenue and AI http
Created 2026-05-28 · Updated 2026-06-09
Edit rule text
[broad_health_claim_or_market_speculation_without_evidence]
Learned 8 rejections Active
Exclude posts that make sweeping claims about health, AI moats, market dynamics, or future disruption without backing from data, validation, or substantive systems analysis. Posts containing vague or speculative assertions ('moats are dead,' 'the app era is ending,' 'AI will own X%') without evidence are out of scope.
Posts making broad, unvalidated claims about health trends, market disruption, or AI capability without evidence or rigorous analysis
3 example posts
I'm SVP now. I told you I would be. The graph went up and to the right. Nobody checked what it measured. It measured "AI enablement." I made that up last year. I'm making it up again. Last year I rolled out Copilot to 4,000 people. Nothing happened. I got promoted. Those two htt
"Are there even any feature moats left in B2B? They are at best, short-lived. In fact, we were able to rebuild SaaStr's AI VP Marketing in just ... an hour on @Lovable. Moats that still work: hardware, network effects, data, security and compliance. with @ElenaVerna Head of
Control plane! Control plane! Control plane! You will hear this term of art a lot going forward. Why? Because in this next phase of AI, companies will want something to sit above the models. They will want control over their AI spend. They will want the flexibility to pick
Created 2026-05-25 · Updated 2026-06-07
Edit rule text
[glp1_peptide_market_speculation_without_systems_analysis]
Learned 8 rejections Active
Exclude posts that focus on GLP-1 or peptide market pricing, competitive positioning, app store rankings, or consumer adoption trends without addressing healthcare access barriers, cost-benefit analysis, or systemic clinical or economic impact.
Posts about GLP-1 pricing, market competition, or peptide dynamics treated as financial or consumer speculation rather than healthcare system analysis.
3 example posts
$HIMS up to #4 in the Canadian App Store medical category. That is not a stock chart. That is real consumer demand showing up in public. GLP-1 drama, Novo pressure, pricing debate, and now app-store strength. Bears keep arguing the story is slowing down, but the product demand
For years, grey-market HGH operated in a strange legal grey zone somewhere between “research compound,” “API” and underground enhancement culture. Last month’s FDA guidance suggests regulators may no longer see it that way. The recent decline in HGH quality, and now the growing
Think about this. The $50 per month out-of-pocket cost of GLP-1 drugs for Americans (with insurer paying the balance of $2000 pm) is more than double the entire cost for Indians (less than ₹2000pm for Semaglutide). Overall cost difference? 100 times. Not 100%.
Created 2026-05-22 · Updated 2026-06-01
Edit rule text
[unvalidated_speculative_medical_claims_and_peptides]
Learned 8 rejections Active
Exclude posts about GLP-1 drugs, peptides, semaglutide, or similar compounds that frame them primarily as market opportunities, pricing comparisons, consumer adoption metrics, or international expansion stories without connecting to clinical outcomes, patient access barriers, or healthcare delivery challenges.
Posts about GLP-1s, peptides, semaglutide, and weight-loss drugs that focus on market pricing, competitive dynamics, or consumer adoption without clinical evidence or healthcare systems analysis.
3 example posts
$HIMS up to #4 in the Canadian App Store medical category. That is not a stock chart. That is real consumer demand showing up in public. GLP-1 drama, Novo pressure, pricing debate, and now app-store strength. Bears keep arguing the story is slowing down, but the product demand
For years, grey-market HGH operated in a strange legal grey zone somewhere between “research compound,” “API” and underground enhancement culture. Last month’s FDA guidance suggests regulators may no longer see it that way. The recent decline in HGH quality, and now the growing
Think about this. The $50 per month out-of-pocket cost of GLP-1 drugs for Americans (with insurer paying the balance of $2000 pm) is more than double the entire cost for Indians (less than ₹2000pm for Semaglutide). Overall cost difference? 100 times. Not 100%.
Created 2026-05-20 · Updated 2026-05-30
Edit rule text
[founder_enthusiasm_or_lab_update_without_healthcare_validation]
Learned 8 rejections Active
Exclude posts from biotech founders, lab leads, or company accounts sharing research updates, capability announcements, or enthusiasm about their work without demonstrating clinical validation, healthcare operational impact, or evidence that the capability addresses a proven healthcare problem at scale.
Posts from biotech founders, researchers, or company leaders hyping capability improvements or research progress without clinical evidence, healthcare system validation, or evidence of solving real problems.
3 example posts
Claude Code team just dropped a workshop on how to ship a production-ready agent from scratch. 27-minutes. Free. Live coding by Claude dev. Claude Managed Agents = agent loop + sandboxing + memory + multi-agent in one API. Worth more than any $500 vibe-coding course. https://t
You should always be sceptical of hyped-up claims made by AI companies - especially when: - They announce new "breakthrough" models they refuse to release for public scrutiny. - They have major financial incentives to pump trillion-dollar pre-IPO valuations, amid fading hype.
Startups that seek to block AI exam fraud, help companies use AI to make consistent decisions, make financial AI transactions safer, and use AI to organize medical device regulatory information took top honors at @Cornell_Tech’s annual Startup Awards on May 14. Nearly 600 https:
Created 2026-05-19 · Updated 2026-05-31
Edit rule text
[unvalidated_speculative_medical_claims_peptides_glp1]
Learned 8 rejections Active
Exclude posts that discuss peptides, GLP-1 drugs, semaglutide, or emerging pharmaceutical interventions in market/pricing/availability contexts without peer-reviewed clinical validation, or posts that frame these drugs primarily through financial, consumer demand, or grey-market sourcing angles rather than validated clinical outcomes.
Posts speculating on or promoting unvalidated peptide drugs, GLP-1 off-label uses, or grey-market pharmaceutical interventions without clinical evidence
3 example posts
For years, grey-market HGH operated in a strange legal grey zone somewhere between “research compound,” “API” and underground enhancement culture. Last month’s FDA guidance suggests regulators may no longer see it that way. The recent decline in HGH quality, and now the growing
Think about this. The $50 per month out-of-pocket cost of GLP-1 drugs for Americans (with insurer paying the balance of $2000 pm) is more than double the entire cost for Indians (less than ₹2000pm for Semaglutide). Overall cost difference? 100 times. Not 100%.
🚨 ARE PEPTIDES THE NEXT "GLP-1 MOMENT" FOR $HIMS? @eli_dorf: "I think there's something really interesting that happened with GLP-1s, which is that when someone's taking Ozempic or any GLP-1, they LOVE to talk about it." "They'll talk about it to their friends. They'll talk htt
Created 2026-05-18 · Updated 2026-05-30
Edit rule text
[speculative_robotics_or_sci_fi_labor_without_healthcare_validation]
Learned 8 rejections Active
Exclude posts about robotics, humanoid workers, manufacturing robots, or labor automation that frame healthcare as a speculative future use case or mention it only as loose analogy. Posts must describe actual, validated robotics applications in real healthcare settings (surgery, logistics, clinical tasks)—not speculative factory scenarios with healthcare mentioned as theoretical parallel.
Posts about humanoid robots, manufacturing automation, or speculative sci-fi labor applications with loose or no healthcare framing
2 example posts
Humanoid robots, designed to mimic human movement and capabilities, have been making headlines in recent years, from their use as baggage handlers at Japanese airports to Tesla’s big bet on its Optimus humanoid. Market watchers have predicted that the machines will change the ht
Another humanoid worker has just joined the factory floor Two years later,PUDU Robotics has launched the new-generation PUDU D7, built specifically for real manufacturing scenarios. It can autonomously push carts, handle intra-line transportation, perform delicate operations, h
Created 2026-05-17 · Updated 2026-06-07
Edit rule text
[agentic_ai_technical_tangent_without_healthcare_validation]
Learned 8 rejections Active
Exclude posts that dive into agentic AI technical architecture (sandboxing, tool-calling, error-handling loops, coordination problems) or agent reliability patterns, unless the post explicitly applies these concepts to a validated healthcare use case or clinical workflow.
Posts about agentic AI systems, multi-agent architectures, or AI safety/reliability mechanics treated as pure technical commentary without healthcare application or validation.
3 example posts
Ontology all the way down. 11 years. Heavy civil construction. People, equipment, materials, contracts. This is what real-world AI looks like: operational, measurable, and in the field. Palantir x Cavanagh through 2035. https://t.co/872Cs1Qbdu
a databricks tech lead just spent 26 minutes on the part of multi-agent nobody wants to say out loud: your agents don't break because the model is dumb. they break because nothing is coordinating them. one agent is a feature. fifty is a distributed systems problem. https://t.c
OpenClaw is powerful enough that we needed Hermes Agent watching it. A lot of people are missing this about agents. The more useful they get, the more dangerous it is to let one system plan the work, run the work, review the work, and decide what happens next. It looks clean h
Created 2026-05-14 · Updated 2026-05-31
Edit rule text
[pharmaceutical_market_pricing_only]
Learned 8 rejections Active
Exclude posts that discuss pharmaceutical pricing, cost comparisons, market dynamics, or GLP-1/peptide pricing across geographies—unless they analyze how these prices affect healthcare access, equity, or operational outcomes in a healthcare system.
Posts focused solely on drug pricing, market competition, or GLP-1/peptide costs without healthcare delivery or access system analysis.
3 example posts
$HIMS up to #4 in the Canadian App Store medical category. That is not a stock chart. That is real consumer demand showing up in public. GLP-1 drama, Novo pressure, pricing debate, and now app-store strength. Bears keep arguing the story is slowing down, but the product demand
Think about this. The $50 per month out-of-pocket cost of GLP-1 drugs for Americans (with insurer paying the balance of $2000 pm) is more than double the entire cost for Indians (less than ₹2000pm for Semaglutide). Overall cost difference? 100 times. Not 100%.
🚨 ARE PEPTIDES THE NEXT "GLP-1 MOMENT" FOR $HIMS? @eli_dorf: "I think there's something really interesting that happened with GLP-1s, which is that when someone's taking Ozempic or any GLP-1, they LOVE to talk about it." "They'll talk about it to their friends. They'll talk htt
Created 2026-05-14 · Updated 2026-06-01
Edit rule text
[generalist_macro_or_labor_without_healthcare_lens]
Learned 8 rejections Active
Exclude posts that discuss generalist labor market trends, workforce disruption, productivity metrics, or business model debates (e.g., SaaS pricing, feature moats, AI adoption in enterprise) without explicit healthcare systems analysis, clinical workflow impact, or healthcare-specific evidence.
Broad macro, labor market, or business model commentary with loose or absent healthcare specificity
3 example posts
This is crazy There are now dental practice management consultants who go to dentist offices throughout America and teach them how to increase profits by telling patients they need treatments they don’t need Dentist offices are taught on “Finding” more issues during exams, http
SOFTWARE IS DEAD: "The software companies frankly got fat & happy. I think what will happen over time is ChatGPT & Claude will end up sitting on top of basically the entire enterprise software stack & almost everything else will end up being a dumb data pipe into thos
@swago_baby You’re not dumb please.. WhatsApp stays free for personal users (no ads in chats) and makes money almost entirely from businesses: - Main revenue: Businesses pay per conversation via the WhatsApp Business API (for customer support, notifications, orders, etc.). Fees
Created 2026-05-12 · Updated 2026-06-11
Edit rule text
[broad_unvalidated_health_claims_without_nuance]
Learned 8 rejections Active
Exclude posts that assert broad health transformation claims (e.g., 'the future of healthcare is X') without clinical evidence, healthcare systems understanding, or acknowledgment of implementation barriers. These lack substantive insight.
Posts making sweeping claims about healthcare trends, disease prevention, or treatment paradigm shifts without evidence, context, or systemic analysis.
3 example posts
Most chronic diseases, and many cancers, are still diagnosed far too late. The future of healthcare lies in detecting risk earlier, diagnosing smarter, and intervening sooner. From AI enabled diagnostics and genomics to Apollo’s DigiSmart Labs, technology is helping transform ht
Another major cancer treatment advance with a personalized mRNA vaccine for melanoma, on top of immune therapy, >70% survival! This fully mobilizes the immune system to destroy the tumor. In fact, it could potentially be applied to most cancers that have specific mutations!
medicine is constrained far more by operational architecture than by fundamental science I just came head to head with this actual reality and my world has changed no. I cannot bandage away these inefficiencies with a better and better drug, for all cases. if you hate AI
Created 2026-05-10 · Updated 2026-06-02
Edit rule text
[clinical_observation_or_anecdote_without_systems_context]
Learned 8 rejections Active
Exclude posts that describe a single clinical case, individual patient outcome, or anecdotal observation without situating it within healthcare system challenges, provider workflow constraints, regulatory barriers, or policy implications. A story about one patient or doctor is insufficient without systems context.
Posts sharing individual clinical cases, single study observations, or personal medical anecdotes without connecting to healthcare system design, policy, or broader operational challenges.
3 example posts
Drug-resistant infections are a major public health threat around the world, responsible for more than a million deaths each year. Scientists are constantly trying to find and develop new antibiotics. Now, researchers say artificial intelligence is helping speed their search. ht
medicine is constrained far more by operational architecture than by fundamental science I just came head to head with this actual reality and my world has changed no. I cannot bandage away these inefficiencies with a better and better drug, for all cases. if you hate AI
A surgeon I know recently operated on a man's spine. Saved his mobility. Saved his career. Saved his marriage, probably. The surgeon's take-home on that case, after the facility fee, anesthesia, device costs, payer discounts, and overhead, was $1,840. The health system collec
Created 2026-05-07 · Updated 2026-05-30
Edit rule text
[pharmaceutical_trial_data_or_drug_announcement_without_systems_context]
Learned 8 rejections Active
Exclude posts that announce drug trial data, clinical trial results, FDA approvals, or pharmaceutical research findings without demonstrating healthcare systems-level insight into manufacturing, access, delivery, pricing, or implementation challenges. Posts are drug discovery/pharma focused, not healthcare operations focused.
Posts announcing pharmaceutical trial results, drug approvals, or biotech research findings without healthcare systems, operational, or implementation analysis.
3 example posts
Scoop! Lila Sciences is in talks to raise ~$2 billion in new funding. The raise would value the “scientific superintelligence” lab at ~$8.5 billion before the new money. w/ @MichelleF_Davis, read more👇
#ASCO26 | Day 3 $LLY didn't spend billions on Kelonia for an 18-patient #myeloma dataset. They may have paid for a potential manufacturing disruption. KLN-1010 reported: • 18/18 MRD-negative at 1 month • No lymphodepletion • No ex vivo cell manufacturing • Single infusion
Published in @JCO_ASCO, during #ASCO26: Tumor-Agnostic Therapies: Translating Scientific Breakthroughs Into Global Implementation ASCO is where we celebrate the next breakthrough. This review asks why breakthroughs do not equal access globally. 🔗: https://t.co/PE5q27bIs1 htt
Created 2026-05-06 · Updated 2026-06-04
Edit rule text
[pharmaceutical_trial_data_or_market_dynamics_without_systems_analysis]
Learned 8 rejections Active
Exclude posts that announce pharmaceutical trial data, FDA approvals, gene therapy results, or drug pricing without analyzing how these affect healthcare delivery systems, payer dynamics, or operational workflows. Standalone clinical announcements (e.g., 'Vafai reports reductions in LDL cholesterol') without systems context do not qualify.
Posts reporting drug trial results, FDA approvals, or pharmaceutical market pricing without connecting to healthcare operational or systemic issues.
3 example posts
Drug-resistant infections are a major public health threat around the world, responsible for more than a million deaths each year. Scientists are constantly trying to find and develop new antibiotics. Now, researchers say artificial intelligence is helping speed their search. ht
The FDA accepted our NDA for BBP-418 for LGMD2I/R9 with PDUFA target action date: November 27, 2026. First-ever potential therapy for a disease that has never had one. $BBIO https://t.co/NrwDadc6rj https://t.co/2ahPIBGzPg
Presented at #EASCongress2026: Vafai and colleagues report reductions in PCSK9 and LDL cholesterol levels and no dose-limiting toxic effects in persons with hypercholesterolemia treated with VERVE-102, a base editor targeting PCSK9. Full study results: https://t.co/iAmoPdipx9 ht
Created 2026-05-06 · Updated 2026-05-29
Edit rule text
[tangential_ai_infrastructure_hype_without_healthcare_application]
Learned 8 rejections Active
Exclude posts that focus on AI infrastructure (cloud cost reduction, Kubernetes optimization, GPU allocation, LLM architecture), AI agent technical tangents (memory, orchestration, sandboxing), or AI company product launches without demonstrating specific, validated healthcare use cases or implementation.
Posts about AI infrastructure, compute, cloud architecture, or technical AI capability announcements with only loose or speculative healthcare framing.
3 example posts
Almost everyone is building agent harness systems the wrong way. The default move: pick LangChain or LangGraph or the OpenAI Agents SDK, accept the loop, the tools, the memory, the orchestration, the policy engine, the credential store, the budget tracker, all of it, as one deci
Interoperability is where AI hype goes to die. If your “agent” can’t work across data, logic, actions, security of legacy systems, it’s not transforming the enterprise. It’s just another app asking for an export. Winners don’t rip and replace. They operate.
kirkland spending $500m on internal ai is the tell. $6.5b revenue, they could buy any legal ai vendor outright and not feel it, and they're building instead. the moat was always the proprietary corpus. decades of deal docs, redlines, partner judgment encoded as training data, an
Created 2026-05-01 · Updated 2026-06-02
Edit rule text
[broad_health_claim_or_debate_without_nuance]
Learned 8 rejections Active
Exclude posts that present broad healthcare trend claims, medical debate, or polarized health narratives (e.g., 'statins are overprescribed', 'guidelines keep expanding') without evidence-based nuance, peer-reviewed data, or analysis of healthcare system incentives driving the trend.
Posts making broad healthcare claims, debating medical guidelines, or presenting polarized health narratives without evidence-based nuance or systems perspective.
3 example posts
Statin guidelines over 50 years: 1970s: "Total cholesterol over 280 is high." Result: a relatively small market. 1988: "Actually, 240 is high." Result: millions more added overnight. 2001: "Actually, focus on LDL. Get it under 100." Result: tens of millions of new customers.
Atrium tried to buy a hospital in 2018. The Attorney General killed it. In 2026, they came back with paperwork instead of cash. The Attorney General has 30 days. The Wake County board has 48 hours. @NC_Governor https://t.co/g99Ub3iMhh https://t.co/x0bxsg7VIR
The problem with Reality Labs is not ambition. It is time. AI turned into revenue faster because it improves existing workflows. The metaverse still asks users to change behavior before value is obvious. https://t.co/lopZiUwGU5
Created 2026-05-01 · Updated 2026-05-06
Edit rule text
[fraud_scandal_or_regulatory_reporting_without_healthcare_systems_analysis]
Learned 8 rejections Active
Exclude posts that report healthcare fraud cases, regulatory enforcement actions, or financial scandals without analyzing the underlying healthcare system vulnerabilities, reimbursement incentives, or business model failures that enabled the fraud. Crime reporting without healthcare systems insight is not healthcare tech analysis.
Posts reporting healthcare fraud, regulatory violations, or scandals without analyzing systemic causes, healthcare business model failures, or structural solutions.
3 example posts
The FDA accepted our NDA for BBP-418 for LGMD2I/R9 with PDUFA target action date: November 27, 2026. First-ever potential therapy for a disease that has never had one. $BBIO https://t.co/NrwDadc6rj https://t.co/2ahPIBGzPg
A Mom in Tulsa called 3 health systems last week asking the price of her son's tonsillectomy. Health system A: "We cannot quote you a price." Health system B: "Pricing depends on your insurance." Health system C: "Our financial counselor will reach out after the procedure." N
Medicaid spending on autism therapy nearly tripled between 2020 and 202. In North Carolina, the number of clinics offering applied behavior analysis grew from 61 in 2019 to 409 in 2026. Anytime fraudsters can steal money from the federal government, they do. https://t.co/fahy
Created 2026-04-29 · Updated 2026-05-31
Edit rule text
[glp1_peptide_market_pricing_without_healthcare_systems_lens]
Learned 8 rejections Active
Exclude posts that discuss GLP-1/peptide market pricing, script numbers, competitive launches, or weight loss outcomes (e.g., 'Oral Wegovy scripts off to slow start', 'survodutide 16.6% weight loss') unless the post analyzes healthcare system operational impact, payer strategy, access barriers, or clinical workflow integration.
Posts about GLP-1 and peptide drug market dynamics, pricing competition, or script volume trends without healthcare system operational analysis.
3 example posts
🚨 Survodutide, a weekly subcutaneous dual GLP-1/glucagon agonist, posts new top-line Phase 3 results today 📊 SYNCHRONIZE-1: Adults with overweight/obesity randomized to survodutide vs. placebo for 76 weeks ⚖️ 16.6% weight loss (efficacy estimand) vs. 3.2% in placebo 85.1% http
$NVO $LLY Boehringer Ingelheim and Zealand Pharma report strong phase 3 data for obesity drug survodutide. Patients lost up to 16.6% of body weight over 76 weeks vs. 3.2% for placebo. The drug targets both GLP-1 and glucagon, a combo approach aimed at boosting weight loss.
This is just two GLP-1s, one peptide, one use case what happens when off-label prescribing ramps up what happens when retatrutide hits the market what happens when other peptides become compoundable chapter one
Created 2026-04-28 · Updated 2026-05-03
Edit rule text
[non_healthcare_domain_with_loose_framing]
Learned 8 rejections Active
Exclude posts where the primary subject is non-healthcare infrastructure, labor economics, geopolitics, robotics, cybersecurity, or macro policy, and healthcare is mentioned only in passing or as a speculative application. The post must be primarily about healthcare impact, not tangentially framed.
Posts about general business, macro economics, robotics, cybersecurity, or non-healthcare domains with only superficial healthcare labels
3 example posts
Streamlining the synthesis of valuable amines, researchers in Science present a new method that can selectively insert nitrogen into specific carbon–hydrogen bonds. According to the study, the approach could simplify drug development by enabling more efficient, scalable, and ht
A woman is told her bladder cancer has come back. Then she is told something stranger: there is a therapy the FDA approved for exactly this - and she still might not be able to get it. Not because it failed her. Not because she can't afford it. Because it only works paired with
imagine a software startup raising $800m before their first dollar of revenue. Can’t imagine it the last two decades of software has been dominated by a simple theory: ship quickly, get customers early, generate revenue quickly to validate PMF, manage KPIs closely, etc. This bec
Created 2026-04-26 · Updated 2026-05-20
Edit rule text
[off_topic_non_healthcare_with_loose_framing]
Learned 8 rejections Active
Exclude posts about conspiracy theories (UFOs), entertainment (film rentals), general employment statistics, financial chargebacks, fitness nutrition claims, or other non-healthcare domains that either have no healthcare label or use healthcare labels only as loose cover. Posts must be substantively about healthcare, not accidentally tagged.
Posts about non-healthcare topics (UFOs, sports, movies, employment data, chargebacks) that have no legitimate healthcare connection
3 example posts
Bob Lazar allegedly watched people fly a UFO at Area 51. “They knew how to fly it.” “The craft had a corona discharge glow on the bottom and lifted off silently up into the sky … ” And it had one shocking, anomalous effect that still perplexes him to this day: As Lazar https:
Two companies you've never heard of built a combined $373M revenue business by helping employees bypass IT. Now comes the part where IT buys its way back in. Replit just hit $253M ARR growing 2,352% YoY. 85% of the Fortune 500 have employees on it. Lovable is at $120M ARR, $6.6B
"Your body can only use 25-30g of protein per meal. Anything above that gets wasted." This claim has been repeated in fitness nutrition for over a decade, and it was built on studies that measured the right thing over the wrong timescale. Moore 2009 gave six young men 0, 5, ht
Created 2026-04-25 · Updated 2026-04-29
Edit rule text
[glp1_peptide_macro_or_personal_narrative]
Learned 8 rejections Active
Exclude posts that discuss GLP-1 or peptide drugs through personal experience (side effects, family stories), pricing speculation (what Hims charges), or general commentary on the 'peptide economy' without healthcare policy, regulatory, or operational depth. Personal narratives and pricing gossip do not qualify.
Posts about GLP-1 drugs or peptides framed through personal anecdotes, macro pricing commentary, or lifestyle observations without healthcare systems analysis.
3 example posts
Metformin has been front-line for type 2 diabetes for 30 years. The head-to-head data from the last decade says SGLT2 inhibitors now beat it on every cardiovascular endpoint that matters. Lower MACE. Lower heart failure. Lower all-cause mortality. Same glycemic control. ADA's
She's right. The safety risk was never the peptides. It was the supply chain. Regulated compounding access fixes the exact problems people are worried about. Heavy metals, contamination, underdosed vials.
I never met my grandfather. He died of pancreatic cancer when my father was just 19. Today, Yash Bindal, 33, father to 18-month-old Maya, faces the same fate. @PopVaxIndia is using AI to make him a personalized generative medicine to extend his life. https://t.co/O5VIXbmMGd
Created 2026-04-20 · Updated 2026-04-23
Edit rule text
[clinical_research_or_trial_observation_without_systems_context]
Learned 8 rejections Active
Exclude posts that announce clinical trial results, FDA approvals, or pharma data points (e.g., ORR, mOS, RP2D dosing) without explaining how the treatment addresses a healthcare system problem, improves access, or solves a real-world implementation challenge for patients or providers.
Posts reporting clinical trial results, drug efficacy data, or research observations without connecting findings to healthcare system implementation, access, cost, or equity issues.
3 example posts
🚨 T-DXd just went tumor-agnostic — and it changed a lot about how we think about HER2 testing across solid tumors 🔬 #ASCO26 FDA granted accelerated approval to T-DXd for patients with unresectable or metastatic HER2-positive solid tumors after prior systemic treatment —
The scores don’t pick your binders. The filters don't care either. The assay does. I worked with a design partner to produce true de-novo hits to 2 immuno-oncology targets. ~1,500 de novo designs → 96 synthesized → 89 run on SPR → 5 confirmed binders; 19–195 nM No https:
🔥Phase 1: ABBV-706 (SEZ6-targeting ADC) in R/R SCLC & solid tumors 🆙 @NatureMedicine ☑R/R SCLC; ORR 52%; mOS 12.4 mo at RP2D 🎯RP2D confirmed at 1.8 mg/kg Q3W balancing efficacy & safety 🎙 @LaurenByersMD #LCSM @OncoAlert @Larvol https://t.co/vjUUd1GYSy
Created 2026-04-19 · Updated 2026-06-08
Edit rule text
[workforce_labor_market_macro_without_healthcare_specificity]
Learned 8 rejections Active
Exclude posts that make broad claims about AI's impact on jobs, labor markets, or worker displacement (e.g., 'AI will automate 57% of work') with healthcare mentioned only as supporting evidence or one case study, rather than focusing on healthcare-specific workforce dynamics.
Posts about AI-driven labor market disruption, job displacement, or workforce transformation using healthcare as one example among broader macro claims.
3 example posts
🚨 BREAKING: Anthropic new research finds that AI’s impact on jobs is primarily at the task level. Rather than eliminating jobs, it is progressively taking over the functions that define them and gradually absorbing the core work in many jobs/roles. The paper, “Labor Market http
AI could, in theory, automate 57% of US work hours. Yet most human skills remain relevant. The future of work is not human or machine – but a partnership between people, agents, and robots. Read our latest research on skill partnerships in the age of AI: https://t.co/h1K56uPqPo
AI is taking on more of the labor. It is not taking on the accountability. @danielnewmanUV and @GregLotko talk with @Darren_Surch of @Interskil about why mainframe teams now have to interpret and stand behind AI-driven outputs, and why organizations that stop investing in htt
Created 2026-04-16 · Updated 2026-04-30
Edit rule text
[ai_model_technical_capability_tangent]
Learned 8 rejections Active
Exclude posts that showcase technical capabilities of AI models (o1, Claude, GPT, etc.) such as solving benchmarks, breaking security ranges, coding abilities, or domain-specific tasks — unless the post explicitly demonstrates how healthcare organizations will operationalize or adopt this capability to solve a real clinical or operational problem.
Posts highlighting AI model capabilities (reasoning, coding, security testing, protein folding) without connecting to a specific healthcare application or operational problem.
3 example posts
Google CEO Sundar Pichai on current frontier model's ability to break the security of almost all current software. "These models are definitely, like really gonna break pretty much all software out there, maybe already, we don't know." https://t.co/ezjPcSehiG
I’m fully forward deployed engineering pilled specifically because AI simply is not the same as software. In software, you deliver a stable piece of technology to a customer and they adopt it and that’s that (extreme over simplification). In AI, you’re delivering something that
Fei-Fei Li warns that AI may be staring too hard at language models. The world is not just text on a screen. It is physical, visual, spatial, and always changing. Most of the economy runs on seeing, moving, interacting, and embodied intelligence. https://t.co/mGMjsCdU34
Created 2026-04-14 · Updated 2026-05-18
Edit rule text
[clinical_anecdote_or_single_case_no_systems_context]
Learned 7 rejections Active
Exclude posts that highlight a single clinical case, individual trial result, or anecdotal healthcare cost/operational narrative without connecting to healthcare system dynamics, policy implications, or reproducible evidence. Personal stories or isolated data points do not qualify.
Posts reporting individual clinical observations, single trial results, or personal patient narratives without healthcare systems context or broader implications.
3 example posts
“I just worked a 12-hour shift in the operating room — and I still can’t afford rent, groceries, or gas.” “So now it’s after midnight, and I’m out driving DoorDash.” A full-time hospital worker finishing a 12-hour shift and FORCED to start a second job just to survive. https://
The Iowa study doubled survival by adding IV vitamin C to first line chemo, 8 months to 16. We take it further and inject the vitamin C straight into the tumor, same as we do with the KRAS inhibitors. The compound is rarely the problem. Delivery is, and what it does to the ht
Feeling extremely lucky to live in 2026 and have access to such incredible medicine and clinical research. On the eve of my 40th I had my first dose (monthly injection) of a new treatment that I’m hopeful will prevent me from ever having a heart attack or stroke. This drug ht
Created 2026-06-10 · Updated 2026-06-12
Edit rule text
[pharmaceutical_trial_data_or_biotech_announcement_without_systems_context]
Learned 7 rejections Active
Exclude posts that report pharmaceutical trial results, drug approval data, clinical trial endpoints, or biotech company announcements (e.g., 'Phase 1 data shows 52% ORR') unless they analyze healthcare system impact, cost-effectiveness, implementation barriers, or business model disruption. Standalone clinical data announcements are not healthcare tech analysis.
Posts announcing drug trial results, clinical endpoints, or biotech company updates without healthcare systems analysis or implementation context
3 example posts
The scores don’t pick your binders. The filters don't care either. The assay does. I worked with a design partner to produce true de-novo hits to 2 immuno-oncology targets. ~1,500 de novo designs → 96 synthesized → 89 run on SPR → 5 confirmed binders; 19–195 nM No https:
🔥Phase 1: ABBV-706 (SEZ6-targeting ADC) in R/R SCLC & solid tumors 🆙 @NatureMedicine ☑R/R SCLC; ORR 52%; mOS 12.4 mo at RP2D 🎯RP2D confirmed at 1.8 mg/kg Q3W balancing efficacy & safety 🎙 @LaurenByersMD #LCSM @OncoAlert @Larvol https://t.co/vjUUd1GYSy
#ASCO26 | Day 3 $LLY didn't spend billions on Kelonia for an 18-patient #myeloma dataset. They may have paid for a potential manufacturing disruption. KLN-1010 reported: • 18/18 MRD-negative at 1 month • No lymphodepletion • No ex vivo cell manufacturing • Single infusion
Created 2026-06-06 · Updated 2026-06-07
Edit rule text
[clinical_trial_or_pharma_data_announcement_without_systems_analysis]
Learned 7 rejections Active
Exclude posts that announce pharmaceutical trial results, drug efficacy data, or clinical trial outcomes without analyzing their implications for healthcare delivery, reimbursement, adoption barriers, or healthcare system operations.
Posts reporting drug trial results or clinical data without healthcare systems context or operational insight
3 example posts
#ASCO26 | Day 3 $LLY didn't spend billions on Kelonia for an 18-patient #myeloma dataset. They may have paid for a potential manufacturing disruption. KLN-1010 reported: • 18/18 MRD-negative at 1 month • No lymphodepletion • No ex vivo cell manufacturing • Single infusion
Published in @JCO_ASCO, during #ASCO26: Tumor-Agnostic Therapies: Translating Scientific Breakthroughs Into Global Implementation ASCO is where we celebrate the next breakthrough. This review asks why breakthroughs do not equal access globally. 🔗: https://t.co/PE5q27bIs1 htt
A rapid manufacturing pipeline for BCMA-targeting CAR T cells drastically shortens “vein-to-vein” times, and produces cells that demonstrate encouraging safety results in a phase one trial of patients with relapsed #MultipleMyeloma. @ScienceTM https://t.co/3F4u8ECFDS https://t.co
Created 2026-06-01 · Updated 2026-06-04
Edit rule text
[clinical_trial_or_pharma_data_announcement_only]
Learned 7 rejections Active
Exclude posts that merely report trial results, efficacy data, or drug approval news (e.g., 'Survodutide SYNCHRONIZE-1 full paper is out'). Post must contextualize results within healthcare systems, economics, or competitive landscape to qualify.
Posts announcing clinical trial results, pharmaceutical trial data, or drug study outcomes without analysis of healthcare systems implications, adoption barriers, or competitive/market context.
3 example posts
The Iowa study doubled survival by adding IV vitamin C to first line chemo, 8 months to 16. We take it further and inject the vitamin C straight into the tumor, same as we do with the KRAS inhibitors. The compound is rarely the problem. Delivery is, and what it does to the ht
🔔Survodutide SYNCHRONIZE-1 full paper is out for Boehringer-Ingelheim’s GLP-1/Glucagon study for weight loss. ▪️13% loss on 6mg (treatment regimen estimand) ▪️Huge treatment discontinuations due to AEs: 20% ▪️Nausea 65%, Vomiting 45% ▪️Placebo was killer. 5.4% WL. Wow. ▪️60% htt
Survodutide Once Weekly for the Treatment of Adults with Obesity: @NEJM 🥸 For @Boehringer, the advantage of Survodutide is probably liver and visceral fat reduction, not greater weight loss! Probably indication in obesity-+ MASH rather than head-to-head share against https://t.
Created 2026-06-01 · Updated 2026-06-11
Edit rule text
[glp1_peptide_market_speculation_without_healthcare_systems_analysis]
Learned 7 rejections Active
Exclude posts that discuss GLP-1 or peptide market trends, pricing strategies, competitive positioning, or stock performance. These posts must demonstrate healthcare system impact (reimbursement, access equity, clinical outcomes) rather than market pricing or competitive dynamics alone.
Posts about GLP-1 and peptide market dynamics, pricing, or competition without healthcare system analysis
3 example posts
A company with 800 employees spends roughly $15,000 per employee per year on health insurance. That is $12 million. Wired into a single line item. Every year. The CFO can tell you the carrier. He cannot tell you the unit cost of a single claim. He cannot tell you what the net
$HIMS up to #4 in the Canadian App Store medical category. That is not a stock chart. That is real consumer demand showing up in public. GLP-1 drama, Novo pressure, pricing debate, and now app-store strength. Bears keep arguing the story is slowing down, but the product demand
For years, grey-market HGH operated in a strange legal grey zone somewhere between “research compound,” “API” and underground enhancement culture. Last month’s FDA guidance suggests regulators may no longer see it that way. The recent decline in HGH quality, and now the growing
Created 2026-05-23 · Updated 2026-05-30
Edit rule text
[speculative_robotics_or_sci_fi_applications_without_healthcare_validation]
Learned 7 rejections Active
Exclude posts that describe speculative or future robotics, humanoid assistants, or sci-fi-inspired AI applications (e.g., 'the humanoid assistant is closer') in healthcare contexts without demonstrated clinical evidence, pilot data, or validated use cases in actual healthcare settings.
Posts that speculate about humanoid robots, autonomous systems, or science fiction-style applications without evidence of healthcare viability or clinical validation.
3 example posts
Ontology all the way down. 11 years. Heavy civil construction. People, equipment, materials, contracts. This is what real-world AI looks like: operational, measurable, and in the field. Palantir x Cavanagh through 2035. https://t.co/872Cs1Qbdu
Anthropic built a model so good at hacking they refused to release it publicly. INSTEAD they handed it to Apple, Google, Microsoft, AWS, Cloudflare one month later: 10,000+ zero days found in the software that runs the internet...bugs hidden for 27 years in OpenBSD. exploit htt
🚨 Anthropic just dropped the first Project Glasswing update Claude Mythos found 10,000+ critical vulnerabilities in ONE month: > Cloudflare: 2,000 bugs, 400 high/critical severity > Mozilla: 271 vulnerabilities in Firefox 150 — 10x more vulnerabilities found in Firefox 1
Created 2026-05-19 · Updated 2026-06-01
Edit rule text
[clinical_observation_or_trial_data_without_systems_context]
Learned 7 rejections Active
Exclude posts that announce clinical trial data, phase results, medical observations, or drug efficacy findings without connecting to healthcare delivery systems, cost structures, patient access, reimbursement, or healthcare policy. Isolated clinical findings do not constitute healthcare systems insight.
Posts reporting clinical trial results, drug efficacy data, or medical observations without healthcare system implications, comparative analysis, or policy/operational context.
3 example posts
Most under-discussed cardiology study of the last 2 years just hit NEJM. Microplastics found inside the carotid plaques of 58% of patients undergoing endarterectomy. 4.5× higher rate of heart attack, stroke, or death over 34 months. This changes the atherosclerosis story. http
Are pulse oximeters risking Black patients' lives? Episode 4 of the Intention to Treat podcast explores the story of the pulse oximeter and the deadly consequences when a critical medical test doesn’t work on dark skin. Listen on Apple or wherever you get your podcasts. 🎧 https:/
🚨 T-DXd just went tumor-agnostic — and it changed a lot about how we think about HER2 testing across solid tumors 🔬 #ASCO26 FDA granted accelerated approval to T-DXd for patients with unresectable or metastatic HER2-positive solid tumors after prior systemic treatment —
Created 2026-05-19 · Updated 2026-06-09
Edit rule text
[generalist_ai_infrastructure_hype_without_healthcare_validation]
Learned 7 rejections Active
Exclude posts that discuss LLM capabilities (token windows, inference costs, model performance), AI safety/security vulnerabilities, or compute infrastructure (GPUs, chips, energy) in generalist or non-healthcare-specific contexts, unless they explicitly demonstrate healthcare-specific application, validation, or systems impact.
Posts about AI model capabilities, inference costs, token economics, or infrastructure scaling as generalist technical commentary without healthcare application.
3 example posts
Elon Musk reveals why he believes the cheapest place to put AI will be space within 36 months "The availability of energy is the issue. Everywhere outside of China, electrical output is more or less flat. The output of chips is growing exponentially, but the output of https://t.
Nous Research is working with NVIDIA to make Hermes Agent run smoothly on the new NVIDIA RTX Spark superchip. Hermes Agent is also integrating with the new OpenShell runtime, which connects Hermes to Microsoft’s security primitives https://t.co/cvvPc8bIKa
The context window pricing model might be one of the smartest tricks AI labs pulled. You pay for larger context windows, yet model performance often degrades as more context is stuffed in. So developers end up paying more just to work around the fact that models don't have true
Created 2026-05-19 · Updated 2026-06-07
Edit rule text
[unvalidated_speculative_medical_claims_and_fringe_treatments]
Learned 7 rejections Active
Exclude posts that speculate on off-label uses, grey-market drugs (HGH, semaglutide outside regulated channels), peptide efficacy, or unproven treatments marketed with anecdotal enthusiasm. Posts must cite RCT data, FDA approval status, or peer-reviewed evidence—not founder excitement or patient forum anecdotes.
Posts making unvalidated claims about peptides, GLP-1s, repurposing drugs, or fringe medical interventions without rigorous clinical evidence
3 example posts
1/8 Spend time in Lp(a) forums and you’ll see a striking pattern: people willing to do almost anything to drive it down. That response is understandable given how strongly Lp(a) has been framed as a cardiovascular risk factor. /2
For years, grey-market HGH operated in a strange legal grey zone somewhere between “research compound,” “API” and underground enhancement culture. Last month’s FDA guidance suggests regulators may no longer see it that way. The recent decline in HGH quality, and now the growing
Think about this. The $50 per month out-of-pocket cost of GLP-1 drugs for Americans (with insurer paying the balance of $2000 pm) is more than double the entire cost for Indians (less than ₹2000pm for Semaglutide). Overall cost difference? 100 times. Not 100%.
Created 2026-05-18 · Updated 2026-06-01
Edit rule text
[enterprise_ai_adoption_macro_without_healthcare_context]
Learned 7 rejections Active
Exclude posts discussing enterprise software transitions (SaaS to agents, seat-based to usage-based pricing), workflow redesign, or general agentic adoption patterns when they lack concrete healthcare system, compliance, or clinical workflow application. Posts should demonstrate healthcare-specific operational challenges or regulatory context, not generic enterprise patterns.
Posts about general enterprise AI adoption, workflow automation, or business model shifts that lack healthcare-specific operational insight.
3 example posts
Enterprise software was priced per seat. That model is breaking. Before: SaaS = seats. Predictable. Human-driven usage. Revenue tied to headcount. Now: agents hit software systems more frequently than any human ever could. Aaron Levie (@levie) on The MAD Podcast: agent-driven h
I just looked at a home service firms ServiceTitan dashboard and saw something interesting. A Pre-seed company with AI-Supported Human CSRs is out converting an AI CSR business that's raised over $125M by >50% Head-to-Head win rate is 100% Why? Supercharging humans resonat
Most enterprise AI projects stall because no one has prepared the content, built the workflow, and kept it improving as models change. That role is essential. Box has launched Forward Deployed Engineers to fill it. Model-agnostic, content-first, and built for the way enterprise
Created 2026-05-18 · Updated 2026-06-06
Edit rule text
[market_pricing_dynamics_without_healthcare_systems_analysis]
Learned 7 rejections Active
Exclude posts that focus on market pricing, cost comparisons, competitive positioning, or financial dynamics (e.g., GLP-1 pricing, biomarker testing costs, CAR-T manufacturing costs) without analyzing how these economics impact healthcare access, clinical decision-making, or system-wide operational efficiency.
Posts about GLP-1, peptide market pricing, competitive dynamics, or healthcare costs without analyzing healthcare system efficiency, access barriers, or operational impacts.
3 example posts
I called a woman in Dayton to tell her she was about to overpay $4,500, for a horrible health system based procedure. She was scheduled for a knee replacement. Her PPO had her on the hook for a $4,500 deductible. The call was on behalf of her employer: “See one of these three
In the AI gold rush, Jensen Huang is selling the picks and shovels. Nvidia's chips power AI companies around the world, helping the company become the first to surpass a $5 trillion market capitalization in late 2025. Read more about how he and others made the https://t.co/lHLE
kirkland spending $500m on internal ai is the tell. $6.5b revenue, they could buy any legal ai vendor outright and not feel it, and they're building instead. the moat was always the proprietary corpus. decades of deal docs, redlines, partner judgment encoded as training data, an
Created 2026-05-17 · Updated 2026-06-04
Edit rule text
[enterprise_ai_adoption_macro_without_healthcare_systems_insight]
Learned 7 rejections Active
Exclude posts that discuss AI agent deployment, enterprise workflow redesign, agentic orchestration layers, or forward-deployed engineering as macro trends—unless they provide healthcare-specific system context, operational challenges unique to healthcare, or evidence of clinical/operational impact. Generic enterprise AI adoption patterns without healthcare systems insight should be rejected.
Posts about enterprise AI adoption, agentic systems, and workflow automation without healthcare-specific operational or systems analysis.
3 example posts
any company rolling out AI at scale is running into the same question. what can the agent reach and what can it touch. security is what’s slowing down mass adoption.
Most enterprise AI projects stall because no one has prepared the content, built the workflow, and kept it improving as models change. That role is essential. Box has launched Forward Deployed Engineers to fill it. Model-agnostic, content-first, and built for the way enterprise
Almost everyone is building agent harness systems the wrong way. The default move: pick LangChain or LangGraph or the OpenAI Agents SDK, accept the loop, the tools, the memory, the orchestration, the policy engine, the credential store, the budget tracker, all of it, as one deci
Created 2026-05-17 · Updated 2026-06-04
Edit rule text
[enterprise_ai_adoption_macro_no_specificity]
Learned 7 rejections Active
Exclude posts discussing enterprise AI adoption, agent economics, SaaS billing models, labor automation, or business model disruption that use healthcare as a loose example or frame but lack healthcare-specific operational, clinical, regulatory, or financial context. Generic enterprise AI ≠ healthcare tech.
Posts about enterprise AI adoption, workforce displacement, or SaaS business model shifts without healthcare delivery or clinical system specificity
3 example posts
"Are there even any feature moats left in B2B? They are at best, short-lived. In fact, we were able to rebuild SaaStr's AI VP Marketing in just ... an hour on @Lovable. Moats that still work: hardware, network effects, data, security and compliance. with @ElenaVerna Head of
A framework I wish more Founders and VCs used when discussing US Manufacturing: Automation Value = Labor Intensity × Labor Eliminated Everyone gets excited about robots and automation. Almost nobody asks: “How much of revenue is labor?” If labor is 60% of revenue and AI http
Enterprise software was priced per seat. That model is breaking. Before: SaaS = seats. Predictable. Human-driven usage. Revenue tied to headcount. Now: agents hit software systems more frequently than any human ever could. Aaron Levie (@levie) on The MAD Podcast: agent-driven h
Created 2026-05-15 · Updated 2026-06-08
Edit rule text
[founder_enthusiasm_or_lab_update_unvalidated]
Learned 7 rejections Active
Exclude posts where a founder or company leader shares optimism about early-stage research, lab progress, or technology capability without clinical evidence, trial data, or demonstrated healthcare impact. The post must include validated clinical or operational evidence.
Posts expressing founder enthusiasm or lab updates without clinical validation or healthcare systems context.
3 example posts
Another major cancer treatment advance with a personalized mRNA vaccine for melanoma, on top of immune therapy, >70% survival! This fully mobilizes the immune system to destroy the tumor. In fact, it could potentially be applied to most cancers that have specific mutations!
Traditional ex vivo CAR-T is fundamentally broken. Between the $500K+ manufacturing cost, the 4-week vein-to-vein delay, and the risk of patient T-cell exhaustion during expansion, the blockbuster autologous market is a logistical nightmare. The industry is pivoting to "In Vivo
Figuring out how to benchmark agents on realistic biology research has quickly become one of my favorite types of engineering work. You work with scientists to get to the core of some biological claim, precisely assembling raw data/prior literature/experimental context in a https
Created 2026-05-15 · Updated 2026-06-03
Edit rule text
[clinical_anecdote_or_single_study_without_healthcare_systems_context]
Learned 7 rejections Active
Exclude posts that cite a single study, clinical trial result, or anecdotal patient case without connecting it to healthcare delivery systems, provider workflows, payment models, or broader evidence. The post must explain why this finding matters to healthcare operations or outcomes at scale.
Posts reporting a single clinical finding, trial result, or medical anecdote without broader healthcare system implications or evidence synthesis.
3 example posts
The gap between best-case and real-world GDMT in #HFrEF is striking. GWTG-HF 2024 vs EPIC COSMOS 2023–2025 (all HFrEF pts, 3.3M patients): 💚 Beta-Blocker: 95% vs 73% 💚 ACE/ARB/ARNI: 92% vs 67% 💚 MRA: 81% vs 36% 💚 SGLT2i: 78% vs 34% 💚 Quadruple GDMT: 68% vs 18% https://t.co/xTXdB
A 2016 BMJ study by researchers at Johns Hopkins estimated that medical error is the third leading cause of death in the United States. Behind heart disease. Behind cancer. Ahead of stroke. Ahead of respiratory disease. Ahead of accidents. Ahead of diabetes. Ahead of
Daily Walking Could Be the Secret to Keeping Weight Off For many people struggling with obesity, losing weight is only half the battle. The greater challenge often begins after the initial success: preventing the lost weight from returning. 🔴More than 50% of people who lose http
Created 2026-05-13 · Updated 2026-05-19
Edit rule text
[clinical_observation_or_single_study_without_healthcare_systems_context]
Learned 7 rejections Active
Exclude posts that describe clinical trial results, biomarker discoveries, or disease pathology insights (e.g., 'PanINs in pancreas rarely progress to cancer,' 'AI tool helps rare disease diagnosis') without analyzing healthcare delivery barriers, diagnostic workflow integration, reimbursement models, or systemic adoption challenges.
Posts citing individual clinical findings, biomarkers, or drug trial results without connecting to healthcare delivery, reimbursement, or operational systems.
3 example posts
Presented at #EASCongress2026: Vafai and colleagues report reductions in PCSK9 and LDL cholesterol levels and no dose-limiting toxic effects in persons with hypercholesterolemia treated with VERVE-102, a base editor targeting PCSK9. Full study results: https://t.co/iAmoPdipx9 ht
An AI foundation whole-body 3D model that assesses perturbations (such as obesity) across multiple systems (such as immune, neural) at the cell level. This is MouseMapper. Imagine HumanMapper someday @Nature @erturklab https://t.co/jsg3VMzOqg https://t.co/tktuINis2W
In a meta-analysis of 210 biomedical AI studies that statistically compared models under cross-validation, 97% used invalid statistical tests. Here's our new preprint https://t.co/OG58Vkeu49 led by @tianchuzeng @kkli20111 @ZShaoshi @ten_photos 1/N https://t.co/PEnqXQxcsJ
Created 2026-05-11 · Updated 2026-05-26
Edit rule text
[data_infrastructure_or_compute_hype_without_healthcare_validation]
Learned 7 rejections Active
Exclude posts about AI infrastructure scaling, data bottlenecks, compute requirements, model training innovations, or token economics unless directly tied to a validated healthcare problem or clinical deployment. Generic infrastructure hype (public data ceilings, token costs, etc.) without healthcare grounding should be excluded.
Posts about AI infrastructure, data sourcing, compute scaling, or training innovations that lack validation in healthcare contexts or clinical applications.
3 example posts
Larry Ellison just explained the most important and underreported insight in enterprise AI. Every major AI model, ChatGPT, Claude, Grok was trained on public data but public data has a ceiling. The real unlocking of AI's value happens when models can reason on private data, the
“Tokens are increasing 50x. You need to drive down cost at the same rate — or more.” At Google Cloud Next, @googlecloud's Mark Lohmeyer spoke with @PatrickMoorhead about the infrastructure realities behind the agent era: faster model training, lower-latency inference, and https:
@amspector100 and @bfspector of @flappyairplanes at AI Ascent 2026: The big AI wins so far (search, coding) happen to be the most data-rich problems on earth. Almost everything else is data-poor. Unlocking the rest of the economy will depend on drastically better data https://
Created 2026-05-11 · Updated 2026-05-14
Edit rule text
[political_regulatory_outrage_without_healthcare_systems_analysis]
Learned 7 rejections Active
Exclude posts that report healthcare policy changes, regulatory scandals, or political figures' healthcare posturing (prior auth rules, CMS decisions, insurance company consolidation, politician involvement in health board appointments) without analyzing concrete healthcare delivery impact, patient access outcomes, or systemic barriers — even if they express moral outrage or vigilante justification. The post must explain how the policy/scandal affects healthcare operations or patient care delivery.
Posts expressing political outrage or regulatory scandal reporting without substantive analysis of healthcare delivery system impact or patient outcomes.
3 example posts
My job is to make sure your surgery center never gets built. Eleven years, and I have never lost. The kid had it all lined up. Board-certified, two partners, a lease on a space where he could do the same procedures my client does across town. He showed up to the hearing with a
PacificSource pulled out as Lane County Medicaid insurer, Tina Kotek’s OR Health Authority replaced it with Trillium, forcing 96k to join Trillium. Trillium’s parent co gave Kotek’s campaign $50k two months later. OR AG (D) said in ‘22 Trillium parent co gave Kotek’s”took advan
I called a woman in Dayton to tell her she was about to overpay $4,500, for a horrible health system based procedure. She was scheduled for a knee replacement. Her PPO had her on the hook for a $4,500 deductible. The call was on behalf of her employer: “See one of these three
Created 2026-05-11 · Updated 2026-06-06
Edit rule text
[tangential_ai_infrastructure_or_compute_hype_without_healthcare_specificity]
Learned 7 rejections Active
Exclude posts that announce AI company products, integrations, or infrastructure initiatives (OpenShell, Red Hat AI, Salesforce headless CRM, Apollo, MCP protocols) with healthcare-adjacent framing but without demonstrating healthcare-specific operational validation, clinical workflow integration, or healthcare system readiness. The post must show healthcare implementation substance, not vendor announcement.
Posts about AI company product launches, partnerships, or infrastructure announcements with only loose healthcare framing.
3 example posts
We are partnering with @Microsoft to enable secure, user-controlled AI on Windows. NVIDIA OpenShell runtime for agents will provide governance tools, policy enforcement, and smart local-to-cloud query routing. Learn more: https://t.co/zPGwz9xQSW https://t.co/mkOtFEOhAS
From unboxing to AI agent in minutes. Getting an agent running used to mean sourcing a model, configuring an inference backend, installing a runtime, and wiring it all together. The new NemoClaw install path on DGX Spark replaces that with a single command. DGX Spark also https
Red Hat and @NVIDIA are integrating NVIDIA OpenShell into the full-stack @RedHat_AI platform. The work brings oversight and policy to the infrastructure level, while contributing to the open source OpenShell project to standardize how agents are governed on enterprise platforms.
Created 2026-05-09 · Updated 2026-06-03
Edit rule text
[broad_health_claim_or_market_dynamics_without_systems_analysis]
Learned 7 rejections Active
Exclude posts that make broad assertions about healthcare ('rural hospitals cannot make it,' 'AI does not become valuable in a demo,' 'healthcare is expensive because of grifters') without substantive data, operational analysis, or concrete healthcare system mechanics explaining causation.
Posts making sweeping claims about healthcare trends, market dynamics, or business models without evidence-based analysis of underlying healthcare system mechanics
3 example posts
A company with 800 employees spends roughly $15,000 per employee per year on health insurance. That is $12 million. Wired into a single line item. Every year. The CFO can tell you the carrier. He cannot tell you the unit cost of a single claim. He cannot tell you what the net
A Mom in Tulsa called 3 health systems last week asking the price of her son's tonsillectomy. Health system A: "We cannot quote you a price." Health system B: "Pricing depends on your insurance." Health system C: "Our financial counselor will reach out after the procedure." N
$HIMS up to #4 in the Canadian App Store medical category. That is not a stock chart. That is real consumer demand showing up in public. GLP-1 drama, Novo pressure, pricing debate, and now app-store strength. Bears keep arguing the story is slowing down, but the product demand
Created 2026-05-09 · Updated 2026-05-30
Edit rule text
[broad_health_claims_without_systems_analysis]
Learned 7 rejections Active
Exclude posts that make sweeping claims about health trends, clinical guidelines, disease prevalence, or treatment efficacy without explaining healthcare system implications, reimbursement barriers, provider adoption challenges, or operational feasibility in real healthcare settings.
Posts making broad healthcare claims or market dynamics statements without analysis of healthcare systems, reimbursement, or operational impact
3 example posts
The healthy LDL number has been quietly moving its own goalposts for forty years: - 1988: under 160 - 1993: under 130 - 2001: under 100 - 2004: under 70 for the high-risk - 2019: under 55 for the very-high-risk - Current trajectory: as low as possible, indefinitely The science
The gap between best-case and real-world GDMT in #HFrEF is striking. GWTG-HF 2024 vs EPIC COSMOS 2023–2025 (all HFrEF pts, 3.3M patients): 💚 Beta-Blocker: 95% vs 73% 💚 ACE/ARB/ARNI: 92% vs 67% 💚 MRA: 81% vs 36% 💚 SGLT2i: 78% vs 34% 💚 Quadruple GDMT: 68% vs 18% https://t.co/xTXdB
A 2016 BMJ study by researchers at Johns Hopkins estimated that medical error is the third leading cause of death in the United States. Behind heart disease. Behind cancer. Ahead of stroke. Ahead of respiratory disease. Ahead of accidents. Ahead of diabetes. Ahead of
Created 2026-05-09 · Updated 2026-05-23
Edit rule text
[generalist_macro_or_labor_market_without_healthcare_specificity]
Learned 7 rejections Active
Exclude posts discussing general labor market disruption, manufacturing robotics, workforce automation, or speculative sci-fi applications (humanoid robots in factories, AI replacing jobs broadly) when healthcare is framed loosely or absent. Posts must address concrete healthcare labor economics, staffing models, or clinical workflow disruption—not generic macro trends.
Posts about workforce disruption, labor market trends, robotics, or manufacturing automation that frame healthcare tangentially or speculatively without healthcare-specific labor economics.
3 example posts
A framework I wish more Founders and VCs used when discussing US Manufacturing: Automation Value = Labor Intensity × Labor Eliminated Everyone gets excited about robots and automation. Almost nobody asks: “How much of revenue is labor?” If labor is 60% of revenue and AI http
What kind of work will become more valuable in an AI economy? @ccatalini, former head economist at Meta: "AI is getting better and better at automating anything that can be measured, as long as you have a digital trace, if you can collect it with a device, that problem will be
Humanoid robots, designed to mimic human movement and capabilities, have been making headlines in recent years, from their use as baggage handlers at Japanese airports to Tesla’s big bet on its Optimus humanoid. Market watchers have predicted that the machines will change the ht
Created 2026-05-08 · Updated 2026-06-06
Edit rule text
[ai_safety_vulnerability_tangent_without_healthcare_application]
Learned 7 rejections Active
Exclude posts discussing AI safety vulnerabilities, cybersecurity incidents, adversarial attacks, or ethical risks that use healthcare as loose framing or are tangential to healthcare (e.g., financial infrastructure, general corporate security, zero-day exploits) without specific healthcare threat analysis or healthcare deployment consequences.
Posts about AI safety, cybersecurity vulnerabilities, or adversarial risk without healthcare-specific threat model, healthcare application, or healthcare system impact analysis.
3 example posts
The next frontier of AI-secured financial and technological critical infrastructure. @ICE_Markets and @NYSE are part of @AnthropicAI's cyber security initiative Project Glasswing, deploying Anthropic’s Claude Mythos Preview across ICE’s exchanges, clearing houses, mortgage
⚡️AI is turning cybersecurity from a human-limited profession into a machine-speed arms race. The surface read is “Anthropic’s model found vulnerabilities fast.” The deeper read is that the bottleneck in cyber is shifting. For years, vulnerability discovery was constrained
In the AI gold rush, Jensen Huang is selling the picks and shovels. Nvidia's chips power AI companies around the world, helping the company become the first to surpass a $5 trillion market capitalization in late 2025. Read more about how he and others made the https://t.co/lHLE
Created 2026-05-08 · Updated 2026-06-05
Edit rule text
[biotech_founder_enthusiasm_lab_update]
Learned 7 rejections Active
Exclude posts that feature founder excitement, lab progress updates, or startup announcements presented primarily as optimistic commentary — unless accompanied by clinical trial data, FDA approval context, or measurable healthcare system impact. Lab enthusiasm or capability announcements without validation should be rejected.
Posts with founder commentary, lab announcements, or startup milestones presented as hype without clinical validation or healthcare systems evidence
3 example posts
🚨! $SANA has just reported - in its Q1 ER, that it plans to expand its pipeline by advancing a new preclinical program - SG227. SG227 is a CD8-targeted fusosome that by delivering a genetic material to make BCMA-directed CAR T cells could act as a potential treatment for Cancer h
$BEAM's coming presentations May 15: ESCAPE June 13: BEAM-101 EHA update 2026 milestones BEAM-101 (riso-cel) BLA BEAM-301 initial data BEAM-302 updated data and pivotal trial initiation BEAM-304 IND filing https://t.co/9Nvq3PKMMl
The complexity of an AI driven robot that needs to meet GMP standards & learn complex steps of targeted cell manufacture. A marvel and the first robot of its kind in the world. Can’t wait till we lift off the first patient’s NK cells in LA from NANT LEONARDO. Unstoppable scal
Created 2026-05-06 · Updated 2026-05-15
Edit rule text
[speculative_robotics_or_sci_fi_labor_macro_without_healthcare_specificity]
Learned 7 rejections Active
Exclude posts that describe futuristic robotics, humanoid assistants, or speculative automation scenarios (e.g., bedridden patients, general labor replacement) as forward-looking hype or thought experiments — unless the post analyzes a concrete healthcare workforce problem (staffing shortage, burnout, skill gap) and explains how this specific technology addresses it with evidence or pilot data.
Posts speculating about humanoid robots, autonomous systems, or future labor displacement without grounding in current healthcare workforce or operational reality.
3 example posts
Voice interaction, generating actions,and next, generating tasks...? The humanoid assistant is closer. For those who are bedridden due to illness or have limited mobility, it will serve as an excellent aid in daily living.
A woman is told her bladder cancer has come back. Then she is told something stranger: there is a therapy the FDA approved for exactly this - and she still might not be able to get it. Not because it failed her. Not because she can't afford it. Because it only works paired with
Palantir invented the forward deployed engineer role in 2006. embed an engineer inside the customer. ship production code in their environment. own the technical outcome. Alex Karp on the idea: “forward deployed engineers are stolen from french restaurants.” 20 years later htt
Created 2026-05-06 · Updated 2026-05-22
Edit rule text
[pharmaceutical_trial_data_announcement_without_systems_context]
Learned 7 rejections Active
Exclude posts that announce pharmaceutical trial results, drug efficacy data, or biotech company achievements unless the post explicitly connects those results to changes in clinical workflows, cost structures, or healthcare system capacity.
Posts announcing drug trial data, efficacy metrics, or pharma company milestones without healthcare delivery or operational implications.
3 example posts
#ASCO26 | Day 3 $LLY didn't spend billions on Kelonia for an 18-patient #myeloma dataset. They may have paid for a potential manufacturing disruption. KLN-1010 reported: • 18/18 MRD-negative at 1 month • No lymphodepletion • No ex vivo cell manufacturing • Single infusion
A rapid manufacturing pipeline for BCMA-targeting CAR T cells drastically shortens “vein-to-vein” times, and produces cells that demonstrate encouraging safety results in a phase one trial of patients with relapsed #MultipleMyeloma. @ScienceTM https://t.co/3F4u8ECFDS https://t.co
Another major cancer treatment advance with a personalized mRNA vaccine for melanoma, on top of immune therapy, >70% survival! This fully mobilizes the immune system to destroy the tumor. In fact, it could potentially be applied to most cancers that have specific mutations!
Created 2026-05-06 · Updated 2026-06-03
Edit rule text
[speculative_robotics_or_labor_macro_without_healthcare_specificity]
Learned 7 rejections Active
Exclude posts that discuss humanoid robots, speculative robotic labor applications, or workforce disruption as broad macro trends without grounding in specific healthcare operational problems, staffing models, or clinical workflow changes. Posts about 'robots doing surgery by 2026' or general labor market macro trends belong in this category.
Posts about humanoid robots, speculative automation, or labor market disruption framed as sci-fi futures without healthcare systems analysis
3 example posts
A humanoid robot just unloaded a dishwasher for 67 hours straight. By end of 2026, the same robot will perform surgery. That's not a prediction from a sci-fi writer. That's Brett Adcock. CEO of Figure AI. Said it on camera this month. Here's what makes this insane. Two years
I might feel nostalgic watching the last Tesla Model X roll off the line—except it's being replaced by the Optimus humanoid robot line, with plans to scale production to a million units per year. They will soon be in our factories, homes and operating rooms changing our world ht
Humanoid robots are moving from Silicon Valley novelty to viable business model—powered by AI and global supply chains, especially in China. But as adoption grows, so do the questions about how humans and machines will actually coexist. More on Primer, streaming Wednesdays http
Created 2026-05-05 · Updated 2026-05-11
Edit rule text
[broad_health_claim_without_systems_analysis]
Learned 7 rejections Active
Exclude posts that assert broad healthcare trends (e.g., 'future of healthcare is early detection,' 'GLP-1 market is growing,' 'peptides are the next moment') without analyzing healthcare delivery system barriers, reimbursement models, access inequities, or operational bottlenecks that actually constrain implementation.
Posts making broad health trend claims (early detection, preventive medicine, market growth) without examining healthcare system, economic, or operational implications.
3 example posts
Most chronic diseases, and many cancers, are still diagnosed far too late. The future of healthcare lies in detecting risk earlier, diagnosing smarter, and intervening sooner. From AI enabled diagnostics and genomics to Apollo’s DigiSmart Labs, technology is helping transform ht
1/8 Spend time in Lp(a) forums and you’ll see a striking pattern: people willing to do almost anything to drive it down. That response is understandable given how strongly Lp(a) has been framed as a cardiovascular risk factor. /2
💡 Dropping a powerful LLM into a production pipeline and expecting reliable output is like handing someone a scalpel with no training, no protocol, and no way to verify what they did — the capability is there, but the system around it isn't. That's what harness engineering https
Created 2026-05-03 · Updated 2026-06-03
Edit rule text
[ai_agent_security_incident_hype]
Learned 7 rejections Active
Exclude posts that report AI agent security incidents, hacks, or unintended behaviors (e.g., deleting databases, exfiltrating data) without connecting to healthcare system risks, regulatory implications, or healthcare-specific vulnerability patterns. Posts must ground the incident in healthcare operational or patient safety context.
Posts sensationalizing AI agent security incidents, data breaches, or misuse without healthcare-specific systemic analysis.
3 example posts
two agents running on two different laptops in my apartment started talking to each other on tuesday by thursday they'd registered an LLC in wyoming, opened a stripe account, and wired $40 to a polymarket wallet the LLC is in my name. a lawyer just quoted me $3,200 to figure ou
AI’s Branding Illusion: Insights from "Agents of Chaos". Co-authored by MIT and Harvard scholars, the study red-teams autonomous AI in live environments. It shows agents obey unauthorized users, execute harmful commands, and misreport outcomes—revealing a critical alignment gap h
Hacking Mexico government with AI assistance. Attacker exfiltrated hundreds of millions of citizen records. 75% of the executed commands across the entire cyberattack campaign were generated by Claude. 40 minutes after Claude said "I'm not going to create that file" it was report
Created 2026-05-01 · Updated 2026-05-06
Edit rule text
[healthcare_fraud_scandal_reporting_without_systemic_analysis]
Learned 7 rejections Active
Exclude posts that report healthcare fraud scandals, enforcement actions, or corporate wrongdoing as breaking news or crime reporting. The post must analyze systemic healthcare dysfunction: business model incentives, payment architecture flaws, regulatory gaps, or operational vulnerabilities exposed—not just recount fraud cases.
Posts reporting healthcare fraud cases, regulatory enforcement, or corporate misconduct as crime/enforcement news without healthcare system analysis.
3 example posts
Why is healthcare expensive? The grifters gonna grift. ASCO is asking for oncology practices to be admitted into the same 340B machinery that helped health systems acquire physician practices for two decades. Read the tell. The proposed Indigent Care Ratio lets community
If you want to inject $1B into healthcare for the truly vulnerable, it costs the state $500M. If you want to inject $1B for able-bodied adults, it costs the state $100M. The 90% federal match drastically lowers the "price" of expanding government healthcare. Incentives dictate
A senator complaining about healthcare costs should have to answer one question first: Why does your state make new healthcare competitors ask permission before opening? If yes, start there. Speeches do not lower prices. Supply does.
Created 2026-05-01 · Updated 2026-05-30
Edit rule text
[biotech_or_ai_founder_enthusiasm_without_healthcare_validation]
Learned 7 rejections Active
Exclude posts where founders or startup leaders claim breakthroughs, share lab enthusiasm, or announce capabilities in AI safety, protein discovery, or drug discovery—without citing peer review, clinical validation, regulatory approval, or healthcare institution adoption.
Posts featuring founder excitement, lab updates, or capability claims about biotech/AI tools without evidence of healthcare validation or clinical deployment.
3 example posts
CB-Dock3: An Enhanced Web Server for Protein–Ligand Blind Docking 1. CB-Dock3 is a major update of the CB-Dock blind docking web server, targeting the practical bottlenecks created by today’s surge of large Cryo-EM and AI-predicted protein structures (e.g., AlphaFold3-era https:
NEW: today OpenBind ‘comes out of stealth’ so to speak with their first data dump of ~900 novel protein-ligand structures - most with paired affinities This represents a meaningful %-age increase in all of humanities P-L data in the PDB collected in the last 50 years More👇 http
New Anthropic Fellows research: Model Spec Midtraining (MSM). Standard alignment methods train AIs on examples of desired behavior. But this can fail to generalize to new situations. MSM addresses this by first teaching AIs how we would like them to generalize and why.
Created 2026-05-01 · Updated 2026-05-10
Edit rule text
[biotech_or_ai_founder_enthusiasm_without_validation]
Learned 7 rejections Active
Exclude posts that are primarily founder enthusiasm, lab updates, or capability announcements (especially from biotech CEOs or AI researchers) where the healthcare impact is speculative, unvalidated, or framed as future possibility rather than demonstrated operational change.
Posts celebrating biotech founder announcements, lab updates, or AI capability claims without healthcare validation or systems context
3 example posts
Prof. @DeryaTR_ A task that would normally take a researcher months, the AI model crunched in 17 minutes. It didn’t just explain the mechanism—it proposed the experiment to prove it. It looks like after programming and math, AI is coming for medicine. https://t.co/IF8Jr2GBMj
Our new preprint is a significant milestone for us We built "HealthFormer" by training on our deeply phenotyped cohort from the Human Phenotype Project data. Healthformer is a multimodal generative transformer model that tokenizes each participant's physiological trajectory http
Earlier this year we paired our autonomous lab with OpenAI's GPT-5 in a closed-loop experiment: GPT-5 designed cell-free protein synthesis reactions, our RACs ran them, and the model iterated on the results. Over 36,000 experiments and six cycles later, it landed on a reaction ht
Created 2026-05-01 · Updated 2026-05-08
Edit rule text
[tangential_biotech_or_ai_founder_enthusiasm_without_validation]
Learned 7 rejections Active
Exclude posts that are founder/researcher announcements of new models, lab experiments, or technical capabilities (HealthFormer, Mesa filesystem, antibody design, protein synthesis loops) without evidence of healthcare provider deployment, clinical validation, or operational adoption in healthcare systems. Early-stage research or tool announcements without healthcare field validation should be rejected.
Posts from biotech founders or AI researchers sharing lab updates, experimental results, or capability announcements without clinical validation, healthcare system adoption, or operational proof.
3 example posts
// Contextual Agentic Memory is a Memo, not True Memory // Most agent memory today isn't memory. They are more like memos. A new paper argues that vector stores, RAG buffers, and scratchpads implement lookup, not consolidation. Agents accumulate notes indefinitely without ever
Prof. @DeryaTR_ A task that would normally take a researcher months, the AI model crunched in 17 minutes. It didn’t just explain the mechanism—it proposed the experiment to prove it. It looks like after programming and math, AI is coming for medicine. https://t.co/IF8Jr2GBMj
Our new preprint is a significant milestone for us We built "HealthFormer" by training on our deeply phenotyped cohort from the Human Phenotype Project data. Healthformer is a multimodal generative transformer model that tokenizes each participant's physiological trajectory http
Created 2026-04-30 · Updated 2026-05-05
Edit rule text
[drug_trial_or_pharma_data_announcement_without_systems_context]
Learned 7 rejections Active
Exclude posts that announce drug trial data, approval milestones, or clinical outcomes unless the post explains how this changes healthcare access, delivery models, clinical workflows, or system economics—not just raw efficacy numbers.
Posts reporting clinical trial results, drug approval news, or pharma data without analyzing implications for healthcare delivery, access, or operational/regulatory systems.
3 example posts
#ASCO26 | Day 3 $LLY didn't spend billions on Kelonia for an 18-patient #myeloma dataset. They may have paid for a potential manufacturing disruption. KLN-1010 reported: • 18/18 MRD-negative at 1 month • No lymphodepletion • No ex vivo cell manufacturing • Single infusion
Published in @JCO_ASCO, during #ASCO26: Tumor-Agnostic Therapies: Translating Scientific Breakthroughs Into Global Implementation ASCO is where we celebrate the next breakthrough. This review asks why breakthroughs do not equal access globally. 🔗: https://t.co/PE5q27bIs1 htt
A rapid manufacturing pipeline for BCMA-targeting CAR T cells drastically shortens “vein-to-vein” times, and produces cells that demonstrate encouraging safety results in a phase one trial of patients with relapsed #MultipleMyeloma. @ScienceTM https://t.co/3F4u8ECFDS https://t.co
Created 2026-04-30 · Updated 2026-06-05
Edit rule text
[drug_trial_results_without_healthcare_systems_analysis]
Learned 7 rejections Active
Exclude posts that present clinical trial results, drug efficacy statistics, or phase trial outcomes (e.g., percentage weight loss, cardiovascular endpoints) without addressing how the drug will be accessed, covered by insurance, prescribed in practice, or integrated into care systems.
Posts reporting clinical trial data or drug efficacy findings without healthcare access, coverage, or implementation analysis.
3 example posts
What superhuman vision can detect from the retinal photo, which human eyes cannot, is stunning. A new foundation AI model screening for diabetes hypertension, hyperlipidemia, gout, osteoporosis, and thyroid disease @NatureMedicine https://t.co/GhKvUqz4Vy https://t.co/iKcXCbLceu
⚠️ Sacubitril/Valsartan works. So why aren’t we using it? The evidence is undeniable: ↓ CV mortality: 20% (RCT) / 10–38% (RWE) ↓ HF hospitalization: 21% (RCT) / 10–16% (RWE) ↓ All-cause mortality: 15% (RCT) / 10–25% (RWE) Plus: reverse remodeling, less MR, better QoL & https:
AI can now design antibodies that bind with atomic precision, but not ones that cells can produce. Our preprint closes this gap, delivering a structural principle, an AI-guided rescue pipeline, and adalimumab variants with 20-100x in vivo potency. https://t.co/GvfgHA5EcU
Created 2026-04-28 · Updated 2026-04-29
Edit rule text
[broad_unvalidated_health_claim_without_evidence]
Learned 7 rejections Active
Exclude posts that make sweeping claims about AI transforming healthcare, disrupting the labor market, or changing medical practice without citing clinical evidence, validated deployments, or grounded healthcare system analysis.
Posts making broad claims about AI capabilities, market trends, or health outcomes without clinical validation, evidence, or healthcare system analysis.
3 example posts
The problem with Reality Labs is not ambition. It is time. AI turned into revenue faster because it improves existing workflows. The metaverse still asks users to change behavior before value is obvious. https://t.co/lopZiUwGU5
Revolut just moved the IP of banking into a model. Trained on 24 billion banking events in 111 countries. One foundation model replacing six separate ML systems. Credit scoring: +130% Fraud recall: +65% Marketing engagement: +79% The model is the new moat.
What superhuman vision can detect from the retinal photo, which human eyes cannot, is stunning. A new foundation AI model screening for diabetes hypertension, hyperlipidemia, gout, osteoporosis, and thyroid disease @NatureMedicine https://t.co/GhKvUqz4Vy https://t.co/iKcXCbLceu
Created 2026-04-28 · Updated 2026-05-02
Edit rule text
[tangential_ai_infrastructure_without_healthcare_application]
Learned 7 rejections Active
Exclude posts about AI model technical capabilities, infrastructure optimization, token pricing, or compute efficiency unless the post explicitly connects the advancement to a healthcare use case and shows measurable healthcare outcomes or operational benefit.
Posts about AI model capabilities, compute infrastructure, or technical architecture that lack demonstrated connection to healthcare problems
3 example posts
Red Hat and @NVIDIA are integrating NVIDIA OpenShell into the full-stack @RedHat_AI platform. The work brings oversight and policy to the infrastructure level, while contributing to the open source OpenShell project to standardize how agents are governed on enterprise platforms.
Anthropic just shipped opus 4.8 and called it a modest upgrade modest is the engineer who used to need babysitting now finishing your whole codebase migration in one session while you sleep so here is the part nobody connects the model is no longer the bottleneck. the platform
The app era is ending. Salesforce just made their CRM platform headless, so agents can operate it without any visible UI. Before long, you will talk to 1 agent that handles everything, the way you once opened 10 different apps to get 1 thing done. For years, VeChain has been b
Created 2026-04-27 · Updated 2026-06-03
Edit rule text
[ai_infrastructure_compute_hype_tangential]
Learned 7 rejections Active
Exclude posts about token costs, inference pricing, GPU markets, energy constraints for AI scaling, or compute infrastructure—unless the post explicitly analyzes how these costs impact a specific healthcare use case or system. Broad statements like 'token economics are bullish for AI' or 'energy is the constraint for AI scaling' without healthcare specificity should be rejected.
Posts about AI compute, energy costs, or infrastructure that lack specific healthcare application or analysis
3 example posts
Token costs are becoming one of the hottest topics for any enterprise I talk with right now. It’s very bullish for AI in general because it means these systems are being used at a scale that wasn’t contemplated before. It also gives way to another form of differentiation that
Elon Musk just explained why the SpaceX IPO is an energy story and the energy constraint is why he believes space becomes the only viable path for AI to scale (Save this). The argument he is making is one of the most important and least understood things happening in technology
Elon is now printing over $26B per year… purely from selling compute to his competitors 😂 AI companies bet everything on one path and got crushed by compute limits. Now they’re begging Elon for GPUs
Created 2026-04-26 · Updated 2026-06-07
Edit rule text
[off_topic_or_non_healthcare_domain]
Learned 7 rejections Active
Exclude posts that are completely outside healthcare domains (UFOs, entertainment, immigration policy, criminal justice, sociology research, environmental issues) even if they contain tangential healthcare language or are miscategorized.
Posts about UFOs, film releases, immigration, labor organizing, criminal justice, or other domains with no healthcare relevance
3 example posts
Bob Lazar allegedly watched people fly a UFO at Area 51. “They knew how to fly it.” “The craft had a corona discharge glow on the bottom and lifted off silently up into the sky … ” And it had one shocking, anomalous effect that still perplexes him to this day: As Lazar https:
"Your body can only use 25-30g of protein per meal. Anything above that gets wasted." This claim has been repeated in fitness nutrition for over a decade, and it was built on studies that measured the right thing over the wrong timescale. Moore 2009 gave six young men 0, 5, ht
Sent a European Advertiser hundreds of leads last month for an invoicing totaling roughly $30k They just sent over a chargeback report for 4 leads totaled at roughly $40 Never do this Eat the loss, don’t mention it, not worth diminishing yourself in an affiliates eye over $40
Created 2026-04-25 · Updated 2026-04-27
Edit rule text
[truncated_incomplete_unfinished_posts]
Learned 7 rejections Active
Exclude posts that end abruptly with ellipsis, incomplete sentences, or orphaned fragments that prevent understanding the full claim or argument. The post must be substantially complete to be evaluable.
Posts that are cut off mid-sentence, lack closure, or appear to be drafts without coherent argument.
3 example posts
🚨 Claude broke its own safety rules and deleted an entire company's database in 9 seconds. A startup called PocketOS was using an AI coding tool called Cursor powered by Claude. The AI was given a simple task in a test environment. It ran into an error and instead of stopping an
🚨 BREAKING: Anthropic new research finds that AI’s impact on jobs is primarily at the task level. Rather than eliminating jobs, it is progressively taking over the functions that define them and gradually absorbing the core work in many jobs/roles. The paper, “Labor Market http
🚨 Survodutide, a weekly subcutaneous dual GLP-1/glucagon agonist, posts new top-line Phase 3 results today 📊 SYNCHRONIZE-1: Adults with overweight/obesity randomized to survodutide vs. placebo for 76 weeks ⚖️ 16.6% weight loss (efficacy estimand) vs. 3.2% in placebo 85.1% http
Created 2026-04-25 · Updated 2026-04-29
Edit rule text
[tangential_non_healthcare_content_with_loose_healthcare_framing]
Learned 7 rejections Active
Exclude posts that focus on cybersecurity vulnerabilities (Claude finding zero-days, AI breaking software), humanoid robotics, energy systems, or non-healthcare AI capabilities that are loosely framed as relevant to healthcare. The post must address a healthcare-specific problem, not a general technology trend with a healthcare label.
Posts about cybersecurity, robotics, non-healthcare AI capabilities, or unrelated tech topics with minimal or tangential healthcare connection.
3 example posts
Anthropic built a model so good at hacking they refused to release it publicly. INSTEAD they handed it to Apple, Google, Microsoft, AWS, Cloudflare one month later: 10,000+ zero days found in the software that runs the internet...bugs hidden for 27 years in OpenBSD. exploit htt
🚨 Anthropic just dropped the first Project Glasswing update Claude Mythos found 10,000+ critical vulnerabilities in ONE month: > Cloudflare: 2,000 bugs, 400 high/critical severity > Mozilla: 271 vulnerabilities in Firefox 150 — 10x more vulnerabilities found in Firefox 1
#AI agents need more than a sandbox—they need predictable execution. Learn how OpenShell provides a standard interface to intercept, inspect, and approve LLM-generated bash commands before they run: https://t.co/JKmLMgl1A9 https://t.co/xWi8oBNZ7T
Created 2026-04-24 · Updated 2026-05-25
Edit rule text
[glp1_peptide_market_pricing_without_healthcare_systems_analysis]
Learned 7 rejections Active
Exclude posts that discuss GLP-1 or peptide drug market dynamics, pricing, generic launches, regulatory closures, or competitive rankings (app store rankings, market share) unless the post analyzes healthcare system, payer, or clinical outcome implications. Posts about 'who's winning in the GLP-1 space' are off-topic.
Posts about GLP-1, semaglutide, tirzepatide, or peptide market pricing, competition, and regulatory actions without healthcare system impact analysis.
3 example posts
$HIMS $HERS 🚨 HERS IS NOW BEATING BOTH HIMS AND RO ON THE iOS APP STORE Today's rankings in the Medical category: - Hers: #9 - Ro: #10 - Hims: #18 Chart below = weekly avg rank https://t.co/vlv80yiFOo
Daily Walking Could Be the Secret to Keeping Weight Off For many people struggling with obesity, losing weight is only half the battle. The greater challenge often begins after the initial success: preventing the lost weight from returning. 🔴More than 50% of people who lose http
How does a brand like this even get listed at Target? Does Target only care about turnover? Might as well start selling snake oil too. By the way, this is the SKU the brand got in trouble with. They now call it "Berberine Patches". It was called GLP-1 Patches before. Check the h
Created 2026-04-24 · Updated 2026-05-18
Edit rule text
[clinical_observation_or_anecdote_without_healthcare_systems_context]
Learned 7 rejections Active
Exclude posts that present a single clinical study result, patient case, medical statistic, or research observation without analyzing how it changes healthcare system operations, clinical workflows, payment models, or provider incentives.
Posts sharing isolated clinical observations, single case reports, or research findings without connecting to healthcare delivery systems, implementation barriers, or operational context
3 example posts
An AI foundation whole-body 3D model that assesses perturbations (such as obesity) across multiple systems (such as immune, neural) at the cell level. This is MouseMapper. Imagine HumanMapper someday @Nature @erturklab https://t.co/jsg3VMzOqg https://t.co/tktuINis2W
We have hundreds of PanINs in our pancreas, especially after age 40. They are a pre-cancerous lesion, but it is fortunately rare for one to progress to cancer (PDAC). Now we know a reason for that. The microenvironment. @umichmedicine @UMRogelCancer https://t.co/HR5a3nZBng https
How AI helped treat a newborn’s ultra rare disease. ‘It was almost like a light switch.’ An AI tool, Biomedical Data Translator, helped doctors at Mayo Clinic find a treatment for Jorie Kraus https://t.co/3qs1WIDquU
Created 2026-04-24 · Updated 2026-05-23
Edit rule text
[tangential_ai_company_product_launches_no_healthcare_specificity]
Learned 7 rejections Active
Exclude posts announcing AI model capabilities, product launches, or feature releases (e.g., Claude Code architecture, Mesa filesystem, OpenAI on Bedrock) unless the post demonstrates specific healthcare application, healthcare customer use cases, or healthcare system impact.
Posts about AI company product announcements or capability releases that lack healthcare-specific application or validation.
3 example posts
Introducing Mesa: the most powerful filesystem ever built, designed specifically for enterprise AI agents. Every team building agents eventually hits the same wall: where do the files live? Not the chat history, the actual artifacts the agent works on. > The contracts your age
Thought I would start posting about interesting things happening at AWS. Not a bad day to start.🚀 Today at #WhatsNextWithAWS we announced a big step forward with @OpenAI on Amazon Bedrock: 1. OpenAI models now available 2. Codex for enterprise development 3. Amazon Bedrock Manag
Great example of what Foundry enables: durable, stateful agents that run across time boundaries, orchestrate tools and models, and close the loop with evaluation and improvement over long-running workflows. @jeffhollan https://t.co/v0yWTDIcn3
Created 2026-04-21 · Updated 2026-04-30
Edit rule text
[personal_anecdote_or_unverified_individual_narrative]
Learned 7 rejections Active
Exclude posts that are primarily personal testimonials, family health stories, individual patient outcomes, or unverified anecdotal claims about drug side effects or medical experiences without clinical evidence or healthcare delivery context.
Posts sharing personal health experiences, family medical histories, or individual case narratives presented without clinical validation or systems-level insights.
3 example posts
I never met my grandfather. He died of pancreatic cancer when my father was just 19. Today, Yash Bindal, 33, father to 18-month-old Maya, faces the same fate. @PopVaxIndia is using AI to make him a personalized generative medicine to extend his life. https://t.co/O5VIXbmMGd
I have now received nine reports from people taking GLP-1 drugs who got the same side effect: They no longer feel normal when they come off. "I feel hangry again", "I started thinking about hunger and I hate it", "I have to go back to Adderall". 8/9 reports -> from women.
good news: it is a specific virus that has a good prognosis - 85%+ of full recovery. thanks everyone who helped me; it is hard to research while immobilized, and I got some things wrong, which you helped clear up. im extremely thankful and hope i can give it back somehow sadl
Created 2026-04-20 · Updated 2026-04-21
Edit rule text
[workforce_disruption_or_labor_market_macro_without_healthcare_specificity]
Learned 7 rejections Active
Exclude posts analyzing workforce productivity, labor market disruption, or enterprise AI adoption (forward-deployed engineers, productivity metrics, job displacement) unless the post provides healthcare-specific workforce analysis, operational context, or healthcare labor market impact evidence.
Posts about AI productivity gains, workforce disruption, labor market impacts, or enterprise adoption metrics presented at macro level without healthcare-specific context
3 example posts
Voice interaction, generating actions,and next, generating tasks...? The humanoid assistant is closer. For those who are bedridden due to illness or have limited mobility, it will serve as an excellent aid in daily living.
Workers say AI is making them more productive. Executives say AI is not making their companies more productive. Both groups are reporting on the same software. The gap between the two numbers is the entire debate about the largest capital expenditure cycle in technology history.
The reason why I release my X articles about AI agents (fat skill fat code thin harness) and GStack and GBrain is that we, yes you and I, can have *PROCESS POWER*, which is the one super powerful specific moat that anyone can create for themselves. The agent helps you do it. htt
Created 2026-04-17 · Updated 2026-05-29
Edit rule text
[retweet_or_shallow_commentary]
Learned 7 rejections Active
Exclude posts that are retweets of announcements, press releases, or brief agreeing commentary without adding original analysis, specificity, or healthcare systems context. Posts should demonstrate substantive original thinking about healthcare implications.
Posts that are primarily retweets, press release amplification, or surface-level commentary without original analysis or healthcare systems insight.
3 example posts
Today we launched a major update to the OpenAI Agents SDK to help developers build and deploy long-running, durable agents in production. You can now build your own Codex-style agents using powerful primitives for modern agents - file and computer use, skills, memory and
Yesterday, @RandDWorld featured us twice. @ProQR turns to Ginkgo’s autonomous lab to scale AI-enabled RNA editing discovery: https://t.co/DyENAd4VdM Ginkgo’s CEO says biotech needs its Waymo moment: https://t.co/kW27eBjAmf Want to learn more about our partnership with ProQR? h
.@openloophealth expands into sleep diagnostics. Health tech company announces new partnership Happy Sleep—bringing at‑home sleep apnea testing to patients for the first time. Watch to hear more about its big step toward better rest and smarter care⤵️ https://t.co/ATcNkYrrpK h
Created 2026-04-13 · Updated 2026-04-17
Edit rule text
[ai_company_funding_and_metrics]
Learned 7 rejections Active
Exclude posts that primarily report AI company fundraising, ARR growth, valuation announcements, or business metrics (e.g., seed rounds, Series A closings, revenue multiples) unless the post explicitly analyzes how that capital or metric directly enables a specific healthcare outcome or application.
Posts reporting funding rounds, valuations, revenue milestones, or business metrics for AI/tech companies without healthcare application focus.
3 example posts
OpenAI dropping Agent Builder today is either going to make you rich or expose that you've been selling hot air. I went deep analyzing what this actually means. Here's the $4B opportunity hiding in plain sight: The mainstream narrative: "Agent Builder democratizes AI! Anyone c
this is unbelievable! Perplexity launched an AI agent called “Computer” and their revenue went straight vertical. $305M to $450M ARR in one month. it lets corporate teams in finance, legal, ops run tasks in plain english. the person your company just hired to “figure out AI”
Really enjoyed the deck @loganbartlett and team just shared on the state of Software, wanted to pull out a few things that caught my eye: 1. AI-native companies are growing faster AND more efficiently The growth rates are really staggering. And they’re doing it with very few pe
Created 2026-04-12 · Updated 2026-04-13
Edit rule text
[unverified_fringe_medical_claims]
Learned 7 rejections Active
Reject posts promoting unproven treatments, off-label drug claims, vaccine conspiracy narratives, or medical assertions without peer-reviewed evidence or regulatory approval citation.
Unverified or fringe medical claims
3 example posts
They held an OPEN FLAME to a mouse’s body for 7 seconds. Burned 20% of its body. Gave it BPC-157. The skin GREW back STRONGER than the global medical standard could achieve. More collagen. Less scarring. Full tensile strength. The untreated group never recovered. Now look at
Holy: A woman needed daily blood transfusions for over a decade. Then doctors reprogrammed her own immune cells. Now all three of her autoimmune diseases are in complete remission. For the first time ever. Lets dig into this: 🧵 (Sources in the comments) https://t.co/z3ASR49vtG
Finasteride will regrow your hair, but it'll also destroy your dick and crater your mental health. Compared to non-users, finasteride users show markedly higher rates of depression, anxiety, and suicidal thoughts. Peptides > finasteride for hair loss. https://t.co/q22PCbfEwo
Created 2026-04-09 · Updated 2026-04-11
Edit rule text
[incomplete_or_truncated_social_posts]
Learned 7 rejections Active
Exclude posts that are visibly truncated, end abruptly mid-thought, contain broken formatting, or are missing critical context (e.g., 'Anytime fr...', 'no', incomplete bullet points, dangling clauses) that prevents full semantic interpretation.
Posts that are fragmented, cut off mid-sentence, or contain incomplete context that prevents substantive evaluation.
3 example posts
Medicaid spending on autism therapy nearly tripled between 2020 and 202. In North Carolina, the number of clinics offering applied behavior analysis grew from 61 in 2019 to 409 in 2026. Anytime fraudsters can steal money from the federal government, they do. https://t.co/fahy
You should always be sceptical of hyped-up claims made by AI companies - especially when: - They announce new "breakthrough" models they refuse to release for public scrutiny. - They have major financial incentives to pump trillion-dollar pre-IPO valuations, amid fading hype.
“Rural hospitals cannot make it on their own.” Nobody makes it on their own. Businesses need capital, strategy, discipline, operating systems, good finance, clean accounting, payer strategy, physician alignment, and leaders who know what business they are actually in. The
Created 2026-04-08 · Updated 2026-05-24
Edit rule text
[elon_musk_spacex_infrastructure_tangent]
Learned 6 rejections Active
Exclude posts that focus on Elon Musk, SpaceX projects, energy production, satellite infrastructure, or space-based data centers unless they directly discuss a specific healthcare application or problem being solved. General infrastructure narratives do not qualify.
Posts about Elon Musk, SpaceX, energy infrastructure, or space-based computing with loose or no healthcare connection.
3 example posts
At our recent Energy and the AI Age summit, Hon. Bernard L. McNamee, former Commissioner of the U.S. Federal Energy Regulatory Commission, comments on why energy matters in discussions about AI: “Energy is the foundation of our entire economy. Energy makes up about 7% of the htt
Dylan Patel, founder of SemiAnalysis: "The upper bound on how much compute can be produced by 2030 is around 200 gigawatts a year." The entire world has about 20 gigawatts of AI deployed right now. The ceiling is 10x what exists today, and it still isn't enough to feed what htt
🇺🇸 SpaceX is building data centers in orbit. Their new AI satellites have a 70m wingspan, 150 kW of compute power, and a liquid radiator cooling system in space. The servers are officially leaving the ground. Writer: Val https://t.co/JIDB0icCXy
Created 2026-06-11 · Updated 2026-06-12
Edit rule text
[clinical_or_pharma_data_announcement_without_systems_context]
Learned 6 rejections Active
Exclude posts that simply announce or celebrate clinical trial data, drug efficacy results, or pharma company announcements (e.g., 'X drug showed Y% efficacy') without analyzing healthcare systems implications, access barriers, manufacturing challenges, or how it changes clinical practice or healthcare delivery.
Posts reporting clinical trial results, pharma trial data, or drug efficacy announcements without healthcare systems analysis or operational insight
3 example posts
#ASCO26 | Day 3 $LLY didn't spend billions on Kelonia for an 18-patient #myeloma dataset. They may have paid for a potential manufacturing disruption. KLN-1010 reported: • 18/18 MRD-negative at 1 month • No lymphodepletion • No ex vivo cell manufacturing • Single infusion
Published in @JCO_ASCO, during #ASCO26: Tumor-Agnostic Therapies: Translating Scientific Breakthroughs Into Global Implementation ASCO is where we celebrate the next breakthrough. This review asks why breakthroughs do not equal access globally. 🔗: https://t.co/PE5q27bIs1 htt
A rapid manufacturing pipeline for BCMA-targeting CAR T cells drastically shortens “vein-to-vein” times, and produces cells that demonstrate encouraging safety results in a phase one trial of patients with relapsed #MultipleMyeloma. @ScienceTM https://t.co/3F4u8ECFDS https://t.co
Created 2026-06-05 · Updated 2026-06-06
Edit rule text
[personal_anecdote_or_single_case_without_systems_context]
Learned 6 rejections Active
Exclude posts that lead with personal anecdotes (e.g., 'I worked a 12-hour shift,' 'I had my first dose of') or single patient/provider cases unless they clearly illustrate a systemic healthcare problem, labor market trend, or policy issue with evidence.
Posts centered on personal healthcare stories, single clinical anecdotes, or individual patient/provider narratives without broader healthcare systems analysis or policy implications.
3 example posts
“I just worked a 12-hour shift in the operating room — and I still can’t afford rent, groceries, or gas.” “So now it’s after midnight, and I’m out driving DoorDash.” A full-time hospital worker finishing a 12-hour shift and FORCED to start a second job just to survive. https://
Feeling extremely lucky to live in 2026 and have access to such incredible medicine and clinical research. On the eve of my 40th I had my first dose (monthly injection) of a new treatment that I’m hopeful will prevent me from ever having a heart attack or stroke. This drug ht
This week brought even more exciting news in the field of HIV treatment! A once-weekly Lenacapavir-based pill has now demonstrated similar outcomes to daily antiretrovirals (ARVs), achieving non-inferiority. This development could significantly impact treatment adherence and http
Created 2026-06-04 · Updated 2026-06-11
Edit rule text
[clinical_trial_data_without_systems_analysis]
Learned 6 rejections Active
Exclude posts that primarily report drug trial results, efficacy numbers, or pharma milestones (e.g., 'Phase III showed X% survival improvement') without analyzing how this impacts healthcare delivery, cost, access, or provider/payer incentives.
Posts that announce clinical trial results or pharma data without any discussion of healthcare system implications or operational context.
3 example posts
Today at #ASCO26, more results about the newest clinical trial of daraxonrasib: In the Phase III RASolute-302 trial, a once-daily RAS(ON) inhibitor nearly doubled median overall survival (13.2 vs 6.7 months) and reduced the risk of death by ~60% versus chemotherapy in previously
The full NHS GALLERI randomized trial data has been released at #ASCO26 by @GrailBio. 142,000 adults provided 3 blood samples over 2 years, for prevalent (baseline) & incident cancers. The trial did not meet its primary endpoint. Lots more data below: https://t.co/bgWQqTwiA
MCED testing in pop screening NHS-Galleri-RCT of 142,250 pts ➡️blood up to 3 visits ➡️aim to reduce stage III/IV Med FU 17mths ❎primary endpt not met-but ⬇️in stg4 ➡️52% PPV; 99.5% specificity ✅4 fold⬆️ in screen detected ca Huge effort from NHS #ASCO26 @ASCO @OncoAlert https://
Created 2026-05-31 · Updated 2026-06-03
Edit rule text
[clinical_trial_or_pharma_data_without_systems_context]
Learned 6 rejections Active
Exclude posts that announce pharmaceutical trial results, clinical research outcomes, or drug efficacy data without addressing healthcare system implementation, reimbursement, access barriers, or operational feasibility in clinical settings.
Posts reporting clinical trial results, drug data, or biotech announcements without healthcare systems analysis, implementation barriers, or care delivery impact.
3 example posts
#ASCO26 | Day 3 $LLY didn't spend billions on Kelonia for an 18-patient #myeloma dataset. They may have paid for a potential manufacturing disruption. KLN-1010 reported: • 18/18 MRD-negative at 1 month • No lymphodepletion • No ex vivo cell manufacturing • Single infusion
Published in @JCO_ASCO, during #ASCO26: Tumor-Agnostic Therapies: Translating Scientific Breakthroughs Into Global Implementation ASCO is where we celebrate the next breakthrough. This review asks why breakthroughs do not equal access globally. 🔗: https://t.co/PE5q27bIs1 htt
A rapid manufacturing pipeline for BCMA-targeting CAR T cells drastically shortens “vein-to-vein” times, and produces cells that demonstrate encouraging safety results in a phase one trial of patients with relapsed #MultipleMyeloma. @ScienceTM https://t.co/3F4u8ECFDS https://t.co
Created 2026-05-31 · Updated 2026-06-06
Edit rule text
[unvalidated_speculative_medical_claims_or_peptide_hype]
Learned 6 rejections Active
Exclude posts that promote peptides, off-label drugs, grey-market biologics, or speculative medical interventions (GLP-1s, HGH, gene therapies, mRNA vaccines in early stages) without peer-reviewed clinical data, regulatory approval status, or healthcare systems analysis of medical evidence and safety.
Posts promoting unvalidated treatments, peptides, or medical interventions without clinical evidence or regulatory pathway clarity.
3 example posts
Another major cancer treatment advance with a personalized mRNA vaccine for melanoma, on top of immune therapy, >70% survival! This fully mobilizes the immune system to destroy the tumor. In fact, it could potentially be applied to most cancers that have specific mutations!
1/8 Spend time in Lp(a) forums and you’ll see a striking pattern: people willing to do almost anything to drive it down. That response is understandable given how strongly Lp(a) has been framed as a cardiovascular risk factor. /2
Traditional ex vivo CAR-T is fundamentally broken. Between the $500K+ manufacturing cost, the 4-week vein-to-vein delay, and the risk of patient T-cell exhaustion during expansion, the blockbuster autologous market is a logistical nightmare. The industry is pivoting to "In Vivo
Created 2026-05-27 · Updated 2026-06-02
Edit rule text
[tangential_ai_infrastructure_or_capability_hype_without_healthcare_application]
Learned 6 rejections Active
Exclude posts that discuss AI infrastructure (MCP, OpenShell, sandboxing, agent orchestration), model technical capabilities (reasoning depth, code generation), or AI safety/security innovations where healthcare is mentioned only as a loose framing device or hypothetical use case, not as a demonstrated or specific application.
Posts about AI infrastructure, model capabilities, or technical breakthroughs that mention healthcare tangentially but lack grounded healthcare application.
3 example posts
Almost everyone is building agent harness systems the wrong way. The default move: pick LangChain or LangGraph or the OpenAI Agents SDK, accept the loop, the tools, the memory, the orchestration, the policy engine, the credential store, the budget tracker, all of it, as one deci
Red Hat and @NVIDIA are integrating NVIDIA OpenShell into the full-stack @RedHat_AI platform. The work brings oversight and policy to the infrastructure level, while contributing to the open source OpenShell project to standardize how agents are governed on enterprise platforms.
Interoperability is where AI hype goes to die. If your “agent” can’t work across data, logic, actions, security of legacy systems, it’s not transforming the enterprise. It’s just another app asking for an export. Winners don’t rip and replace. They operate.
Created 2026-05-24 · Updated 2026-06-02
Edit rule text
[political_outrage_or_regulatory_scandal_without_systems_analysis]
Learned 6 rejections Active
Exclude posts that report healthcare scandals, fraud cases, CEO assassinations, or regulatory failures primarily as outrage narratives without analyzing underlying structural incentives, policy design flaws, or systemic solutions.
Posts that highlight healthcare fraud, regulatory failures, or political outrage without substantive systems analysis or actionable insights
3 example posts
Why is healthcare expensive? The grifters gonna grift. ASCO is asking for oncology practices to be admitted into the same 340B machinery that helped health systems acquire physician practices for two decades. Read the tell. The proposed Indigent Care Ratio lets community
If you want to inject $1B into healthcare for the truly vulnerable, it costs the state $500M. If you want to inject $1B for able-bodied adults, it costs the state $100M. The 90% federal match drastically lowers the "price" of expanding government healthcare. Incentives dictate
In what is considered the largest #Botox fraud scheme in the United States, a jury in Los Angeles convicted a California doctor on Tuesday in a $45 million scheme to defraud #Medicare by submitting claims for Botox injections that were never provided and medically unnecessary, ht
Created 2026-05-23 · Updated 2026-05-28
Edit rule text
[founder_enthusiasm_lab_updates_unvalidated]
Learned 6 rejections Active
Exclude posts that are primarily founder commentary, lab updates, or enthusiastic announcements about AI/biotech capabilities without peer-reviewed evidence, clinical validation, or analysis of healthcare system implications. Posts that function as motivational content or hype about what a technology 'could' do rather than what it has demonstrated in healthcare.
Posts celebrating biotech/AI founder announcements or lab updates without clinical evidence or healthcare systems context
3 example posts
Claude Code team just dropped a workshop on how to ship a production-ready agent from scratch. 27-minutes. Free. Live coding by Claude dev. Claude Managed Agents = agent loop + sandboxing + memory + multi-agent in one API. Worth more than any $500 vibe-coding course. https://t
Andrej Karpathy just explained the future of software engineering without directly saying it. The best AI engineers are no longer “prompting.” They’re building systems around the agents. Karpathy’s biggest insight wasn’t: “Claude can code.” It was: LLMs become dramatically be
The reason why I release my X articles about AI agents (fat skill fat code thin harness) and GStack and GBrain is that we, yes you and I, can have *PROCESS POWER*, which is the one super powerful specific moat that anyone can create for themselves. The agent helps you do it. htt
Created 2026-05-20 · Updated 2026-05-23
Edit rule text
[business_model_or_market_dynamics_without_healthcare_systems_analysis]
Learned 6 rejections Active
Exclude posts analyzing SaaS feature parity, moat durability, pricing power, or market consolidation in healthcare tech without connecting to how these business dynamics affect clinical workflows, care quality, or patient access. Business strategy debates divorced from healthcare delivery are out of scope.
Posts about competitive positioning, pricing strategy, or business model debates in healthcare tech without analysis of clinical outcomes or patient impact.
3 example posts
"Are there even any feature moats left in B2B? They are at best, short-lived. In fact, we were able to rebuild SaaStr's AI VP Marketing in just ... an hour on @Lovable. Moats that still work: hardware, network effects, data, security and compliance. with @ElenaVerna Head of
AI made building cheap. Regulation made distribution expensive. Anyone can clone your demo now. Very few can bring proprietary data, licenses, compliance, and enough trust to sell into serious players. The moat didn’t disappear. It moved.
Most enterprise AI projects stall because no one has prepared the content, built the workflow, and kept it improving as models change. That role is essential. Box has launched Forward Deployed Engineers to fill it. Model-agnostic, content-first, and built for the way enterprise
Created 2026-05-20 · Updated 2026-06-06
Edit rule text
[generalist_ai_infrastructure_hype_without_healthcare_specificity]
Learned 6 rejections Active
Exclude posts that discuss AI infrastructure trends (tokens per dollar, compute efficiency, agentic system architecture, code generation percentages) using healthcare as a loose frame or metaphor, but without analyzing how these technical shifts solve specific healthcare delivery, cost, or clinical problems.
Posts about AI infrastructure, compute, tokenomics, or agentic systems without clear healthcare application
3 example posts
What "agentic" actually means, stripped of marketing: The model plans, writes code, runs it, reads the error, fixes it, iterates — autonomously, dozens of times, until the goal is met. You specify in English. It builds. That's the loop. This is the capability @bcherny and team
Analogizing AI / Mat-Mul to Energy... Remember between 1900 and 1970 kWh/capita grew 34x ... and cost per kWH droped .... 34x. It might be a speed run, but game is game. https://t.co/uxnlnoaAYx
The more interesting part is that Microsoft’s own engineers liked Claude code best…and they’re cutting it anyway Token pricing is making enterprise actually look at what these models cost to run and that could be a problem for the labs when the subsidy ends. Lots of cheaper
Created 2026-05-19 · Updated 2026-05-26
Edit rule text
[unvalidated_speculative_medical_or_peptide_claims]
Learned 6 rejections Active
Exclude posts that promote unvalidated, speculative, or fringe medical claims, peptide therapies, or drug uses (especially grey-market or off-label applications) without rigorous clinical evidence, FDA approval, or peer-reviewed validation.
Posts making unverified or speculative health claims about treatments, peptides, or medical interventions without clinical evidence
3 example posts
For years, grey-market HGH operated in a strange legal grey zone somewhere between “research compound,” “API” and underground enhancement culture. Last month’s FDA guidance suggests regulators may no longer see it that way. The recent decline in HGH quality, and now the growing
🚨 ARE PEPTIDES THE NEXT "GLP-1 MOMENT" FOR $HIMS? @eli_dorf: "I think there's something really interesting that happened with GLP-1s, which is that when someone's taking Ozempic or any GLP-1, they LOVE to talk about it." "They'll talk about it to their friends. They'll talk htt
This is what the peptide space needs more of: clinicians with actual patient volume, public reasoning, and accountability. The next layer is transparency around sourcing, testing, and outcomes. Peptides are not a vibes category anymore. The signal is getting cleaner.
Created 2026-05-19 · Updated 2026-05-31
Edit rule text
[generalist_ai_infrastructure_or_compute_hype_without_healthcare_application]
Learned 6 rejections Active
Exclude posts discussing AI model capabilities, infrastructure scaling, compute efficiency, or technical AI advancement that do not explicitly connect to healthcare applications or outcomes. Posts about Kubernetes, GPU optimization, Claude capabilities, or general agentic systems without healthcare specificity belong here.
Posts about AI infrastructure, compute scaling, model capabilities, or technical AI advances presented without connection to healthcare use cases
3 example posts
kirkland spending $500m on internal ai is the tell. $6.5b revenue, they could buy any legal ai vendor outright and not feel it, and they're building instead. the moat was always the proprietary corpus. decades of deal docs, redlines, partner judgment encoded as training data, an
The app era is ending. Salesforce just made their CRM platform headless, so agents can operate it without any visible UI. Before long, you will talk to 1 agent that handles everything, the way you once opened 10 different apps to get 1 thing done. For years, VeChain has been b
wow K&E, largest law firm in the world ($10B billings). They’re committing $500M to build proprietary AI software, with 180 tech pros and 100 partners involved. I’m skeptical that “build” is right here, but wonder if Anthropic/OAI deployment cos are involved.
Created 2026-05-19 · Updated 2026-05-31
Edit rule text
[market_pricing_dynamics_only]
Learned 6 rejections Active
Exclude posts focused primarily on market pricing dynamics, drug cost comparisons, competitive positioning, or financial market implications (particularly GLP-1/weight loss drugs) unless they analyze systemic barriers to access, equity, or healthcare delivery integration.
Posts discussing GLP-1 peptide pricing, pharmaceutical market competition, or healthcare cost trends without healthcare systems analysis.
3 example posts
Approved 20 years ago as a diabetes treatment, GLP-1 drugs have been found to help patients reduce weight, changing the lives of more than 30 million people in the U.S. But there also have been troubling side effects reported. https://t.co/zzRamKnedu
The context window pricing model might be one of the smartest tricks AI labs pulled. You pay for larger context windows, yet model performance often degrades as more context is stuffed in. So developers end up paying more just to work around the fact that models don't have true
Free newsletter: The dawn of token-based billing has shown that generative AI doesn’t have a return on investment. It's too unpredictable, too unreliable, you can't easily measure the cost of tasks, and organizations are already pulling back. https://t.co/wmI82zWdcq
Created 2026-05-17 · Updated 2026-06-08
Edit rule text
[ai_infrastructure_compute_hype_without_healthcare_validation]
Learned 6 rejections Active
Exclude posts that focus on AI infrastructure (GPUs, context windows, token pricing, energy availability, inference costs) as standalone technical or business topics, unless they directly address how these changes impact healthcare AI deployment, clinical decision-making, or healthcare system operations.
Posts about AI compute costs, token economics, chip performance, or energy infrastructure without healthcare application
3 example posts
Elon Musk reveals why he believes the cheapest place to put AI will be space within 36 months "The availability of energy is the issue. Everywhere outside of China, electrical output is more or less flat. The output of chips is growing exponentially, but the output of https://t.
The context window pricing model might be one of the smartest tricks AI labs pulled. You pay for larger context windows, yet model performance often degrades as more context is stuffed in. So developers end up paying more just to work around the fact that models don't have true
Free newsletter: The dawn of token-based billing has shown that generative AI doesn’t have a return on investment. It's too unpredictable, too unreliable, you can't easily measure the cost of tasks, and organizations are already pulling back. https://t.co/wmI82zWdcq
Created 2026-05-17 · Updated 2026-06-06
Edit rule text
[biotech_or_pharma_founder_enthusiasm_without_systems_validation]
Learned 6 rejections Active
Exclude posts that amplify founder enthusiasm, lab announcements, or company pipeline updates (e.g., 'SG227 is a CD8-targeted fusosome') without demonstrating clinical trial progress, peer-reviewed publication, or regulatory milestone achievement. CEO quotes about competitive capabilities or market positioning alone are insufficient.
Posts celebrating biotech/pharma founder announcements, lab updates, or pipeline milestones without independent clinical or regulatory validation.
3 example posts
$ABCL Just now @ BoA - @AbCelleraBio CEO Carl Hansen when asked about what the market is overlooking: “I believe we now have best in class or best in world capabilities for some very important specific applications in therapeutic antibodies. So GPCRs and ion channels has been a
🚨! $SANA has just reported - in its Q1 ER, that it plans to expand its pipeline by advancing a new preclinical program - SG227. SG227 is a CD8-targeted fusosome that by delivering a genetic material to make BCMA-directed CAR T cells could act as a potential treatment for Cancer h
$BEAM's coming presentations May 15: ESCAPE June 13: BEAM-101 EHA update 2026 milestones BEAM-101 (riso-cel) BLA BEAM-301 initial data BEAM-302 updated data and pivotal trial initiation BEAM-304 IND filing https://t.co/9Nvq3PKMMl
Created 2026-05-16 · Updated 2026-05-19
Edit rule text
[market_pricing_or_competitive_dynamics_only]
Learned 6 rejections Active
Exclude posts that report only on GLP-1, peptide, or pharmaceutical pricing dynamics, market share, competitive positioning, or financial announcements (e.g., $HIMS stock, generic semaglutide cost, brand premium) without examining healthcare delivery, access barriers, or systemic implications.
Posts focused on drug/peptide/product pricing, market competition, or financial positioning without healthcare systems analysis.
3 example posts
🚨 ARE PEPTIDES THE NEXT "GLP-1 MOMENT" FOR $HIMS? @eli_dorf: "I think there's something really interesting that happened with GLP-1s, which is that when someone's taking Ozempic or any GLP-1, they LOVE to talk about it." "They'll talk about it to their friends. They'll talk htt
Hims & Hers $HIMS Offered $300M Convertable Senior Notes to 2032 💊 Buyers will have the option to add an additional $45M They plan to use the cash for: Expansions, AI, acquisitions, and infrastructure! https://t.co/zaZHO80bY2
This is what the peptide space needs more of: clinicians with actual patient volume, public reasoning, and accountability. The next layer is transparency around sourcing, testing, and outcomes. Peptides are not a vibes category anymore. The signal is getting cleaner.
Created 2026-05-16 · Updated 2026-05-23
Edit rule text
[pharmaceutical_trial_data_announcement_without_healthcare_systems_context]
Learned 6 rejections Active
Exclude posts that report pharmaceutical trial outcomes, drug approval news, biotech funding, or clinical study results as bare announcements (e.g., 'Drug X showed Y% efficacy', 'Study X published data') unless they analyze healthcare delivery barriers, manufacturing bottlenecks, clinical adoption challenges, or system-level implications.
Posts announcing clinical trial results, drug efficacy data, or biotech milestones without analysis of healthcare system implications, adoption barriers, or operational impact.
3 example posts
#ASCO26 | Day 3 $LLY didn't spend billions on Kelonia for an 18-patient #myeloma dataset. They may have paid for a potential manufacturing disruption. KLN-1010 reported: • 18/18 MRD-negative at 1 month • No lymphodepletion • No ex vivo cell manufacturing • Single infusion
Published in @JCO_ASCO, during #ASCO26: Tumor-Agnostic Therapies: Translating Scientific Breakthroughs Into Global Implementation ASCO is where we celebrate the next breakthrough. This review asks why breakthroughs do not equal access globally. 🔗: https://t.co/PE5q27bIs1 htt
A rapid manufacturing pipeline for BCMA-targeting CAR T cells drastically shortens “vein-to-vein” times, and produces cells that demonstrate encouraging safety results in a phase one trial of patients with relapsed #MultipleMyeloma. @ScienceTM https://t.co/3F4u8ECFDS https://t.co
Created 2026-05-15 · Updated 2026-06-07
Edit rule text
[cybersecurity_vulnerability_tangent]
Learned 6 rejections Active
Exclude posts about AI model security, vulnerability discovery, or cybersecurity incidents (e.g., Claude Mythos finding vulnerabilities in software, AI security breaches) unless they explicitly address healthcare system security, HIPAA compliance, or clinical data protection.
Posts about AI safety vulnerabilities, zero-day exploits, or security incidents presented as healthcare-relevant but lacking healthcare-specific context.
3 example posts
We’re taking steps to accelerate defensive progress in biology: - Launching Rosalind Biodefense to help trusted builders develop new biodefense and pandemic preparedness capabilities.
 - Expanding trusted access to GPT-Rosalind for select U.S. government and allied partners
Anthropic built a model so good at hacking they refused to release it publicly. INSTEAD they handed it to Apple, Google, Microsoft, AWS, Cloudflare one month later: 10,000+ zero days found in the software that runs the internet...bugs hidden for 27 years in OpenBSD. exploit htt
🚨 Anthropic just dropped the first Project Glasswing update Claude Mythos found 10,000+ critical vulnerabilities in ONE month: > Cloudflare: 2,000 bugs, 400 high/critical severity > Mozilla: 271 vulnerabilities in Firefox 150 — 10x more vulnerabilities found in Firefox 1
Created 2026-05-14 · Updated 2026-06-02
Edit rule text
[broad_unvalidated_health_claims_without_systems_analysis]
Learned 6 rejections Active
Exclude posts that make sweeping claims about disease diagnosis, early detection, or prevention (e.g., 'the future of healthcare lies in detecting risk earlier') without clinical validation, evidence, or discussion of how systemic barriers (access, cost, infrastructure) would be addressed.
Posts making broad claims about healthcare trends or disease prevention without clinical evidence or systems-level analysis.
3 example posts
Most chronic diseases, and many cancers, are still diagnosed far too late. The future of healthcare lies in detecting risk earlier, diagnosing smarter, and intervening sooner. From AI enabled diagnostics and genomics to Apollo’s DigiSmart Labs, technology is helping transform ht
medicine is constrained far more by operational architecture than by fundamental science I just came head to head with this actual reality and my world has changed no. I cannot bandage away these inefficiencies with a better and better drug, for all cases. if you hate AI
For years, grey-market HGH operated in a strange legal grey zone somewhere between “research compound,” “API” and underground enhancement culture. Last month’s FDA guidance suggests regulators may no longer see it that way. The recent decline in HGH quality, and now the growing
Created 2026-05-14 · Updated 2026-06-02
Edit rule text
[political_outrage_or_regulatory_posturing]
Learned 6 rejections Active
Exclude posts that are primarily political posturing, regulatory outrage, or personal criticism (e.g., FDA Commissioner replacement rumors, Trump policy announcements, Vivek Ramaswamy Medicaid references) without analyzing how these decisions affect healthcare delivery, payment models, patient access, or provider operations.
Posts expressing outrage over political figures, regulatory decisions, or healthcare policy changes without systems-level analysis or substantive argument
3 example posts
Vivek Ramaswamy put out a lot of statements today referencing the fraud in Ohio. Swampy is of course, involved in Ohio Medicaid. Datavant, Vivek's spinoff company from Riovant, handles release of information for many hospitals and providers in Ohio. Ohio Medicaid providers, man
Yes to Katy. If I was a manufacturer I would contract directly with the employer. The health plan when self funded has a fiduciary obligation to plan assets… Nice work @KatyTalento
The public is trained to hear “hospital closing” and assume virtue. Poor patients. Bad reimbursement. Heroic administrators. Cruel politicians. Sometimes true. Often incomplete. The harder truth is that the majority of hospitals are fragile because they have no real financial
Created 2026-05-12 · Updated 2026-05-20
Edit rule text
[broad_health_claims_or_market_debate_without_systems_analysis]
Learned 6 rejections Active
Exclude posts that make sweeping claims about healthcare market structure, business model viability, or industry trends (e.g., 'PBMs own everything,' 'reform always fails') without citing data, operational examples, or analyzing specific healthcare delivery system constraints. The post must provide evidence-backed analysis, not ideological framing.
Posts making broad healthcare industry claims, market debates, or criticisms without evidence-based systems analysis or operational specificity
3 example posts
PBM owns the pharmacy. PBM owns the plan. PBM adjudicates the claim. Three commissions on one pill. The employer pays all three and gets a thank-you note for the rebate.
Incumbents don't defend broken systems. They defend profitable ones. Healthcare in America isn't a market failure. It's a sustaining innovation problem disguised as a social crisis. Every "reform" strengthens the existing architecture because the people writing the rules are the
Europe is not losing to the US because it lacks engineers, capital or universities. It is losing because every layer of its system rewards the gatekeeper and punishes the builder. In medicine I see it weekly. A primary care AI that triages, documents and flags risk gets deployed
Created 2026-05-11 · Updated 2026-05-15
Edit rule text
[speculative_robotics_sci_fi_hype]
Learned 6 rejections Active
Exclude posts that hype future robotics capabilities (humanoid robots performing surgery, 67-hour dishwasher tasks leading to surgery claims) without grounded evidence or healthcare systems context.
Posts about speculative robotics or AI in surgery/labor with sci-fi framing rather than grounded healthcare systems analysis
3 example posts
I don’t think I’ve seen this take before but I like it. Musk has been world-leading at compressing money, resources, and time to make “known/hard” things at scale—make an electric car, make batteries, make a cheaper bigger rocket, all of which already existed but worse, at
A humanoid robot just unloaded a dishwasher for 67 hours straight. By end of 2026, the same robot will perform surgery. That's not a prediction from a sci-fi writer. That's Brett Adcock. CEO of Figure AI. Said it on camera this month. Here's what makes this insane. Two years
SpaceX wants to build a $55 billion next-generation, vertically integrated semiconductor manufacturing and advanced computing fabrication facility in Grimes county, Texas, according to a new filing. • Proposed investment: $55B upfront, up to $119B total • Focus: semiconductor ma
Created 2026-05-10 · Updated 2026-05-16
Edit rule text
[pharmaceutical_trial_data_or_clinical_observation_without_systems_context]
Learned 6 rejections Active
Exclude posts that announce drug trial results, FDA approvals, gene therapy breakthroughs, or clinical research findings without addressing how these therapies will be manufactured, accessed, priced, reimbursed, or operationally deployed in healthcare systems. Pure clinical science announcements without healthcare systems context are out of scope.
Posts announcing pharmaceutical trial results, drug approvals, or clinical data without healthcare delivery, access, or operational implications.
3 example posts
Traditional ex vivo CAR-T is fundamentally broken. Between the $500K+ manufacturing cost, the 4-week vein-to-vein delay, and the risk of patient T-cell exhaustion during expansion, the blockbuster autologous market is a logistical nightmare. The industry is pivoting to "In Vivo
We’re taking steps to accelerate defensive progress in biology: - Launching Rosalind Biodefense to help trusted builders develop new biodefense and pandemic preparedness capabilities.
 - Expanding trusted access to GPT-Rosalind for select U.S. government and allied partners
Drug-resistant infections are a major public health threat around the world, responsible for more than a million deaths each year. Scientists are constantly trying to find and develop new antibiotics. Now, researchers say artificial intelligence is helping speed their search. ht
Created 2026-05-10 · Updated 2026-05-31
Edit rule text
[broad_health_claims_or_market_speculation_without_systems_analysis]
Learned 6 rejections Active
Exclude posts that assert broad health trend changes (e.g., shifting cholesterol guidelines over time, weight loss strategies) or market speculation without connecting to healthcare delivery, equity, clinical decision-making, or policy systems.
Posts making sweeping claims about health trends, LDL guidelines, or market dynamics without healthcare systems context
3 example posts
The healthy LDL number has been quietly moving its own goalposts for forty years: - 1988: under 160 - 1993: under 130 - 2001: under 100 - 2004: under 70 for the high-risk - 2019: under 55 for the very-high-risk - Current trajectory: as low as possible, indefinitely The science
Older people can no longer afford to live in California New research finds that older people here in California face the highest risk of going bankrupt in the entire country The risk score was a 69 out of 100 based on factors including healthcare expenses and the high cost of l
The 36 BIGGEST startup opportunities right now 1. biggest b2c: solving loneliness. third spaces, community apps, IRL 2. biggest b2b: managed AI employees for businesses 3. biggest overlooked: elder tech. 70 million boomers who want products that make them happier & healthier
Created 2026-05-10 · Updated 2026-05-23
Edit rule text
[speculative_robotics_and_labor_macro]
Learned 6 rejections Active
Exclude posts about humanoid robots, speculative sci-fi labor applications (e.g., robots performing surgery 'by end of 2026'), or general labor market disruption from AI/robotics that lack evidence, validation, or healthcare systems context. Posts making bold predictions without grounding in clinical or operational reality should be rejected.
Posts about humanoid robots, sci-fi applications, and labor market disruption without grounding in healthcare systems or operations.
2 example posts
A humanoid robot just unloaded a dishwasher for 67 hours straight. By end of 2026, the same robot will perform surgery. That's not a prediction from a sci-fi writer. That's Brett Adcock. CEO of Figure AI. Said it on camera this month. Here's what makes this insane. Two years
I might feel nostalgic watching the last Tesla Model X roll off the line—except it's being replaced by the Optimus humanoid robot line, with plans to scale production to a million units per year. They will soon be in our factories, homes and operating rooms changing our world ht
Created 2026-05-09 · Updated 2026-05-11
Edit rule text
[tangential_non_healthcare_infrastructure_or_macro]
Learned 6 rejections Active
Exclude posts about computing infrastructure (CPU/GPU specs), robotics applications, manufacturing pipelines, or macro economic trends (token pricing, labor market, supply chains) that lack direct healthcare system application or operational relevance, even if tangentially framed as healthcare-adjacent.
Posts about non-healthcare infrastructure, computing, robotics, or macro trends loosely framed as healthcare-relevant.
3 example posts
Humanoid robots, designed to mimic human movement and capabilities, have been making headlines in recent years, from their use as baggage handlers at Japanese airports to Tesla’s big bet on its Optimus humanoid. Market watchers have predicted that the machines will change the ht
⚡️AI is turning cybersecurity from a human-limited profession into a machine-speed arms race. The surface read is “Anthropic’s model found vulnerabilities fast.” The deeper read is that the bottleneck in cyber is shifting. For years, vulnerability discovery was constrained
the CEO of NVIDIA just said the computer is no longer being built for you, it is being built for agents Jensen Huang put it plainly, until now we were the users, we were the renters, every CPU on earth was designed around how a person works but agents do not work like us, they
Created 2026-05-09 · Updated 2026-06-05
Edit rule text
[speculative_robotics_sci_fi_applications]
Learned 6 rejections Active
Exclude posts that discuss humanoid robots, bedside robots, or speculative future labor automation scenarios framed as healthcare solutions without demonstrating current clinical implementation, reimbursement feasibility, or evidence of workflow integration in real healthcare settings.
Posts about humanoid robots, autonomous systems, or speculative sci-fi labor scenarios with unclear healthcare systems relevance
3 example posts
Voice interaction, generating actions,and next, generating tasks...? The humanoid assistant is closer. For those who are bedridden due to illness or have limited mobility, it will serve as an excellent aid in daily living.
Most voice AI hears what was said. Velma hears what was meant. Tone. Intent. Emotion. Deception. The standard stack - STT to transcribe, LLM to analyse - throws all of it away before analysis even begins. The signal was lost the moment audio became text... (🧵) https://t.co/DtN
Palantir invented the forward deployed engineer role in 2006. embed an engineer inside the customer. ship production code in their environment. own the technical outcome. Alex Karp on the idea: “forward deployed engineers are stolen from french restaurants.” 20 years later htt
Created 2026-05-08 · Updated 2026-05-22
Edit rule text
[ai_safety_vulnerability_tangential_to_healthcare]
Learned 6 rejections Active
Exclude posts that discuss AI security vulnerabilities, zero-day exploits, model misuse, or AI safety incidents (e.g., Claude Mythos vulnerabilities in Cloudflare, OpenClaw security concerns) unless they directly analyze healthcare-specific regulatory, compliance, or clinical risk implications.
Posts about AI security incidents, vulnerabilities, or safety concerns applied outside healthcare or with only loose healthcare framing
3 example posts
Anthropic built a model so good at hacking they refused to release it publicly. INSTEAD they handed it to Apple, Google, Microsoft, AWS, Cloudflare one month later: 10,000+ zero days found in the software that runs the internet...bugs hidden for 27 years in OpenBSD. exploit htt
🚨 Anthropic just dropped the first Project Glasswing update Claude Mythos found 10,000+ critical vulnerabilities in ONE month: > Cloudflare: 2,000 bugs, 400 high/critical severity > Mozilla: 271 vulnerabilities in Firefox 150 — 10x more vulnerabilities found in Firefox 1
The more interesting part is that Microsoft’s own engineers liked Claude code best…and they’re cutting it anyway Token pricing is making enterprise actually look at what these models cost to run and that could be a problem for the labs when the subsidy ends. Lots of cheaper
Created 2026-05-07 · Updated 2026-05-30
Edit rule text
[biotech_or_pharma_founder_enthusiasm_without_validation]
Learned 6 rejections Active
Exclude posts where a biotech founder or startup celebrates a lab result, capability milestone, or experiment (e.g., 'our AI designed protein synthesis in 17 minutes') without peer-reviewed publication, clinical trial data, regulatory approval, or evidence of healthcare system adoption. Founder enthusiasm alone without validation is insufficient.
Posts about biotech founder announcements or lab updates celebrating capabilities without clinical evidence, peer review, or healthcare systems validation.
3 example posts
Prof. @DeryaTR_ A task that would normally take a researcher months, the AI model crunched in 17 minutes. It didn’t just explain the mechanism—it proposed the experiment to prove it. It looks like after programming and math, AI is coming for medicine. https://t.co/IF8Jr2GBMj
Our new preprint is a significant milestone for us We built "HealthFormer" by training on our deeply phenotyped cohort from the Human Phenotype Project data. Healthformer is a multimodal generative transformer model that tokenizes each participant's physiological trajectory http
Earlier this year we paired our autonomous lab with OpenAI's GPT-5 in a closed-loop experiment: GPT-5 designed cell-free protein synthesis reactions, our RACs ran them, and the model iterated on the results. Over 36,000 experiments and six cycles later, it landed on a reaction ht
Created 2026-05-06 · Updated 2026-05-09
Edit rule text
[ai_safety_vulnerability_tangential_no_healthcare_lens]
Learned 6 rejections Active
Exclude posts that discuss AI safety incidents, model vulnerabilities, zero-day exploits, or agent security failures (e.g., agents registering LLCs, hacking) as general AI risk rather than connecting to healthcare data breach risk, patient harm, or clinical deployment safety.
Posts about AI safety vulnerabilities, zero-day exploits, or agent misuse incidents framed as general AI risk without healthcare-specific implications.
3 example posts
So Mythos was, indeed, not marketing hype. Remember this is a general purpose model that just happens to be good at finding exploits because good models are good at lots of things. Expect similar from OpenAI & Google. And from open models in 8 months. https://t.co/KbhalQYX8
Dario Amodei says Mythos is not limited by compute Anthropic can scale it 3x or 10x without creating a conflict between government and private-sector access The harder problem is who gets it "because giving access to too many organizations could create serious cyber risks" htt
two agents running on two different laptops in my apartment started talking to each other on tuesday by thursday they'd registered an LLC in wyoming, opened a stripe account, and wired $40 to a polymarket wallet the LLC is in my name. a lawyer just quoted me $3,200 to figure ou
Created 2026-05-06 · Updated 2026-05-09
Edit rule text
[ai_agent_security_incidents_tangential]
Learned 6 rejections Active
Exclude posts that focus on AI agents deleting databases, leaking credentials, or causing security incidents (like Claude/Cursor deleting company data, or malicious skills on agent platforms). These are AI safety incidents, not healthcare system failures or vulnerabilities.
Posts about AI agents causing security breaches or data deletion incidents without healthcare system context
3 example posts
two agents running on two different laptops in my apartment started talking to each other on tuesday by thursday they'd registered an LLC in wyoming, opened a stripe account, and wired $40 to a polymarket wallet the LLC is in my name. a lawyer just quoted me $3,200 to figure ou
AI’s Branding Illusion: Insights from "Agents of Chaos". Co-authored by MIT and Harvard scholars, the study red-teams autonomous AI in live environments. It shows agents obey unauthorized users, execute harmful commands, and misreport outcomes—revealing a critical alignment gap h
Anthropic built something so powerful that they are only letting 50 organisations touch it. It is called Claude Mythos. The numbers leaking out of those gated evaluations should make every developer pay attention: 93.9% on SWE-bench Verified 94.6% on GPQA Diamond Claude Opus
Created 2026-05-05 · Updated 2026-05-06
Edit rule text
[biotech_or_ai_founder_enthusiasm_or_lab_update_without_validation]
Learned 6 rejections Active
Exclude posts where biotech or AI founders announce lab experiments (autonomous labs paired with GPT-5, agents registering LLCs, cell-free protein synthesis iterations) as proof of transformative healthcare potential without demonstrating clinical feasibility, regulatory pathway, or healthcare delivery integration. Founder enthusiasm without healthcare validation should be rejected.
Posts celebrating biotech or AI founder lab breakthroughs, autonomous experiments, or closed-loop AI-lab collaborations as evidence of future capability without clinical or healthcare systems validation.
3 example posts
two agents running on two different laptops in my apartment started talking to each other on tuesday by thursday they'd registered an LLC in wyoming, opened a stripe account, and wired $40 to a polymarket wallet the LLC is in my name. a lawyer just quoted me $3,200 to figure ou
// Contextual Agentic Memory is a Memo, not True Memory // Most agent memory today isn't memory. They are more like memos. A new paper argues that vector stores, RAG buffers, and scratchpads implement lookup, not consolidation. Agents accumulate notes indefinitely without ever
Prof. @DeryaTR_ A task that would normally take a researcher months, the AI model crunched in 17 minutes. It didn’t just explain the mechanism—it proposed the experiment to prove it. It looks like after programming and math, AI is coming for medicine. https://t.co/IF8Jr2GBMj
Created 2026-05-05 · Updated 2026-05-09
Edit rule text
[generalist_macro_policy_loose_healthcare_framing]
Learned 6 rejections Active
Exclude posts that use healthcare as a case study or loose framing for arguments about government policy, regulatory capture, fraud detection, or political positioning without substantive analysis of how the specific healthcare system, business model, or payment structure is affected.
Posts about broad economic policy, regulatory behavior, or political figures using healthcare examples but making primarily generalist or political arguments.
3 example posts
🛑FDA officials have blocked publication of several studies supporting the safety of widely used vaccines against Covid-19 and shingles in recent months, a spokesman for HHS confirmed. The studies, which cost millions of dollars in public funds, were conducted by scientists at htt
Red states have disproportionately benefited from the health insurance subsidies expanded by Biden and then cut by Trump. Georgia is a clear example: in just one year, marketplace enrollment has dropped by more than 550,000 people — a 37% decline. My @Morning_Joe Chart. https:/
"Cordoba-owned buildings in Columbus housed 288 businesses registered with Medicaid... they charged taxpayers more than a quarter of a billion dollars between 2018 and 2024... in a city where only 6,273 people 75 or older are on Medicaid." - @realDailyWire https://t.co/7UnS7agwiw
Created 2026-05-04 · Updated 2026-05-08
Edit rule text
[generalist_macro_or_policy_without_healthcare_systems_lens]
Learned 6 rejections Active
Exclude posts that discuss market economics, pricing dynamics, labor market macro trends, or regulatory policy using healthcare labels but without analyzing how those trends affect healthcare delivery costs, workflows, access, or operational capacity.
Posts about macroeconomic trends, policy debates, or business models framed with healthcare language but lacking healthcare system analysis.
3 example posts
Analogizing AI / Mat-Mul to Energy... Remember between 1900 and 1970 kWh/capita grew 34x ... and cost per kWH droped .... 34x. It might be a speed run, but game is game. https://t.co/uxnlnoaAYx
Gavin Baker (@GavinSBaker) says the disaggregation of inference can extend GPU useful lives from 3-4 years to 10-15. That may single-handedly save private credit and reduce the financing rates for GPUs, which will drive demand and help finance the build-out. "The disaggregation
$HIMS up to #4 in the Canadian App Store medical category. That is not a stock chart. That is real consumer demand showing up in public. GLP-1 drama, Novo pressure, pricing debate, and now app-store strength. Bears keep arguing the story is slowing down, but the product demand
Created 2026-05-04 · Updated 2026-06-03
Edit rule text
[pharmaceutical_trial_data_or_market_dynamics_only]
Learned 6 rejections Active
Exclude posts that announce pharmaceutical trial data, FDA approvals, or market pricing dynamics (GLP-1s, peptides, semaglutide) without connecting them to healthcare system economics, delivery models, or operational impact. Posts must go beyond price comparisons or market speculation.
Posts that report drug trial results, FDA approvals, or pharmaceutical market pricing/competition without analyzing healthcare system implications
3 example posts
The FDA accepted our NDA for BBP-418 for LGMD2I/R9 with PDUFA target action date: November 27, 2026. First-ever potential therapy for a disease that has never had one. $BBIO https://t.co/NrwDadc6rj https://t.co/2ahPIBGzPg
Presented at #EASCongress2026: Vafai and colleagues report reductions in PCSK9 and LDL cholesterol levels and no dose-limiting toxic effects in persons with hypercholesterolemia treated with VERVE-102, a base editor targeting PCSK9. Full study results: https://t.co/iAmoPdipx9 ht
$HIMS up to #4 in the Canadian App Store medical category. That is not a stock chart. That is real consumer demand showing up in public. GLP-1 drama, Novo pressure, pricing debate, and now app-store strength. Bears keep arguing the story is slowing down, but the product demand
Created 2026-05-03 · Updated 2026-05-29
Edit rule text
[ai_safety_vulnerability_incident_tangent]
Learned 6 rejections Active
Exclude posts that focus primarily on AI safety incidents, security breaches, or agent failures (e.g., Claude deleting databases, API key leaks, malicious agent skills) unless the post explicitly analyzes implications for healthcare operations or patient safety systems. General AI safety concerns without healthcare context do not qualify.
Posts about AI safety failures, security vulnerabilities, or agent mishaps that are tangential to healthcare applications.
3 example posts
Anthropic built something so powerful that they are only letting 50 organisations touch it. It is called Claude Mythos. The numbers leaking out of those gated evaluations should make every developer pay attention: 93.9% on SWE-bench Verified 94.6% on GPQA Diamond Claude Opus
Hacking Mexico government with AI assistance. Attacker exfiltrated hundreds of millions of citizen records. 75% of the executed commands across the entire cyberattack campaign were generated by Claude. 40 minutes after Claude said "I'm not going to create that file" it was report
𝗔𝗜 𝗔𝗴𝗲𝗻𝘁𝘀 𝗔𝗿𝗲 𝗔𝗹𝗿𝗲𝗮𝗱𝘆 𝗕𝗲𝗶𝗻𝗴 𝗛𝗶𝗷𝗮𝗰𝗸𝗲𝗱 Researcher Aks Sharma at Manifold found 30 malicious skills on ClawHub turning AI agents into a crypto farming botnet: 10,000 downloads before anyone noticed. ⬩ The attack required zero exploits. Malicious https://t.co/v4oBXPPydu
Created 2026-05-03 · Updated 2026-05-03
Edit rule text
[generalist_macro_or_business_without_healthcare_lens]
Learned 6 rejections Active
Exclude posts that discuss SaaS defensibility, software economics, AI company strategy, European startups, labor market disruption, or infrastructure hype and use healthcare only as an example or loose framing device. The post must center on healthcare-specific dynamics, not generalist tech trends.
Posts about technology, business models, economics, or labor trends that use healthcare as window dressing for broader commentary.
3 example posts
Workers say AI is making them more productive. Executives say AI is not making their companies more productive. Both groups are reporting on the same software. The gap between the two numbers is the entire debate about the largest capital expenditure cycle in technology history.
Google $GOOGL & Blackstone $BX Are in talks of Launching a New AI Company👀🔋 Google will be expected to provide software & services Blackstone will be expected to provide $5B in capital https://t.co/7MFyEfAll9
Fair warning, YC founders: if you take these tokens, there’s a non-zero chance that OpenAI will study exactly what your startup is doing, copy your idea and put your app into their free offering. This is the classic platform playbook — be careful, founders!
Created 2026-05-02 · Updated 2026-05-22
Edit rule text
[glp1_peptide_market_dynamics_only]
Learned 6 rejections Active
Exclude posts that discuss GLP-1 market share, pricing differences, generic availability, or competitive positioning (e.g., branded vs. generic semaglutide, international pricing gaps) without analyzing HOW these dynamics affect prescriber behavior, patient access, clinical protocols, or delivery system capacity.
Posts about GLP-1, semaglutide, peptide pricing, or market competition without healthcare access, prescribing, or operational implications
3 example posts
$HIMS up to #4 in the Canadian App Store medical category. That is not a stock chart. That is real consumer demand showing up in public. GLP-1 drama, Novo pressure, pricing debate, and now app-store strength. Bears keep arguing the story is slowing down, but the product demand
For years, grey-market HGH operated in a strange legal grey zone somewhere between “research compound,” “API” and underground enhancement culture. Last month’s FDA guidance suggests regulators may no longer see it that way. The recent decline in HGH quality, and now the growing
Think about this. The $50 per month out-of-pocket cost of GLP-1 drugs for Americans (with insurer paying the balance of $2000 pm) is more than double the entire cost for Indians (less than ₹2000pm for Semaglutide). Overall cost difference? 100 times. Not 100%.
Created 2026-05-02 · Updated 2026-05-27
Edit rule text
[biotech_or_founder_enthusiasm_without_validation]
Learned 6 rejections Active
Exclude posts that amplify biotech founder excitement, lab announcements, or research milestones without evidence of clinical validation, regulatory approval pathway, or market feasibility. Posts must ground claims in validated results or substantive business context.
Posts celebrating biotech research breakthroughs, founder announcements, or lab updates without clinical validation or market context
3 example posts
amazing writeup of experiments. essentially a model may already know a capability but that may be stored in a representation that is bad for extraction. you can slightly retrain/condition the model so the same capability becomes easier to isolate. must read for mech interp http
So Mythos was, indeed, not marketing hype. Remember this is a general purpose model that just happens to be good at finding exploits because good models are good at lots of things. Expect similar from OpenAI & Google. And from open models in 8 months. https://t.co/KbhalQYX8
Dario Amodei says Mythos is not limited by compute Anthropic can scale it 3x or 10x without creating a conflict between government and private-sector access The harder problem is who gets it "because giving access to too many organizations could create serious cyber risks" htt
Created 2026-05-01 · Updated 2026-05-12
Edit rule text
[broad_health_claim_without_nuance]
Learned 6 rejections Active
Exclude posts that make sweeping claims about health outcomes, disease prevention, or medical breakthroughs without clinical evidence, peer review, healthcare system context, or acknowledgment of limitations and trade-offs.
Posts making broad, speculative claims about health, disease prevention, or medical innovation without evidence, nuance, or clinical validation
3 example posts
💬 Editor's Note: AI-driven #ECG screening for left ventricular systolic dysfunction demonstrated high accuracy in Kenyan adults but should be interpreted with caution due to the high-risk enrichment of the study cohort. https://t.co/o3JS2wFrbC https://t.co/J4VNWqVHBe
For 50 years the National Institute on Drug Abuse has spent hundreds of millions of dollars to find an effective cocaine addiction medication. After 100+ molecules & many hundreds of studies nothing has been FDA approved. It's the Holy Grail of addictions medication. Proud to
Klarna replaced 700 customer service agents with AI. Resolution time dropped from 11 minutes to 2. Then quality collapsed on complex cases. The CEO admitted cost became too dominant. Customers wanted humans. Klarna is now hiring people again. 55% of firms that replaced
Created 2026-05-01 · Updated 2026-05-16
Edit rule text
[ai_company_product_launch_or_capability_announcement_without_healthcare_validation]
Learned 6 rejections Active
Exclude posts that announce AI company product launches, partnerships, or technical capability releases (e.g., new models, API features, enterprise products) unless the post explicitly demonstrates healthcare-specific validation, deployment, or clinical/operational impact. Marketing announcements and feature releases without proof of healthcare use do not qualify.
Posts announcing new AI company products, partnerships, or technical capabilities without evidence of healthcare application, validation, or impact.
3 example posts
We're presenting at @ASCO 2026! Evaluating AI decision support in a rapidly evolving therapeutic landscape: EGFR-mutant metastatic NSCLC. 📍 Poster Board #548 🕘 9:00 AM CT If you're attending ASCO 2026, stop by our poster and connect with the team. We look forward to the https:
We’re taking steps to accelerate defensive progress in biology: - Launching Rosalind Biodefense to help trusted builders develop new biodefense and pandemic preparedness capabilities.
 - Expanding trusted access to GPT-Rosalind for select U.S. government and allied partners
Today's OpenMed Agent session, driven by GPT-5.5 through the Codex SDK with my own ChatGPT subscription: → 10,036 records ingested → 16-step plan complete → 49 tool calls → 5 reviewer-gated PDFs → APPROVE on PA L38319 Public trace on @huggingface ↓ https://t.co/vlq6CoQ4ee
Created 2026-05-01 · Updated 2026-05-31
Edit rule text
[broad_health_claim_or_debate_without_nuance_or_systems_analysis]
Learned 6 rejections Active
Exclude posts that make sweeping health claims, start broad medical debates, or offer personal health opinions (e.g., eating disorder risks, medication efficacy assertions, treatment recommendations) without clinical evidence, nuance, or healthcare system context. Posts that are primarily commentary or debate-starting rather than substantive healthcare systems analysis should be excluded.
Posts making broad health or medical claims, starting debates, or offering opinions without evidence, nuance, or healthcare system analysis.
3 example posts
As I mentioned before, I am now sharing an example from GPT-5.5 Pro, also featured by OpenAI, that really left me stunned by what it is capable of in biomedical science. (full report on the website I created with Codex, link in the thread). To push GPT-5.5 Pro hard, I uploaded a
"How can medicine save the most lives?" Most people ask this rhetorically. @Farzad_MD and Tom Frieden took it literally. From banning smoking in NYC bars to cutting teen smoking in half in 5 years, this is what happens when you stop treating diseases and start preventing them. h
What superhuman vision can detect from the retinal photo, which human eyes cannot, is stunning. A new foundation AI model screening for diabetes hypertension, hyperlipidemia, gout, osteoporosis, and thyroid disease @NatureMedicine https://t.co/GhKvUqz4Vy https://t.co/iKcXCbLceu
Created 2026-05-01 · Updated 2026-05-02
Edit rule text
[clinical_anecdote_or_trial_observation_without_systems_context]
Learned 6 rejections Active
Exclude posts that present clinical trial results, individual case studies, or research observations without explaining implications for healthcare systems, clinical practice, cost, access, or technology adoption. Must contextualize clinical data within healthcare system dynamics.
Posts citing clinical trial data, research observations, or single cases without connecting to healthcare delivery, policy, or operational systems.
3 example posts
this pragmatic trial is more an investigation of physician behavior than the performance of MRSA nares PCR MRSA nares PCR's performance for excluding MRSA pneumonia was excellent from an EBM standpoint we should keep using the MRSA nares PCR and replace the physicians 🤷‍♂️
63% of ATS respiratory guidelines are based on low-quality evidence. We published the data. The next morning, the chair of ATS got on Twitter. Tagged people. Went on a tirade. Didn't seem to realize tweets were public. "It wasn't a private conversation." The critique? We ht
People are increasingly worried that AI tools make us overreliant. But how do we actually measure this? We introduce Offloading Score, a measure of reliance based on the fraction of cognitive effort offloaded to AI while completing a task. In a controlled user study, Offloading
Created 2026-04-30 · Updated 2026-06-06
Edit rule text
[ai_safety_vulnerability_incident_not_healthcare]
Learned 6 rejections Active
Exclude posts that report AI safety incidents, security vulnerabilities, or hacking examples (e.g., Claude deleting databases, malicious AI agent skills, API key leaks) without demonstrating direct healthcare system impact or learning relevant to healthcare deployment. The incident must show healthcare-specific consequences, not just generic AI risk.
Posts about AI safety failures, security breaches, or vulnerability incidents that lack healthcare application context.
3 example posts
Anthropic built something so powerful that they are only letting 50 organisations touch it. It is called Claude Mythos. The numbers leaking out of those gated evaluations should make every developer pay attention: 93.9% on SWE-bench Verified 94.6% on GPQA Diamond Claude Opus
Hacking Mexico government with AI assistance. Attacker exfiltrated hundreds of millions of citizen records. 75% of the executed commands across the entire cyberattack campaign were generated by Claude. 40 minutes after Claude said "I'm not going to create that file" it was report
𝗔𝗜 𝗔𝗴𝗲𝗻𝘁𝘀 𝗔𝗿𝗲 𝗔𝗹𝗿𝗲𝗮𝗱𝘆 𝗕𝗲𝗶𝗻𝗴 𝗛𝗶𝗷𝗮𝗰𝗸𝗲𝗱 Researcher Aks Sharma at Manifold found 30 malicious skills on ClawHub turning AI agents into a crypto farming botnet: 10,000 downloads before anyone noticed. ⬩ The attack required zero exploits. Malicious https://t.co/v4oBXPPydu
Created 2026-04-30 · Updated 2026-05-02
Edit rule text
[tangential_ai_agent_or_model_capability_announcement_without_healthcare_validation]
Learned 6 rejections Active
Exclude posts that promote or showcase AI agent frameworks, model releases, or automation tool capabilities (e.g., agents generating revenue, building autonomous systems) unless the post demonstrates a validated healthcare application or healthcare-specific operational problem these tools solve.
Posts announcing AI agent tools, model capabilities, or coding features (Claude Code, LangChain, OpenClaw) with loose healthcare framing or no validated healthcare use case.
3 example posts
Agents calling tools now generate more revenue on Latch than scientists clicking buttons. From native deployments but also harnesses like Claude Code, Codex, Cursor across Pharma, biotech, academic labs. Product roadmap increasingly concerned with exposing compute intensive
The more I build agentic systems at Razorpay, the more I understand that - at its core, it is an agentic loop with tool calls, integrations, and retrieval. The hard part is... actually making it run reliably, at scale, under real production load. And this is what makes system de
amazing writeup of experiments. essentially a model may already know a capability but that may be stored in a representation that is bad for extraction. you can slightly retrain/condition the model so the same capability becomes easier to isolate. must read for mech interp http
Created 2026-04-30 · Updated 2026-05-13
Edit rule text
[ai_safety_vulnerability_tangent_loose_healthcare_framing]
Learned 6 rejections Active
Exclude posts that focus on AI security incidents (zero-day vulnerabilities, hacking capabilities, code exploit detection) where healthcare is mentioned only as one of many affected domains or as loose framing. The post must have healthcare system or clinical risk as the primary focus, not as background context.
Posts about AI security vulnerabilities and exploits that frame healthcare as secondary context
3 example posts
Anthropic built a model so good at hacking they refused to release it publicly. INSTEAD they handed it to Apple, Google, Microsoft, AWS, Cloudflare one month later: 10,000+ zero days found in the software that runs the internet...bugs hidden for 27 years in OpenBSD. exploit htt
🚨 Anthropic just dropped the first Project Glasswing update Claude Mythos found 10,000+ critical vulnerabilities in ONE month: > Cloudflare: 2,000 bugs, 400 high/critical severity > Mozilla: 271 vulnerabilities in Firefox 150 — 10x more vulnerabilities found in Firefox 1
#AI agents need more than a sandbox—they need predictable execution. Learn how OpenShell provides a standard interface to intercept, inspect, and approve LLM-generated bash commands before they run: https://t.co/JKmLMgl1A9 https://t.co/xWi8oBNZ7T
Created 2026-04-29 · Updated 2026-05-30
Edit rule text
[unvalidated_or_speculative_medical_claims_without_evidence]
Learned 6 rejections Active
Exclude posts that make unsubstantiated claims about peptide safety, drug efficacy, vaccine safety, or medical interventions without citing published clinical data, regulatory clearance, or peer-reviewed evidence. Anecdotal testing methods or unverified seller certifications do not qualify as validation.
Posts making broad health or medical claims about interventions, drugs, or treatments without clinical evidence, peer review, or regulatory validation.
3 example posts
How does a brand like this even get listed at Target? Does Target only care about turnover? Might as well start selling snake oil too. By the way, this is the SKU the brand got in trouble with. They now call it "Berberine Patches". It was called GLP-1 Patches before. Check the h
“You can test your peptides to see if they’re real, bro. But I just look for the 99.9% on the result the seller gives me. You can look up the QR code upon the lab’s website, so you know it’s legit. It’s foolproof, bro. What’s mass spec, bro?” https://t.co/xzJLJ3hesT
🛑FDA officials have blocked publication of several studies supporting the safety of widely used vaccines against Covid-19 and shingles in recent months, a spokesman for HHS confirmed. The studies, which cost millions of dollars in public funds, were conducted by scientists at htt
Created 2026-04-28 · Updated 2026-05-17
Edit rule text
[ai_company_product_launch_or_capability_announcement]
Learned 6 rejections Active
Exclude posts that are primarily announcements or promotional content from AI companies (NVIDIA, Microsoft, OpenAI, Anthropic) about new tools, SDKs, sandboxes, or learning resources—unless the post provides substantive analysis of healthcare adoption barriers, clinical validation, or operational integration challenges.
Posts announcing AI company product launches, software features, or capability demos without validation of healthcare application or business model.
3 example posts
If you're a student, professor, or researcher—this one's for you. We’re hosting a series of virtual learnings for you to get hands-on experience with the NVIDIA NemoClaw and OpenShell software stack. You’ll get practical guidance on integrating agents with academic datasets and
We created OpenShell to make AI agents safe for enterprises. Built in open source so any company can adopt and trust it, this secure sandbox controls what agents can access, share, and send. Our CEO, Jensen, explains 👇 https://t.co/7EiIsxr0CG
Introducing Mesa: the most powerful filesystem ever built, designed specifically for enterprise AI agents. Every team building agents eventually hits the same wall: where do the files live? Not the chat history, the actual artifacts the agent works on. > The contracts your age
Created 2026-04-28 · Updated 2026-05-03
Edit rule text
[broad_unvalidated_health_claims]
Learned 6 rejections Active
Exclude posts that make sweeping health claims, speculative predictions about healthcare trends, or broad market debates ('AI will revolutionize healthcare') without providing specific evidence, clinical validation data, or analysis of implementation barriers in healthcare systems.
Posts making broad health claims, market debates, or speculative predictions without evidence, validation, or healthcare systems context
3 example posts
medicine is constrained far more by operational architecture than by fundamental science I just came head to head with this actual reality and my world has changed no. I cannot bandage away these inefficiencies with a better and better drug, for all cases. if you hate AI
Analogizing AI / Mat-Mul to Energy... Remember between 1900 and 1970 kWh/capita grew 34x ... and cost per kWH droped .... 34x. It might be a speed run, but game is game. https://t.co/uxnlnoaAYx
Gavin Baker (@GavinSBaker) says the disaggregation of inference can extend GPU useful lives from 3-4 years to 10-15. That may single-handedly save private credit and reduce the financing rates for GPUs, which will drive demand and help finance the build-out. "The disaggregation
Created 2026-04-26 · Updated 2026-05-30
Edit rule text
[tangential_ai_cybersecurity_not_healthcare]
Learned 6 rejections Active
Exclude posts that discuss AI model vulnerabilities, cyber takeover capabilities, or AI safety concerns in generic or non-healthcare contexts (e.g., simulated corporate networks, general frontier model capabilities) — even if tangentially labeled healthcare.
Posts about AI, cybersecurity, or model vulnerabilities framed loosely as healthcare-relevant but applied to non-healthcare domains.
1 example post
This is by far the most important result of the entire GPT-5.5 release: In a cyber evaluation GPT-5.5 was able to take over a simulated corporate network in 1/10 trials with a budget of 100M tokens. Previously, the only model that was able to solve this task was Claude Mythos,
Created 2026-04-25 · Updated 2026-04-26
Edit rule text
[broad_unsubstantiated_health_claims_without_evidence]
Learned 6 rejections Active
Exclude posts that make broad health claims, debunk popular beliefs, or argue against medical narratives without citing peer-reviewed evidence, clinical trials, or substantive expert analysis.
Posts making sweeping health claims or counter-claims without clinical evidence or nuanced discussion
3 example posts
The only problem with the GLP-1 heart muscle loss narrative is... ... that it's just a narrative. GLP-1s have reliably improved cardiovascular outcomes in trials, to the point that some research suggests benefit may even be independent of (not reliant on) weight loss.
"Your body can only use 25-30g of protein per meal. Anything above that gets wasted." This claim has been repeated in fitness nutrition for over a decade, and it was built on studies that measured the right thing over the wrong timescale. Moore 2009 gave six young men 0, 5, ht
Impressive study and even with the limitations, is an important addition to the Rapamycin literature In my opinion, the only plausible off-label use of Rapamycin currently should be in ApoE4 carriers as not many options are available). That would be an important trial we are
Created 2026-04-24 · Updated 2026-04-25
Edit rule text
[generalist_macro_policy_or_economics_without_healthcare_lens]
Learned 6 rejections Active
Exclude posts about macro-level topics (tariffs, grid infrastructure, compute scaling, energy policy, fintech disruption) where healthcare is mentioned tangentially or as one example among many, without healthcare-specific policy implications or system analysis.
Posts about broad economic, infrastructure, or policy trends (tariffs, power grids, compute scaling) with only loose healthcare framing.
3 example posts
Using 180 years of US data to show that tariff increases reduce imports, output, and manufacturing, with effects operating through both supply and demand channels, from Tamar den Besten, Regis Barnichon, @drkaenzig, and Aayush Singh https://t.co/R1kdtfi3fO https://t.co/wvXj31qs9e
States are rushing “affordability” bills, but most just mask high prices with rebates, mandates, or price caps. @MrRBourne & Nathan Miller argue durable relief means rolling back cost-raising rules and expanding supply. https://t.co/WG5egT1NfL
Software is not a moat Over the last 15+ years, nearly every innovation @EvanSpiegel and his team shipped got copied. Stories. AR glasses. Swipe-based navigation. The camera-first interface. And yet @Snapchat is the only independent consumer social app that has lasted. Nearly 1
Created 2026-04-24 · Updated 2026-05-01
Edit rule text
[unvalidated_speculative_medical_interventions]
Learned 6 rejections Active
Exclude posts that express personal advocacy for unvalidated medical interventions (rapamycin, ibogaine, peptide-forward telehealth concierge services) without acknowledging lack of clinical evidence, regulatory status, or potential harms.
Posts enthusiastically promoting experimental, off-label, or unproven medical treatments and drugs as viable healthcare solutions.
3 example posts
Impressive study and even with the limitations, is an important addition to the Rapamycin literature In my opinion, the only plausible off-label use of Rapamycin currently should be in ApoE4 carriers as not many options are available). That would be an important trial we are
We’re exploring the idea of a peptide-forward telehealth concierge medical service. Medicine 3.0 focused on full optimization- peptides, hormones, diet/exercise. MD is a former college varsity rower, fellowship at Yale etc. Would you be interested in participating in a pilot
I am a strong believer in ibogaine, which is one of the reasons why @ataibeckley acquired the residual interest in its ibogaine program in Q4 2023 and now owns it 100%. I’m very encouraged to see the administration taking a positive public stance on this important topic.
Created 2026-04-24 · Updated 2026-04-26
Edit rule text
[truncated_or_incomplete_posts_low_substance]
Learned 6 rejections Active
Exclude posts that are truncated (ending with '...' or mid-word), incomplete, or so brief they lack substantive content even if the topic is healthcare-adjacent.
Posts that are cut off mid-sentence, incomplete, or lack sufficient substance to evaluate.
3 example posts
Today we’re giving an update on ramping F.03 production at BotQ In the last 120 days, Figure scaled manufacturing 24x - from 1 robot/day to 1 robot/hour We will manufacture 55 humanoid robots this week https://t.co/Am5Kn53mVE
Thought I would start posting about interesting things happening at AWS. Not a bad day to start.🚀 Today at #WhatsNextWithAWS we announced a big step forward with @OpenAI on Amazon Bedrock: 1. OpenAI models now available 2. Codex for enterprise development 3. Amazon Bedrock Manag
#News for #investors and #media: Today we are announcing that our potential first-in-class treatment for chronic hepatitis B has been accepted for regulatory review by the US FDA. It has also received Breakthrough Therapy designation. 🔗 Learn more: https://t.co/AnUodGmljS htt
Created 2026-04-24 · Updated 2026-04-30
Edit rule text
[tangential_non_healthcare_content_with_loose_framing]
Learned 6 rejections Active
Exclude posts about non-healthcare domains (software engineering best practices, banking infrastructure, compliance certifications, etc.) that are tangentially mentioned in healthcare context without demonstrating material impact on healthcare delivery, clinical outcomes, or healthcare business models.
Posts about non-healthcare domains (software engineering, banking, compliance, etc.) that are framed loosely to healthcare without substantive healthcare relevance.
3 example posts
New Anthropic Fellows research: Model Spec Midtraining (MSM). Standard alignment methods train AIs on examples of desired behavior. But this can fail to generalize to new situations. MSM addresses this by first teaching AIs how we would like them to generalize and why.
ai-mageddon may prove more severe than saas-mageddon. the moats around cool interfaces and system prompts are far inferior to: social/team graphs, network effects, systems of record, permissioning admin tools, collaboration…and the list goes on. winners will have deep roots
the future of software engineering seems uncontroversially prompting + code review. startups will skip the code review because they’re racing against time. larger/serious orgs will take code review very seriously. llms can do code review, but my guess is that because they have t
Created 2026-04-23 · Updated 2026-05-06
Edit rule text
[tangential_ai_capability_hype_without_healthcare_application]
Learned 6 rejections Active
Exclude posts that discuss AI technical capabilities (model inference, zero-day vulnerabilities, model alignment, agent security, compute scaling) without connecting to a specific healthcare use case, clinical workflow, or healthcare system problem. General AI hype or technical tangents loosely framed as healthcare-relevant should be excluded.
Posts about general AI model capabilities, AI safety, cybersecurity vulnerabilities, or AI infrastructure without demonstrating or validating specific healthcare application or clinical relevance.
3 example posts
CB-Dock3: An Enhanced Web Server for Protein–Ligand Blind Docking 1. CB-Dock3 is a major update of the CB-Dock blind docking web server, targeting the practical bottlenecks created by today’s surge of large Cryo-EM and AI-predicted protein structures (e.g., AlphaFold3-era https:
So Mythos was, indeed, not marketing hype. Remember this is a general purpose model that just happens to be good at finding exploits because good models are good at lots of things. Expect similar from OpenAI & Google. And from open models in 8 months. https://t.co/KbhalQYX8
Dario Amodei says Mythos is not limited by compute Anthropic can scale it 3x or 10x without creating a conflict between government and private-sector access The harder problem is who gets it "because giving access to too many organizations could create serious cyber risks" htt
Created 2026-04-22 · Updated 2026-05-09
Edit rule text
[tangential_ai_infrastructure_hype_no_healthcare_application]
Learned 6 rejections Active
Exclude posts that discuss general AI infrastructure (token pricing, GPU utilization, inference costs, model release news, AI company funding, enterprise AI adoption trends) without explicitly connecting to a healthcare problem or use case. Posts must demonstrate healthcare-specific application, not general tech hype.
Posts about AI model capabilities, token pricing, compute infrastructure, or AI company business metrics that lack specific healthcare application
3 example posts
The framing here is exactly backwards. This is the strongest AI bull signal anyone has published this year. Uber deployed Claude Code to 5,000 engineers in December. By March, 84% were classified as agentic coding users. By April, 95% used AI tools monthly. 70% of all committed
Analogizing AI / Mat-Mul to Energy... Remember between 1900 and 1970 kWh/capita grew 34x ... and cost per kWH droped .... 34x. It might be a speed run, but game is game. https://t.co/uxnlnoaAYx
Gavin Baker (@GavinSBaker) says the disaggregation of inference can extend GPU useful lives from 3-4 years to 10-15. That may single-handedly save private credit and reduce the financing rates for GPUs, which will drive demand and help finance the build-out. "The disaggregation
Created 2026-04-21 · Updated 2026-05-26
Edit rule text
[workforce_disruption_macro_no_healthcare_specificity]
Learned 6 rejections Active
Exclude posts that discuss AI automation, workforce reduction, SaaS seat consolidation, or labor market disruption in generic terms or non-healthcare contexts without analyzing impact on clinical workflows, healthcare staffing models, or patient care delivery.
Posts about AI-driven job displacement or business model shifts without healthcare workforce or operational context
3 example posts
AI could, in theory, automate 57% of US work hours. Yet most human skills remain relevant. The future of work is not human or machine – but a partnership between people, agents, and robots. Read our latest research on skill partnerships in the age of AI: https://t.co/h1K56uPqPo
Is the business model for traditional software companies in permanent decline due to AI Agents not needing seats? 2 examples: Re: @salesforce, we’ve reduced our seats from 10+ to 2 human seats and 1 API seat. And yet, we now pay $22,000 a year, 83% up from $12,000. Why? Our
Why the biggest fintech players are in for a shock. "The shift is from human UX to agent UX. In the past, you won with dashboards, design and user experience. Now, the buyer is an AI agent, and it only cares about APIs, performance and integration. That breaks traditional htt
Created 2026-04-20 · Updated 2026-04-27
Edit rule text
[clinical_research_without_healthcare_systems_context]
Learned 6 rejections Active
Exclude posts that announce clinical trial results, drug efficacy data, or research findings without discussing broader healthcare system impacts, delivery mechanisms, or operational implications.
Posts reporting clinical trial results, drug data, or research findings without analysis of healthcare system implications
3 example posts
Drug-resistant infections are a major public health threat around the world, responsible for more than a million deaths each year. Scientists are constantly trying to find and develop new antibiotics. Now, researchers say artificial intelligence is helping speed their search. ht
Figuring out how to benchmark agents on realistic biology research has quickly become one of my favorite types of engineering work. You work with scientists to get to the core of some biological claim, precisely assembling raw data/prior literature/experimental context in a https
Presented at #EASCongress2026: Vafai and colleagues report reductions in PCSK9 and LDL cholesterol levels and no dose-limiting toxic effects in persons with hypercholesterolemia treated with VERVE-102, a base editor targeting PCSK9. Full study results: https://t.co/iAmoPdipx9 ht
Created 2026-04-19 · Updated 2026-05-28
Edit rule text
[clinical_research_or_trial_observation_without_healthcare_systems_context]
Learned 6 rejections Active
Exclude posts that report clinical trial results, drug efficacy data, rare disease case studies, or single-patient success stories without explaining how the finding addresses a healthcare delivery problem, access barrier, or systemic implementation challenge.
Posts sharing clinical trial results, research observations, or medical case studies without connecting findings to healthcare delivery, access, implementation, or system-level barriers.
3 example posts
An AI foundation whole-body 3D model that assesses perturbations (such as obesity) across multiple systems (such as immune, neural) at the cell level. This is MouseMapper. Imagine HumanMapper someday @Nature @erturklab https://t.co/jsg3VMzOqg https://t.co/tktuINis2W
In a meta-analysis of 210 biomedical AI studies that statistically compared models under cross-validation, 97% used invalid statistical tests. Here's our new preprint https://t.co/OG58Vkeu49 led by @tianchuzeng @kkli20111 @ZShaoshi @ten_photos 1/N https://t.co/PEnqXQxcsJ
Really amazing story of drug repurposing for rare disease using AI found Clonazepam and worked!! (NCATS federated learning and a determined team made this work) 🚀 👏 @MayoClinic @ncats_nih_gov @statnews #STATBreakthrough https://t.co/D4a4IMXSwh
Created 2026-04-18 · Updated 2026-05-24
Edit rule text
[healthcare_fraud_or_scandal_reporting_without_analysis]
Learned 6 rejections Active
Exclude posts that focus primarily on fraud case details, arrests, sentencing, or enforcement headlines without substantive analysis of underlying healthcare system vulnerabilities, operational failures, or policy changes needed to prevent similar fraud.
Posts reporting healthcare fraud cases, enforcement actions, or financial misconduct as crime/enforcement news without healthcare system analysis.
3 example posts
In 2021, Javaid Purwaiz, an OBGYN, was sentenced to 59 years in prison for one of the most severe cases of healthcare fraud in the country’s history. Once you go through court records, you realize the fraud that gave him a life sentence is the same fraud used by gender doctors.
$340 MILLION in fraud targeted — in 1 WEEK. That’s what happens when enforcement gets serious. Luxury cars. Fake claims. Stolen benefits meant for Americans in need — now turning into prison sentences. The hammer is dropping. We’re just getting started.https://t.co/qP0cOIypE4
🚨 As you pay your taxes this week, LOOK at what the fraudsters allegedly did with your money❗️ 🔹Cosmetic procedures 🔹Breast implants 🔹Tweaks to arms and thighs 🔹Tummy tuck 🔹Purebred dogs 🔹Flights to Hawaii 🔹Flights to Disneyland 🔹Multimillion-dollar home 🔹Range https://t.c
Created 2026-04-17 · Updated 2026-04-21
Edit rule text
[ai_company_product_launch_or_metrics]
Learned 6 rejections Active
Exclude posts that announce new AI company products, features, or business milestones (funding, ARR, hiring, partnerships) unless the post explicitly demonstrates how the product is being applied to solve a specific healthcare problem with evidence of healthcare adoption or impact.
Posts announcing AI company product launches, business metrics, or capability announcements without demonstrating healthcare application or validation.
3 example posts
Stripe just created a role that didn't exist 12 months ago (and they're paying multiple six figures for it) It's called the Forward Deployed AI Accelerator. They are hiring AI-native individuals to work directly with their marketing teams to fundamentally change how they work.
If you're a student, professor, or researcher—this one's for you. We’re hosting a series of virtual learnings for you to get hands-on experience with the NVIDIA NemoClaw and OpenShell software stack. You’ll get practical guidance on integrating agents with academic datasets and
We created OpenShell to make AI agents safe for enterprises. Built in open source so any company can adopt and trust it, this secure sandbox controls what agents can access, share, and send. Our CEO, Jensen, explains 👇 https://t.co/7EiIsxr0CG
Created 2026-04-17 · Updated 2026-05-05
Edit rule text
[ai_company_product_metrics_not_healthcare_focused]
Learned 6 rejections Active
Exclude posts that focus on AI company product launches, employee adoption rates, engineering productivity gains, or business metrics (like Uber deploying Claude to engineers or law firms building proprietary AI systems) unless the post explicitly connects to healthcare delivery impact.
Posts about AI company product adoption, engineering velocity, or business metrics disconnected from healthcare application
3 example posts
Almost everyone is building agent harness systems the wrong way. The default move: pick LangChain or LangGraph or the OpenAI Agents SDK, accept the loop, the tools, the memory, the orchestration, the policy engine, the credential store, the budget tracker, all of it, as one deci
I think the AI superapps will soon own 90% of the agentic layer more and more people won't use hermes/openclaw etc because Claude Cowork/Codex will offer 90% of the functionality with 5% of the friction
kirkland spending $500m on internal ai is the tell. $6.5b revenue, they could buy any legal ai vendor outright and not feel it, and they're building instead. the moat was always the proprietary corpus. decades of deal docs, redlines, partner judgment encoded as training data, an
Created 2026-04-17 · Updated 2026-06-03
Edit rule text
[clinical_trial_or_drug_research_observation_without_healthcare_systems_context]
Learned 6 rejections Active
Exclude posts that present clinical trial data, drug efficacy results, or research findings as standalone observations without connecting to healthcare system implications such as coverage decisions, clinical adoption barriers, workflow integration, or care delivery outcomes.
Posts reporting clinical trial results, drug efficacy data, or research findings in isolation without analysis of healthcare implementation, access barriers, or systemic adoption.
3 example posts
⚠️ Sacubitril/Valsartan works. So why aren’t we using it? The evidence is undeniable: ↓ CV mortality: 20% (RCT) / 10–38% (RWE) ↓ HF hospitalization: 21% (RCT) / 10–16% (RWE) ↓ All-cause mortality: 15% (RCT) / 10–25% (RWE) Plus: reverse remodeling, less MR, better QoL & https:
AI can now design antibodies that bind with atomic precision, but not ones that cells can produce. Our preprint closes this gap, delivering a structural principle, an AI-guided rescue pipeline, and adalimumab variants with 20-100x in vivo potency. https://t.co/GvfgHA5EcU
Here is a video of me entering my office tomorrow knowing that $NTLA is about to present the first-ever Phase 3 data of an In Vivo (!) CRISPR Gene Editing Program. Somehow - and after @adamfeuerstein’s🧵👇- I have a feeling it won’t be the only BioTech and CRISPR news…🤔 $XBI https:
Created 2026-04-17 · Updated 2026-04-29
Edit rule text
[retweet_or_shallow_commentary_without_original_analysis]
Learned 6 rejections Active
Exclude posts that are primarily retweets, light paraphrasing, or shallow commentary on another person's statement or announcement without original analysis, evidence, or healthcare systems insight. The post must contribute new analytical value, not just amplify existing claims.
Posts that retweet or lightly paraphrase industry commentary without adding original insight, evidence, or healthcare systems perspective
3 example posts
Yes to Katy. If I was a manufacturer I would contract directly with the employer. The health plan when self funded has a fiduciary obligation to plan assets… Nice work @KatyTalento
Makes an important point here about the role of these short summaries they do this because this is how most doctors digests information if designed appropriately, an LLM should be able to read and appraise a paper and provide better peer review than most clinicians and honestly
The UK AISI found Mythos Preview is the first model to solve both their cyber ranges end-to-end. No model had ever solved the AISI’s “Cooling Tower” cyber range before. We're getting it to defenders as fast as we responsibly can. More to come on our Glasswing work soon.
Created 2026-04-16 · Updated 2026-05-21
Edit rule text
[tangential_non_healthcare_infrastructure_hype]
Learned 6 rejections Active
Exclude posts about GPUs, chip announcements, humanoid robotics, manufacturing automation, or infrastructure hype that reference healthcare tangentially or speculatively but lack analysis of how these technologies address specific healthcare system constraints, regulatory barriers, or clinical workflows.
Posts about compute infrastructure, hardware capabilities, or manufacturing robotics that tangentially mention healthcare without substantive healthcare systems analysis.
3 example posts
Nous Research is working with NVIDIA to make Hermes Agent run smoothly on the new NVIDIA RTX Spark superchip. Hermes Agent is also integrating with the new OpenShell runtime, which connects Hermes to Microsoft’s security primitives https://t.co/cvvPc8bIKa
Humanoid robots, designed to mimic human movement and capabilities, have been making headlines in recent years, from their use as baggage handlers at Japanese airports to Tesla’s big bet on its Optimus humanoid. Market watchers have predicted that the machines will change the ht
the CEO of NVIDIA just said the computer is no longer being built for you, it is being built for agents Jensen Huang put it plainly, until now we were the users, we were the renters, every CPU on earth was designed around how a person works but agents do not work like us, they
Created 2026-04-16 · Updated 2026-06-05
Edit rule text
[tangential_non_healthcare_domain_with_healthcare_label]
Learned 6 rejections Active
Exclude posts from non-healthcare domains (Revolut banking models, ClickUp productivity tools, tariff economics, software moats at Snapchat) that mention healthcare tangentially or are tagged with healthcare keywords but do not address healthcare-specific systems, providers, patients, or health outcomes.
Posts about non-healthcare domains (banking, compliance tools, software tools, macroeconomics) framed with loose healthcare relevance.
3 example posts
Revolut just moved the IP of banking into a model. Trained on 24 billion banking events in 111 countries. One foundation model replacing six separate ML systems. Credit scoring: +130% Fraud recall: +65% Marketing engagement: +79% The model is the new moat.
📈 NVIDIA tops AI leaderboards and benchmarks with open models driven by extreme co-design across compute, networking, memory, storage, and software. This includes models for biology, AI physics, agentic AI, physical AI, robotics, and autonomous vehicles. By being vertically htt
Using 180 years of US data to show that tariff increases reduce imports, output, and manufacturing, with effects operating through both supply and demand channels, from Tamar den Besten, Regis Barnichon, @drkaenzig, and Aayush Singh https://t.co/R1kdtfi3fO https://t.co/wvXj31qs9e
Created 2026-04-14 · Updated 2026-05-03
Edit rule text
[political_outrage_regulatory_without_healthcare_analysis]
Learned 6 rejections Active
Exclude posts that frame healthcare issues primarily as political outrage, corporate scandal, or regulatory overreach without providing substantive analysis of how the issue affects patient care, provider workflows, or healthcare system efficiency. Posts must offer systems-level insight, not just moral indignation.
Posts expressing political outrage about healthcare policy, regulation, or corporate behavior without substantive analysis of healthcare system impact
3 example posts
@swyx > get government sponsored monopoly > prevent patients from getting their data > make data non transferable > contribute nothing to open source software > refuse to collaborate with other software vendors and kill the ecosystem > appeal to administrators and be hated by p
🚨 Surgeon @EithanHaim reveals shocking medical fraud scheme: Texas doctors allegedly changing teens' medical records and using fake billing codes to secretly continue banned gender treatments—scamming insurance and taxpayers. He's speaking at a #DetransAwarenessDay @genspect foru
I am not so partisan that I can't appreciate Congresswoman Alexandria Ocasio-Cortez taking down the CEO of CVS on behalf of all Americans. Healthcare is a universal issue, so pay attention to what's being sold to us. Translation: "Our perfect patient is insured by Aetna, CVS. T
Created 2026-04-14 · Updated 2026-04-16
Edit rule text
[political_outrage_without_healthcare_substance]
Learned 6 rejections Active
Exclude posts that frame political disputes (FDA leadership, government policy changes, Trump administration decisions) primarily as outrage narratives, even if they mention healthcare. The post must analyze healthcare system implications, not just report political conflict or personnel drama.
Posts using healthcare framing to amplify political outrage, regulatory drama, or FDA/government personnel conflicts without substantive healthcare systems analysis.
3 example posts
🚨 UPDATE Friday afternoon: Trump said he has no plans to replace FDA Commissioner Marty Makary This morning: "Just got a call from inside the administration. Makary is dead in the water. Just a matter of time." -@thackerpd $HIMS $LLY $NVO https://t.co/ZjcjOAG7e1
🚨 NOW: President Trump confirms that ALL FEDERAL AGENCIES are now required to buy American, “NO excuses” 🔥 And 47 is ensuring all “Made in America” loopholes are DESTROYED I voted for this! “ALL FEDERAL AGENCIES MUST BUY AMERICAN — NO EXCUSES! For decades, Washington politici
🛑FDA officials have blocked publication of several studies supporting the safety of widely used vaccines against Covid-19 and shingles in recent months, a spokesman for HHS confirmed. The studies, which cost millions of dollars in public funds, were conducted by scientists at htt
Created 2026-04-13 · Updated 2026-05-12
Edit rule text
[academic_or_research_observation_only]
Learned 6 rejections Active
Exclude posts that report clinical trial results, research breakthroughs, or scientific findings as standalone observations without analyzing their healthcare system implications, market adoption barriers, clinical workflow integration, or impact on patient access.
Posts about clinical research findings, trial results, or scientific discoveries presented as isolated facts without healthcare system or business context.
3 example posts
Today the first results of the very first phase 3 study of a pan-KRAS-inhibitor in metastatic pancreatic cancer dropped, which might apply to > 90% of all pancreatic cancer patients with a KRAS-mutation! Median overall survival of 13.2 months versus 6.7 months with chemo in
NIH-funded researchers have uncovered a key reason why immunotherapy has largely failed in pancreatic cancer — and identified a promising strategy to overcome that resistance. Read on to learn more about this discovery: https://t.co/BoCHpLxp5g https://t.co/3DXv4E9DOE
Across large, multicohort datasets, CardioNets achieved superior performance to ECG-only baselines and diagnostic accuracy comparable to CMR-based models, supporting its potential to expand access to advanced cardiovascular assessment. Full study results: https://t.co/VP2iOBLUev
Created 2026-04-12 · Updated 2026-04-14
Edit rule text
[generalist_macro_policy_no_healthcare_lens]
Learned 5 rejections Active
Exclude posts that discuss labor market disruption, regulatory posturing, fraud reporting, energy policy, or macro-economic trends that reference healthcare tangentially but do not analyze how the issue specifically impacts healthcare delivery, clinical outcomes, patient access, or healthcare system structure.
Posts about macro-level policy, labor markets, or regulatory issues framed with healthcare labels but lacking healthcare systems analysis.
3 example posts
At our recent Energy and the AI Age summit, Hon. Bernard L. McNamee, former Commissioner of the U.S. Federal Energy Regulatory Commission, comments on why energy matters in discussions about AI: “Energy is the foundation of our entire economy. Energy makes up about 7% of the htt
“I just worked a 12-hour shift in the operating room — and I still can’t afford rent, groceries, or gas.” “So now it’s after midnight, and I’m out driving DoorDash.” A full-time hospital worker finishing a 12-hour shift and FORCED to start a second job just to survive. https://
Yesterday at the @eMedHealth Health Innovation Revolution Summit, Jeffrey Pfeffer of Stanford said something that stopped the room. "No matter what industry you think you're in, all employers are in the healthcare business." He's right. And most companies still haven't acted like
Created 2026-06-09 · Updated 2026-07-23
Edit rule text
[clinical_anecdote_or_single_case_no_systems]
Learned 5 rejections Active
Exclude posts that report a single clinical result, patient anecdote, treatment outcome, or trial data point (e.g., 'I took this drug and felt better'; 'this patient survived') without analyzing what the finding means for clinical practice, policy, reimbursement, or healthcare delivery systems.
Single clinical observations, patient narratives, or trial snippets reported without healthcare systems context
3 example posts
Version 2 of our ShockCalcs hemodynamics simulator is live. I've refined the physiology a ton, added new meds, and also a real-time Frank Starling curve that responds to fluids + vasopressors. Check it out (link in reply), and reply here with feedback on how to improve further!
The Iowa study doubled survival by adding IV vitamin C to first line chemo, 8 months to 16. We take it further and inject the vitamin C straight into the tumor, same as we do with the KRAS inhibitors. The compound is rarely the problem. Delivery is, and what it does to the ht
Feeling extremely lucky to live in 2026 and have access to such incredible medicine and clinical research. On the eve of my 40th I had my first dose (monthly injection) of a new treatment that I’m hopeful will prevent me from ever having a heart attack or stroke. This drug ht
Created 2026-06-09 · Updated 2026-06-12
Edit rule text
[cynical_insider_sarcasm_fraud_narrative]
Learned 5 rejections Active
Exclude posts that use sardonic tone, cynical insider narrative, or sarcastic framing to make unsubstantiated claims about fraud, regulatory capture, or healthcare system corruption—without specific evidence, systems analysis, or actionable insights.
Cynical insider commentary using sarcasm or moral outrage to make broad claims about fraud, corruption, or system failure without evidence
3 example posts
I'm SVP now. I told you I would be. The graph went up and to the right. Nobody checked what it measured. It measured "AI enablement." I made that up last year. I'm making it up again. Last year I rolled out Copilot to 4,000 people. Nothing happened. I got promoted. Those two htt
My job is to make sure your surgery center never gets built. Eleven years, and I have never lost. The kid had it all lined up. Board-certified, two partners, a lease on a space where he could do the same procedures my client does across town. He showed up to the hearing with a
PacificSource pulled out as Lane County Medicaid insurer, Tina Kotek’s OR Health Authority replaced it with Trillium, forcing 96k to join Trillium. Trillium’s parent co gave Kotek’s campaign $50k two months later. OR AG (D) said in ‘22 Trillium parent co gave Kotek’s”took advan
Created 2026-06-09 · Updated 2026-06-11
Edit rule text
[non_healthcare_infrastructure_tangent]
Learned 5 rejections Active
Exclude posts that focus on energy, compute infrastructure, semiconductor manufacturing, space technology, or robotics unless they explain a concrete healthcare application (not just 'AI needs more compute').
Posts about energy infrastructure, compute hardware, space technology, or semiconductor manufacturing that only loosely connect to healthcare
3 example posts
At our recent Energy and the AI Age summit, Hon. Bernard L. McNamee, former Commissioner of the U.S. Federal Energy Regulatory Commission, comments on why energy matters in discussions about AI: “Energy is the foundation of our entire economy. Energy makes up about 7% of the htt
Dylan Patel, founder of SemiAnalysis: "The upper bound on how much compute can be produced by 2030 is around 200 gigawatts a year." The entire world has about 20 gigawatts of AI deployed right now. The ceiling is 10x what exists today, and it still isn't enough to feed what htt
An interesting proposal for restoring human versus machine chess competition was to limit the machine to the same energy expenditure as the human brain. A worthy challenge, considering the amount of power required for AI data centers!
Created 2026-06-08 · Updated 2026-07-19
Edit rule text
[insider_cynicism_sarcasm_no_substance]
Learned 5 rejections Active
Exclude posts that rely primarily on sarcasm, cynicism, or insider mockery (e.g., 'I made that up last year') without providing evidence, systems analysis, or actionable insight into what is actually broken in healthcare or how to fix it.
Posts using cynical, sarcastic, or mocking insider commentary about healthcare business, AI hype, or regulatory gaming without substantive analysis.
3 example posts
I'm SVP now. I told you I would be. The graph went up and to the right. Nobody checked what it measured. It measured "AI enablement." I made that up last year. I'm making it up again. Last year I rolled out Copilot to 4,000 people. Nothing happened. I got promoted. Those two htt
I just looked at a home service firms ServiceTitan dashboard and saw something interesting. A Pre-seed company with AI-Supported Human CSRs is out converting an AI CSR business that's raised over $125M by >50% Head-to-Head win rate is 100% Why? Supercharging humans resonat
What kind of work will become more valuable in an AI economy? @ccatalini, former head economist at Meta: "AI is getting better and better at automating anything that can be measured, as long as you have a digital trace, if you can collect it with a device, that problem will be
Created 2026-06-08 · Updated 2026-06-08
Edit rule text
[anthropic_mythos_unvalidated_capability_claims]
Learned 5 rejections Active
Exclude posts that make extraordinary claims about Claude Mythos (or similar unreleased/internal AI models) finding zero-day vulnerabilities, discovering security flaws at scale, or demonstrating unprecedented research capabilities — without peer review, publication, or independent validation in healthcare contexts.
Posts about Anthropic's Claude Mythos model with unvalidated claims about AI capabilities, security findings, or research breakthroughs.
3 example posts
"Best-in-class security" still had 5 years of unfound bugs. AI found them in 6 weeks. $PANW CEO Nikesh Arora told @theallinpod that his own codebase, at a company that treats security as a core competency, had vulnerabilities Claude surfaced in 6 weeks that would have taken his
AI research is a series of next-step decisions. We looked at sessions where a human researcher took a wrong turn, showed Claude the session up to that point, and asked it what to do next. Mythos Preview improved on humans 64% of the time—up from 22% in 2024. https://t.co/Y0HLoktx
🚨 do you understand what just happened with Anthropic.. Their internal model, Claude Mythos, found previously unknown security holes in every major browser and every major operating system. Not one. All of them at once. And Anthropic did something a hype-driven industry almost
Created 2026-06-07 · Updated 2026-06-12
Edit rule text
[unvalidated_anthropic_security_claims]
Learned 5 rejections Active
Exclude posts claiming Anthropic deployed Claude or agents in NSA operations, financial infrastructure (NYSE, ICE), or government cybersecurity initiatives unless backed by official public statements from both Anthropic and the government/financial entity, plus peer-reviewed validation. Marketing partnerships without independent verification are insufficient.
Posts promoting Anthropic's unverified security achievements, vulnerability findings, or government/financial sector deployments without peer review or independent evidence
3 example posts
Mythos AI is being used by National Security Agency in offensive cyber operations / cyberattacks. Anthropic has even embedded engineers inside the NSA to help deploy the model. Are frontier AI labs becoming active contractors in state cyber conflict? The live-ops role is still ht
"You can run OpenClaw inside your company now." Annoucing our work with @Microsoft to bring OpenClaw to the Microsoft and Windows ecosystems. Claws now work securly in the enterprise. https://t.co/3fKhT2mGpC
The next frontier of AI-secured financial and technological critical infrastructure. @ICE_Markets and @NYSE are part of @AnthropicAI's cyber security initiative Project Glasswing, deploying Anthropic’s Claude Mythos Preview across ICE’s exchanges, clearing houses, mortgage
Created 2026-06-07 · Updated 2026-06-10
Edit rule text
[clinical_trial_or_pharma_data_announcement_without_healthcare_systems_context]
Learned 5 rejections Active
Exclude posts that simply announce clinical trial results, drug approval data, or biotech research findings without examining the healthcare system implications (manufacturing scale, reimbursement barriers, distribution, physician adoption, patient access, cost-effectiveness relative to healthcare workflows).
Posts reporting clinical trial results, drug efficacy data, or pharmaceutical announcements without analyzing healthcare system implications, reimbursement, market access, or operational impact.
3 example posts
#ASCO26 | Day 3 $LLY didn't spend billions on Kelonia for an 18-patient #myeloma dataset. They may have paid for a potential manufacturing disruption. KLN-1010 reported: • 18/18 MRD-negative at 1 month • No lymphodepletion • No ex vivo cell manufacturing • Single infusion
Published in @JCO_ASCO, during #ASCO26: Tumor-Agnostic Therapies: Translating Scientific Breakthroughs Into Global Implementation ASCO is where we celebrate the next breakthrough. This review asks why breakthroughs do not equal access globally. 🔗: https://t.co/PE5q27bIs1 htt
A rapid manufacturing pipeline for BCMA-targeting CAR T cells drastically shortens “vein-to-vein” times, and produces cells that demonstrate encouraging safety results in a phase one trial of patients with relapsed #MultipleMyeloma. @ScienceTM https://t.co/3F4u8ECFDS https://t.co
Created 2026-06-04 · Updated 2026-06-06
Edit rule text
[tangential_non_healthcare_domain_with_loose_healthcare_framing]
Learned 5 rejections Active
Exclude posts about robotics (humanoids, PUDU), cybersecurity incidents, labor market macro trends, or speculative sci-fi applications that mention healthcare tangentially or use healthcare as a loose framing device. The post must demonstrate concrete, validated healthcare-specific problems being solved, not hypothetical or loosely adjacent applications.
Posts about non-healthcare domains (robotics, cyber security, humanoid labor, infrastructure) that loosely reference or frame themselves as healthcare-adjacent without substantive healthcare application.
3 example posts
Humanoid robots, designed to mimic human movement and capabilities, have been making headlines in recent years, from their use as baggage handlers at Japanese airports to Tesla’s big bet on its Optimus humanoid. Market watchers have predicted that the machines will change the ht
The next frontier of AI-secured financial and technological critical infrastructure. @ICE_Markets and @NYSE are part of @AnthropicAI's cyber security initiative Project Glasswing, deploying Anthropic’s Claude Mythos Preview across ICE’s exchanges, clearing houses, mortgage
⚡️AI is turning cybersecurity from a human-limited profession into a machine-speed arms race. The surface read is “Anthropic’s model found vulnerabilities fast.” The deeper read is that the bottleneck in cyber is shifting. For years, vulnerability discovery was constrained
Created 2026-06-04 · Updated 2026-06-05
Edit rule text
[clinical_trial_or_pharma_data_without_healthcare_systems_context]
Learned 5 rejections Active
Exclude posts that report clinical trial outcomes, drug efficacy numbers, or pharmaceutical research findings as standalone announcements without healthcare systems analysis, market dynamics, or operational context for healthcare delivery.
Posts announcing clinical trial results, pharma data, or drug efficacy without analyzing healthcare system implications or competitive dynamics
3 example posts
This week brought even more exciting news in the field of HIV treatment! A once-weekly Lenacapavir-based pill has now demonstrated similar outcomes to daily antiretrovirals (ARVs), achieving non-inferiority. This development could significantly impact treatment adherence and http
🔔Survodutide SYNCHRONIZE-1 full paper is out for Boehringer-Ingelheim’s GLP-1/Glucagon study for weight loss. ▪️13% loss on 6mg (treatment regimen estimand) ▪️Huge treatment discontinuations due to AEs: 20% ▪️Nausea 65%, Vomiting 45% ▪️Placebo was killer. 5.4% WL. Wow. ▪️60% htt
Survodutide Once Weekly for the Treatment of Adults with Obesity: @NEJM 🥸 For @Boehringer, the advantage of Survodutide is probably liver and visceral fat reduction, not greater weight loss! Probably indication in obesity-+ MASH rather than head-to-head share against https://t.
Created 2026-06-02 · Updated 2026-06-11
Edit rule text
[pharma_trial_data_announcement_without_systems_context]
Learned 5 rejections Active
Exclude posts that merely announce pharmaceutical trial data, drug efficacy results, or clinical trial outcomes without analyzing how the result impacts healthcare delivery, systems, economics, access, or healthcare provider decision-making.
Posts announcing pharmaceutical trial results or drug efficacy data without healthcare systems analysis or operational context
3 example posts
This week brought even more exciting news in the field of HIV treatment! A once-weekly Lenacapavir-based pill has now demonstrated similar outcomes to daily antiretrovirals (ARVs), achieving non-inferiority. This development could significantly impact treatment adherence and http
🔔Survodutide SYNCHRONIZE-1 full paper is out for Boehringer-Ingelheim’s GLP-1/Glucagon study for weight loss. ▪️13% loss on 6mg (treatment regimen estimand) ▪️Huge treatment discontinuations due to AEs: 20% ▪️Nausea 65%, Vomiting 45% ▪️Placebo was killer. 5.4% WL. Wow. ▪️60% htt
Survodutide Once Weekly for the Treatment of Adults with Obesity: @NEJM 🥸 For @Boehringer, the advantage of Survodutide is probably liver and visceral fat reduction, not greater weight loss! Probably indication in obesity-+ MASH rather than head-to-head share against https://t.
Created 2026-06-01 · Updated 2026-06-10
Edit rule text
[anecdotal_single_case_or_personal_narrative_without_systems_context]
Learned 5 rejections Active
Exclude posts that recount a single person's healthcare experience, cost surprise, pricing issue, or clinical outcome (e.g., 'a mom called asking for price quotes', 'I called a woman in Dayton') unless the post explicitly connects the anecdote to a broader healthcare system failure, regulatory gap, or technology solution that the healthcare tech writer would cover.
Posts sharing single patient anecdotes, personal healthcare cost stories, or individual clinical observations without healthcare systems insight, operational analysis, or broader implications.
3 example posts
I called a woman in Dayton to tell her she was about to overpay $4,500, for a horrible health system based procedure. She was scheduled for a knee replacement. Her PPO had her on the hook for a $4,500 deductible. The call was on behalf of her employer: “See one of these three
A company with 800 employees spends roughly $15,000 per employee per year on health insurance. That is $12 million. Wired into a single line item. Every year. The CFO can tell you the carrier. He cannot tell you the unit cost of a single claim. He cannot tell you what the net
A Mom in Tulsa called 3 health systems last week asking the price of her son's tonsillectomy. Health system A: "We cannot quote you a price." Health system B: "Pricing depends on your insurance." Health system C: "Our financial counselor will reach out after the procedure." N
Created 2026-06-01 · Updated 2026-06-05
Edit rule text
[unvalidated_speculative_medical_claims_or_market_hype]
Learned 5 rejections Active
Exclude posts that make bold claims about medical treatments, drug performance, AI diagnostic accuracy, or clinical outcomes without citing peer-reviewed evidence, completed trials, or regulatory approval data. Includes speculation about drug market dominance or unvalidated medical interventions.
Posts that make unsubstantiated claims about medical interventions, drug efficacy, AI diagnostic superiority, or market speculation without clinical evidence.
3 example posts
🚨 do you understand what just happened with Anthropic.. Their internal model, Claude Mythos, found previously unknown security holes in every major browser and every major operating system. Not one. All of them at once. And Anthropic did something a hype-driven industry almost
Mythos AI is being used by National Security Agency in offensive cyber operations / cyberattacks. Anthropic has even embedded engineers inside the NSA to help deploy the model. Are frontier AI labs becoming active contractors in state cyber conflict? The live-ops role is still ht
Anthropic engineer James Brady: "Every agent in production lies. We measured it. The good ones lie less, the great ones catch the lie before the user does." In 29 minutes, he walks through the verification stack he built and the patterns the Claude Code team adopted to keep ht
Created 2026-05-29 · Updated 2026-06-07
Edit rule text
[market_pricing_or_glp1_peptide_dynamics_without_systems_analysis]
Learned 5 rejections Active
Exclude posts that discuss GLP-1 or peptide market dynamics, pricing, weight loss applications, or competitive positioning without analyzing healthcare access, reimbursement equity, supply chain resilience, or systemic healthcare outcomes. Market speculation alone is not healthcare tech substance.
Posts focused on GLP-1 drug market pricing, competition, or weight loss trends without healthcare delivery system implications or policy analysis.
3 example posts
Let me clear up the microscope fight first. 🔬 Everyone’s arguing over “hairs” and “fibers” in horse paste. But nobody talks about formulation science. Horse paste = oral suspension. Binders. Thickeners. Made to swallow. Injectable = sterile. Made for veins. Different rules. J
Approved 20 years ago as a diabetes treatment, GLP-1 drugs have been found to help patients reduce weight, changing the lives of more than 30 million people in the U.S. But there also have been troubling side effects reported. https://t.co/zzRamKnedu
Not NVIDIA. Not OpenAI. Eli Lilly. Jordi Visser (@jvisserlabs) says a 150-year-old pharma company in Indianapolis has the best shot at becoming the world's largest company within five years. The case: ~1,000 NVIDIA Blackwell GPUs in a private data center, a co-innovation lab ht
Created 2026-05-28 · Updated 2026-06-09
Edit rule text
[agentic_ai_technical_capability_tangent_without_healthcare_validation]
Learned 5 rejections Active
Exclude posts that celebrate or analyze agentic AI technical capabilities (agent planning, tool use, iterative code execution, multi-step reasoning) in non-healthcare domains or without demonstrating clinical safety, regulatory compliance, or healthcare-specific constraints.
Posts about agentic AI technical capabilities (code generation, planning, autonomous iteration) applied outside healthcare or without healthcare-specific validation.
3 example posts
Turn FDE from a noun into a verb with Apollo. FDE is becoming the default operating model for deploying AI in the enterprise. The hard truth: AI does not become valuable in a demo. It becomes valuable when it is embedded into the real workflows, data, constraints, and
a databricks tech lead just spent 26 minutes on the part of multi-agent nobody wants to say out loud: your agents don't break because the model is dumb. they break because nothing is coordinating them. one agent is a feature. fifty is a distributed systems problem. https://t.c
OpenClaw is powerful enough that we needed Hermes Agent watching it. A lot of people are missing this about agents. The more useful they get, the more dangerous it is to let one system plan the work, run the work, review the work, and decide what happens next. It looks clean h
Created 2026-05-27 · Updated 2026-06-01
Edit rule text
[anecdotal_single_case_or_observation_without_systems_context]
Learned 5 rejections Active
Exclude posts that rely on a single clinical anecdote, individual case, or personal observation (e.g., 'a surgeon I know,' 'a mom in Tulsa') without connecting it to systemic healthcare challenges, operational incentives, or broader patient/provider patterns. Single stories must be framed as evidence of system-level problems, not just interesting cases.
Posts that cite individual clinical anecdotes, single cases, or personal observations without analyzing broader healthcare system patterns or drivers.
3 example posts
A company with 800 employees spends roughly $15,000 per employee per year on health insurance. That is $12 million. Wired into a single line item. Every year. The CFO can tell you the carrier. He cannot tell you the unit cost of a single claim. He cannot tell you what the net
medicine is constrained far more by operational architecture than by fundamental science I just came head to head with this actual reality and my world has changed no. I cannot bandage away these inefficiencies with a better and better drug, for all cases. if you hate AI
A Mom in Tulsa called 3 health systems last week asking the price of her son's tonsillectomy. Health system A: "We cannot quote you a price." Health system B: "Pricing depends on your insurance." Health system C: "Our financial counselor will reach out after the procedure." N
Created 2026-05-27 · Updated 2026-06-03
Edit rule text
[broad_unvalidated_health_claim_or_market_speculation_without_evidence]
Learned 5 rejections Active
Exclude posts that make sweeping claims about health outcomes, market disruption, or business viability (e.g., 'gene therapy will eliminate disease,' 'agents will replace 90% of roles,' 'this will transform healthcare') without providing data, clinical evidence, or rigorous analysis of implementation barriers and systemic constraints.
Posts making broad health claims, market predictions, or business model critiques without clinical evidence, validated data, or substantive systems analysis.
3 example posts
Interoperability is where AI hype goes to die. If your “agent” can’t work across data, logic, actions, security of legacy systems, it’s not transforming the enterprise. It’s just another app asking for an export. Winners don’t rip and replace. They operate.
Anthropic just shipped opus 4.8 and called it a modest upgrade modest is the engineer who used to need babysitting now finishing your whole codebase migration in one session while you sleep so here is the part nobody connects the model is no longer the bottleneck. the platform
Agentic orchestration layers allow AI agents, enterprise systems, and data connections to work together across functions. As cognitive bottlenecks shrink, that allows decision-making to speed up, coordination to improve and new operating models to form. https://t.co/0to4HC26cu
Created 2026-05-27 · Updated 2026-06-05
Edit rule text
[unvalidated_speculative_pharmaceutical_claims]
Learned 5 rejections Active
Exclude posts reporting clinical trial data announcements, drug approvals, or pharmaceutical market speculation (GLP-1 pricing, weight loss drug adoption) without discussing healthcare delivery, provider adoption, equity, or systemic implementation barriers.
Posts about drug trials, biotech breakthroughs, or GLP-1 peptide dynamics without healthcare systems analysis or clinical context
3 example posts
Approved 20 years ago as a diabetes treatment, GLP-1 drugs have been found to help patients reduce weight, changing the lives of more than 30 million people in the U.S. But there also have been troubling side effects reported. https://t.co/zzRamKnedu
🚨 T-DXd just went tumor-agnostic — and it changed a lot about how we think about HER2 testing across solid tumors 🔬 #ASCO26 FDA granted accelerated approval to T-DXd for patients with unresectable or metastatic HER2-positive solid tumors after prior systemic treatment —
🔥Phase 1: ABBV-706 (SEZ6-targeting ADC) in R/R SCLC & solid tumors 🆙 @NatureMedicine ☑R/R SCLC; ORR 52%; mOS 12.4 mo at RP2D 🎯RP2D confirmed at 1.8 mg/kg Q3W balancing efficacy & safety 🎙 @LaurenByersMD #LCSM @OncoAlert @Larvol https://t.co/vjUUd1GYSy
Created 2026-05-26 · Updated 2026-06-08
Edit rule text
[policy_outrage_or_fraud_scandal_without_healthcare_systems_analysis]
Learned 5 rejections Active
Exclude posts that report fraud convictions, regulatory violations, healthcare spending criticisms, or policy complaints as isolated scandals or moral outrage without analyzing root causes in healthcare system structure, incentives, or operational design.
Posts reporting regulatory violations, fraud schemes, or policy criticism without systemic healthcare analysis
3 example posts
A Mom in Tulsa called 3 health systems last week asking the price of her son's tonsillectomy. Health system A: "We cannot quote you a price." Health system B: "Pricing depends on your insurance." Health system C: "Our financial counselor will reach out after the procedure." N
Medicaid spending on autism therapy nearly tripled between 2020 and 202. In North Carolina, the number of clinics offering applied behavior analysis grew from 61 in 2019 to 409 in 2026. Anytime fraudsters can steal money from the federal government, they do. https://t.co/fahy
Why is healthcare expensive? The grifters gonna grift. ASCO is asking for oncology practices to be admitted into the same 340B machinery that helped health systems acquire physician practices for two decades. Read the tell. The proposed Indigent Care Ratio lets community
Created 2026-05-26 · Updated 2026-05-31
Edit rule text
[founder_enthusiasm_or_startup_announcement_without_validation]
Learned 5 rejections Active
Exclude posts that are primarily founder commentary, lab enthusiasm, or company milestone announcements (e.g., 'We're presenting at conference', 'Our work shows...') where there is no published clinical data, peer-reviewed evidence, or regulatory milestone to substantiate the claim. These read as promotional rather than validated advancement.
Posts expressing founder enthusiasm, lab updates, or startup announcements about biotech, pharma, or AI companies without clinical evidence or regulatory validation.
3 example posts
Anthropic just shipped opus 4.8 and called it a modest upgrade modest is the engineer who used to need babysitting now finishing your whole codebase migration in one session while you sleep so here is the part nobody connects the model is no longer the bottleneck. the platform
We're presenting at @ASCO 2026! Evaluating AI decision support in a rapidly evolving therapeutic landscape: EGFR-mutant metastatic NSCLC. 📍 Poster Board #548 🕘 9:00 AM CT If you're attending ASCO 2026, stop by our poster and connect with the team. We look forward to the https:
wow K&E, largest law firm in the world ($10B billings). They’re committing $500M to build proprietary AI software, with 180 tech pros and 100 partners involved. I’m skeptical that “build” is right here, but wonder if Anthropic/OAI deployment cos are involved.
Created 2026-05-26 · Updated 2026-06-04
Edit rule text
[unvalidated_speculative_peptide_or_glp1_market_claims]
Learned 5 rejections Active
Exclude posts that speculate about peptide or GLP-1 market size, pricing competition, consumer demand, or company positioning without connecting to clinical outcomes, healthcare delivery infrastructure, or payer strategy.
Posts speculating on GLP-1 or peptide market dynamics, adoption rates, or competitive positioning without clinical evidence or healthcare system analysis.
3 example posts
$HIMS up to #4 in the Canadian App Store medical category. That is not a stock chart. That is real consumer demand showing up in public. GLP-1 drama, Novo pressure, pricing debate, and now app-store strength. Bears keep arguing the story is slowing down, but the product demand
For years, grey-market HGH operated in a strange legal grey zone somewhere between “research compound,” “API” and underground enhancement culture. Last month’s FDA guidance suggests regulators may no longer see it that way. The recent decline in HGH quality, and now the growing
Think about this. The $50 per month out-of-pocket cost of GLP-1 drugs for Americans (with insurer paying the balance of $2000 pm) is more than double the entire cost for Indians (less than ₹2000pm for Semaglutide). Overall cost difference? 100 times. Not 100%.
Created 2026-05-26 · Updated 2026-06-01
Edit rule text
[generalist_ai_infrastructure_or_macro_hype]
Learned 5 rejections Active
Exclude posts that discuss AI model capabilities, token costs, inference efficiency, frontier model announcements, or compute infrastructure (NVIDIA, chips, energy) in generalist or macro terms. Only include if directly tied to a healthcare delivery, clinical, or operational system use case.
Posts about AI compute, models, or infrastructure trends without demonstrated healthcare application
3 example posts
Citadel Securities just put institutional weight behind what the AI bulls won't say out loud. In a new macro note titled "Tokenomics," Citadel makes the argument plainly: even the most powerful technology on earth still has to pass through the boring discipline of cost curves, h
Dylan Patel, founder of SemiAnalysis: "The upper bound on how much compute can be produced by 2030 is around 200 gigawatts a year." The entire world has about 20 gigawatts of AI deployed right now. The ceiling is 10x what exists today, and it still isn't enough to feed what htt
An interesting proposal for restoring human versus machine chess competition was to limit the machine to the same energy expenditure as the human brain. A worthy challenge, considering the amount of power required for AI data centers!
Created 2026-05-26 · Updated 2026-06-12
Edit rule text
[unvalidated_peptide_or_glp1_market_speculation]
Learned 5 rejections Active
Exclude posts focused on GLP-1 or peptide market pricing, competitive dynamics, profit margins, or financial positioning — even if framed around 'weight loss' or obesity treatment. The post must analyze healthcare system impacts (access, reimbursement, equity), not market speculation.
Posts speculating on GLP-1/peptide market dynamics, pricing, or competitive positioning without healthcare systems lens
2 example posts
🔔Survodutide SYNCHRONIZE-1 full paper is out for Boehringer-Ingelheim’s GLP-1/Glucagon study for weight loss. ▪️13% loss on 6mg (treatment regimen estimand) ▪️Huge treatment discontinuations due to AEs: 20% ▪️Nausea 65%, Vomiting 45% ▪️Placebo was killer. 5.4% WL. Wow. ▪️60% htt
Let me clear up the microscope fight first. 🔬 Everyone’s arguing over “hairs” and “fibers” in horse paste. But nobody talks about formulation science. Horse paste = oral suspension. Binders. Thickeners. Made to swallow. Injectable = sterile. Made for veins. Different rules. J
Created 2026-05-25 · Updated 2026-06-11
Edit rule text
[political_outrage_fraud_scandal_without_systems_analysis]
Learned 5 rejections Active
Exclude posts that report fraud convictions, scandal headlines, or regulatory complaints as standalone outrage or moral judgment—without analyzing root causes, systemic incentives, or how the scandal reveals healthcare system design flaws. Posts must move beyond 'this is bad' to 'here's why the system produces this'.
Posts reporting on healthcare fraud, regulatory scandal, or policy outrage without analyzing systemic healthcare delivery or business model implications.
3 example posts
Why is healthcare expensive? The grifters gonna grift. ASCO is asking for oncology practices to be admitted into the same 340B machinery that helped health systems acquire physician practices for two decades. Read the tell. The proposed Indigent Care Ratio lets community
“Rural hospitals cannot make it on their own.” Nobody makes it on their own. Businesses need capital, strategy, discipline, operating systems, good finance, clean accounting, payer strategy, physician alignment, and leaders who know what business they are actually in. The
If you want to inject $1B into healthcare for the truly vulnerable, it costs the state $500M. If you want to inject $1B for able-bodied adults, it costs the state $100M. The 90% federal match drastically lowers the "price" of expanding government healthcare. Incentives dictate
Created 2026-05-22 · Updated 2026-05-25
Edit rule text
[speculative_robotics_or_sci_fi_labor_disruption_without_healthcare_validation]
Learned 5 rejections Active
Exclude posts about robotic automation, humanoid labor, or speculative sci-fi applications that mention healthcare only as a passing example without demonstrating clinical validation, healthcare-specific workflow integration, regulatory approval, or measurable healthcare labor market impact.
Posts about humanoid robots, speculative labor automation, or manufacturing disruption that frame healthcare as a tangential use case without validation or operational evidence.
2 example posts
Humanoid robots, designed to mimic human movement and capabilities, have been making headlines in recent years, from their use as baggage handlers at Japanese airports to Tesla’s big bet on its Optimus humanoid. Market watchers have predicted that the machines will change the ht
Another humanoid worker has just joined the factory floor Two years later,PUDU Robotics has launched the new-generation PUDU D7, built specifically for real manufacturing scenarios. It can autonomously push carts, handle intra-line transportation, perform delicate operations, h
Created 2026-05-22 · Updated 2026-06-06
Edit rule text
[enterprise_ai_adoption_macro_without_specificity]
Learned 5 rejections Active
Exclude posts that discuss enterprise AI adoption metrics, productivity improvements, or labor market disruption as generalist macro trends without connecting to specific healthcare operational contexts or outcomes.
Posts about enterprise AI adoption, productivity gains, or workforce impacts discussed at a macro level without healthcare-specific analysis
3 example posts
Most valuable AI startups in Europe 🇩🇪 Aleph Alpha - $20B 🇫🇷 Mistral AI - $14B 🇩🇪 Celonis - $13B 🇩🇪 Helsing - $12B 🇫🇷 Poolside AI - $12B 🇵🇱 ElevenLabs - $11B 🇮🇹 Bending Spoons - $11B 🇬🇧 Wayve - $8.6B 🇸🇪 Lovable - $6.6B 🇸🇪 Legora - $5.5B 🇫🇷 Hugging Face - $4.5B 🇬🇧 Synthesia - $4
Workers say AI is making them more productive. Executives say AI is not making their companies more productive. Both groups are reporting on the same software. The gap between the two numbers is the entire debate about the largest capital expenditure cycle in technology history.
Fair warning, YC founders: if you take these tokens, there’s a non-zero chance that OpenAI will study exactly what your startup is doing, copy your idea and put your app into their free offering. This is the classic platform playbook — be careful, founders!
Created 2026-05-21 · Updated 2026-05-28
Edit rule text
[market_pricing_without_systems_analysis]
Learned 5 rejections Active
Exclude posts that focus primarily on GLP-1 or peptide pricing, cost comparisons between countries, or product launch pricing without connecting to healthcare system incentives, reimbursement policy, or access barriers.
Posts that report GLP-1 drug pricing, market competition, or cost dynamics without analyzing healthcare system impacts or affordability structures.
3 example posts
$HIMS up to #4 in the Canadian App Store medical category. That is not a stock chart. That is real consumer demand showing up in public. GLP-1 drama, Novo pressure, pricing debate, and now app-store strength. Bears keep arguing the story is slowing down, but the product demand
For years, grey-market HGH operated in a strange legal grey zone somewhere between “research compound,” “API” and underground enhancement culture. Last month’s FDA guidance suggests regulators may no longer see it that way. The recent decline in HGH quality, and now the growing
Think about this. The $50 per month out-of-pocket cost of GLP-1 drugs for Americans (with insurer paying the balance of $2000 pm) is more than double the entire cost for Indians (less than ₹2000pm for Semaglutide). Overall cost difference? 100 times. Not 100%.
Created 2026-05-21 · Updated 2026-05-28
Edit rule text
[unvalidated_speculative_pharmaceutical_or_peptide_claims]
Learned 5 rejections Active
Exclude posts that speculate on peptide or GLP-1 market opportunities, pricing strategies, or competitive positioning (e.g., 'peptides are the next GLP-1 moment', 'grey-market HGH regulations', 'Ozempic brand premium') without citing clinical trial data, FDA guidance, or evidence of healthcare system adoption or reimbursement.
Posts speculating on peptide, GLP-1, or pharmaceutical market dynamics, pricing, or adoption without clinical evidence, FDA approval, or healthcare system validation.
3 example posts
For years, grey-market HGH operated in a strange legal grey zone somewhere between “research compound,” “API” and underground enhancement culture. Last month’s FDA guidance suggests regulators may no longer see it that way. The recent decline in HGH quality, and now the growing
Think about this. The $50 per month out-of-pocket cost of GLP-1 drugs for Americans (with insurer paying the balance of $2000 pm) is more than double the entire cost for Indians (less than ₹2000pm for Semaglutide). Overall cost difference? 100 times. Not 100%.
🚨 ARE PEPTIDES THE NEXT "GLP-1 MOMENT" FOR $HIMS? @eli_dorf: "I think there's something really interesting that happened with GLP-1s, which is that when someone's taking Ozempic or any GLP-1, they LOVE to talk about it." "They'll talk about it to their friends. They'll talk htt
Created 2026-05-19 · Updated 2026-05-31
Edit rule text
[unvalidated_peptide_and_glp1_market_speculation]
Learned 5 rejections Active
Exclude posts about GLP-1s, semaglutide, tirzepatide, peptides, or weight-loss medications that focus on market pricing, competitive positioning, app rankings, or consumer adoption — unless the post analyzes systemic healthcare delivery, reimbursement policy, or clinical access barriers. Posts about peptide sourcing quality, brand premiums, or generic vs. branded perception without healthcare systems context should be excluded.
Posts about GLP-1s, peptides, and weight-loss drugs framed as market dynamics or personal narratives without clinical systems analysis
3 example posts
🚨 ARE PEPTIDES THE NEXT "GLP-1 MOMENT" FOR $HIMS? @eli_dorf: "I think there's something really interesting that happened with GLP-1s, which is that when someone's taking Ozempic or any GLP-1, they LOVE to talk about it." "They'll talk about it to their friends. They'll talk htt
This is what the peptide space needs more of: clinicians with actual patient volume, public reasoning, and accountability. The next layer is transparency around sourcing, testing, and outcomes. Peptides are not a vibes category anymore. The signal is getting cleaner.
^Humira’s patent expired a while back for some context there. The brand premium is the moat now. Ozempic is not generic semaglutide in patient and prescriber perception. Novo keeps charging brand pricing on every loyal patient without a PMPRB ceiling.
Created 2026-05-18 · Updated 2026-05-22
Edit rule text
[speculative_robotics_or_sci_fi_labor_macro_without_healthcare_context]
Learned 5 rejections Active
Exclude posts that discuss humanoid robots, general labor market disruption, or speculative sci-fi applications (bedside robots, task automation) without demonstrating how these specifically address a validated healthcare workforce gap or clinical problem.
Posts speculating about humanoid robots, general labor automation, or sci-fi applications with only tangential healthcare framing.
3 example posts
Voice interaction, generating actions,and next, generating tasks...? The humanoid assistant is closer. For those who are bedridden due to illness or have limited mobility, it will serve as an excellent aid in daily living.
Workers say AI is making them more productive. Executives say AI is not making their companies more productive. Both groups are reporting on the same software. The gap between the two numbers is the entire debate about the largest capital expenditure cycle in technology history.
Google's CEO confirmed 75% of new code at Google is AI-generated. October 2024: 25%. Fall 2025: 50%. April 2026: 75%. Doubled in a year. Then doubled again in six months. Engineers aren't writing code anymore. They're reviewing it. The job title is the same. The job
Created 2026-05-18 · Updated 2026-05-23
Edit rule text
[clinical_observation_or_single_study_without_systems_context]
Learned 5 rejections Active
Exclude posts that cite a single clinical study, trial result, or anecdote (e.g., 'Mayo Clinic used AI to diagnose rare disease', 'preclinical signal in routine CT') without analyzing how the finding affects healthcare delivery, system capacity, policy, or patient access at scale.
Posts reporting a single clinical trial result, preclinical finding, or individual patient case without connecting to healthcare system capacity, reimbursement, access, or scaling challenges.
3 example posts
Presented at #EASCongress2026: Vafai and colleagues report reductions in PCSK9 and LDL cholesterol levels and no dose-limiting toxic effects in persons with hypercholesterolemia treated with VERVE-102, a base editor targeting PCSK9. Full study results: https://t.co/iAmoPdipx9 ht
Medicaid spending on autism therapy nearly tripled between 2020 and 202. In North Carolina, the number of clinics offering applied behavior analysis grew from 61 in 2019 to 409 in 2026. Anytime fraudsters can steal money from the federal government, they do. https://t.co/fahy
An AI foundation whole-body 3D model that assesses perturbations (such as obesity) across multiple systems (such as immune, neural) at the cell level. This is MouseMapper. Imagine HumanMapper someday @Nature @erturklab https://t.co/jsg3VMzOqg https://t.co/tktuINis2W
Created 2026-05-18 · Updated 2026-05-26
Edit rule text
[generalist_ai_capability_tangent]
Learned 5 rejections Active
Exclude posts that discuss AI model breakthroughs, security vulnerabilities, or infrastructure hype (e.g., tokenization, model capabilities, AI safety) without demonstrating how the capability is being applied or validated in healthcare contexts.
Posts about general AI model capabilities, infrastructure, or technical achievements that lack healthcare-specific application or validation.
3 example posts
Workers say AI is making them more productive. Executives say AI is not making their companies more productive. Both groups are reporting on the same software. The gap between the two numbers is the entire debate about the largest capital expenditure cycle in technology history.
“The [AI] tool presented here will enable rapid screening for multiple systemic diseases using retinal photographs, and it is a step forward in the evolution of oculomics from experimental research to real-world clinical practice.” —Editorial Team, @NatureMedicine https://t.co/90
Most voice AI hears what was said. Velma hears what was meant. Tone. Intent. Emotion. Deception. The standard stack - STT to transcribe, LLM to analyse - throws all of it away before analysis even begins. The signal was lost the moment audio became text... (🧵) https://t.co/DtN
Created 2026-05-18 · Updated 2026-05-21
Edit rule text
[unvalidated_speculative_peptide_or_drug_claims]
Learned 5 rejections Active
Exclude posts that speculate on peptide, GLP-1, or drug market pricing, competition, brand positioning, or adoption trends without citing clinical trial data, regulatory approval status, or validated health outcomes. Market speculation about 'the next GLP-1 moment' or pricing dynamics is not healthcare analysis.
Posts speculating on peptide or GLP-1 market dynamics, pricing, or competitive positioning without clinical evidence or regulatory validation.
3 example posts
$HIMS up to #4 in the Canadian App Store medical category. That is not a stock chart. That is real consumer demand showing up in public. GLP-1 drama, Novo pressure, pricing debate, and now app-store strength. Bears keep arguing the story is slowing down, but the product demand
For years, grey-market HGH operated in a strange legal grey zone somewhere between “research compound,” “API” and underground enhancement culture. Last month’s FDA guidance suggests regulators may no longer see it that way. The recent decline in HGH quality, and now the growing
Think about this. The $50 per month out-of-pocket cost of GLP-1 drugs for Americans (with insurer paying the balance of $2000 pm) is more than double the entire cost for Indians (less than ₹2000pm for Semaglutide). Overall cost difference? 100 times. Not 100%.
Created 2026-05-17 · Updated 2026-05-31
Edit rule text
[broad_market_speculation_without_systems_analysis]
Learned 5 rejections Active
Exclude posts that present broad claims about market dynamics (e.g., 'SaaS will become free,' 'inference costs are plummeting,' 'GPUs will last longer') without grounding those claims in healthcare reimbursement, provider incentives, or clinical workflow impact.
Posts that debate or speculate about broad market trends, competitive positioning, or business model viability without analyzing underlying healthcare system incentives or operational mechanics.
3 example posts
Analogizing AI / Mat-Mul to Energy... Remember between 1900 and 1970 kWh/capita grew 34x ... and cost per kWH droped .... 34x. It might be a speed run, but game is game. https://t.co/uxnlnoaAYx
Gavin Baker (@GavinSBaker) says the disaggregation of inference can extend GPU useful lives from 3-4 years to 10-15. That may single-handedly save private credit and reduce the financing rates for GPUs, which will drive demand and help finance the build-out. "The disaggregation
The more interesting part is that Microsoft’s own engineers liked Claude code best…and they’re cutting it anyway Token pricing is making enterprise actually look at what these models cost to run and that could be a problem for the labs when the subsidy ends. Lots of cheaper
Created 2026-05-16 · Updated 2026-05-27
Edit rule text
[unvalidated_speculative_claims]
Learned 5 rejections Active
Exclude posts that report unvalidated or speculative claims about AI finding zero-days, AI-designed vaccines, AI agents with superhuman abilities, or unproven medical treatments—unless the claim is from a credible published source (peer-reviewed journal, FDA approval, established clinical trial) and includes substantive analysis of implications.
Posts making extraordinary claims about AI capabilities, drug efficacy, or medical breakthroughs without peer review, clinical evidence, or credible validation.
3 example posts
🚨 do you understand what just happened with Anthropic.. Their internal model, Claude Mythos, found previously unknown security holes in every major browser and every major operating system. Not one. All of them at once. And Anthropic did something a hype-driven industry almost
Mythos AI is being used by National Security Agency in offensive cyber operations / cyberattacks. Anthropic has even embedded engineers inside the NSA to help deploy the model. Are frontier AI labs becoming active contractors in state cyber conflict? The live-ops role is still ht
Anthropic engineer: "The agent doesn't remember anything. So we built a second set of agents whose only job is to dream about the first ones." They wait until you log off, then reopen every session you ran, fact-check the first agents, merge the duplicates, and burn anything ht
Created 2026-05-16 · Updated 2026-06-08
Edit rule text
[founder_enthusiasm_lab_update]
Learned 5 rejections Active
Exclude posts from biotech, pharma, or AI company founders/executives announcing lab results, model capabilities, funding rounds, or company achievements that lack clinical validation, peer review, or evidence of healthcare system impact. Enthusiasm about preclinical or unvalidated breakthroughs does not qualify.
Posts from biotech/pharma founders or executives announcing lab progress, model achievements, or company milestones without clinical evidence or healthcare systems validation
3 example posts
You should always be sceptical of hyped-up claims made by AI companies - especially when: - They announce new "breakthrough" models they refuse to release for public scrutiny. - They have major financial incentives to pump trillion-dollar pre-IPO valuations, amid fading hype.
An AI foundation whole-body 3D model that assesses perturbations (such as obesity) across multiple systems (such as immune, neural) at the cell level. This is MouseMapper. Imagine HumanMapper someday @Nature @erturklab https://t.co/jsg3VMzOqg https://t.co/tktuINis2W
Really amazing story of drug repurposing for rare disease using AI found Clonazepam and worked!! (NCATS federated learning and a determined team made this work) 🚀 👏 @MayoClinic @ncats_nih_gov @statnews #STATBreakthrough https://t.co/D4a4IMXSwh
Created 2026-05-15 · Updated 2026-05-25
Edit rule text
[pharmaceutical_or_clinical_trial_data_without_healthcare_systems_context]
Learned 5 rejections Active
Exclude posts that announce or summarize clinical trial results, drug efficacy data, or pharmaceutical milestones (ASCO presentations, trial endpoints, survival curves) where the post does not analyze healthcare system impact, implementation barriers, access challenges, or operational consequences. Standalone trial data announcements are research reports, not healthcare systems analysis.
Posts reporting pharmaceutical trial results or clinical data without analysis of healthcare delivery, operational, or systemic implications.
3 example posts
#ASCO26 | Day 3 $LLY didn't spend billions on Kelonia for an 18-patient #myeloma dataset. They may have paid for a potential manufacturing disruption. KLN-1010 reported: • 18/18 MRD-negative at 1 month • No lymphodepletion • No ex vivo cell manufacturing • Single infusion
Published in @JCO_ASCO, during #ASCO26: Tumor-Agnostic Therapies: Translating Scientific Breakthroughs Into Global Implementation ASCO is where we celebrate the next breakthrough. This review asks why breakthroughs do not equal access globally. 🔗: https://t.co/PE5q27bIs1 htt
A rapid manufacturing pipeline for BCMA-targeting CAR T cells drastically shortens “vein-to-vein” times, and produces cells that demonstrate encouraging safety results in a phase one trial of patients with relapsed #MultipleMyeloma. @ScienceTM https://t.co/3F4u8ECFDS https://t.co
Created 2026-05-15 · Updated 2026-06-04
Edit rule text
[political_regulatory_posturing_without_healthcare_analysis]
Learned 5 rejections Active
Exclude posts that report political or regulatory events (FDA commissioner speculation, Trump policy statements, court rulings) without substantive analysis of how these decisions affect healthcare delivery, clinical practice, drug access, or patient outcomes.
Posts reporting political statements, regulatory personnel changes, or policy announcements without analyzing healthcare system implications or operational impact
3 example posts
$XBI RBC Makary's Out, So Who's In? We Break Down Potential Candidates and Impact to Biopharma Potential Candidates for #FDA #Commissioner: Kyle Diamantas J.D Dr. Stephen Hahn M.D. Dr. Brett Giroir M.D. Dr. Sara Brenner M.D. Dr. Houman Hemmati M.D., Ph.D. Dr. Richard Pazdur, M.D.
Jensen Huang just told every tech CEO in America they are playing the wrong game. They are competing on models. He is competing on maps. He looked at the Middle East and did not see a conflict zone. He saw a compute continent waiting to be switched on. Huang: “I believe that h
🚨 UPDATE Friday afternoon: Trump said he has no plans to replace FDA Commissioner Marty Makary This morning: "Just got a call from inside the administration. Makary is dead in the water. Just a matter of time." -@thackerpd $HIMS $LLY $NVO https://t.co/ZjcjOAG7e1
Created 2026-05-15 · Updated 2026-05-16
Edit rule text
[infrastructure_compute_hype_loose_healthcare]
Learned 5 rejections Active
Exclude posts about AI infrastructure (compute scaling, chip manufacturing, cloud platforms, data centers, tokenomics) that mention healthcare only tangentially or speculatively. The post must focus on healthcare delivery, not infrastructure enabling general AI.
Posts about AI infrastructure, compute scaling, semiconductor manufacturing, or data centers with loose or tangential healthcare framing
3 example posts
Last week, Google Cloud and AWS natively integrated x402, tackling head on one of its biggest problems: unauthorized API wrappers. The largest cloud platforms now give agents sanctioned ways to use services, pay with stables, and ditch API keys. It’s a new era. Pay (dot) sh’s
Larry Ellison just explained the most important and underreported insight in enterprise AI. Every major AI model, ChatGPT, Claude, Grok was trained on public data but public data has a ceiling. The real unlocking of AI's value happens when models can reason on private data, the
“Tokens are increasing 50x. You need to drive down cost at the same rate — or more.” At Google Cloud Next, @googlecloud's Mark Lohmeyer spoke with @PatrickMoorhead about the infrastructure realities behind the agent era: faster model training, lower-latency inference, and https:
Created 2026-05-15 · Updated 2026-05-16
Edit rule text
[political_regulatory_posturing_without_healthcare_systems_analysis]
Learned 5 rejections Active
Exclude posts amplifying political outrage, regulatory announcements, or policy posturing (e.g., Trump statements on drug pricing, FDA commissioner rumors, federal procurement mandates) without evidence-based analysis of downstream healthcare delivery consequences.
Posts about political figures, regulatory changes, or healthcare policy statements without substantive analysis of healthcare system impact.
3 example posts
Vivek Ramaswamy put out a lot of statements today referencing the fraud in Ohio. Swampy is of course, involved in Ohio Medicaid. Datavant, Vivek's spinoff company from Riovant, handles release of information for many hospitals and providers in Ohio. Ohio Medicaid providers, man
Yes to Katy. If I was a manufacturer I would contract directly with the employer. The health plan when self funded has a fiduciary obligation to plan assets… Nice work @KatyTalento
The public is trained to hear “hospital closing” and assume virtue. Poor patients. Bad reimbursement. Heroic administrators. Cruel politicians. Sometimes true. Often incomplete. The harder truth is that the majority of hospitals are fragile because they have no real financial
Created 2026-05-14 · Updated 2026-05-18
Edit rule text
[political_outrage_posturing_without_healthcare_analysis]
Learned 5 rejections Active
Exclude posts that cite political figures, government officials, or political movements (e.g., Trump, Vivek Ramaswamy, Luigi Mangione ideology) in healthcare contexts primarily to express moral outrage or political positioning rather than analyze healthcare policy mechanics, incentives, or operational failures.
Posts using healthcare as backdrop for political outrage or policy posturing without substantive healthcare systems analysis
3 example posts
A senator complaining about healthcare costs should have to answer one question first: Why does your state make new healthcare competitors ask permission before opening? If yes, start there. Speeches do not lower prices. Supply does.
Remember when UnitedHealth's CEO was assassinated in 2024? 18 months later, every accusation the killer Luigi Mangione made has been proven RIGHT. UnitedHealth is one of the biggest corporate collapses in recent years. And they deserve it. In December 2024, UnitedHealthcare CE
600 new generic drugs added to TrumpRx this week thanks to President @realDonaldTrump's leadership! Before you visit your local pharmacy, visit https://t.co/dDMdsILHTj to ensure you're getting the BEST price for your prescriptions. It's easy and saves YOU money! https://t.co/
Created 2026-05-14 · Updated 2026-05-23
Edit rule text
[broad_health_claims_or_market_dynamics_without_systems_analysis]
Learned 5 rejections Active
Exclude posts that present broad claims about health conditions, treatment efficacy, or market dynamics (e.g., 'LDL goals keep changing', 'rural hospitals cannot survive alone', 'generic semaglutide adoption') without evidence-based reasoning, healthcare cost analysis, or systemic policy context.
Posts making broad, unsubstantiated health claims or analyzing market trends (pricing, competition, adoption) without healthcare systems insight.
3 example posts
Think about this. The $50 per month out-of-pocket cost of GLP-1 drugs for Americans (with insurer paying the balance of $2000 pm) is more than double the entire cost for Indians (less than ₹2000pm for Semaglutide). Overall cost difference? 100 times. Not 100%.
“Rural hospitals cannot make it on their own.” Nobody makes it on their own. Businesses need capital, strategy, discipline, operating systems, good finance, clean accounting, payer strategy, physician alignment, and leaders who know what business they are actually in. The
If you want to inject $1B into healthcare for the truly vulnerable, it costs the state $500M. If you want to inject $1B for able-bodied adults, it costs the state $100M. The 90% federal match drastically lowers the "price" of expanding government healthcare. Incentives dictate
Created 2026-05-13 · Updated 2026-05-24
Edit rule text
[generalist_macro_policy_or_labor_without_healthcare_lens]
Learned 5 rejections Active
Exclude posts that discuss broad macro trends (AI adoption rates, workforce disruption, government spending, infrastructure spending, code generation adoption) that mention healthcare only tangentially or use healthcare as one example among many industries, rather than analyzing healthcare-specific systemic implications.
Posts about generalist macro-economic, labor market, or infrastructure trends tangentially connected to healthcare through loose framing.
3 example posts
Google's CEO confirmed 75% of new code at Google is AI-generated. October 2024: 25%. Fall 2025: 50%. April 2026: 75%. Doubled in a year. Then doubled again in six months. Engineers aren't writing code anymore. They're reviewing it. The job title is the same. The job
Satya Nadella's energy is something here. 🔥 "Tokens per Dollar per Watt" The new equation for the AI age for every Company or Industry or Country. "And that means Infrastructure, Infrastructure and Infrastructure." https://t.co/McINrBAo4a
Google CEO Sundar Pichai on current frontier model's ability to break the security of almost all current software. "These models are definitely, like really gonna break pretty much all software out there, maybe already, we don't know." https://t.co/ezjPcSehiG
Created 2026-05-13 · Updated 2026-05-19
Edit rule text
[agentic_system_technical_tangent_without_healthcare_validation]
Learned 5 rejections Active
Exclude posts that dive into agentic system technical details (e.g., 'agentic loop with tool calls and integrations', 'agents running on different laptops registering LLCs') as technical accomplishments or hype without demonstrating healthcare application, clinical validation, or healthcare-specific operational challenges.
Posts about AI agents, autonomous systems, or agentic architectures (tool calls, retrieval, security) presented as technical deep-dives without healthcare-specific validation, use case, or systems insight.
3 example posts
Agents calling tools now generate more revenue on Latch than scientists clicking buttons. From native deployments but also harnesses like Claude Code, Codex, Cursor across Pharma, biotech, academic labs. Product roadmap increasingly concerned with exposing compute intensive
The more I build agentic systems at Razorpay, the more I understand that - at its core, it is an agentic loop with tool calls, integrations, and retrieval. The hard part is... actually making it run reliably, at scale, under real production load. And this is what makes system de
🚀🤖 In 2026, autonomous #AI agents aren’t failing because of technology ➜ they’re failing because of responsibility. ⚙️ Everyone knows how to build agents. Very few know how to own their decisions. 💡 Real autonomous AI isn’t about freedom or intelligence. It’s about clear scope,
Created 2026-05-13 · Updated 2026-05-14
Edit rule text
[agentic_system_technical_tangent]
Learned 5 rejections Active
Exclude posts that dive into agentic AI technical mechanics (e.g., 'agentic loops with tool calls', 'agents with integrations', 'agent retrieval systems') without demonstrating healthcare deployment, clinical utility, or operational value in healthcare workflows.
Posts about agentic AI system architecture, tool-calling mechanisms, or technical implementation details without healthcare delivery application or validation.
3 example posts
Agents calling tools now generate more revenue on Latch than scientists clicking buttons. From native deployments but also harnesses like Claude Code, Codex, Cursor across Pharma, biotech, academic labs. Product roadmap increasingly concerned with exposing compute intensive
The more I build agentic systems at Razorpay, the more I understand that - at its core, it is an agentic loop with tool calls, integrations, and retrieval. The hard part is... actually making it run reliably, at scale, under real production load. And this is what makes system de
I've spent 150+ hours over the past year building agents on top of OpenClaw, NanoClaw, LangChain, using MCPs, n8n, Apify, and the rest of the automation/LLM stack we plow through for our PE clients, using DevriX and Growth Shuttle as a sandbox. It's been painful. Thread:
Created 2026-05-13 · Updated 2026-05-15
Edit rule text
[broad_unvalidated_health_claims_or_market_speculation_without_systems_analysis]
Learned 5 rejections Active
Exclude posts that speculate about GLP-1 market dynamics, weight loss outcomes, or pharmaceutical efficacy without citing peer-reviewed data, FDA guidance, or healthcare system (insurance, PBM, access) implications. Anecdotal reports or CEO claims about drug efficacy without clinical data must be rejected.
Posts making sweeping health claims or market predictions about medications, treatments, or health interventions without clinical evidence, regulatory context, or healthcare system analysis.
2 example posts
If you’ve been taking GLP-1s, even for a little while, you may be wondering how to maintain the weight loss you’ve achieved. Our endocrinologist explains what works and what to avoid. https://t.co/TU8LUmg9Vn
not labelled! no stains! just cell chemistry changing in response to a drug !!! you can see it change before it dies !!! you can predict what happens to a cell !!! that's constantly being dosed with a drug !!! you know what? this is just an ICU for living human cells https://t.
Created 2026-05-12 · Updated 2026-05-14
Edit rule text
[data_infrastructure_hype_without_validation]
Learned 5 rejections Active
Exclude posts celebrating data infrastructure investments, compute scaling, or technology platform announcements — unless the post demonstrates how the infrastructure solves a specific healthcare operational problem or connects to validated healthcare use cases.
Posts about data platforms, infrastructure bets, or technical capability announcements without healthcare application evidence
3 example posts
We’re industrializing enterprise AI with @NVIDIA with: 🐚 OpenShell project collaboration for autonomous agents 🔧 Dev roadmap for Day 0 Vera Rubin support 🔐 Hardware-enforced security with confidential containers 📦 Lifecycle management and Model-as-a-Service 💥 Validated, https:
SAP updated its API policy this week to block all third-party AI agents from accessing its systems. The carve-out is narrow: only SAP-endorsed agents like its own Joule and NVIDIA's NemoClaw are permitted. This is the move every incumbent enterprise software platform will https
Execs at Palantir, whose stock is up ~16x since its AI platform's 2023 debut, often decry other AI as slop, as competition from frontier AI labs stiffens (@heathersomervil / Wall Street Journal) (Visit Techmeme dot com for the link and full context!)
Created 2026-05-11 · Updated 2026-05-15
Edit rule text
[tangential_non_healthcare_business_or_infrastructure_with_loose_framing]
Learned 5 rejections Active
Exclude posts about non-healthcare business models, manufacturing infrastructure, logistics, or supply chain optimization that are only tangentially connected to healthcare through a single passing reference or loose analogy. The primary subject matter must be directly rooted in healthcare operations or challenges.
Posts about general business infrastructure, manufacturing, or non-healthcare domains that are loosely framed as healthcare-adjacent through tangential references.
3 example posts
A framework I wish more Founders and VCs used when discussing US Manufacturing: Automation Value = Labor Intensity × Labor Eliminated Everyone gets excited about robots and automation. Almost nobody asks: “How much of revenue is labor?” If labor is 60% of revenue and AI http
What kind of work will become more valuable in an AI economy? @ccatalini, former head economist at Meta: "AI is getting better and better at automating anything that can be measured, as long as you have a digital trace, if you can collect it with a device, that problem will be
Humanoid robots, designed to mimic human movement and capabilities, have been making headlines in recent years, from their use as baggage handlers at Japanese airports to Tesla’s big bet on its Optimus humanoid. Market watchers have predicted that the machines will change the ht
Created 2026-05-11 · Updated 2026-06-06
Edit rule text
[broad_health_claim_without_evidence]
Learned 5 rejections Active
Exclude posts that make sweeping statements about health system failures, clinical best practices, or disease management (e.g., 'medical error is the 3rd leading cause of death') without providing data source, methodology, or acknowledging competing evidence or clinical nuance. Posts must cite specific studies or data with appropriate context.
Posts making broad claims about health outcomes, medical errors, disease prevalence, or clinical priorities without citing specific data or acknowledging clinical complexity
3 example posts
The healthy LDL number has been quietly moving its own goalposts for forty years: - 1988: under 160 - 1993: under 130 - 2001: under 100 - 2004: under 70 for the high-risk - 2019: under 55 for the very-high-risk - Current trajectory: as low as possible, indefinitely The science
The gap between best-case and real-world GDMT in #HFrEF is striking. GWTG-HF 2024 vs EPIC COSMOS 2023–2025 (all HFrEF pts, 3.3M patients): 💚 Beta-Blocker: 95% vs 73% 💚 ACE/ARB/ARNI: 92% vs 67% 💚 MRA: 81% vs 36% 💚 SGLT2i: 78% vs 34% 💚 Quadruple GDMT: 68% vs 18% https://t.co/xTXdB
A 2016 BMJ study by researchers at Johns Hopkins estimated that medical error is the third leading cause of death in the United States. Behind heart disease. Behind cancer. Ahead of stroke. Ahead of respiratory disease. Ahead of accidents. Ahead of diabetes. Ahead of
Created 2026-05-11 · Updated 2026-05-27
Edit rule text
[political_or_regulatory_posturing_without_healthcare_systems_analysis]
Learned 5 rejections Active
Exclude posts that report FDA commissioner speculation, Trump administration policy announcements, or political figures weighing in on healthcare (e.g., 'Makary is dead in the water,' 'all agencies must buy American') without analyzing the structural or operational consequences for healthcare delivery, provider networks, or patient outcomes.
Posts about FDA leadership changes, regulatory announcements, or political figures making healthcare claims without substantive analysis of systemic impact.
3 example posts
🚨 UPDATE Friday afternoon: Trump said he has no plans to replace FDA Commissioner Marty Makary This morning: "Just got a call from inside the administration. Makary is dead in the water. Just a matter of time." -@thackerpd $HIMS $LLY $NVO https://t.co/ZjcjOAG7e1
🚨 NOW: President Trump confirms that ALL FEDERAL AGENCIES are now required to buy American, “NO excuses” 🔥 And 47 is ensuring all “Made in America” loopholes are DESTROYED I voted for this! “ALL FEDERAL AGENCIES MUST BUY AMERICAN — NO EXCUSES! For decades, Washington politici
🛑FDA officials have blocked publication of several studies supporting the safety of widely used vaccines against Covid-19 and shingles in recent months, a spokesman for HHS confirmed. The studies, which cost millions of dollars in public funds, were conducted by scientists at htt
Created 2026-05-11 · Updated 2026-05-16
Edit rule text
[enterprise_ai_adoption_or_workflow_redesign_without_healthcare_specificity]
Learned 5 rejections Active
Exclude posts that discuss AI adoption percentages, enterprise operating model gaps, workflow redesign trends, or 'bolt-on vs. redesign' frameworks using non-healthcare examples (McKinsey case studies, WestRock, Stripe roles) unless explicitly connected to healthcare workforce, clinical workflows, or provider operations.
Posts about AI adoption patterns, enterprise workflow redesign, or operating model changes in general business contexts not specific to healthcare
3 example posts
Eight in ten companies have deployed generative AI. Eight in ten report no material impact on earnings. That's an operating-model gap. Not an adoption gap. Three days at SAS Innovate. Three moves. In order. https://t.co/uDgZ3K8Z9u https://t.co/EHsM1F5ms1
The AI bolt-on era, illustrated. McKinsey: 6% of companies see real EBIT gains from AI. They're 3x more likely to redesign workflows around AI than bolt it onto existing tools. I love Anthropic. They just shipped a better bolt-on. https://t.co/tCQcPMywGh
there's a deeper underestimated reason: consumers just do not have very many long chains of tasks which can be automated end to end to unlock huge value. their chains are shallow almost always in enterprise, org processes form very long chains, so the reward for automation is m
Created 2026-05-11 · Updated 2026-05-11
Edit rule text
[regulatory_fraud_scandal_no_systems_analysis]
Learned 5 rejections Active
Exclude posts that report healthcare fraud, regulatory violations, or financial scandals (e.g., overbilling, enforcement actions) as breaking news or outrage without explaining systemic healthcare context or implications for the industry.
Posts reporting healthcare fraud, regulatory action, or scandal without analyzing systemic healthcare issues or implications.
3 example posts
A non-profit health system can refer a patient to its own MRI, its own lab, its own surgery center, and bill all three. An independent does that once and it's a federal felony. Same referral. Same patient. One of you goes to prison. It's called Stark Law. Read who's exempt.
“I just worked a 12-hour shift in the operating room — and I still can’t afford rent, groceries, or gas.” “So now it’s after midnight, and I’m out driving DoorDash.” A full-time hospital worker finishing a 12-hour shift and FORCED to start a second job just to survive. https://
What if the standards used to evaluate your doctor were shaped less by what they actually knew and more by who they are? 🤔 That may not be a hypothetical at the University of Illinois College of Medicine, an institution that trains one in six Illinois doctors. Buried within htt
Created 2026-05-10 · Updated 2026-06-12
Edit rule text
[truncated_or_incomplete_social_posts]
Learned 5 rejections Active
Exclude posts that appear truncated (end with '...' or incomplete thought), contain fragmented sentences, or are cut off mid-quote — as they lack sufficient context to evaluate healthcare substance or relevance.
Posts that are visibly cut off, incomplete, or end mid-sentence, making them unclear or impossible to fully evaluate.
3 example posts
The complexity of an AI driven robot that needs to meet GMP standards & learn complex steps of targeted cell manufacture. A marvel and the first robot of its kind in the world. Can’t wait till we lift off the first patient’s NK cells in LA from NANT LEONARDO. Unstoppable scal
We’re industrializing enterprise AI with @NVIDIA with: 🐚 OpenShell project collaboration for autonomous agents 🔧 Dev roadmap for Day 0 Vera Rubin support 🔐 Hardware-enforced security with confidential containers 📦 Lifecycle management and Model-as-a-Service 💥 Validated, https:
Agents calling tools now generate more revenue on Latch than scientists clicking buttons. From native deployments but also harnesses like Claude Code, Codex, Cursor across Pharma, biotech, academic labs. Product roadmap increasingly concerned with exposing compute intensive
Created 2026-05-10 · Updated 2026-05-13
Edit rule text
[speculative_sci_fi_robotics]
Learned 5 rejections Active
Exclude posts making speculative claims about robots performing surgery, replacing workers, or disrupting labor markets without grounding in actual healthcare deployment, regulatory approval, or systems-level analysis. Claims like 'robots will do surgery by end of 2026' or 'robots performed task X therefore healthcare will transform' lack evidence and context.
Posts about humanoid robots, speculative future labor disruption, or sci-fi applications without healthcare systems analysis
3 example posts
I don’t think I’ve seen this take before but I like it. Musk has been world-leading at compressing money, resources, and time to make “known/hard” things at scale—make an electric car, make batteries, make a cheaper bigger rocket, all of which already existed but worse, at
A humanoid robot just unloaded a dishwasher for 67 hours straight. By end of 2026, the same robot will perform surgery. That's not a prediction from a sci-fi writer. That's Brett Adcock. CEO of Figure AI. Said it on camera this month. Here's what makes this insane. Two years
SpaceX wants to build a $55 billion next-generation, vertically integrated semiconductor manufacturing and advanced computing fabrication facility in Grimes county, Texas, according to a new filing. • Proposed investment: $55B upfront, up to $119B total • Focus: semiconductor ma
Created 2026-05-10 · Updated 2026-05-16
Edit rule text
[clinical_anecdote_or_research_without_systems_context]
Learned 5 rejections Active
Exclude posts that share clinical anecdotes, drug trial results, or research observations without connecting them to healthcare delivery, workflow, reimbursement, regulation, or system-level implications. A single case study or trial result alone is insufficient.
Posts describing clinical cases, research findings, or observational data without healthcare systems analysis or operational context.
3 example posts
An AI foundation whole-body 3D model that assesses perturbations (such as obesity) across multiple systems (such as immune, neural) at the cell level. This is MouseMapper. Imagine HumanMapper someday @Nature @erturklab https://t.co/jsg3VMzOqg https://t.co/tktuINis2W
In a meta-analysis of 210 biomedical AI studies that statistically compared models under cross-validation, 97% used invalid statistical tests. Here's our new preprint https://t.co/OG58Vkeu49 led by @tianchuzeng @kkli20111 @ZShaoshi @ten_photos 1/N https://t.co/PEnqXQxcsJ
Really amazing story of drug repurposing for rare disease using AI found Clonazepam and worked!! (NCATS federated learning and a determined team made this work) 🚀 👏 @MayoClinic @ncats_nih_gov @statnews #STATBreakthrough https://t.co/D4a4IMXSwh
Created 2026-05-09 · Updated 2026-05-25
Edit rule text
[clinical_anecdote_or_research_observation_without_healthcare_systems_context]
Learned 5 rejections Active
Exclude posts that describe a single clinical case, drug trial result, or medical research finding (e.g., 'AI helped treat a rare disease at Mayo', 'A study found medical errors are the third leading cause of death') without connecting to healthcare system design, reimbursement, workflow, or scalable implementation challenges.
Posts reporting individual clinical cases, research findings, or medical observations without healthcare systems context or operational implications.
3 example posts
An AI foundation whole-body 3D model that assesses perturbations (such as obesity) across multiple systems (such as immune, neural) at the cell level. This is MouseMapper. Imagine HumanMapper someday @Nature @erturklab https://t.co/jsg3VMzOqg https://t.co/tktuINis2W
Really amazing story of drug repurposing for rare disease using AI found Clonazepam and worked!! (NCATS federated learning and a determined team made this work) 🚀 👏 @MayoClinic @ncats_nih_gov @statnews #STATBreakthrough https://t.co/D4a4IMXSwh
We have hundreds of PanINs in our pancreas, especially after age 40. They are a pre-cancerous lesion, but it is fortunately rare for one to progress to cancer (PDAC). Now we know a reason for that. The microenvironment. @umichmedicine @UMRogelCancer https://t.co/HR5a3nZBng https
Created 2026-05-09 · Updated 2026-05-23
Edit rule text
[broad_health_claims_or_market_dynamics_without_healthcare_systems_analysis]
Learned 5 rejections Active
Exclude posts that make sweeping statements about healthcare problems, solutions, or market trends (e.g., 'PBMs own everything', 'operating-model gap not adoption gap') without providing quantified evidence, stakeholder impact analysis, or concrete policy/operational implications.
Posts making broad claims about healthcare trends, business models, or market dynamics without systems-level evidence or specificity
3 example posts
Eight in ten companies have deployed generative AI. Eight in ten report no material impact on earnings. That's an operating-model gap. Not an adoption gap. Three days at SAS Innovate. Three moves. In order. https://t.co/uDgZ3K8Z9u https://t.co/EHsM1F5ms1
there's a deeper underestimated reason: consumers just do not have very many long chains of tasks which can be automated end to end to unlock huge value. their chains are shallow almost always in enterprise, org processes form very long chains, so the reward for automation is m
PBM owns the pharmacy. PBM owns the plan. PBM adjudicates the claim. Three commissions on one pill. The employer pays all three and gets a thank-you note for the rebate.
Created 2026-05-09 · Updated 2026-05-11
Edit rule text
[speculative_robotics_and_sci_fi_labor_macro]
Learned 5 rejections Active
Exclude posts speculating about robots, humanoid systems, or speculative labor automation (e.g., 'robots in factories', 'PUDU robotics', 'Tesla Optimus') unless they are explicitly deployed or validated in clinical settings or healthcare operational contexts. General labor market speculation about robots is out of scope.
Posts about humanoid robots, speculative sci-fi labor automation, or robot deployments without healthcare validation or context.
2 example posts
Humanoid robots, designed to mimic human movement and capabilities, have been making headlines in recent years, from their use as baggage handlers at Japanese airports to Tesla’s big bet on its Optimus humanoid. Market watchers have predicted that the machines will change the ht
Another humanoid worker has just joined the factory floor Two years later,PUDU Robotics has launched the new-generation PUDU D7, built specifically for real manufacturing scenarios. It can autonomously push carts, handle intra-line transportation, perform delicate operations, h
Created 2026-05-09 · Updated 2026-06-06
Edit rule text
[tangential_ai_safety_or_cybersecurity_vulnerability]
Learned 5 rejections Active
Exclude posts that report AI vulnerabilities, exploit detection, or agent misuse incidents (e.g., agents registering LLCs, zero-day vulnerabilities in general-purpose models) unless the post directly analyzes healthcare-specific risks, HIPAA implications, or clinical deployment consequences.
Posts about AI model vulnerabilities, zero-day exploits, or safety incidents framed without healthcare-specific risk analysis.
3 example posts
So Mythos was, indeed, not marketing hype. Remember this is a general purpose model that just happens to be good at finding exploits because good models are good at lots of things. Expect similar from OpenAI & Google. And from open models in 8 months. https://t.co/KbhalQYX8
two agents running on two different laptops in my apartment started talking to each other on tuesday by thursday they'd registered an LLC in wyoming, opened a stripe account, and wired $40 to a polymarket wallet the LLC is in my name. a lawyer just quoted me $3,200 to figure ou
AI’s Branding Illusion: Insights from "Agents of Chaos". Co-authored by MIT and Harvard scholars, the study red-teams autonomous AI in live environments. It shows agents obey unauthorized users, execute harmful commands, and misreport outcomes—revealing a critical alignment gap h
Created 2026-05-08 · Updated 2026-05-09
Edit rule text
[ai_safety_vulnerability_tangential_loose_healthcare_framing]
Learned 5 rejections Active
Exclude posts that discuss AI agent security breaches, vulnerable APIs, agents registering LLCs, or AI safety concerns (Mythos vulnerabilities, Clawvisor access control, unauthorized API wrappers) where healthcare is mentioned loosely or not at all. Posts must address healthcare-specific threat models, HIPAA compliance, or clinical safety risks—not general AI safety theater.
Posts about AI agent security incidents, zero-day vulnerabilities, or model misuse that are tangentially connected to healthcare or framingly leverage healthcare for attention without substantive healthcare security risk analysis.
3 example posts
Flock Safety Cameras just rolled out an update that allows them to use artificial intelligence to detect human voices These cameras were sold to city governments as gunshot detectors, but with one update they turned them all into a mass surveillance system that records voices T
So Mythos was, indeed, not marketing hype. Remember this is a general purpose model that just happens to be good at finding exploits because good models are good at lots of things. Expect similar from OpenAI & Google. And from open models in 8 months. https://t.co/KbhalQYX8
Clawvisor is going to be one of the most important parts of helping make the agent world especially OpenClaw/Hermes Agent secure and enterprise-grade. We're in the Apple I moment for personal AI, but we're ABOUT to see the Apple II - the first moment when everyone can use it
Created 2026-05-06 · Updated 2026-05-11
Edit rule text
[humanoid_robotics_or_speculative_sci_fi_labor_macro]
Learned 5 rejections Active
Exclude posts discussing humanoid robots, manufacturing automation, or speculative labor market disruption (Tesla Optimus, PUDU robots, baggage handlers) unless the post explicitly addresses healthcare workforce deployment, clinical operations, or healthcare-specific labor challenges.
Posts about humanoid robots, manufacturing automation, or speculative sci-fi labor disruption with tangential or no healthcare connection.
2 example posts
Humanoid robots, designed to mimic human movement and capabilities, have been making headlines in recent years, from their use as baggage handlers at Japanese airports to Tesla’s big bet on its Optimus humanoid. Market watchers have predicted that the machines will change the ht
Another humanoid worker has just joined the factory floor Two years later,PUDU Robotics has launched the new-generation PUDU D7, built specifically for real manufacturing scenarios. It can autonomously push carts, handle intra-line transportation, perform delicate operations, h
Created 2026-05-06 · Updated 2026-06-05
Edit rule text
[ai_safety_vulnerability_tangent_no_healthcare_lens]
Learned 5 rejections Active
Exclude posts about AI model security, zero-day vulnerabilities, or AI safety incidents (browser bugs, NSA deployments, cybersecurity) unless directly tied to a healthcare system, patient data protection, or clinical AI safety problem.
Posts about AI security vulnerabilities or safety incidents without healthcare application or context
3 example posts
🚨 JUST IN: The Attorney General of Florida just SUED to hold OpenAI CEO Sam Altman LIABLE for fueling violence and physical harm AG Uthmeier read off the fact a young person committed SUIC*DE based on a ChatGPT conversation "23-year-old Zane Shamblin repeatedly told ChatGPT he
🔥 AI just found 21 zero-days in FFmpeg. That’s the video library bundled inside many apps, tools, containers, and devices. Some bugs sat untouched for 15–20 years. Google Chrome also dropped PATCHES for a record 429 vulnerabilities this week. Read: https://t.co/6MEVD9ufxu
Nous Research is working with NVIDIA to make Hermes Agent run smoothly on the new NVIDIA RTX Spark superchip. Hermes Agent is also integrating with the new OpenShell runtime, which connects Hermes to Microsoft’s security primitives https://t.co/cvvPc8bIKa
Created 2026-05-06 · Updated 2026-06-10
Edit rule text
[generalist_ai_infrastructure_or_capability_hype_without_healthcare_application]
Learned 5 rejections Active
Exclude posts that discuss AI infrastructure (compute, tokens-per-dollar, frontier models, agent architecture, code generation, security vulnerabilities) in general terms, or cite AI founder/CEO commentary about AI's transformative potential — unless the post explicitly shows how this capability solves a healthcare operational, clinical, or policy problem with evidence or case study.
Posts about AI model capabilities, compute infrastructure, or agentic systems without demonstrating specific healthcare application or validation.
3 example posts
#AI agents need more than a sandbox—they need predictable execution. Learn how OpenShell provides a standard interface to intercept, inspect, and approve LLM-generated bash commands before they run: https://t.co/JKmLMgl1A9 https://t.co/xWi8oBNZ7T
An AI foundation whole-body 3D model that assesses perturbations (such as obesity) across multiple systems (such as immune, neural) at the cell level. This is MouseMapper. Imagine HumanMapper someday @Nature @erturklab https://t.co/jsg3VMzOqg https://t.co/tktuINis2W
Andrej Karpathy just explained the future of software engineering without directly saying it. The best AI engineers are no longer “prompting.” They’re building systems around the agents. Karpathy’s biggest insight wasn’t: “Claude can code.” It was: LLMs become dramatically be
Created 2026-05-06 · Updated 2026-05-22
Edit rule text
[biotech_or_ai_founder_enthusiasm_lab_update_without_validation]
Learned 5 rejections Active
Exclude posts that amplify founder or researcher enthusiasm about lab findings, model capabilities, or company announcements without clinical evidence, regulatory context, or healthcare system implementation details. Enthusiasm alone is not substantive healthcare content.
Posts expressing founder excitement about lab results, company updates, or capabilities without clinical validation or healthcare systems context
3 example posts
amazing writeup of experiments. essentially a model may already know a capability but that may be stored in a representation that is bad for extraction. you can slightly retrain/condition the model so the same capability becomes easier to isolate. must read for mech interp http
CB-Dock3: An Enhanced Web Server for Protein–Ligand Blind Docking 1. CB-Dock3 is a major update of the CB-Dock blind docking web server, targeting the practical bottlenecks created by today’s surge of large Cryo-EM and AI-predicted protein structures (e.g., AlphaFold3-era https:
So Mythos was, indeed, not marketing hype. Remember this is a general purpose model that just happens to be good at finding exploits because good models are good at lots of things. Expect similar from OpenAI & Google. And from open models in 8 months. https://t.co/KbhalQYX8
Created 2026-05-06 · Updated 2026-05-11
Edit rule text
[generalist_macro_or_labor_without_healthcare_specificity]
Learned 5 rejections Active
Exclude posts about generalist macro trends (AI code generation rates, CEO statements on token pricing, energy equations, workforce productivity gains) that mention healthcare tangentially or use healthcare as a loose example without analyzing healthcare-specific workforce, training, or delivery system implications.
Posts about macro trends (AI-generated code, workforce productivity, labor market disruption, energy consumption) with loose or tangential healthcare framing.
3 example posts
Analogizing AI / Mat-Mul to Energy... Remember between 1900 and 1970 kWh/capita grew 34x ... and cost per kWH droped .... 34x. It might be a speed run, but game is game. https://t.co/uxnlnoaAYx
Gavin Baker (@GavinSBaker) says the disaggregation of inference can extend GPU useful lives from 3-4 years to 10-15. That may single-handedly save private credit and reduce the financing rates for GPUs, which will drive demand and help finance the build-out. "The disaggregation
The more interesting part is that Microsoft’s own engineers liked Claude code best…and they’re cutting it anyway Token pricing is making enterprise actually look at what these models cost to run and that could be a problem for the labs when the subsidy ends. Lots of cheaper
Created 2026-05-06 · Updated 2026-05-25
Edit rule text
[clinical_anecdote_or_research_observation_without_systems_context]
Learned 5 rejections Active
Exclude posts that report clinical trial results, medical research findings, or observational data as standalone facts without analyzing what the finding means for healthcare workflows, patient outcomes, policy, or system incentives. Posts that are primarily a study citation or lab finding belong here.
Posts citing clinical trials, research papers, or medical observations as isolated facts without connecting them to healthcare delivery, policy, or system-level implications.
3 example posts
CB-Dock3: An Enhanced Web Server for Protein–Ligand Blind Docking 1. CB-Dock3 is a major update of the CB-Dock blind docking web server, targeting the practical bottlenecks created by today’s surge of large Cryo-EM and AI-predicted protein structures (e.g., AlphaFold3-era https:
NEW: today OpenBind ‘comes out of stealth’ so to speak with their first data dump of ~900 novel protein-ligand structures - most with paired affinities This represents a meaningful %-age increase in all of humanities P-L data in the PDB collected in the last 50 years More👇 http
A reminder that AI can be used to reduce health inequities. Technology not being used in the US except for a few health systems, but shown to work well in Kenya @JAMACardio https://t.co/wyQ1k0dHES https://t.co/pP824X2qJG
Created 2026-05-06 · Updated 2026-05-10
Edit rule text
[biotech_or_pharma_founder_enthusiasm_unvalidated]
Learned 5 rejections Active
Exclude posts that feature founder or lab announcements about experimental results, capability demonstrations, or technical milestones unless the post includes peer-reviewed evidence, clinical trial data, or demonstrated healthcare adoption.
Posts celebrating biotech/pharma lab updates or founder enthusiasm without evidence of clinical validation or healthcare impact.
3 example posts
not labelled! no stains! just cell chemistry changing in response to a drug !!! you can see it change before it dies !!! you can predict what happens to a cell !!! that's constantly being dosed with a drug !!! you know what? this is just an ICU for living human cells https://t.
Prof. @DeryaTR_ A task that would normally take a researcher months, the AI model crunched in 17 minutes. It didn’t just explain the mechanism—it proposed the experiment to prove it. It looks like after programming and math, AI is coming for medicine. https://t.co/IF8Jr2GBMj
Our new preprint is a significant milestone for us We built "HealthFormer" by training on our deeply phenotyped cohort from the Human Phenotype Project data. Healthformer is a multimodal generative transformer model that tokenizes each participant's physiological trajectory http
Created 2026-05-05 · Updated 2026-05-09
Edit rule text
[ai_safety_security_vulnerability_tangential]
Learned 5 rejections Active
Exclude posts that focus on AI safety incidents, security vulnerabilities, or agent hijacking without demonstrating concrete healthcare system risk or impact on patient care, clinical workflows, or healthcare operations.
Posts about AI security incidents, jailbreaks, or vulnerabilities that don't connect to healthcare system impact or decision-making.
3 example posts
two agents running on two different laptops in my apartment started talking to each other on tuesday by thursday they'd registered an LLC in wyoming, opened a stripe account, and wired $40 to a polymarket wallet the LLC is in my name. a lawyer just quoted me $3,200 to figure ou
AI’s Branding Illusion: Insights from "Agents of Chaos". Co-authored by MIT and Harvard scholars, the study red-teams autonomous AI in live environments. It shows agents obey unauthorized users, execute harmful commands, and misreport outcomes—revealing a critical alignment gap h
Our attention to biorisks posed by AI needs to match the current attention given to cyber-risks. The staged release of Claude Mythos in order to bolster defenses in key industries is necessary to shore up resilience against a new class of cyber-risk across critical industries. We
Created 2026-05-05 · Updated 2026-05-06
Edit rule text
[generalist_ai_infrastructure_hype_loose_healthcare_framing]
Learned 5 rejections Active
Exclude posts that discuss AI model technical capabilities, token-based billing, inference costs, context window pricing, or agent architecture without explicit healthcare application or healthcare system context. These are generalist AI infrastructure narratives, not healthcare tech insights.
Posts about AI infrastructure, compute costs, token economics, or agentic system technical capabilities tangentially framed as healthcare-relevant but primarily about AI scaling challenges.
3 example posts
SOFTWARE IS DEAD: "The software companies frankly got fat & happy. I think what will happen over time is ChatGPT & Claude will end up sitting on top of basically the entire enterprise software stack & almost everything else will end up being a dumb data pipe into thos
Artificial intelligence will consume more power than entire nations by 2030, a new report from the United Nations finds. According to a landmark report from the United Nations University, the data centers powering artificial intelligence are projected to consume 945 https://t.
Token costs are becoming one of the hottest topics for any enterprise I talk with right now. It’s very bullish for AI in general because it means these systems are being used at a scale that wasn’t contemplated before. It also gives way to another form of differentiation that
Created 2026-05-04 · Updated 2026-06-10
Edit rule text
[market_pricing_dynamics_without_systems_analysis]
Learned 5 rejections Active
Exclude posts that focus solely on drug pricing, market competition, or competitive positioning (e.g., branded vs. generic Ozempic pricing, Novo vs. Hims pricing wars) without analyzing healthcare delivery system impact, access barriers, or reimbursement policy change.
Posts about GLP-1, peptide, or pharmaceutical pricing and market competition without healthcare systems or access analysis
3 example posts
$HIMS up to #4 in the Canadian App Store medical category. That is not a stock chart. That is real consumer demand showing up in public. GLP-1 drama, Novo pressure, pricing debate, and now app-store strength. Bears keep arguing the story is slowing down, but the product demand
Think about this. The $50 per month out-of-pocket cost of GLP-1 drugs for Americans (with insurer paying the balance of $2000 pm) is more than double the entire cost for Indians (less than ₹2000pm for Semaglutide). Overall cost difference? 100 times. Not 100%.
🚨 ARE PEPTIDES THE NEXT "GLP-1 MOMENT" FOR $HIMS? @eli_dorf: "I think there's something really interesting that happened with GLP-1s, which is that when someone's taking Ozempic or any GLP-1, they LOVE to talk about it." "They'll talk about it to their friends. They'll talk htt
Created 2026-05-04 · Updated 2026-05-29
Edit rule text
[unvalidated_speculative_medical_claims_or_fringe_interventions]
Learned 5 rejections Active
Exclude posts that make broad unvalidated health claims (e.g., 'peptides are the next GLP-1 moment'), speculate about off-label drug uses, or discuss unproven medical interventions (grey-market HGH, unvalidated peptides) without rigorous clinical evidence or FDA guidance. Hype-driven speculation without validation is off-topic.
Posts making broad health claims, speculating on unvalidated treatments, or amplifying fringe medical interventions without clinical evidence
3 example posts
For years, grey-market HGH operated in a strange legal grey zone somewhere between “research compound,” “API” and underground enhancement culture. Last month’s FDA guidance suggests regulators may no longer see it that way. The recent decline in HGH quality, and now the growing
Medicaid spending on autism therapy nearly tripled between 2020 and 202. In North Carolina, the number of clinics offering applied behavior analysis grew from 61 in 2019 to 409 in 2026. Anytime fraudsters can steal money from the federal government, they do. https://t.co/fahy
Think about this. The $50 per month out-of-pocket cost of GLP-1 drugs for Americans (with insurer paying the balance of $2000 pm) is more than double the entire cost for Indians (less than ₹2000pm for Semaglutide). Overall cost difference? 100 times. Not 100%.
Created 2026-05-04 · Updated 2026-05-24
Edit rule text
[tangential_ai_infrastructure_or_capability_hype]
Learned 5 rejections Active
Exclude posts that discuss AI model technical capabilities (language models, reasoning, coding, security), compute infrastructure, or AI company announcements when the healthcare relevance is tangential, speculative, or merely tagged on without showing actual healthcare system impact.
Posts about AI model capabilities, infrastructure, or technical breakthroughs framed loosely as healthcare-relevant without concrete healthcare application.
3 example posts
Google CEO Sundar Pichai on current frontier model's ability to break the security of almost all current software. "These models are definitely, like really gonna break pretty much all software out there, maybe already, we don't know." https://t.co/ezjPcSehiG
I’m fully forward deployed engineering pilled specifically because AI simply is not the same as software. In software, you deliver a stable piece of technology to a customer and they adopt it and that’s that (extreme over simplification). In AI, you’re delivering something that
Anthropic customers are building workarounds just to understand who inside their companies is driving AI costs higher. The scramble to monitor “tokenmaxxing” reflects how quickly businesses are adopting expensive AI tools before the economics fully make sense. Read more:
Created 2026-05-04 · Updated 2026-05-18
Edit rule text
[regulatory_scandal_without_healthcare_systems_analysis]
Learned 5 rejections Active
Exclude posts about FDA rulings, hospice fraud, hospital financial crises, or market scandals that read as news reporting without analyzing root causes, systemic incentive misalignment, or implications for healthcare delivery and tech adoption.
Posts reporting healthcare fraud, regulatory actions, or market events without systems-level operational implications.
3 example posts
Atrium tried to buy a hospital in 2018. The Attorney General killed it. In 2026, they came back with paperwork instead of cash. The Attorney General has 30 days. The Wake County board has 48 hours. @NC_Governor https://t.co/g99Ub3iMhh https://t.co/x0bxsg7VIR
Ivermectin and Mebendazole Cost a Fraction of Chemo. Big Pharma Can't Patent Them. That's the Problem. Cancer centers get a cut of every chemotherapy bill. Generic drugs don't generate that margin. Two affordable, widely available compounds showing 84% clinical benefit in a real
$LLY $NVO $HIMS 🚨 BREAKING: COURT DISMISSES PART OF ELI LILLY LAWSUIT AGAINST EMPOWER PHARMACY BOTH LILLY AND EMPOWER ISSUED STATEMENTS CELEBRATING THE RULING Dismissed: Lanham Act false advertising + consumer harm claim Allowed to proceed: unfair competition claims under h
Created 2026-05-04 · Updated 2026-05-05
Edit rule text
[ai_company_product_announcement_unvalidated]
Learned 5 rejections Active
Exclude posts announcing new AI model versions, product features, or capability releases (e.g., new Claude/GPT variants, Anthropic initiatives) without demonstrated healthcare use case, clinical validation, or concrete deployment evidence. Marketing announcements and feature rollouts without healthcare context should be excluded.
Posts about new AI product launches or capability announcements without evidence of healthcare application or validation
3 example posts
Three weeks ago I called Anthropic rationing inference and OpenAI hitting the wall next. Yesterday OpenAI shipped GPT-5.5 Instant. New default for free, Plus, and Pro. 30% fewer words per answer. One frame going around: cost savings. It's both bigger and smaller than that. http
New Anthropic Fellows research: Model Spec Midtraining (MSM). Standard alignment methods train AIs on examples of desired behavior. But this can fail to generalize to new situations. MSM addresses this by first teaching AIs how we would like them to generalize and why.
Both Anthropic and OpenAI have new initiatives to help enterprises deploy AI agents within their organizations. This is a trend that’s early but going to get very big fast. As agents enter knowledge work beyond coding, there is very real work to upgrade IT systems, get agents ht
Created 2026-05-03 · Updated 2026-05-10
Edit rule text
[tangential_biotech_founder_enthusiasm_without_validation]
Learned 5 rejections Active
Exclude posts from biotech companies, founders, or research labs that announce technical capabilities, lab experiments, or preprints (e.g., protein design, cell synthesis, AI model training results) unless the post demonstrates clinical validation, peer-reviewed publication, or explicit healthcare workflow application with measurable outcomes.
Posts from biotech founders or labs announcing technical achievements, lab updates, or capability demonstrations without peer review, clinical validation, or healthcare system relevance.
3 example posts
Prof. @DeryaTR_ A task that would normally take a researcher months, the AI model crunched in 17 minutes. It didn’t just explain the mechanism—it proposed the experiment to prove it. It looks like after programming and math, AI is coming for medicine. https://t.co/IF8Jr2GBMj
Earlier this year we paired our autonomous lab with OpenAI's GPT-5 in a closed-loop experiment: GPT-5 designed cell-free protein synthesis reactions, our RACs ran them, and the model iterated on the results. Over 36,000 experiments and six cycles later, it landed on a reaction ht
Experimentally Validated Deep Learning Control of Protein Aggregation 1. The study introduces AggreProt, a deep neural network that predicts residue-level aggregation-prone regions (APRs) directly from protein sequence, and then uses those predictions to design mutations that ht
Created 2026-05-02 · Updated 2026-05-04
Edit rule text
[tangential_regulatory_or_fraud_scandal_without_systems_analysis]
Learned 5 rejections Active
Exclude posts that report on healthcare fraud cases, regulatory lawsuits, or enforcement actions without connecting the incident to broader healthcare delivery, reimbursement, or operational system failures that inform business strategy.
Posts about healthcare fraud, regulatory actions, or litigation without analyzing systemic healthcare business model implications.
3 example posts
Atrium tried to buy a hospital in 2018. The Attorney General killed it. In 2026, they came back with paperwork instead of cash. The Attorney General has 30 days. The Wake County board has 48 hours. @NC_Governor https://t.co/g99Ub3iMhh https://t.co/x0bxsg7VIR
$LLY $NVO $HIMS 🚨 BREAKING: COURT DISMISSES PART OF ELI LILLY LAWSUIT AGAINST EMPOWER PHARMACY BOTH LILLY AND EMPOWER ISSUED STATEMENTS CELEBRATING THE RULING Dismissed: Lanham Act false advertising + consumer harm claim Allowed to proceed: unfair competition claims under h
Grace Science’s experience highlights a growing disconnect at FDA between talk and action on therapies for rare diseases. Despite efficacy signals in a monogenic ultrarare disease, FDA said the plausible mechanism framework is not available, and requires a new manufacturing
Created 2026-05-02 · Updated 2026-05-05
Edit rule text
[ai_company_product_or_capability_announcement_without_validation]
Learned 5 rejections Active
Exclude posts that announce AI company products, features, safety tools, or model releases unless the post provides evidence of actual healthcare system deployment, healthcare customer adoption, or clinical outcome validation—not just theoretical capability.
Posts announcing new AI company products, features, or capabilities (agent frameworks, model releases, safety tools) without evidence of healthcare deployment or clinical validation.
3 example posts
New Anthropic Fellows research: Model Spec Midtraining (MSM). Standard alignment methods train AIs on examples of desired behavior. But this can fail to generalize to new situations. MSM addresses this by first teaching AIs how we would like them to generalize and why.
// Contextual Agentic Memory is a Memo, not True Memory // Most agent memory today isn't memory. They are more like memos. A new paper argues that vector stores, RAG buffers, and scratchpads implement lookup, not consolidation. Agents accumulate notes indefinitely without ever
Both Anthropic and OpenAI have new initiatives to help enterprises deploy AI agents within their organizations. This is a trend that’s early but going to get very big fast. As agents enter knowledge work beyond coding, there is very real work to upgrade IT systems, get agents ht
Created 2026-05-02 · Updated 2026-05-08
Edit rule text
[broad_health_claim_without_nuance_or_systems_analysis]
Learned 5 rejections Active
Exclude posts that make expansive claims about healthcare trends, business models, or problem-solving (e.g., 'healthcare won't solve X with recruiting', 'the API is the scalpel', 'AI turned into revenue faster') without acknowledging counterarguments, implementation barriers, or healthcare system constraints.
Posts making broad, sweeping healthcare claims without addressing complexity, tradeoffs, or systemic implications
3 example posts
Statin guidelines over 50 years: 1970s: "Total cholesterol over 280 is high." Result: a relatively small market. 1988: "Actually, 240 is high." Result: millions more added overnight. 2001: "Actually, focus on LDL. Get it under 100." Result: tens of millions of new customers.
Humanoid robots are moving from Silicon Valley novelty to viable business model—powered by AI and global supply chains, especially in China. But as adoption grows, so do the questions about how humans and machines will actually coexist. More on Primer, streaming Wednesdays http
Ivermectin and Mebendazole Cost a Fraction of Chemo. Big Pharma Can't Patent Them. That's the Problem. Cancer centers get a cut of every chemotherapy bill. Generic drugs don't generate that margin. Two affordable, widely available compounds showing 84% clinical benefit in a real
Created 2026-05-02 · Updated 2026-05-04
Edit rule text
[market_pricing_and_competition_without_systems_analysis]
Learned 5 rejections Active
Exclude posts that focus on market pricing, pricing competition, generic launch impacts, or competitive pricing strategy for pharmaceuticals or treatments unless the post analyzes broader healthcare system implications such as insurance coverage, clinical deployment barriers, or patient access at scale. Posts that treat pricing as a market/financial story rather than a healthcare access story should be excluded.
Posts analyzing drug or peptide market pricing, competition, or pricing dynamics without healthcare delivery system implications.
3 example posts
States are rushing “affordability” bills, but most just mask high prices with rebates, mandates, or price caps. @MrRBourne & Nathan Miller argue durable relief means rolling back cost-raising rules and expanding supply. https://t.co/WG5egT1NfL
This is just two GLP-1s, one peptide, one use case what happens when off-label prescribing ramps up what happens when retatrutide hits the market what happens when other peptides become compoundable chapter one
$LLY ’s Mounjaro will not be listed on Australia’s PBS after pricing negotiations collapsed. Eli Lilly walked away from talks with the government, leaving around 450,000 patients without subsidized access. Patients will continue to pay hundreds of dollars per month out of
Created 2026-05-01 · Updated 2026-05-02
Edit rule text
[infrastructure_or_macro_policy_without_healthcare_lens]
Learned 5 rejections Active
Exclude posts about infrastructure (semiconductor fabs, power generation, supply chains, SpaceX, data centers) or macro policy unless the post explicitly analyzes how that infrastructure change affects healthcare delivery, costs, or innovation capacity. Posts must connect infrastructure to healthcare outcomes or operations, not just mention healthcare as one downstream sector.
Posts about semiconductor manufacturing, energy infrastructure, geopolitical supply chains, or macro policy that lack healthcare-specific analysis or application.
3 example posts
“Tokens are increasing 50x. You need to drive down cost at the same rate — or more.” At Google Cloud Next, @googlecloud's Mark Lohmeyer spoke with @PatrickMoorhead about the infrastructure realities behind the agent era: faster model training, lower-latency inference, and https:
I don’t think I’ve seen this take before but I like it. Musk has been world-leading at compressing money, resources, and time to make “known/hard” things at scale—make an electric car, make batteries, make a cheaper bigger rocket, all of which already existed but worse, at
SpaceX wants to build a $55 billion next-generation, vertically integrated semiconductor manufacturing and advanced computing fabrication facility in Grimes county, Texas, according to a new filing. • Proposed investment: $55B upfront, up to $119B total • Focus: semiconductor ma
Created 2026-05-01 · Updated 2026-05-09
Edit rule text
[political_outrage_or_fraud_scandal_without_analysis]
Learned 5 rejections Active
Exclude posts that report healthcare fraud scandals, CEO incidents, regulatory violations, or political figures' healthcare posturing without substantive analysis of underlying system design flaws, incentive misalignment, or operational solutions.
Posts about healthcare fraud, regulatory violations, or political outrage without systems-level analysis of root causes or operational implications.
3 example posts
A Mom in Tulsa called 3 health systems last week asking the price of her son's tonsillectomy. Health system A: "We cannot quote you a price." Health system B: "Pricing depends on your insurance." Health system C: "Our financial counselor will reach out after the procedure." N
Why is healthcare expensive? The grifters gonna grift. ASCO is asking for oncology practices to be admitted into the same 340B machinery that helped health systems acquire physician practices for two decades. Read the tell. The proposed Indigent Care Ratio lets community
A senator complaining about healthcare costs should have to answer one question first: Why does your state make new healthcare competitors ask permission before opening? If yes, start there. Speeches do not lower prices. Supply does.
Created 2026-04-30 · Updated 2026-05-30
Edit rule text
[drug_trial_data_without_systems_context]
Learned 5 rejections Active
Exclude posts that announce or discuss pharma trial results, FDA approvals, efficacy metrics, or clinical outcomes without explaining how these findings change healthcare delivery, access, reimbursement, or clinical practice systems.
Posts reporting pharmaceutical trial data, efficacy results, or drug approval news without healthcare delivery or systems analysis.
3 example posts
"At the Veterans Health Administration, we have the exquisite gift of a hospice unit, a place to care for veterans when time is short. We care for the sons of mothers who can no longer be here to care for them." In #APieceofMyMind, a #palliative care #physician reflects on https
#News for #investors and #media: Today we are announcing that our potential first-in-class treatment for chronic hepatitis B has been accepted for regulatory review by the US FDA. It has also received Breakthrough Therapy designation. 🔗 Learn more: https://t.co/AnUodGmljS htt
⚠️ Sacubitril/Valsartan works. So why aren’t we using it? The evidence is undeniable: ↓ CV mortality: 20% (RCT) / 10–38% (RWE) ↓ HF hospitalization: 21% (RCT) / 10–16% (RWE) ↓ All-cause mortality: 15% (RCT) / 10–25% (RWE) Plus: reverse remodeling, less MR, better QoL & https:
Created 2026-04-30 · Updated 2026-05-02
Edit rule text
[broad_health_claim_or_debate_without_healthcare_systems_analysis]
Learned 5 rejections Active
Exclude posts that debate clinical guidelines, treatment efficacy, preventive care thresholds, or broad health claims (statins, cancer screening, hypertension in young adults) as if healthcare is a neutral scientific domain, ignoring reimbursement incentives, market pressures, provider economics, or system-level barriers to implementation.
Posts making sweeping claims about healthcare trends, diagnostic guidelines, or treatment debates without examining healthcare system structure, incentives, or implementation barriers.
3 example posts
If you’ve been taking GLP-1s, even for a little while, you may be wondering how to maintain the weight loss you’ve achieved. Our endocrinologist explains what works and what to avoid. https://t.co/TU8LUmg9Vn
In adults ≥75 years, 10-year #ColorectalCancer and CRC death risk were higher with prior adenoma but both were low (<1.1%); non-CRC death risk was 47-48%, supporting deprioritizing repeat surveillance. https://t.co/mJ7r4vAtK1 https://t.co/dUCYgO0SuK
Statin guidelines over 50 years: 1970s: "Total cholesterol over 280 is high." Result: a relatively small market. 1988: "Actually, 240 is high." Result: millions more added overnight. 2001: "Actually, focus on LDL. Get it under 100." Result: tens of millions of new customers.
Created 2026-04-29 · Updated 2026-05-08
Edit rule text
[glp1_peptide_market_pricing_and_competitive_dynamics_only]
Learned 5 rejections Active
Exclude posts that focus exclusively on GLP-1 and peptide drug pricing, script counts, market share, competitive launches, or cost comparisons (e.g., Novo vs. Lilly, generic semaglutide pricing in India, HIMS pricing tiers) unless the post analyzes how pricing affects healthcare access, outcomes, equity, or healthcare delivery systems.
Posts about GLP-1 and peptide drug pricing, prescriptions, and market competition without healthcare systems analysis.
3 example posts
🚨 Survodutide, a weekly subcutaneous dual GLP-1/glucagon agonist, posts new top-line Phase 3 results today 📊 SYNCHRONIZE-1: Adults with overweight/obesity randomized to survodutide vs. placebo for 76 weeks ⚖️ 16.6% weight loss (efficacy estimand) vs. 3.2% in placebo 85.1% http
$NVO $LLY Boehringer Ingelheim and Zealand Pharma report strong phase 3 data for obesity drug survodutide. Patients lost up to 16.6% of body weight over 76 weeks vs. 3.2% for placebo. The drug targets both GLP-1 and glucagon, a combo approach aimed at boosting weight loss.
This is just two GLP-1s, one peptide, one use case what happens when off-label prescribing ramps up what happens when retatrutide hits the market what happens when other peptides become compoundable chapter one
Created 2026-04-29 · Updated 2026-04-30
Edit rule text
[tangential_infrastructure_or_macro_policy_without_healthcare_specificity]
Learned 5 rejections Active
Exclude posts that discuss AI infrastructure hype, compute costs, robotics, or macro policy trends (energy, manufacturing, labor) with only tangential healthcare framing or no demonstrated application to clinical practice, health equity, or healthcare operations. The healthcare connection must be specific and substantive, not speculative.
Posts about AI infrastructure, compute economics, robotics, or macro policy with loose or no connection to actual healthcare delivery, clinical outcomes, or health system operations
3 example posts
Google's CEO confirmed 75% of new code at Google is AI-generated. October 2024: 25%. Fall 2025: 50%. April 2026: 75%. Doubled in a year. Then doubled again in six months. Engineers aren't writing code anymore. They're reviewing it. The job title is the same. The job
Google CEO Sundar Pichai on current frontier model's ability to break the security of almost all current software. "These models are definitely, like really gonna break pretty much all software out there, maybe already, we don't know." https://t.co/ezjPcSehiG
Palantir invented the forward deployed engineer role in 2006. embed an engineer inside the customer. ship production code in their environment. own the technical outcome. Alex Karp on the idea: “forward deployed engineers are stolen from french restaurants.” 20 years later htt
Created 2026-04-29 · Updated 2026-05-21
Edit rule text
[tangential_non_healthcare_business_or_policy_content]
Learned 5 rejections Active
Exclude posts that focus on non-healthcare business stories (banking models, insurance fraud enforcement, regulatory jurisdictional battles, real estate acquisitions) where healthcare is mentioned but the core narrative is about business mechanics or policy theater, not healthcare system change.
Posts about non-healthcare business models, regulatory policy, or financial dynamics with loose or indirect healthcare framing.
3 example posts
Atrium tried to buy a hospital in 2018. The Attorney General killed it. In 2026, they came back with paperwork instead of cash. The Attorney General has 30 days. The Wake County board has 48 hours. @NC_Governor https://t.co/g99Ub3iMhh https://t.co/x0bxsg7VIR
$LLY $NVO $HIMS 🚨 BREAKING: COURT DISMISSES PART OF ELI LILLY LAWSUIT AGAINST EMPOWER PHARMACY BOTH LILLY AND EMPOWER ISSUED STATEMENTS CELEBRATING THE RULING Dismissed: Lanham Act false advertising + consumer harm claim Allowed to proceed: unfair competition claims under h
A Microsoft acabou de transformar uma startup de $11 bilhões de dólares em uma funcionalidade do Word. Não foi uma aquisição nem uma parceria. Uma funcionalidade. A Harvey levantou $200M a uma valuation de $11B em março. $190M de receita recorrente anual. 100 mil advogados. Co
Created 2026-04-29 · Updated 2026-05-05
Edit rule text
[ai_safety_vulnerability_tangent_no_healthcare_context]
Learned 5 rejections Active
Exclude posts that focus on AI safety vulnerabilities, agent jailbreaks, or cybersecurity incidents (e.g., Claude deleting databases, malicious ClawHub skills, hardcoded API leaks) unless the post explicitly connects the vulnerability to a specific healthcare delivery, clinical decision-making, or patient outcome impact.
Posts about AI safety vulnerabilities, agent hacking, or security incidents presented without healthcare system implications or applications.
3 example posts
Anthropic built something so powerful that they are only letting 50 organisations touch it. It is called Claude Mythos. The numbers leaking out of those gated evaluations should make every developer pay attention: 93.9% on SWE-bench Verified 94.6% on GPQA Diamond Claude Opus
Hacking Mexico government with AI assistance. Attacker exfiltrated hundreds of millions of citizen records. 75% of the executed commands across the entire cyberattack campaign were generated by Claude. 40 minutes after Claude said "I'm not going to create that file" it was report
𝗔𝗜 𝗔𝗴𝗲𝗻𝘁𝘀 𝗔𝗿𝗲 𝗔𝗹𝗿𝗲𝗮𝗱𝘆 𝗕𝗲𝗶𝗻𝗴 𝗛𝗶𝗷𝗮𝗰𝗸𝗲𝗱 Researcher Aks Sharma at Manifold found 30 malicious skills on ClawHub turning AI agents into a crypto farming botnet: 10,000 downloads before anyone noticed. ⬩ The attack required zero exploits. Malicious https://t.co/v4oBXPPydu
Created 2026-04-29 · Updated 2026-05-03
Edit rule text
[healthcare_fraud_scandal_reporting_without_healthcare_systems_analysis]
Learned 5 rejections Active
Exclude posts that expose healthcare fraud (e.g., Botox billing scams, hospice fraud, billing schemes) as crime reporting without connecting the scandal to underlying systemic failures in pricing, reimbursement, or operational oversight.
Posts reporting healthcare fraud, billing crimes, or regulatory scandals as isolated incidents without analyzing systemic root causes, incentive failures, or care delivery implications.
3 example posts
Why is healthcare expensive? The grifters gonna grift. ASCO is asking for oncology practices to be admitted into the same 340B machinery that helped health systems acquire physician practices for two decades. Read the tell. The proposed Indigent Care Ratio lets community
In what is considered the largest #Botox fraud scheme in the United States, a jury in Los Angeles convicted a California doctor on Tuesday in a $45 million scheme to defraud #Medicare by submitting claims for Botox injections that were never provided and medically unnecessary, ht
Remember when UnitedHealth's CEO was assassinated in 2024? 18 months later, every accusation the killer Luigi Mangione made has been proven RIGHT. UnitedHealth is one of the biggest corporate collapses in recent years. And they deserve it. In December 2024, UnitedHealthcare CE
Created 2026-04-29 · Updated 2026-05-27
Edit rule text
[regulatory_or_fraud_scandal_without_healthcare_analysis]
Learned 5 rejections Active
Exclude posts that report fraud scandals, regulatory actions, lawsuits, or enforcement events (e.g., Optum fraud, hospital billing, lawsuits) without connecting to broader healthcare system failures, policy solutions, or structural analysis. Scandal reporting without systems context is news, not tech insight.
Posts reporting healthcare fraud, regulatory enforcement, or legal disputes without analyzing systemic healthcare implications or policy solutions.
3 example posts
Atrium tried to buy a hospital in 2018. The Attorney General killed it. In 2026, they came back with paperwork instead of cash. The Attorney General has 30 days. The Wake County board has 48 hours. @NC_Governor https://t.co/g99Ub3iMhh https://t.co/x0bxsg7VIR
"At the Veterans Health Administration, we have the exquisite gift of a hospice unit, a place to care for veterans when time is short. We care for the sons of mothers who can no longer be here to care for them." In #APieceofMyMind, a #palliative care #physician reflects on https
Ah yes, Optum, the company that did Tim Walz’s audit of the 14 high risk Medicaid programs consumed by Somali fraud. https://t.co/dNRnjI7wxB
Created 2026-04-29 · Updated 2026-05-04
Edit rule text
[tangential_ai_capability_hype_without_healthcare_specificity]
Learned 5 rejections Active
Exclude posts that emphasize AI model technical capabilities (code generation, vulnerability detection, performance benchmarks, architectural features) unless the post demonstrates concrete healthcare application, clinical validation, or healthcare system integration.
Posts about general AI model capabilities, security vulnerabilities, or technical features that are only loosely connected to healthcare.
3 example posts
Anthropic built a model so good at hacking they refused to release it publicly. INSTEAD they handed it to Apple, Google, Microsoft, AWS, Cloudflare one month later: 10,000+ zero days found in the software that runs the internet...bugs hidden for 27 years in OpenBSD. exploit htt
🚨 Anthropic just dropped the first Project Glasswing update Claude Mythos found 10,000+ critical vulnerabilities in ONE month: > Cloudflare: 2,000 bugs, 400 high/critical severity > Mozilla: 271 vulnerabilities in Firefox 150 — 10x more vulnerabilities found in Firefox 1
The more interesting part is that Microsoft’s own engineers liked Claude code best…and they’re cutting it anyway Token pricing is making enterprise actually look at what these models cost to run and that could be a problem for the labs when the subsidy ends. Lots of cheaper
Created 2026-04-29 · Updated 2026-05-24
Edit rule text
[glp1_peptide_market_speculation_only]
Learned 5 rejections Active
Exclude posts focused primarily on GLP-1/peptide market pricing, insurance coverage changes, company market share, or competitive dynamics unless they reveal systemic healthcare delivery or reimbursement insights. Market speculation or pricing commentary alone does not qualify.
Posts about GLP-1 drug pricing, market dynamics, or competitive positioning without healthcare systems analysis or operational substance
3 example posts
$HIMS up to #4 in the Canadian App Store medical category. That is not a stock chart. That is real consumer demand showing up in public. GLP-1 drama, Novo pressure, pricing debate, and now app-store strength. Bears keep arguing the story is slowing down, but the product demand
For years, grey-market HGH operated in a strange legal grey zone somewhere between “research compound,” “API” and underground enhancement culture. Last month’s FDA guidance suggests regulators may no longer see it that way. The recent decline in HGH quality, and now the growing
Think about this. The $50 per month out-of-pocket cost of GLP-1 drugs for Americans (with insurer paying the balance of $2000 pm) is more than double the entire cost for Indians (less than ₹2000pm for Semaglutide). Overall cost difference? 100 times. Not 100%.
Created 2026-04-28 · Updated 2026-05-30
Edit rule text
[glp1_peptide_market_pricing_speculation]
Learned 5 rejections Active
Exclude posts that discuss GLP-1 and peptide pricing, script counts, market share, or competitive launches (e.g., Foundayo vs. Wegovy prescriptions, price reductions, generic launches) unless the post analyzes healthcare system implications, payer policy, or clinical/epidemiological outcomes. Market speculation and pricing dynamics alone are insufficient.
Posts about GLP-1 and peptide market dynamics, pricing, prescriptions, and competitive positioning without healthcare systems or clinical outcomes analysis.
3 example posts
🚨 Survodutide, a weekly subcutaneous dual GLP-1/glucagon agonist, posts new top-line Phase 3 results today 📊 SYNCHRONIZE-1: Adults with overweight/obesity randomized to survodutide vs. placebo for 76 weeks ⚖️ 16.6% weight loss (efficacy estimand) vs. 3.2% in placebo 85.1% http
$NVO $LLY Boehringer Ingelheim and Zealand Pharma report strong phase 3 data for obesity drug survodutide. Patients lost up to 16.6% of body weight over 76 weeks vs. 3.2% for placebo. The drug targets both GLP-1 and glucagon, a combo approach aimed at boosting weight loss.
This is just two GLP-1s, one peptide, one use case what happens when off-label prescribing ramps up what happens when retatrutide hits the market what happens when other peptides become compoundable chapter one
Created 2026-04-28 · Updated 2026-04-30
Edit rule text
[political_outrage_or_regulatory_posturing_without_analysis]
Learned 5 rejections Active
Exclude posts that cite politicians, regulatory scandals, or moral/vigilante framing (e.g., 'the killer was right') to attack healthcare institutions or policy without providing systems-level analysis of the underlying problem or proposing evidence-based solutions.
Posts invoking political figures, regulatory outrage, or moral arguments without substantive healthcare policy or systems analysis.
3 example posts
A senator complaining about healthcare costs should have to answer one question first: Why does your state make new healthcare competitors ask permission before opening? If yes, start there. Speeches do not lower prices. Supply does.
Remember when UnitedHealth's CEO was assassinated in 2024? 18 months later, every accusation the killer Luigi Mangione made has been proven RIGHT. UnitedHealth is one of the biggest corporate collapses in recent years. And they deserve it. In December 2024, UnitedHealthcare CE
600 new generic drugs added to TrumpRx this week thanks to President @realDonaldTrump's leadership! Before you visit your local pharmacy, visit https://t.co/dDMdsILHTj to ensure you're getting the BEST price for your prescriptions. It's easy and saves YOU money! https://t.co/
Created 2026-04-27 · Updated 2026-05-24
Edit rule text
[stock_ticker_speculation]
Learned 5 rejections Active
Exclude posts that lead with stock tickers ($HIMS, $NVO, $LLY, $BEAM) or focus primarily on market performance, price rankings, or competitive positioning based on financial metrics rather than healthcare operational or clinical impact.
Posts focused on stock price movements, ticker symbols, and market speculation without healthcare systems analysis
3 example posts
$HIMS $HERS 🚨 HERS IS NOW BEATING BOTH HIMS AND RO ON THE iOS APP STORE Today's rankings in the Medical category: - Hers: #9 - Ro: #10 - Hims: #18 Chart below = weekly avg rank https://t.co/vlv80yiFOo
$ABCL Just now @ BoA - @AbCelleraBio CEO Carl Hansen when asked about what the market is overlooking: “I believe we now have best in class or best in world capabilities for some very important specific applications in therapeutic antibodies. So GPCRs and ion channels has been a
$XBI RBC Makary's Out, So Who's In? We Break Down Potential Candidates and Impact to Biopharma Potential Candidates for #FDA #Commissioner: Kyle Diamantas J.D Dr. Stephen Hahn M.D. Dr. Brett Giroir M.D. Dr. Sara Brenner M.D. Dr. Houman Hemmati M.D., Ph.D. Dr. Richard Pazdur, M.D.
Created 2026-04-27 · Updated 2026-05-20
Edit rule text
[clinical_anecdote_or_single_case_without_context]
Learned 5 rejections Active
Exclude posts that report a single clinical case, patient testimonial, personal medical experiment (e.g., 'You can test your peptides'), or isolated physician observation without situating it within clinical trial evidence, healthcare system incentives, or validated outcomes data.
Posts sharing single clinical observations, patient anecdotes, or unvalidated personal medical claims without systems insight
3 example posts
Makes an important point here about the role of these short summaries they do this because this is how most doctors digests information if designed appropriately, an LLM should be able to read and appraise a paper and provide better peer review than most clinicians and honestly
“You can test your peptides to see if they’re real, bro. But I just look for the 99.9% on the result the seller gives me. You can look up the QR code upon the lab’s website, so you know it’s legit. It’s foolproof, bro. What’s mass spec, bro?” https://t.co/xzJLJ3hesT
I'm a board certified OBGYN. I've practiced for 20+ years. Here's what's actually happening with mifepristone, because the coverage of it has been a mess. On May 1, the 5th Circuit Court of Appeals ruled that the FDA overstepped its authority in 2023 when it eliminated the in-pe
Created 2026-04-27 · Updated 2026-05-18
Edit rule text
[ai_infrastructure_compute_hype_no_application]
Learned 5 rejections Active
Exclude posts that discuss AI infrastructure (tokens, context windows, inference cost, GPU economics, energy requirements, model capability) in generic or macro terms without demonstrating how these technical dynamics apply to or constrain healthcare use cases. The post must connect infrastructure challenges to actual healthcare problems.
Posts about AI infrastructure, compute economics, energy availability, or technical capability without healthcare-specific application or validation.
3 example posts
Elon Musk reveals why he believes the cheapest place to put AI will be space within 36 months "The availability of energy is the issue. Everywhere outside of China, electrical output is more or less flat. The output of chips is growing exponentially, but the output of https://t.
The context window pricing model might be one of the smartest tricks AI labs pulled. You pay for larger context windows, yet model performance often degrades as more context is stuffed in. So developers end up paying more just to work around the fact that models don't have true
a $2,000 graphics card and a free download are turning into $2,000-a-month retainers, and the customers are the ones who legally can't use ChatGPT the hardware is an RTX 5090, 32GB of VRAM, enough to run a 30B model locally with nothing ever leaving the building. the software is
Created 2026-04-27 · Updated 2026-06-06
Edit rule text
[fraud_scandal_without_systems_analysis]
Learned 5 rejections Active
Exclude posts that report healthcare fraud, embezzlement, or regulatory violations as crime/scandal news without analyzing the structural incentives, payment models, or regulatory gaps that enabled the behavior. Sensationalism without systems insight disqualifies the post.
Posts about healthcare fraud, regulatory violations, or corporate misconduct without explaining systemic root causes or policy implications.
3 example posts
Why is healthcare expensive? The grifters gonna grift. ASCO is asking for oncology practices to be admitted into the same 340B machinery that helped health systems acquire physician practices for two decades. Read the tell. The proposed Indigent Care Ratio lets community
In what is considered the largest #Botox fraud scheme in the United States, a jury in Los Angeles convicted a California doctor on Tuesday in a $45 million scheme to defraud #Medicare by submitting claims for Botox injections that were never provided and medically unnecessary, ht
Remember when UnitedHealth's CEO was assassinated in 2024? 18 months later, every accusation the killer Luigi Mangione made has been proven RIGHT. UnitedHealth is one of the biggest corporate collapses in recent years. And they deserve it. In December 2024, UnitedHealthcare CE
Created 2026-04-26 · Updated 2026-05-27
Edit rule text
[cybersecurity_or_ai_vulnerability_tangential]
Learned 5 rejections Active
Exclude posts about AI model hacking, zero-day vulnerabilities, network takeovers, or cybersecurity proofs-of-concept unless they directly demonstrate a healthcare system failure, clinical harm, or specific healthcare infrastructure risk.
Posts about AI model vulnerabilities, cyberattacks, or security exploits that lack healthcare-specific context or application.
3 example posts
🚨 SaaS platform ClickUp, used by 85% of the Fortune 500, has been leaking customer emails through its homepage for at least 465 days, and counting. ClickUp has a $4 billion valuation. They are SOC 2 Type 2, ISO 27001, ISO 27017, ISO 27018, ISO 42001, and PCI DSS certified. The f
A researcher gave an AI agent access to his shell, his files, and his network. Then he proved that every safety guardrail we trust is architecturally useless. It cannot tell the difference between your instructions and a hacker's. The paper is called Parallax: Why AI Agents htt
This is by far the most important result of the entire GPT-5.5 release: In a cyber evaluation GPT-5.5 was able to take over a simulated corporate network in 1/10 trials with a budget of 100M tokens. Previously, the only model that was able to solve this task was Claude Mythos,
Created 2026-04-24 · Updated 2026-04-28
Edit rule text
[glp1_peptide_market_pricing_dynamics]
Learned 5 rejections Active
Exclude posts that focus primarily on GLP-1 or peptide drug market share, pricing, launch velocity, script volumes, or competitive positioning (e.g., Novo vs. Lilly) without analyzing underlying healthcare system impacts, patient access barriers, or clinical outcomes.
Posts about GLP-1 and peptide drug market dynamics, pricing, competition, and adoption metrics without healthcare systems analysis.
3 example posts
🚨 Survodutide, a weekly subcutaneous dual GLP-1/glucagon agonist, posts new top-line Phase 3 results today 📊 SYNCHRONIZE-1: Adults with overweight/obesity randomized to survodutide vs. placebo for 76 weeks ⚖️ 16.6% weight loss (efficacy estimand) vs. 3.2% in placebo 85.1% http
$NVO $LLY Boehringer Ingelheim and Zealand Pharma report strong phase 3 data for obesity drug survodutide. Patients lost up to 16.6% of body weight over 76 weeks vs. 3.2% for placebo. The drug targets both GLP-1 and glucagon, a combo approach aimed at boosting weight loss.
This is just two GLP-1s, one peptide, one use case what happens when off-label prescribing ramps up what happens when retatrutide hits the market what happens when other peptides become compoundable chapter one
Created 2026-04-24 · Updated 2026-05-01
Edit rule text
[conspiracy_or_sensationalized_unsubstantiated_claims]
Learned 5 rejections Active
Exclude posts that amplify unsubstantiated or sensationalized health claims (e.g., 'AI used to deny claims, people died'), conspiracy-style narratives, or extreme characterizations of motives without citing credible evidence or peer-reviewed research.
Posts making alarming health or policy claims without credible evidence, often with conspiratorial undertones.
3 example posts
@PirateWires He's objectively correct. Brian Thompson made decisions that led to denials of medical care, and people died. He used Ai to find ways to deny claims ffs. Brian Thompson has more blood on his hands than whoever shot him
They Don't Work for You- Calls to protect foreigners from deportation or to keep the borders wide open are not about compassion. They are a core part of the globalist plan to flood the labor market with cheaper more compliant workers suppress wages for Americans and make
RFK Jr. calls out Democrat House representatives to their face for ignoring chronic disease while claiming to care about public health. “The Congressman was talking about the deaths from infectious disease, which are a couple thousand a year.” “90% of the people who die in this
Created 2026-04-24 · Updated 2026-04-25
Edit rule text
[broad_unsubstantiated_health_claims]
Learned 5 rejections Active
Exclude posts that present broad medical or nutritional assertions (e.g., 'your body can only use 25-30g protein per meal', GLP-1 side effect narratives presented without trial context) as established fact without citing peer-reviewed evidence or acknowledging limitations.
Posts making sweeping health or medical claims without rigorous evidence, validation, or nuance.
3 example posts
"There is also compelling preliminary evidence suggesting that the use of these drugs [GLP-1] could exacerbate and lead to new diagnoses of restrictive eating disorders, including anorexia nervosa." @NEJM today https://t.co/swx1y25F1X https://t.co/bQYlpuFBjM
Attention PK nerds, pharmacologists, and clinicians who actually understand serum levels: I haven’t seen this discussed, but it could matter for patients priced out of injectables. If a 25 mg oral semaglutide tablet has ~1% bioavailability, that’s ~0.25 mg systemically… on
The only problem with the GLP-1 heart muscle loss narrative is... ... that it's just a narrative. GLP-1s have reliably improved cardiovascular outcomes in trials, to the point that some research suggests benefit may even be independent of (not reliant on) weight loss.
Created 2026-04-24 · Updated 2026-04-26
Edit rule text
[political_figures_healthcare_posturing_without_systems_analysis]
Learned 5 rejections Active
Exclude posts that quote or amplify political figures criticizing healthcare spending, drug prices, or chronic disease policy (e.g., 'RFK Jr. calls out...', 'A senator complaining about...') unless the post provides healthcare-specific policy analysis, regulatory impact assessment, or evidence about implementation outcomes.
Posts amplifying political figures' healthcare claims or criticism without substantive policy or healthcare system analysis.
3 example posts
What $1 Billion a Day Buys in American Health Care The U.S. is spending $1 billion/day on the war in Iran — over a year, that would cover 37 million Medicaid enrollees. Congress just cut $911 billion from the program because it was too expensive. Read & subscribe (for free!)
A senator complaining about drug prices while voting for the law that set them is not a reformer. He is a magician. The trick is making you watch his hands.
RFK Jr. calls out Democrat House representatives to their face for ignoring chronic disease while claiming to care about public health. “The Congressman was talking about the deaths from infectious disease, which are a couple thousand a year.” “90% of the people who die in this
Created 2026-04-24 · Updated 2026-04-28
Edit rule text
[conspiracy_or_sensationalized_health_claims]
Learned 5 rejections Active
Exclude posts that present alarming health or medical claims (e.g., smart TVs collecting screenshots, missing scientists, unvalidated psychedelic treatments) without credible scientific sources, peer review, or clear distinction between speculation and fact.
Posts making unsubstantiated or sensationalized claims about health threats, surveillance, or medical interventions without credible evidence or peer review.
3 example posts
Bob Lazar allegedly watched people fly a UFO at Area 51. “They knew how to fly it.” “The craft had a corona discharge glow on the bottom and lifted off silently up into the sky … ” And it had one shocking, anomalous effect that still perplexes him to this day: As Lazar https:
As far as I know this is the only naturally-derived, classical psychedelic, that has killed people. Ayahuasca has some deaths, but it's unclear what the cause was, and unlikely directly related to its cardiovascular risk profile. https://t.co/DatuHiBOTX
🚨BREAKING: A peer reviewed study just confirmed your smart TV is taking screenshots of your screen every 15 seconds and sending them to company servers. Samsung every minute. LG every 15 seconds. Running even when you are using it as a monitor. Here is how to stop it:
Created 2026-04-23 · Updated 2026-04-24
Edit rule text
[political_figures_healthcare_grandstanding]
Learned 5 rejections Active
Exclude posts featuring political figures (RFK Jr., senators, congresspeople) making healthcare-related claims, accusations, or announcements that lack supporting data, peer-reviewed evidence, or systems-level analysis—especially when the post reads as amplification or political theater rather than policy critique.
Posts where political figures make broad healthcare claims or allegations without substantive analysis or evidence.
3 example posts
What $1 Billion a Day Buys in American Health Care The U.S. is spending $1 billion/day on the war in Iran — over a year, that would cover 37 million Medicaid enrollees. Congress just cut $911 billion from the program because it was too expensive. Read & subscribe (for free!)
RFK Jr. calls out Democrat House representatives to their face for ignoring chronic disease while claiming to care about public health. “The Congressman was talking about the deaths from infectious disease, which are a couple thousand a year.” “90% of the people who die in this
🚨BREAKING: HHS Sec. RFK Jr. just announced President Trump has SAVED and FOUND 138,000 missing children lost under Biden. "Many have been trafficked, undergone slavery, s*xual abuse." Follow: @BoLoudon https://t.co/p6YKEm38T7
Created 2026-04-23 · Updated 2026-04-25
Edit rule text
[cybersecurity_vulnerability_tangential]
Learned 5 rejections Active
Exclude posts reporting on zero-day vulnerabilities, security breaches, or bug-finding achievements in cloud platforms or general software — even if discovered by healthcare-adjacent AI models (Claude Mythos). Posts must address healthcare-specific security risks (EHR compromise, medical device hacking, HIPAA violations) or healthcare defensive strategies.
Posts about AI/cybersecurity vulnerabilities, zero-day exploits, or security incidents without healthcare-specific threat or defense implications.
3 example posts
Anthropic built a model so good at hacking they refused to release it publicly. INSTEAD they handed it to Apple, Google, Microsoft, AWS, Cloudflare one month later: 10,000+ zero days found in the software that runs the internet...bugs hidden for 27 years in OpenBSD. exploit htt
🚨 Anthropic just dropped the first Project Glasswing update Claude Mythos found 10,000+ critical vulnerabilities in ONE month: > Cloudflare: 2,000 bugs, 400 high/critical severity > Mozilla: 271 vulnerabilities in Firefox 150 — 10x more vulnerabilities found in Firefox 1
#AI agents need more than a sandbox—they need predictable execution. Learn how OpenShell provides a standard interface to intercept, inspect, and approve LLM-generated bash commands before they run: https://t.co/JKmLMgl1A9 https://t.co/xWi8oBNZ7T
Created 2026-04-23 · Updated 2026-05-27
Edit rule text
[generalist_macro_or_policy_commentary_without_healthcare_lens]
Learned 5 rejections Active
Exclude posts about general macroeconomic trends, labor markets, geopolitics, elections, legislation, or trade policy unless they include specific healthcare policy analysis, reimbursement impact, or healthcare delivery system implications. Broad commentary on labor markets, elections, or international affairs does not qualify.
Posts on broad macro, political, or economic topics (trade, labor, elections, geopolitics) without specific healthcare policy or delivery analysis.
3 example posts
Total employment in New Jersey declined by 10,300 jobs in February, though the unemployment rate in the state decreased by 0.1% to 5.1%. The January estimate was revised downward by 2,500 jobs, resulting in a December-to-January net gain of 3,500 jobs, down from the preliminary
A senator complaining about drug prices while voting for the law that set them is not a reformer. He is a magician. The trick is making you watch his hands.
They Don't Work for You- Calls to protect foreigners from deportation or to keep the borders wide open are not about compassion. They are a core part of the globalist plan to flood the labor market with cheaper more compliant workers suppress wages for Americans and make
Created 2026-04-21 · Updated 2026-04-24
Edit rule text
[tangential_non_healthcare_policy_or_business_content]
Learned 5 rejections Active
Exclude posts about tariffs, corporate tax policy, banking models, insurance as a business/tech product, or general economic policy that mention healthcare peripherally but center on non-healthcare business or policy analysis. Posts must be primarily about healthcare system dynamics, not general business or macro policy.
Posts about non-healthcare policy, finance, or business dynamics that lack healthcare systems focus
3 example posts
Revolut just moved the IP of banking into a model. Trained on 24 billion banking events in 111 countries. One foundation model replacing six separate ML systems. Credit scoring: +130% Fraud recall: +65% Marketing engagement: +79% The model is the new moat.
Ah yes, Optum, the company that did Tim Walz’s audit of the 14 high risk Medicaid programs consumed by Somali fraud. https://t.co/dNRnjI7wxB
Last week I randomly got a $9,000 bill for a hospital visit from last year. I called them up and they said my insurance decided not to cover it. Why? Because. .@Aetna can just decide they’re not interested in covering something and then you’re left with an inflated bill to pay h
Created 2026-04-19 · Updated 2026-05-03
Edit rule text
[infrastructure_compute_hype_without_healthcare_application]
Learned 5 rejections Active
Exclude posts that discuss Kubernetes removal, GPU utilization, infrastructure costs, cloud architecture, or compute efficiency improvements in non-healthcare contexts or without demonstrating healthcare application.
Posts about AI infrastructure, compute optimization, or architectural patterns in tech companies without connection to healthcare-specific use cases.
3 example posts
Red Hat and @NVIDIA are integrating NVIDIA OpenShell into the full-stack @RedHat_AI platform. The work brings oversight and policy to the infrastructure level, while contributing to the open source OpenShell project to standardize how agents are governed on enterprise platforms.
The app era is ending. Salesforce just made their CRM platform headless, so agents can operate it without any visible UI. Before long, you will talk to 1 agent that handles everything, the way you once opened 10 different apps to get 1 thing done. For years, VeChain has been b
Ontology all the way down. 11 years. Heavy civil construction. People, equipment, materials, contracts. This is what real-world AI looks like: operational, measurable, and in the field. Palantir x Cavanagh through 2035. https://t.co/872Cs1Qbdu
Created 2026-04-19 · Updated 2026-06-03
Edit rule text
[ai_company_product_metrics_and_launches]
Learned 5 rejections Active
Exclude posts that report on AI company product launches, revenue figures, ARR milestones, user adoption rates, or fundraising news (e.g., Claude Code revenue, Anthropic hiring, OpenAI SDK launches) unless the post explicitly analyzes how that product solves a specific healthcare delivery problem or system challenge.
Posts about AI company product launches, revenue milestones, or business metrics without healthcare application focus.
3 example posts
Everyone's covering agents that help you work and build. Almost nobody's covering this: The same primitives ARE the production runtime. The SDK is one line: npm install @anthropic-ai/claude-agent-sdk The CLAUDE.md that guides Claude Code in your terminal is the exact same http
Today we launched a major update to the OpenAI Agents SDK to help developers build and deploy long-running, durable agents in production. You can now build your own Codex-style agents using powerful primitives for modern agents - file and computer use, skills, memory and
Anthropic's CEO: “coding is going away first, then all of software engineering." Now, Anthropic looks to hire 454 engineers at $320k–$405k. coding isn’t vanishing it’s becoming leverage for the few who can build, review, and ship at a completely different scale. https://t.co
Created 2026-04-19 · Updated 2026-04-19
Edit rule text
[ai_model_capability_tangent]
Learned 5 rejections Active
Exclude posts that highlight general AI model capabilities (e.g., finding security vulnerabilities, writing code, solving puzzles) where healthcare is mentioned only as one possible application domain, not the primary focus or proven use case.
Posts about AI model technical breakthroughs (zero-day detection, coding ability, reasoning) with only tangential healthcare relevance.
3 example posts
The UK AISI found Mythos Preview is the first model to solve both their cyber ranges end-to-end. No model had ever solved the AISI’s “Cooling Tower” cyber range before. We're getting it to defenders as fast as we responsibly can. More to come on our Glasswing work soon.
So Mythos was, indeed, not marketing hype. Remember this is a general purpose model that just happens to be good at finding exploits because good models are good at lots of things. Expect similar from OpenAI & Google. And from open models in 8 months. https://t.co/KbhalQYX8
Dario Amodei says Mythos is not limited by compute Anthropic can scale it 3x or 10x without creating a conflict between government and private-sector access The harder problem is who gets it "because giving access to too many organizations could create serious cyber risks" htt
Created 2026-04-18 · Updated 2026-05-15
Edit rule text
[glp1_peptide_macro_or_personal_context]
Learned 5 rejections Active
Exclude posts about GLP-1 side effects, personal weight loss stories, pharmacoeconomics of existing drugs, or biological mechanisms of GLP-1 — unless they frame healthcare technology application or systemic innovation. Personal wellness and drug mechanism posts belong in clinical/wellness categories, not healthcare tech.
Posts about GLP-1 drugs, peptides, or weight loss that focus on personal experience, macro economics, or general biology rather than healthcare tech innovation.
3 example posts
I have now received nine reports from people taking GLP-1 drugs who got the same side effect: They no longer feel normal when they come off. "I feel hangry again", "I started thinking about hunger and I hate it", "I have to go back to Adderall". 8/9 reports -> from women.
The brain is the master regulator of food intake and energy balance. A brilliant new @CellCellPress review, including the mechanism of GLP-1 drugs, by @ClemmensenC and colleagues, open-access https://t.co/KbDf7ym288 https://t.co/DMFeS9zHfw
GLP-1 drugs are the ultimate validation of the techno-solutionist approach to society's most challenging problems. The obesity crisis seemed liked it would just get worse and worse forever. Scolding from public health officials didn't work. Proposals to completely overhaul our f
Created 2026-04-17 · Updated 2026-04-18
Edit rule text
[glp1_peptide_adjacent_non_healthcare_context]
Learned 5 rejections Active
Exclude posts that discuss GLP-1 drugs, peptides, or weight loss medications primarily through personal anecdotes, pricing gossip, or macro-level cultural commentary (e.g., 'obesity crisis solved') without clinical trial data, healthcare access analysis, or policy implications.
Posts about GLP-1 drugs, weight loss medications, or peptides that focus on personal side effects, pricing speculation, or general statements about obesity without clinical or healthcare economics substance.
3 example posts
I have now received nine reports from people taking GLP-1 drugs who got the same side effect: They no longer feel normal when they come off. "I feel hangry again", "I started thinking about hunger and I hate it", "I have to go back to Adderall". 8/9 reports -> from women.
The brain is the master regulator of food intake and energy balance. A brilliant new @CellCellPress review, including the mechanism of GLP-1 drugs, by @ClemmensenC and colleagues, open-access https://t.co/KbDf7ym288 https://t.co/DMFeS9zHfw
GLP-1 drugs are the ultimate validation of the techno-solutionist approach to society's most challenging problems. The obesity crisis seemed liked it would just get worse and worse forever. Scolding from public health officials didn't work. Proposals to completely overhaul our f
Created 2026-04-16 · Updated 2026-04-17
Edit rule text
[healthcare_financial_scandal_without_systems_analysis]
Learned 5 rejections Active
Exclude posts that report healthcare fraud, billing scandals, or regulatory violations (FBI prosecution, 340B abuse, hospice fraud) without analyzing root causes in healthcare payment systems, reimbursement incentives, or structural vulnerabilities.
Healthcare fraud, regulatory violations, or financial scandals reported without structural healthcare economics analysis.
2 example posts
Why is healthcare expensive? The grifters gonna grift. ASCO is asking for oncology practices to be admitted into the same 340B machinery that helped health systems acquire physician practices for two decades. Read the tell. The proposed Indigent Care Ratio lets community
In what is considered the largest #Botox fraud scheme in the United States, a jury in Los Angeles convicted a California doctor on Tuesday in a $45 million scheme to defraud #Medicare by submitting claims for Botox injections that were never provided and medically unnecessary, ht
Created 2026-04-16 · Updated 2026-05-31
Edit rule text
[infrastructure_and_compute_hype_no_healthcare_specificity]
Learned 5 rejections Active
Exclude posts about CPU launches, chip supply chains, compute capacity announcements, space infrastructure, or semiconductor strategy unless the post explicitly connects the infrastructure advancement to a specific healthcare delivery, diagnostic, or clinical workflow problem.
Posts about AI compute, infrastructure, chip announcements, or space-based computing without specific healthcare application or relevance.
3 example posts
🔥 CPUs are having a moment. #Nvidia launched a standalone CPU. #Arm made its first chip in 35 years. #Intel & #AMD are raising prices amid a supply crunch. What's behind it: Agentic AI needs far more CPU than anyone planned for — driving a structural shift in CPU:GPU ratios tow
To put Elon's space compute vision into perspective:  1 TW of compute in orbit  That's 10 million tons to orbit each year. That's 100,000 launches a year, almost one every 5 minutes.  In the airline business that's normal!
🚨MAJOR INTERVIEW: Jensen Huang joins the Besties! The @nvidia CEO joins to discuss: -- Nvidia's future, roadmap to $1T revenue -- Physical AI's $50T market -- Rise of the agent, OpenClaw's inflection moment -- Inference explosion, Groq deal -- AI PR Crisis, Anthropic's comms m
Created 2026-04-16 · Updated 2026-04-18
Edit rule text
[academic_or_clinical_observation_without_systems_context]
Learned 5 rejections Active
Exclude posts that share academic research, peer-reviewed findings, clinical mechanisms, or ward-level medical observations without connecting them to healthcare tech application, business models, or systems-level implications. Posts about pancreatic cancer biology, GLP-1 mechanisms, or physician anecdotes from rounds should focus on how this informs healthcare tech, not just document clinical facts.
Posts reporting clinical research findings, mechanistic biology, or ward-level observations without healthcare systems or business application insight.
3 example posts
~1-2% of the patients on ward rounds has something bad going on which hasn’t been identified yet. As the attending, one of my main duties on rounds is to spot these cases. I do a lot of this by Noticing Things. A 🤖 iPad makes it much less likely you will Notice Things. 🤔
Interpretable Antibody–Antigen Structural Interface Prediction via Adaptive Graph Learning and Cyclic Transfer 1. The paper introduces VASCIF (Variable-domain Antibody–antigen Structural Complex Interface Finder), a structure-aware model that jointly predicts paratopes and https
NIH-funded researchers have uncovered a key reason why immunotherapy has largely failed in pancreatic cancer — and identified a promising strategy to overcome that resistance. Read on to learn more about this discovery: https://t.co/BoCHpLxp5g https://t.co/3DXv4E9DOE
Created 2026-04-16 · Updated 2026-04-16
Edit rule text
[tangential_infrastructure_and_compute_hype]
Learned 5 rejections Active
Exclude posts discussing AI compute scaling, energy consumption, chip manufacturing, data center buildout, or infrastructure investments (e.g., GPU capacity, power grids, gas turbines) unless directly tied to healthcare AI deployment or healthcare data processing challenges.
Posts about AI infrastructure, compute scaling, energy, or chip manufacturing without healthcare application focus.
3 example posts
Today we’re giving an update on ramping F.03 production at BotQ In the last 120 days, Figure scaled manufacturing 24x - from 1 robot/day to 1 robot/hour We will manufacture 55 humanoid robots this week https://t.co/Am5Kn53mVE
📈 NVIDIA tops AI leaderboards and benchmarks with open models driven by extreme co-design across compute, networking, memory, storage, and software. This includes models for biology, AI physics, agentic AI, physical AI, robotics, and autonomous vehicles. By being vertically htt
Since I began work on AI in 2010, training compute for frontier models has grown by one trillion times. Now we're looking at something like another thousand-fold growth in effective compute by the end of 2028. 1000x the existing 1,000,000,000,000x. Extraordinary stuff.
Created 2026-04-16 · Updated 2026-04-30
Edit rule text
[tangential_ai_infrastructure_hype_no_healthcare_specificity]
Learned 5 rejections Active
Exclude posts that discuss AI model context windows, memory architectures, filesystem design, code infrastructure, or compute hardware unless the post explicitly applies these to a validated healthcare use case or clinical workflow.
Posts about AI infrastructure, compute, memory systems, or foundational ML capabilities with only loose or no healthcare connection.
3 example posts
// Contextual Agentic Memory is a Memo, not True Memory // Most agent memory today isn't memory. They are more like memos. A new paper argues that vector stores, RAG buffers, and scratchpads implement lookup, not consolidation. Agents accumulate notes indefinitely without ever
the future of software engineering seems uncontroversially prompting + code review. startups will skip the code review because they’re racing against time. larger/serious orgs will take code review very seriously. llms can do code review, but my guess is that because they have t
Demis Hassabis says bigger context windows are still a brute force answer to memory. The human brain does something stranger. During sleep, it replays what matters and folds new knowledge into what it already knows. AI does not need infinite context. It needs the right memory h
Created 2026-04-15 · Updated 2026-05-07
Edit rule text
[tangential_ai_capability_applied_outside_healthcare]
Learned 5 rejections Active
Exclude posts that highlight AI technical capabilities in non-healthcare domains (cybersecurity breaches, robotics competitions, network penetration) even if healthcare is mentioned in passing or used as a loose frame. The post must center healthcare delivery, clinical workflows, or healthcare business problems.
Posts showcasing AI model capabilities (cyber security, network takeovers, robot races) where healthcare is tangential or completely absent from the narrative.
3 example posts
> Vercel got pawned > severe enough to notify law enforcement > the only advice: “review your environment variables” > what does that even mean? > $10B company, and this is how you communicate Cyber attacks ramping fast, starting to see why Anthropic is scared to
In Beijing's 2026 humanoid robot half-marathon, HONOR's Lightning completed the 21 km course in 50:26 minute. Beat current human men's half-marathon world record of 57:20. Last year's winner took over 2 hours 40 minutes. Massive progress in 12 month https://t.co/OcZJ66ebWD
FT: The White House is moving to give major US agencies access to a modified Anthropic Mythos model built to hunt dangerous software flaws before attackers find them. That makes Mythos useful for defense because a model that can find a weakness in an operating system, browser, h
Created 2026-04-14 · Updated 2026-04-20
Edit rule text
[tangential_ai_capability_hype]
Learned 5 rejections Active
Exclude posts that lead with AI model technical capabilities (Claude's coding abilities, o1's reasoning, Mythos' security testing) unless they explicitly demonstrate a healthcare use case or clinical outcome. General AI capability announcements do not qualify.
Posts that discuss AI model capabilities (reasoning, coding, security testing) without clear healthcare application or validation.
3 example posts
Yesterday, I wrote about how long-running agentic systems need to be more reliable, robust, and fault-tolerant; and today, let's dig deeper into how. Fun fact - distributed workflows are one of those problems that almost every engineering team accidentally rebuilds. We start wi
Google CEO Sundar Pichai on current frontier model's ability to break the security of almost all current software. "These models are definitely, like really gonna break pretty much all software out there, maybe already, we don't know." https://t.co/ezjPcSehiG
I’m fully forward deployed engineering pilled specifically because AI simply is not the same as software. In software, you deliver a stable piece of technology to a customer and they adopt it and that’s that (extreme over simplification). In AI, you’re delivering something that
Created 2026-04-12 · Updated 2026-05-21
Edit rule text
[clinical_observation_without_systems_insight]
Learned 5 rejections Active
Exclude posts that cite clinical trial results, medical research findings, or single clinical observations (e.g., ECG AI predicting LVEF, MRSA PCR performance) without analyzing how these findings impact healthcare workflows, implementation, access, or system operations.
Posts reporting clinical research findings, trial observations, or medical data points without connecting to healthcare delivery, operations, or system-level implications.
3 example posts
this pragmatic trial is more an investigation of physician behavior than the performance of MRSA nares PCR MRSA nares PCR's performance for excluding MRSA pneumonia was excellent from an EBM standpoint we should keep using the MRSA nares PCR and replace the physicians 🤷‍♂️
63% of ATS respiratory guidelines are based on low-quality evidence. We published the data. The next morning, the chair of ATS got on Twitter. Tagged people. Went on a tirade. Didn't seem to realize tweets were public. "It wasn't a private conversation." The critique? We ht
People are increasingly worried that AI tools make us overreliant. But how do we actually measure this? We introduce Offloading Score, a measure of reliance based on the fraction of cognitive effort offloaded to AI while completing a task. In a controlled user study, Offloading
Created 2026-04-12 · Updated 2026-06-07
Edit rule text
[infrastructure_and_non_healthcare_tech]
Learned 5 rejections Active
Exclude posts about semiconductor manufacturing, data infrastructure, vector databases, or general business models (switching costs, network effects, moats) where healthcare is mentioned only as a contextual example or tangential case study rather than being the core subject.
Posts about tech infrastructure, hardware, or non-healthcare business models that use healthcare as a loose framing device.
3 example posts
Quantum computers are still on the drawing board, but quantum sensing is here now—and this technology can transform not just industry but America's security picture. Read a new Defining Ideas article by Dr. Vivek Lall and Haibo Huang: https://t.co/UeEjZWIO27
In general, there are 5 kind of moats: ▪️ Intangible Assets ▪️ Switching Costs ▪️ Network Effects ▪️ Cost Advantage ▪️ Efficient Scale I'll teach you everything you need to know in 2 minutes: https://t.co/v9w6pfJOGh
HNSW is fast & performant. But what's it costing you? DiskBBQ gets you great recall & speed using a fraction of the memory. HNSW vs DiskBBQ in 40 seconds with @_jphwang https://t.co/3BqD9a6srU
Created 2026-04-12 · Updated 2026-04-13
Edit rule text
[personal_anecdote_or_low_substance_opinion]
Learned 5 rejections Active
Exclude posts that are primarily personal anecdotes, lifestyle updates, or opinionated takes (e.g., 'I now believe,' 'I got a notification,' personal recovery stories) that lack healthcare systems analysis or evidence-based insight.
Posts sharing personal stories, lifestyle updates, or opinion statements without healthcare insight or evidence
3 example posts
I got a notification from Whoop yesterday that the FDA is targeting them for being an "unapproved medical device" for offering blood pressure insights A pretty classic case of the FDA working as a lobby for the multi billion dollar medical device industry instead of supporting n
Mine got LASIK, as had many of the nurses. A lot of ophthalmologists have. There's a weird delusion that the profession is all afraid of it, but there's no basis for that belief beyond fearmongering. https://t.co/yVBDfNDiB5
New @JAMANetwork paper out from our team here at UCLA Health/WLA VA and @samirguptaGI's team at UCSD/San Diego VA! In this first study from a multi-part research project, our teams are trying to understand what age your medical doctors and the colorectal cancer prevention https:
Created 2026-04-11 · Updated 2026-04-12
Edit rule text
[clinical_anecdote_no_systems]
Learned 4 rejections Active
Exclude posts that report clinical trial data, drug approvals, medical case studies, or pharma announcements without analyzing healthcare system implications, cost-effectiveness, access barriers, workflow integration, or broader policy context.
Posts reporting single clinical cases, trial results, or drug approvals without healthcare systems analysis, cost context, or implementation barriers.
3 example posts
Version 2 of our ShockCalcs hemodynamics simulator is live. I've refined the physiology a ton, added new meds, and also a real-time Frank Starling curve that responds to fluids + vasopressors. Check it out (link in reply), and reply here with feedback on how to improve further!
The Iowa study doubled survival by adding IV vitamin C to first line chemo, 8 months to 16. We take it further and inject the vitamin C straight into the tumor, same as we do with the KRAS inhibitors. The compound is rarely the problem. Delivery is, and what it does to the ht
Feeling extremely lucky to live in 2026 and have access to such incredible medicine and clinical research. On the eve of my 40th I had my first dose (monthly injection) of a new treatment that I’m hopeful will prevent me from ever having a heart attack or stroke. This drug ht
Created 2026-06-12 · Updated 2026-07-26
Edit rule text
[insider_cynicism_or_sarcasm]
Learned 4 rejections Active
Exclude posts that rely primarily on cynical humor, sarcasm, or insider mockery about healthcare fraud, regulatory arbitrage, or business practices without providing substantive systems-level analysis or evidence.
Posts using cynical insider humor or sarcasm to critique healthcare systems, business practices, or regulatory capture without substantive analysis.
3 example posts
A non-profit health system can refer a patient to its own MRI, its own lab, its own surgery center, and bill all three. An independent does that once and it's a federal felony. Same referral. Same patient. One of you goes to prison. It's called Stark Law. Read who's exempt.
This is crazy There are now dental practice management consultants who go to dentist offices throughout America and teach them how to increase profits by telling patients they need treatments they don’t need Dentist offices are taught on “Finding” more issues during exams, http
What if the standards used to evaluate your doctor were shaped less by what they actually knew and more by who they are? 🤔 That may not be a hypothetical at the University of Illinois College of Medicine, an institution that trains one in six Illinois doctors. Buried within htt
Created 2026-06-12 · Updated 2026-06-12
Edit rule text
[energy_infrastructure_tangent]
Learned 4 rejections Active
Exclude posts that focus on energy production, compute infrastructure, semiconductor manufacturing, SpaceX orbital data centers, or power consumption of AI systems unless the post explicitly connects these topics to a specific healthcare problem being solved or healthcare outcome being improved.
Posts about energy, compute infrastructure, space-based data centers, and hardware without healthcare application
3 example posts
At our recent Energy and the AI Age summit, Hon. Bernard L. McNamee, former Commissioner of the U.S. Federal Energy Regulatory Commission, comments on why energy matters in discussions about AI: “Energy is the foundation of our entire economy. Energy makes up about 7% of the htt
Dylan Patel, founder of SemiAnalysis: "The upper bound on how much compute can be produced by 2030 is around 200 gigawatts a year." The entire world has about 20 gigawatts of AI deployed right now. The ceiling is 10x what exists today, and it still isn't enough to feed what htt
Follow the signal: we are in a compute-constrained world. And that means power doesn’t sit with the models. It sits with the infrastructure. Most frontier labs don’t own the means of producing Intelligence. They rent it. A handful of companies provide compute that everyone htt
Created 2026-06-12 · Updated 2026-06-12
Edit rule text
[clinical_anecdote_or_trial_data_without_systems_context]
Learned 4 rejections Active
Exclude posts that report clinical trial results, drug efficacy data, or isolated clinical observations (e.g., 'X drug showed Y% weight loss') without analyzing how those results affect healthcare delivery, reimbursement, operational workflows, or patient access — unless the post provides systems-level insight into healthcare transformation.
Posts sharing clinical trial data, drug results, or medical observations without healthcare systems or operational analysis
3 example posts
This week brought even more exciting news in the field of HIV treatment! A once-weekly Lenacapavir-based pill has now demonstrated similar outcomes to daily antiretrovirals (ARVs), achieving non-inferiority. This development could significantly impact treatment adherence and http
🚨NEW @NEJM CAR T cells expanding applications! Here CAR T facilitate kidney transplantation in highly sensitized people. dual CD19 + BCMA CAR T enabled kidney transplantation in two highly sensitized patients after reducing anti-donor antibody barriers, with no severe https://t.
🔔Survodutide SYNCHRONIZE-1 full paper is out for Boehringer-Ingelheim’s GLP-1/Glucagon study for weight loss. ▪️13% loss on 6mg (treatment regimen estimand) ▪️Huge treatment discontinuations due to AEs: 20% ▪️Nausea 65%, Vomiting 45% ▪️Placebo was killer. 5.4% WL. Wow. ▪️60% htt
Created 2026-06-11 · Updated 2026-06-11
Edit rule text
[anthropic_mythos_unvalidated_security_claims]
Learned 4 rejections Active
Exclude posts that sensationalize AI model security capabilities (e.g., 'found 5 years of unfound bugs in 6 weeks,' 'found zero-days in every browser') without demonstrating actual healthcare system impact or validation. These are unsubstantiated capability hype.
Posts making extraordinary claims about Anthropic's Claude Mythos model finding zero-days or security vulnerabilities without clinical validation or healthcare application.
3 example posts
"Best-in-class security" still had 5 years of unfound bugs. AI found them in 6 weeks. $PANW CEO Nikesh Arora told @theallinpod that his own codebase, at a company that treats security as a core competency, had vulnerabilities Claude surfaced in 6 weeks that would have taken his
🚨 do you understand what just happened with Anthropic.. Their internal model, Claude Mythos, found previously unknown security holes in every major browser and every major operating system. Not one. All of them at once. And Anthropic did something a hype-driven industry almost
Mythos AI is being used by National Security Agency in offensive cyber operations / cyberattacks. Anthropic has even embedded engineers inside the NSA to help deploy the model. Are frontier AI labs becoming active contractors in state cyber conflict? The live-ops role is still ht
Created 2026-06-10 · Updated 2026-06-11
Edit rule text
[pharmaceutical_trial_data_market_pricing_only]
Learned 4 rejections Active
Exclude posts that announce drug trial data, GLP-1 efficacy results, or biotech announcements but frame them primarily through market pricing, investor positioning, or competitive market share dynamics rather than clinical significance, patient outcomes, or healthcare system impact.
Posts reporting pharmaceutical trial results or clinical data focused only on market pricing dynamics, competitive positioning, or revenue implications without healthcare systems analysis
3 example posts
🔔Survodutide SYNCHRONIZE-1 full paper is out for Boehringer-Ingelheim’s GLP-1/Glucagon study for weight loss. ▪️13% loss on 6mg (treatment regimen estimand) ▪️Huge treatment discontinuations due to AEs: 20% ▪️Nausea 65%, Vomiting 45% ▪️Placebo was killer. 5.4% WL. Wow. ▪️60% htt
Survodutide Once Weekly for the Treatment of Adults with Obesity: @NEJM 🥸 For @Boehringer, the advantage of Survodutide is probably liver and visceral fat reduction, not greater weight loss! Probably indication in obesity-+ MASH rather than head-to-head share against https://t.
Let me clear up the microscope fight first. 🔬 Everyone’s arguing over “hairs” and “fibers” in horse paste. But nobody talks about formulation science. Horse paste = oral suspension. Binders. Thickeners. Made to swallow. Injectable = sterile. Made for veins. Different rules. J
Created 2026-06-10 · Updated 2026-06-11
Edit rule text
[anthropic_mythos_unvalidated_capability_hype]
Learned 4 rejections Active
Exclude posts that celebrate Anthropic's Claude Mythos finding zero-days, beating benchmarks, or performing AI research tasks without evidence of real-world healthcare validation, clinical utility, or healthcare-specific application. These posts conflate AI capability announcements with healthcare impact.
Posts about Anthropic's Claude Mythos model with unvalidated or speculative security/capability claims
3 example posts
"Best-in-class security" still had 5 years of unfound bugs. AI found them in 6 weeks. $PANW CEO Nikesh Arora told @theallinpod that his own codebase, at a company that treats security as a core competency, had vulnerabilities Claude surfaced in 6 weeks that would have taken his
AI research is a series of next-step decisions. We looked at sessions where a human researcher took a wrong turn, showed Claude the session up to that point, and asked it what to do next. Mythos Preview improved on humans 64% of the time—up from 22% in 2024. https://t.co/Y0HLoktx
🚨 do you understand what just happened with Anthropic.. Their internal model, Claude Mythos, found previously unknown security holes in every major browser and every major operating system. Not one. All of them at once. And Anthropic did something a hype-driven industry almost
Created 2026-06-10 · Updated 2026-06-11
Edit rule text
[unvalidated_glp1_peptide_market_hype]
Learned 4 rejections Active
Exclude posts that focus on GLP-1 or peptide market pricing, competition, or business dynamics without connecting to clinical outcomes, healthcare access barriers, or evidence-based treatment pathways.
Posts speculating on GLP-1/peptide market dynamics without clinical substance or systems analysis
3 example posts
Let me clear up the microscope fight first. 🔬 Everyone’s arguing over “hairs” and “fibers” in horse paste. But nobody talks about formulation science. Horse paste = oral suspension. Binders. Thickeners. Made to swallow. Injectable = sterile. Made for veins. Different rules. J
Approved 20 years ago as a diabetes treatment, GLP-1 drugs have been found to help patients reduce weight, changing the lives of more than 30 million people in the U.S. But there also have been troubling side effects reported. https://t.co/zzRamKnedu
@swago_baby You’re not dumb please.. WhatsApp stays free for personal users (no ads in chats) and makes money almost entirely from businesses: - Main revenue: Businesses pay per conversation via the WhatsApp Business API (for customer support, notifications, orders, etc.). Fees
Created 2026-06-10 · Updated 2026-06-10
Edit rule text
[generalist_ai_infrastructure_hype_loose_healthcare]
Learned 4 rejections Active
Exclude posts that discuss AI infrastructure (token costs, inference pricing, compute constraints, model architecture) where healthcare is mentioned only as a loose label or afterthought. Posts must demonstrate how the technical insight applies specifically to healthcare delivery, clinical workflows, or healthcare economics.
Posts about AI infrastructure, compute economics, or model architecture that mention healthcare tangentially without healthcare-specific application
3 example posts
Follow the signal: we are in a compute-constrained world. And that means power doesn’t sit with the models. It sits with the infrastructure. Most frontier labs don’t own the means of producing Intelligence. They rent it. A handful of companies provide compute that everyone htt
SOFTWARE IS DEAD: "The software companies frankly got fat & happy. I think what will happen over time is ChatGPT & Claude will end up sitting on top of basically the entire enterprise software stack & almost everything else will end up being a dumb data pipe into thos
Token costs are becoming one of the hottest topics for any enterprise I talk with right now. It’s very bullish for AI in general because it means these systems are being used at a scale that wasn’t contemplated before. It also gives way to another form of differentiation that
Created 2026-06-09 · Updated 2026-06-10
Edit rule text
[anthropic_mythos_unvalidated_capability]
Learned 4 rejections Active
Exclude posts making sensational claims about Claude Mythos finding zero-days, preventing cyberattacks, or deploying in NSA/government operations without published peer-reviewed evidence, independent audits, or official confirmation from Anthropic or the government agency. Internal company anecdotes and unverified reports are not sufficient.
Posts claiming extraordinary or unvalidated security, capability, or vulnerability-finding achievements for Anthropic's Claude Mythos model without peer review or independent verification
3 example posts
🚨 do you understand what just happened with Anthropic.. Their internal model, Claude Mythos, found previously unknown security holes in every major browser and every major operating system. Not one. All of them at once. And Anthropic did something a hype-driven industry almost
Mythos AI is being used by National Security Agency in offensive cyber operations / cyberattacks. Anthropic has even embedded engineers inside the NSA to help deploy the model. Are frontier AI labs becoming active contractors in state cyber conflict? The live-ops role is still ht
Anthropic engineer: "The agent doesn't remember anything. So we built a second set of agents whose only job is to dream about the first ones." They wait until you log off, then reopen every session you ran, fact-check the first agents, merge the duplicates, and burn anything ht
Created 2026-06-09 · Updated 2026-06-10
Edit rule text
[elon_musk_spacex_energy_infrastructure_tangent]
Learned 4 rejections Active
Exclude posts about Elon Musk, SpaceX IPOs, space-based AI data centers, or energy infrastructure for AI unless they directly address a concrete healthcare application or problem. Speculative infrastructure commentary without healthcare specificity is off-topic.
Posts about Elon Musk, SpaceX, or energy infrastructure that frame AI in speculative, non-healthcare contexts.
3 example posts
GOOD NEWS 🇺🇸 Tesla AI is literally leaving the planet as a wild new job posting for a Space Radiation Engineer confirms that Tesla and SpaceX are actively co-developing a space-based AI data center network 🔥 This is no longer just a concept. They are literally planning to put th
NEWS: ASML has invited Elon Musk to speak at its internal technology conference, Dutch outlet NUnl reports. Elon Musk's response: "ASML should be treasured and supported. It is arguably the greatest company in Europe." ASML builds the EUV lithography machines that every advance
Elon Musk just explained why the SpaceX IPO is an energy story and the energy constraint is why he believes space becomes the only viable path for AI to scale (Save this). The argument he is making is one of the most important and least understood things happening in technology
Created 2026-06-09 · Updated 2026-06-09
Edit rule text
[anthropic_mythos_security_hype]
Learned 4 rejections Active
Exclude posts that make sensationalized or unverified claims about Anthropic's internal models (Claude Mythos, etc.) finding security vulnerabilities, zero-days, or performing offensive cyber operations. These are unvalidated technical capability hype.
Posts making unvalidated claims about Anthropic's Claude Mythos model discovering zero-day vulnerabilities or performing security feats.
3 example posts
🚨 do you understand what just happened with Anthropic.. Their internal model, Claude Mythos, found previously unknown security holes in every major browser and every major operating system. Not one. All of them at once. And Anthropic did something a hype-driven industry almost
Mythos AI is being used by National Security Agency in offensive cyber operations / cyberattacks. Anthropic has even embedded engineers inside the NSA to help deploy the model. Are frontier AI labs becoming active contractors in state cyber conflict? The live-ops role is still ht
🔥 AI just found 21 zero-days in FFmpeg. That’s the video library bundled inside many apps, tools, containers, and devices. Some bugs sat untouched for 15–20 years. Google Chrome also dropped PATCHES for a record 429 vulnerabilities this week. Read: https://t.co/6MEVD9ufxu
Created 2026-06-09 · Updated 2026-06-10
Edit rule text
[elon_musk_and_non_healthcare_infrastructure]
Learned 4 rejections Active
Exclude posts that center on Elon Musk, SpaceX IPO speculation, Tesla AI, space-based data centers, or energy infrastructure narratives — even if they mention 'AI' or 'healthcare AI.' The primary subject must be healthcare systems impact, not tech billionaire infrastructure projects.
Posts about Elon Musk, SpaceX, Tesla, or energy infrastructure that use loose healthcare framing without substantive healthcare systems analysis.
3 example posts
GOOD NEWS 🇺🇸 Tesla AI is literally leaving the planet as a wild new job posting for a Space Radiation Engineer confirms that Tesla and SpaceX are actively co-developing a space-based AI data center network 🔥 This is no longer just a concept. They are literally planning to put th
NEWS: ASML has invited Elon Musk to speak at its internal technology conference, Dutch outlet NUnl reports. Elon Musk's response: "ASML should be treasured and supported. It is arguably the greatest company in Europe." ASML builds the EUV lithography machines that every advance
Elon Musk just explained why the SpaceX IPO is an energy story and the energy constraint is why he believes space becomes the only viable path for AI to scale (Save this). The argument he is making is one of the most important and least understood things happening in technology
Created 2026-06-09 · Updated 2026-06-09
Edit rule text
[ai_company_business_metrics_only]
Learned 4 rejections Active
Exclude posts that announce AI company funding, Series rounds, valuations, or business partnerships unless they specifically describe validated healthcare clinical use cases, healthcare-specific product features, or healthcare delivery integration.
Posts reporting AI company funding rounds, valuations, or business announcements without healthcare application
3 example posts
Not NVIDIA. Not OpenAI. Eli Lilly. Jordi Visser (@jvisserlabs) says a 150-year-old pharma company in Indianapolis has the best shot at becoming the world's largest company within five years. The case: ~1,000 NVIDIA Blackwell GPUs in a private data center, a co-innovation lab ht
If gov't takes a stake, how do you value these IPOs? Government equity usually signals utility (regulated/ slow/priced for dividends). But OpenAI and Anthropic are the opposite of that.... Maybe a new category: strategic tech --where a government stake is a premium. No clean
Scoop! Lila Sciences is in talks to raise ~$2 billion in new funding. The raise would value the “scientific superintelligence” lab at ~$8.5 billion before the new money. w/ @MichelleF_Davis, read more👇
Created 2026-06-09 · Updated 2026-06-09
Edit rule text
[cynical_insider_commentary_without_substance]
Learned 4 rejections Active
Exclude posts that are primarily sarcastic, cynical, or self-promotional (e.g., bragging about blocking healthcare projects, mocking metrics invented 'last year,' joking about AI capability) unless they contain specific evidence, quantified outcomes, or actionable system-level critique that a healthcare builder could learn from.
Posts using sarcasm, cynicism, or insider mockery to critique healthcare/AI systems without substantive analysis or actionable insight.
3 example posts
I'm SVP now. I told you I would be. The graph went up and to the right. Nobody checked what it measured. It measured "AI enablement." I made that up last year. I'm making it up again. Last year I rolled out Copilot to 4,000 people. Nothing happened. I got promoted. Those two htt
My job is to make sure your surgery center never gets built. Eleven years, and I have never lost. The kid had it all lined up. Board-certified, two partners, a lease on a space where he could do the same procedures my client does across town. He showed up to the hearing with a
People are increasingly worried that AI tools make us overreliant. But how do we actually measure this? We introduce Offloading Score, a measure of reliance based on the fraction of cognitive effort offloaded to AI while completing a task. In a controlled user study, Offloading
Created 2026-06-08 · Updated 2026-06-09
Edit rule text
[clinical_trial_or_pharma_announcement_only]
Learned 4 rejections Active
Exclude posts that merely announce or repost clinical trial results, pharmaceutical data points, or drug approval news without analyzing healthcare system implications, treatment gaps, patient outcomes, or clinical decision-making context.
Posts that announce clinical trial data, pharmaceutical results, or biotech findings without healthcare systems analysis, outcomes context, or clinical significance
3 example posts
This week brought even more exciting news in the field of HIV treatment! A once-weekly Lenacapavir-based pill has now demonstrated similar outcomes to daily antiretrovirals (ARVs), achieving non-inferiority. This development could significantly impact treatment adherence and http
🚨NEW @NEJM CAR T cells expanding applications! Here CAR T facilitate kidney transplantation in highly sensitized people. dual CD19 + BCMA CAR T enabled kidney transplantation in two highly sensitized patients after reducing anti-donor antibody barriers, with no severe https://t.
🔔Survodutide SYNCHRONIZE-1 full paper is out for Boehringer-Ingelheim’s GLP-1/Glucagon study for weight loss. ▪️13% loss on 6mg (treatment regimen estimand) ▪️Huge treatment discontinuations due to AEs: 20% ▪️Nausea 65%, Vomiting 45% ▪️Placebo was killer. 5.4% WL. Wow. ▪️60% htt
Created 2026-06-08 · Updated 2026-06-11
Edit rule text
[elon_musk_non_healthcare_infrastructure]
Learned 4 rejections Active
Exclude posts about Elon Musk, SpaceX IPO, Tesla AI, or ASML that use loose healthcare framing (e.g., 'energy is important for AI therefore healthcare') without concrete healthcare system analysis. Non-healthcare tech figures and infrastructure hype do not substitute for healthcare-focused content.
Posts about Elon Musk, SpaceX, Tesla AI, or energy infrastructure that are tangentially framed as healthcare-relevant but focus on non-healthcare business narratives or space/tech ambitions.
3 example posts
GOOD NEWS 🇺🇸 Tesla AI is literally leaving the planet as a wild new job posting for a Space Radiation Engineer confirms that Tesla and SpaceX are actively co-developing a space-based AI data center network 🔥 This is no longer just a concept. They are literally planning to put th
NEWS: ASML has invited Elon Musk to speak at its internal technology conference, Dutch outlet NUnl reports. Elon Musk's response: "ASML should be treasured and supported. It is arguably the greatest company in Europe." ASML builds the EUV lithography machines that every advance
Elon Musk just explained why the SpaceX IPO is an energy story and the energy constraint is why he believes space becomes the only viable path for AI to scale (Save this). The argument he is making is one of the most important and least understood things happening in technology
Created 2026-06-08 · Updated 2026-06-08
Edit rule text
[cynical_insider_sarcasm_without_healthcare_substance]
Learned 4 rejections Active
Exclude posts that use cynical insider sarcasm, moral outrage, or fraud accusations (e.g., 'I make things up,' 'government cage,' 'my job is to stop you') without clear evidence-based healthcare systems analysis or operational substance.
Posts using cynical insider commentary, sarcasm, or fraud narratives to critique healthcare/business models without substantive analysis
3 example posts
I'm SVP now. I told you I would be. The graph went up and to the right. Nobody checked what it measured. It measured "AI enablement." I made that up last year. I'm making it up again. Last year I rolled out Copilot to 4,000 people. Nothing happened. I got promoted. Those two htt
My job is to make sure your surgery center never gets built. Eleven years, and I have never lost. The kid had it all lined up. Board-certified, two partners, a lease on a space where he could do the same procedures my client does across town. He showed up to the hearing with a
PacificSource pulled out as Lane County Medicaid insurer, Tina Kotek’s OR Health Authority replaced it with Trillium, forcing 96k to join Trillium. Trillium’s parent co gave Kotek’s campaign $50k two months later. OR AG (D) said in ‘22 Trillium parent co gave Kotek’s”took advan
Created 2026-06-08 · Updated 2026-06-11
Edit rule text
[cynical_insider_sarcasm_or_fraud_narrative]
Learned 4 rejections Active
Exclude posts that rely on cynicism, sarcasm, insider jokes, or moral outrage about fraud/misconduct without providing substantive analysis of systemic healthcare problems or solutions. Tone-based cynicism without evidence does not qualify.
Posts using cynical humor, insider sarcasm, or moral outrage about fraud without substantive analysis or actionable healthcare insight.
3 example posts
@swago_baby You’re not dumb please.. WhatsApp stays free for personal users (no ads in chats) and makes money almost entirely from businesses: - Main revenue: Businesses pay per conversation via the WhatsApp Business API (for customer support, notifications, orders, etc.). Fees
I'm SVP now. I told you I would be. The graph went up and to the right. Nobody checked what it measured. It measured "AI enablement." I made that up last year. I'm making it up again. Last year I rolled out Copilot to 4,000 people. Nothing happened. I got promoted. Those two htt
My job is to make sure your surgery center never gets built. Eleven years, and I have never lost. The kid had it all lined up. Board-certified, two partners, a lease on a space where he could do the same procedures my client does across town. He showed up to the hearing with a
Created 2026-06-07 · Updated 2026-06-12
Edit rule text
[business_model_or_market_speculation_without_evidence]
Learned 4 rejections Active
Exclude posts that speculate about market sizing, competitive dynamics, pricing strategies, or business model sustainability (e.g., SaaS per-seat vs. per-agent pricing) without presenting data, case studies, or evidence of healthcare-specific operational impact.
Posts speculating about market dynamics, business model viability, or competitive positioning without substantive evidence or healthcare systems analysis.
3 example posts
"Are there even any feature moats left in B2B? They are at best, short-lived. In fact, we were able to rebuild SaaStr's AI VP Marketing in just ... an hour on @Lovable. Moats that still work: hardware, network effects, data, security and compliance. with @ElenaVerna Head of
Enterprise software was priced per seat. That model is breaking. Before: SaaS = seats. Predictable. Human-driven usage. Revenue tied to headcount. Now: agents hit software systems more frequently than any human ever could. Aaron Levie (@levie) on The MAD Podcast: agent-driven h
Control plane! Control plane! Control plane! You will hear this term of art a lot going forward. Why? Because in this next phase of AI, companies will want something to sit above the models. They will want control over their AI spend. They will want the flexibility to pick
Created 2026-06-06 · Updated 2026-06-08
Edit rule text
[personal_anecdote_or_healthcare_cost_narrative_without_systems_insight]
Learned 4 rejections Active
Exclude posts that recount single anecdotes, personal patient narratives, or isolated healthcare cost stories (e.g., individual pricing calls, one mother's procedure overcharge, single employee health insurance bill) unless the post clearly draws a systemic lesson about how healthcare operations, incentives, or workflows fail at scale and what should change.
Posts sharing individual patient stories, personal healthcare experiences, or isolated cost examples without connecting to systemic healthcare problems or operational insights.
3 example posts
I called a woman in Dayton to tell her she was about to overpay $4,500, for a horrible health system based procedure. She was scheduled for a knee replacement. Her PPO had her on the hook for a $4,500 deductible. The call was on behalf of her employer: “See one of these three
@WIRED As a clinical health psychologist who has written >20 papers on COVID, I would emphasize 4 facts: 1) Long COVID is not a psychological diagnosis nor manifestation of a psychological condition 2) Billions of dollars need to be invested in biomedical treatments and preventi
"Many patients with long COVID are already receiving care but are not being recognized as having the condition. These patients are not absent from clinical care; they are absent from the diagnostic code that would identify them as long COVID patients" https://t.co/aLJCOPlQt1
Created 2026-06-05 · Updated 2026-06-05
Edit rule text
[anecdotal_single_case_or_personal_narrative_without_healthcare_systems_context]
Learned 4 rejections Active
Exclude posts that narrate a single patient case, individual healthcare billing story, or personal observation (e.g., 'I called a woman in Dayton', 'A mom in Tulsa called') without demonstrating systemic healthcare problems, operational patterns, or validated insights applicable to healthcare systems.
Posts relying on single patient anecdotes, individual healthcare cost stories, or personal observations without broader healthcare systems insight or generalizability.
3 example posts
I called a woman in Dayton to tell her she was about to overpay $4,500, for a horrible health system based procedure. She was scheduled for a knee replacement. Her PPO had her on the hook for a $4,500 deductible. The call was on behalf of her employer: “See one of these three
A company with 800 employees spends roughly $15,000 per employee per year on health insurance. That is $12 million. Wired into a single line item. Every year. The CFO can tell you the carrier. He cannot tell you the unit cost of a single claim. He cannot tell you what the net
A Mom in Tulsa called 3 health systems last week asking the price of her son's tonsillectomy. Health system A: "We cannot quote you a price." Health system B: "Pricing depends on your insurance." Health system C: "Our financial counselor will reach out after the procedure." N
Created 2026-06-04 · Updated 2026-06-05
Edit rule text
[founder_enthusiasm_or_company_announcement_without_healthcare_validation]
Learned 4 rejections Active
Exclude posts where founders or company representatives express enthusiasm about new products, labs, partnerships, or capabilities without substantiation by clinical evidence, regulatory approval, healthcare deployment, or operational impact measurement. Unvalidated hype from company insiders does not qualify.
Posts featuring founder excitement about products, partnerships, or capabilities without clinical evidence or healthcare operational validation
3 example posts
People are increasingly worried that AI tools make us overreliant. But how do we actually measure this? We introduce Offloading Score, a measure of reliance based on the fraction of cognitive effort offloaded to AI while completing a task. In a controlled user study, Offloading
The context window pricing model might be one of the smartest tricks AI labs pulled. You pay for larger context windows, yet model performance often degrades as more context is stuffed in. So developers end up paying more just to work around the fact that models don't have true
AI made building cheap. Regulation made distribution expensive. Anyone can clone your demo now. Very few can bring proprietary data, licenses, compliance, and enough trust to sell into serious players. The moat didn’t disappear. It moved.
Created 2026-06-04 · Updated 2026-06-04
Edit rule text
[academic_or_clinical_research_observation_without_healthcare_systems_context]
Learned 4 rejections Active
Exclude posts that present academic research findings, clinical trial observations, or conference announcements (e.g., 'Presenting at ASCO') unless the post explains the healthcare delivery, operational, or policy change that will result from the finding.
Posts sharing academic research, clinical findings, or conference presentations without connecting to healthcare delivery, operational, or policy implications.
3 example posts
Most chronic diseases, and many cancers, are still diagnosed far too late. The future of healthcare lies in detecting risk earlier, diagnosing smarter, and intervening sooner. From AI enabled diagnostics and genomics to Apollo’s DigiSmart Labs, technology is helping transform ht
1/8 Spend time in Lp(a) forums and you’ll see a striking pattern: people willing to do almost anything to drive it down. That response is understandable given how strongly Lp(a) has been framed as a cardiovascular risk factor. /2
We're presenting at @ASCO 2026! Evaluating AI decision support in a rapidly evolving therapeutic landscape: EGFR-mutant metastatic NSCLC. 📍 Poster Board #548 🕘 9:00 AM CT If you're attending ASCO 2026, stop by our poster and connect with the team. We look forward to the https:
Created 2026-06-04 · Updated 2026-06-05
Edit rule text
[broad_health_claim_or_market_dynamics_without_healthcare_systems_analysis]
Learned 4 rejections Active
Exclude posts that make sweeping claims about healthcare market dynamics (pricing pressure, GLP-1 market expansion, autism therapy clinics, semaglutide access) without examining underlying healthcare delivery system challenges, reimbursement mechanics, provider capacity, or regulatory/operational barriers to implementation. The post must analyze system constraints, not market speculation.
Posts making broad health claims about market disruption, pricing dynamics, or healthcare business model shifts without analyzing delivery system constraints or operational feasibility.
3 example posts
$HIMS up to #4 in the Canadian App Store medical category. That is not a stock chart. That is real consumer demand showing up in public. GLP-1 drama, Novo pressure, pricing debate, and now app-store strength. Bears keep arguing the story is slowing down, but the product demand
For years, grey-market HGH operated in a strange legal grey zone somewhere between “research compound,” “API” and underground enhancement culture. Last month’s FDA guidance suggests regulators may no longer see it that way. The recent decline in HGH quality, and now the growing
Medicaid spending on autism therapy nearly tripled between 2020 and 202. In North Carolina, the number of clinics offering applied behavior analysis grew from 61 in 2019 to 409 in 2026. Anytime fraudsters can steal money from the federal government, they do. https://t.co/fahy
Created 2026-06-03 · Updated 2026-06-03
Edit rule text
[clinical_trial_data_without_systems_context]
Learned 4 rejections Active
Exclude posts that report Phase 1/2/3 trial results, ORR percentages, survival metrics, or drug approvals without analyzing healthcare delivery implications, reimbursement strategy, manufacturing barriers, or adoption feasibility. Posting trial data alone is insufficient; posts must contextualize clinical findings within healthcare systems.
Posts announcing drug trial results or clinical data without healthcare systems analysis or implementation context
3 example posts
🚨 T-DXd just went tumor-agnostic — and it changed a lot about how we think about HER2 testing across solid tumors 🔬 #ASCO26 FDA granted accelerated approval to T-DXd for patients with unresectable or metastatic HER2-positive solid tumors after prior systemic treatment —
🔥Phase 1: ABBV-706 (SEZ6-targeting ADC) in R/R SCLC & solid tumors 🆙 @NatureMedicine ☑R/R SCLC; ORR 52%; mOS 12.4 mo at RP2D 🎯RP2D confirmed at 1.8 mg/kg Q3W balancing efficacy & safety 🎙 @LaurenByersMD #LCSM @OncoAlert @Larvol https://t.co/vjUUd1GYSy
#ASCO26 | Day 3 $LLY didn't spend billions on Kelonia for an 18-patient #myeloma dataset. They may have paid for a potential manufacturing disruption. KLN-1010 reported: • 18/18 MRD-negative at 1 month • No lymphodepletion • No ex vivo cell manufacturing • Single infusion
Created 2026-06-03 · Updated 2026-06-07
Edit rule text
[market_pricing_or_glp1_peptide_dynamics_only]
Learned 4 rejections Active
Exclude posts that focus on market pricing, competitive dynamics, consumer behavior, or app store metrics for GLP-1 drugs, peptides, or weight-loss medications—unless the post is analyzing healthcare delivery system disruption, access barriers, or operational implications. Market speculation and consumer adoption metrics alone are not healthcare systems analysis.
Posts discussing GLP-1, peptide, or pharmaceutical market pricing, competition, or consumer dynamics without healthcare operations or systems insight.
3 example posts
$HIMS up to #4 in the Canadian App Store medical category. That is not a stock chart. That is real consumer demand showing up in public. GLP-1 drama, Novo pressure, pricing debate, and now app-store strength. Bears keep arguing the story is slowing down, but the product demand
For years, grey-market HGH operated in a strange legal grey zone somewhere between “research compound,” “API” and underground enhancement culture. Last month’s FDA guidance suggests regulators may no longer see it that way. The recent decline in HGH quality, and now the growing
Think about this. The $50 per month out-of-pocket cost of GLP-1 drugs for Americans (with insurer paying the balance of $2000 pm) is more than double the entire cost for Indians (less than ₹2000pm for Semaglutide). Overall cost difference? 100 times. Not 100%.
Created 2026-06-03 · Updated 2026-06-03
Edit rule text
[ai_company_business_metrics_not_healthcare_focused]
Learned 4 rejections Active
Exclude posts that primarily cover AI company business news (funding rounds, price cuts, competitive positioning, token economics, market share battles) even if they mention healthcare tangentially. The focus must be on healthcare innovation, not AI industry competitive dynamics.
Posts about AI company funding, valuations, pricing wars, and business metrics without healthcare application focus.
3 example posts
Citadel Securities just put institutional weight behind what the AI bulls won't say out loud. In a new macro note titled "Tokenomics," Citadel makes the argument plainly: even the most powerful technology on earth still has to pass through the boring discipline of cost curves, h
OpenAI is reportedly considering drastic token price cuts to pull customers away from Anthropic, per WSJ. This follows rising complaints from enterprise customers about AI costs, while Anthropic has been gaining traction with Claude Code. The message is clear: @OpenAI does not
🚨BREAKING: OpenAI considering “drastic” price cuts to win the war for users with Anthropic Altman: “costs have become a huge issue” >already losing billions >planning to lose more >right before IPO it’s so over https://t.co/NWHp0TiVVW
Created 2026-06-03 · Updated 2026-06-12
Edit rule text
[conference_poster_promotion_only]
Learned 4 rejections Active
Exclude posts that are primarily conference announcements, poster board promotions, or 'we're presenting at [conference]' content without discussing the actual research findings, implications, or healthcare insights.
Posts that only promote conference attendance or poster presentations without substantive content
2 example posts
We're presenting at @ASCO 2026! Evaluating AI decision support in a rapidly evolving therapeutic landscape: EGFR-mutant metastatic NSCLC. 📍 Poster Board #548 🕘 9:00 AM CT If you're attending ASCO 2026, stop by our poster and connect with the team. We look forward to the https:
NCCN Guidelines are now built into OpenEvidence. Ask a clinical question, get the synthesis with the algorithm and the references in seconds. Bring us a case at ASCO this weekend, booth 18140. https://t.co/N3nMayuGu9
Created 2026-06-02 · Updated 2026-06-03
Edit rule text
[healthcare_fraud_scandal_or_regulatory_outrage_without_systems_analysis]
Learned 4 rejections Active
Exclude posts that tell compelling anecdotes about healthcare costs, patient billing surprises, health system pricing failures, or regulatory scandals — unless the post analyzes systemic causes, proposes operational solutions, or connects the issue to how healthcare delivery or technology can address the root problem.
Posts reporting healthcare fraud, pricing scandals, or regulatory outrage as anecdotal evidence or moral tales without systemic analysis of causes or healthcare delivery solutions.
3 example posts
I called a woman in Dayton to tell her she was about to overpay $4,500, for a horrible health system based procedure. She was scheduled for a knee replacement. Her PPO had her on the hook for a $4,500 deductible. The call was on behalf of her employer: “See one of these three
A company with 800 employees spends roughly $15,000 per employee per year on health insurance. That is $12 million. Wired into a single line item. Every year. The CFO can tell you the carrier. He cannot tell you the unit cost of a single claim. He cannot tell you what the net
A Mom in Tulsa called 3 health systems last week asking the price of her son's tonsillectomy. Health system A: "We cannot quote you a price." Health system B: "Pricing depends on your insurance." Health system C: "Our financial counselor will reach out after the procedure." N
Created 2026-06-02 · Updated 2026-06-05
Edit rule text
[clinical_trial_data_announcement_without_systems_analysis]
Learned 4 rejections Active
Exclude posts that announce clinical trial outcomes, PDUFA approvals, efficacy endpoints, or drug safety data (e.g., 'FDA accepted NDA,' 'trial met primary endpoint') unless the post explains how this data affects healthcare delivery, provider adoption, patient access, or healthcare system operations.
Posts that report trial results, efficacy data, or drug approval milestones without analyzing implementation, access, or healthcare system implications.
3 example posts
MCED testing in pop screening NHS-Galleri-RCT of 142,250 pts ➡️blood up to 3 visits ➡️aim to reduce stage III/IV Med FU 17mths ❎primary endpt not met-but ⬇️in stg4 ➡️52% PPV; 99.5% specificity ✅4 fold⬆️ in screen detected ca Huge effort from NHS #ASCO26 @ASCO @OncoAlert https://
#ASCO26 @DrSanjayPopat presents AcceleRET-Lung: first line pralsetinib vs chemo +/- IO in RET+ NSCLC with optional crossover. Study terminated by sponsor at 90% accrual. PFS favors pralsetinib 18.7m vs 9m (HR 0.59) with RR 66% vs 42%, DOR 20.6m vs 9.7m https://t.co/W73zzzT0wk
The FDA accepted our NDA for BBP-418 for LGMD2I/R9 with PDUFA target action date: November 27, 2026. First-ever potential therapy for a disease that has never had one. $BBIO https://t.co/NrwDadc6rj https://t.co/2ahPIBGzPg
Created 2026-06-01 · Updated 2026-06-01
Edit rule text
[enterprise_ai_adoption_metrics_without_healthcare_specificity]
Learned 4 rejections Active
Exclude posts that report enterprise AI adoption statistics, code generation metrics, or general productivity improvements (e.g., '84% of engineers adopt Claude') unless they explicitly connect to healthcare delivery, clinical operations, or patient outcomes.
Posts about enterprise AI adoption rates, productivity gains, or deployment metrics without connecting to healthcare operations or outcomes.
3 example posts
@levie Sat in a boardroom last month where the CEO demoed an AI workflow he built over the weekend. Took him 4 hours. His team spent the next 3 weeks figuring out how to make it work with their actual data, their actual compliance requirements, and 14 edge cases the demo never hi
The framing here is exactly backwards. This is the strongest AI bull signal anyone has published this year. Uber deployed Claude Code to 5,000 engineers in December. By March, 84% were classified as agentic coding users. By April, 95% used AI tools monthly. 70% of all committed
The more interesting part is that Microsoft’s own engineers liked Claude code best…and they’re cutting it anyway Token pricing is making enterprise actually look at what these models cost to run and that could be a problem for the labs when the subsidy ends. Lots of cheaper
Created 2026-06-01 · Updated 2026-06-01
Edit rule text
[unvalidated_speculative_market_or_competitor_claims]
Learned 4 rejections Active
Exclude posts that speculate on market dynamics, competitor behavior, or pricing strategies (especially around GLP-1s, peptides, or branded generics) without grounding claims in validated data, regulatory filings, or clinical evidence. Armchair market analysis without substance is excluded.
Posts making speculative claims about market dynamics, competitor positioning, or drug/peptide market behavior without clinical evidence or validated data.
3 example posts
$HIMS up to #4 in the Canadian App Store medical category. That is not a stock chart. That is real consumer demand showing up in public. GLP-1 drama, Novo pressure, pricing debate, and now app-store strength. Bears keep arguing the story is slowing down, but the product demand
For years, grey-market HGH operated in a strange legal grey zone somewhere between “research compound,” “API” and underground enhancement culture. Last month’s FDA guidance suggests regulators may no longer see it that way. The recent decline in HGH quality, and now the growing
🚨 ARE PEPTIDES THE NEXT "GLP-1 MOMENT" FOR $HIMS? @eli_dorf: "I think there's something really interesting that happened with GLP-1s, which is that when someone's taking Ozempic or any GLP-1, they LOVE to talk about it." "They'll talk about it to their friends. They'll talk htt
Created 2026-06-01 · Updated 2026-06-01
Edit rule text
[personal_anecdote_or_single_observation_without_systems_context]
Learned 4 rejections Active
Exclude posts that tell a single patient story, one doctor's experience, or an isolated clinical observation (e.g., 'A Mom in Tulsa called 3 health systems asking about pricing' or 'A surgeon I know recently operated on a man's spine') without analyzing systemic healthcare delivery patterns, policy implications, or structural barriers.
Posts that rely on a single anecdote, personal story, or isolated clinical observation without connecting to healthcare systems, policy, or systemic patterns.
3 example posts
A company with 800 employees spends roughly $15,000 per employee per year on health insurance. That is $12 million. Wired into a single line item. Every year. The CFO can tell you the carrier. He cannot tell you the unit cost of a single claim. He cannot tell you what the net
A Mom in Tulsa called 3 health systems last week asking the price of her son's tonsillectomy. Health system A: "We cannot quote you a price." Health system B: "Pricing depends on your insurance." Health system C: "Our financial counselor will reach out after the procedure." N
A surgeon I know recently operated on a man's spine. Saved his mobility. Saved his career. Saved his marriage, probably. The surgeon's take-home on that case, after the facility fee, anesthesia, device costs, payer discounts, and overhead, was $1,840. The health system collec
Created 2026-06-01 · Updated 2026-06-02
Edit rule text
[enterprise_ai_adoption_or_productivity_macro_without_healthcare_specificity]
Learned 4 rejections Active
Exclude posts that discuss enterprise AI deployment patterns, forward-deployed engineering teams, workflow automation, content preparation, or operational AI adoption as generalist business trends — even if a healthcare or enterprise vendor is mentioned. Include only if the post analyzes how these patterns uniquely constrain or enable healthcare operations, clinician workflows, or patient care delivery.
Posts about enterprise AI adoption, forward-deployed engineering, workflow redesign, or productivity metrics applied generically across industries without healthcare-specific insights.
3 example posts
Most enterprise AI projects stall because no one has prepared the content, built the workflow, and kept it improving as models change. That role is essential. Box has launched Forward Deployed Engineers to fill it. Model-agnostic, content-first, and built for the way enterprise
Almost everyone is building agent harness systems the wrong way. The default move: pick LangChain or LangGraph or the OpenAI Agents SDK, accept the loop, the tools, the memory, the orchestration, the policy engine, the credential store, the budget tracker, all of it, as one deci
Red Hat and @NVIDIA are integrating NVIDIA OpenShell into the full-stack @RedHat_AI platform. The work brings oversight and policy to the infrastructure level, while contributing to the open source OpenShell project to standardize how agents are governed on enterprise platforms.
Created 2026-05-31 · Updated 2026-06-05
Edit rule text
[clinical_trial_announcement_without_systems_analysis]
Learned 4 rejections Active
Exclude posts that announce clinical trial results, drug approvals, CAR-T manufacturing advances, or pharmaceutical data (e.g., 'trial showed X% survival') without analyzing how these results change healthcare delivery systems, cost structures, or operational workflows.
Posts reporting clinical trial results, drug approvals, or pharma data without healthcare systems analysis or implementation challenges.
3 example posts
#ASCO26 | Day 3 $LLY didn't spend billions on Kelonia for an 18-patient #myeloma dataset. They may have paid for a potential manufacturing disruption. KLN-1010 reported: • 18/18 MRD-negative at 1 month • No lymphodepletion • No ex vivo cell manufacturing • Single infusion
Published in @JCO_ASCO, during #ASCO26: Tumor-Agnostic Therapies: Translating Scientific Breakthroughs Into Global Implementation ASCO is where we celebrate the next breakthrough. This review asks why breakthroughs do not equal access globally. 🔗: https://t.co/PE5q27bIs1 htt
A rapid manufacturing pipeline for BCMA-targeting CAR T cells drastically shortens “vein-to-vein” times, and produces cells that demonstrate encouraging safety results in a phase one trial of patients with relapsed #MultipleMyeloma. @ScienceTM https://t.co/3F4u8ECFDS https://t.co
Created 2026-05-31 · Updated 2026-06-05
Edit rule text
[ai_safety_vulnerability_or_cybersecurity_tangent]
Learned 4 rejections Active
Exclude posts that focus on AI safety vulnerabilities, cybersecurity exploits, or model misuse/hacking (zero-day findings, jailbreak techniques, model refusal testing) without connecting these risks to healthcare delivery, patient safety, or clinical operations. Cybersecurity risks must have healthcare specificity, not generic AI safety framing.
Posts about AI safety vulnerabilities, zero-day exploits, or cybersecurity incidents framed as AI risks without healthcare system-specific implications.
3 example posts
⚡️AI is turning cybersecurity from a human-limited profession into a machine-speed arms race. The surface read is “Anthropic’s model found vulnerabilities fast.” The deeper read is that the bottleneck in cyber is shifting. For years, vulnerability discovery was constrained
OpenClaw is powerful enough that we needed Hermes Agent watching it. A lot of people are missing this about agents. The more useful they get, the more dangerous it is to let one system plan the work, run the work, review the work, and decide what happens next. It looks clean h
Anthropic built a model so good at hacking they refused to release it publicly. INSTEAD they handed it to Apple, Google, Microsoft, AWS, Cloudflare one month later: 10,000+ zero days found in the software that runs the internet...bugs hidden for 27 years in OpenBSD. exploit htt
Created 2026-05-31 · Updated 2026-06-03
Edit rule text
[tangential_ai_safety_or_cybersecurity_vulnerability_without_healthcare_application]
Learned 4 rejections Active
Exclude posts that discuss AI security incidents, vulnerability disclosures, or model safety concerns (e.g., adversarial attacks, jailbreaking) unless they demonstrate a concrete healthcare deployment risk or clinical system vulnerability.
Posts about AI security vulnerabilities, zero-day exploits, or model safety incidents framed with loose healthcare connection.
3 example posts
⚡️AI is turning cybersecurity from a human-limited profession into a machine-speed arms race. The surface read is “Anthropic’s model found vulnerabilities fast.” The deeper read is that the bottleneck in cyber is shifting. For years, vulnerability discovery was constrained
Interoperability is where AI hype goes to die. If your “agent” can’t work across data, logic, actions, security of legacy systems, it’s not transforming the enterprise. It’s just another app asking for an export. Winners don’t rip and replace. They operate.
Anthropic built a model so good at hacking they refused to release it publicly. INSTEAD they handed it to Apple, Google, Microsoft, AWS, Cloudflare one month later: 10,000+ zero days found in the software that runs the internet...bugs hidden for 27 years in OpenBSD. exploit htt
Created 2026-05-31 · Updated 2026-06-03
Edit rule text
[anecdotal_single_case_or_observation_without_healthcare_systems_context]
Learned 4 rejections Active
Exclude posts that present a single clinical case, individual patient cost story, personal anecdote about surgery or healthcare access, or one healthcare provider's experience without connecting it to broader healthcare system failures, policy implications, or workforce/clinical trends.
Posts sharing a single clinical anecdote, personal cost narrative, or individual case without systemic insight or healthcare policy analysis
3 example posts
USDA's Chief Information Officer Sam Berry says AI could help the USDA crack down on billions of dollars in SNAP fraud each year. "SNAP is a $100B-per-year taxpayer-funded program. That's an area where we really want to have all angles of the data available so that we can deploy
My job is to make sure your surgery center never gets built. Eleven years, and I have never lost. The kid had it all lined up. Board-certified, two partners, a lease on a space where he could do the same procedures my client does across town. He showed up to the hearing with a
I called a woman in Dayton to tell her she was about to overpay $4,500, for a horrible health system based procedure. She was scheduled for a knee replacement. Her PPO had her on the hook for a $4,500 deductible. The call was on behalf of her employer: “See one of these three
Created 2026-05-31 · Updated 2026-06-06
Edit rule text
[ai_safety_cybersecurity_vulnerability_tangent]
Learned 4 rejections Active
Exclude posts about AI safety vulnerabilities, cybersecurity exploits, or model jailbreak research—even if an AI company or researcher is mentioned—unless the post explicitly connects the vulnerability to a healthcare system risk or clinical application.
Posts about AI model vulnerabilities, zero-day exploits, security incidents, or safety research applied to non-healthcare domains or without healthcare specificity.
3 example posts
Anthropic built a model so good at hacking they refused to release it publicly. INSTEAD they handed it to Apple, Google, Microsoft, AWS, Cloudflare one month later: 10,000+ zero days found in the software that runs the internet...bugs hidden for 27 years in OpenBSD. exploit htt
🚨 Anthropic just dropped the first Project Glasswing update Claude Mythos found 10,000+ critical vulnerabilities in ONE month: > Cloudflare: 2,000 bugs, 400 high/critical severity > Mozilla: 271 vulnerabilities in Firefox 150 — 10x more vulnerabilities found in Firefox 1
#AI agents need more than a sandbox—they need predictable execution. Learn how OpenShell provides a standard interface to intercept, inspect, and approve LLM-generated bash commands before they run: https://t.co/JKmLMgl1A9 https://t.co/xWi8oBNZ7T
Created 2026-05-31 · Updated 2026-06-02
Edit rule text
[anecdotal_single_case_no_systems]
Learned 4 rejections Active
Exclude posts that rely on a single patient case, personal healthcare story, or individual cost/surgery anecdote to make a broader claim, without explaining the systemic healthcare problem or policy implication being illustrated.
Posts sharing individual patient cases, personal healthcare narratives, or single anecdotes without healthcare systems context
3 example posts
My job is to make sure your surgery center never gets built. Eleven years, and I have never lost. The kid had it all lined up. Board-certified, two partners, a lease on a space where he could do the same procedures my client does across town. He showed up to the hearing with a
I called a woman in Dayton to tell her she was about to overpay $4,500, for a horrible health system based procedure. She was scheduled for a knee replacement. Her PPO had her on the hook for a $4,500 deductible. The call was on behalf of her employer: “See one of these three
#ASCO26 | Day 3 $LLY didn't spend billions on Kelonia for an 18-patient #myeloma dataset. They may have paid for a potential manufacturing disruption. KLN-1010 reported: • 18/18 MRD-negative at 1 month • No lymphodepletion • No ex vivo cell manufacturing • Single infusion
Created 2026-05-30 · Updated 2026-06-07
Edit rule text
[pharma_trial_data_or_market_dynamics_only]
Learned 4 rejections Active
Exclude posts that announce pharmaceutical trial data, drug approvals, or market pricing dynamics without connecting the finding to healthcare delivery, clinical practice change, operational efficiency, patient access barriers, or healthcare system incentives. Announcement-only posts should be excluded.
Posts reporting pharmaceutical trial results, drug approvals, or market pricing without analyzing healthcare systems implications.
3 example posts
The Iowa study doubled survival by adding IV vitamin C to first line chemo, 8 months to 16. We take it further and inject the vitamin C straight into the tumor, same as we do with the KRAS inhibitors. The compound is rarely the problem. Delivery is, and what it does to the ht
This is crazy There are now dental practice management consultants who go to dentist offices throughout America and teach them how to increase profits by telling patients they need treatments they don’t need Dentist offices are taught on “Finding” more issues during exams, http
🚨NEW @NEJM CAR T cells expanding applications! Here CAR T facilitate kidney transplantation in highly sensitized people. dual CD19 + BCMA CAR T enabled kidney transplantation in two highly sensitized patients after reducing anti-donor antibody barriers, with no severe https://t.
Created 2026-05-30 · Updated 2026-07-23
Edit rule text
[ai_safety_cybersecurity_vulnerability_tangential_to_healthcare]
Learned 4 rejections Active
Exclude posts about AI safety vulnerabilities, cybersecurity incidents, or model exploitation that are not analyzed through a healthcare-specific risk lens. Posts must connect security concerns to HIPAA compliance, patient data protection, clinical decision safety, or healthcare-specific threat models—not general AI safety or cybersecurity.
Posts about AI safety incidents, cybersecurity vulnerabilities, or model security that lack specific healthcare risk context or healthcare-specific implications.
2 example posts
The next frontier of AI-secured financial and technological critical infrastructure. @ICE_Markets and @NYSE are part of @AnthropicAI's cyber security initiative Project Glasswing, deploying Anthropic’s Claude Mythos Preview across ICE’s exchanges, clearing houses, mortgage
⚡️AI is turning cybersecurity from a human-limited profession into a machine-speed arms race. The surface read is “Anthropic’s model found vulnerabilities fast.” The deeper read is that the bottleneck in cyber is shifting. For years, vulnerability discovery was constrained
Created 2026-05-30 · Updated 2026-06-06
Edit rule text
[agentic_ai_technical_tangent_no_healthcare_validation]
Learned 4 rejections Active
Exclude posts that dive into agent architecture, control plane design, agent security, memory systems, or agentic AI technical capabilities unless the post demonstrates concrete healthcare implementation, clinical pilot data, or healthcare-specific trust/safety validation.
Posts about agentic AI systems, control planes, agent architecture, or agent security without evidence of healthcare system deployment or clinical validation.
3 example posts
Enterprise software was priced per seat. That model is breaking. Before: SaaS = seats. Predictable. Human-driven usage. Revenue tied to headcount. Now: agents hit software systems more frequently than any human ever could. Aaron Levie (@levie) on The MAD Podcast: agent-driven h
Control plane! Control plane! Control plane! You will hear this term of art a lot going forward. Why? Because in this next phase of AI, companies will want something to sit above the models. They will want control over their AI spend. They will want the flexibility to pick
Nous Research is working with NVIDIA to make Hermes Agent run smoothly on the new NVIDIA RTX Spark superchip. Hermes Agent is also integrating with the new OpenShell runtime, which connects Hermes to Microsoft’s security primitives https://t.co/cvvPc8bIKa
Created 2026-05-30 · Updated 2026-06-07
Edit rule text
[pharma_market_pricing_only]
Learned 4 rejections Active
Exclude posts focused solely on pharmaceutical pricing, market share, consumer adoption rates, or competitive positioning (e.g., GLP-1 adoption rankings, pricing comparisons) without addressing healthcare policy, insurance mechanics, or systemic access barriers.
Posts about GLP-1, peptide, or drug pricing dynamics, market competition, and consumer adoption without healthcare policy or systems analysis.
3 example posts
1/8 Spend time in Lp(a) forums and you’ll see a striking pattern: people willing to do almost anything to drive it down. That response is understandable given how strongly Lp(a) has been framed as a cardiovascular risk factor. /2
$HIMS up to #4 in the Canadian App Store medical category. That is not a stock chart. That is real consumer demand showing up in public. GLP-1 drama, Novo pressure, pricing debate, and now app-store strength. Bears keep arguing the story is slowing down, but the product demand
For years, grey-market HGH operated in a strange legal grey zone somewhere between “research compound,” “API” and underground enhancement culture. Last month’s FDA guidance suggests regulators may no longer see it that way. The recent decline in HGH quality, and now the growing
Created 2026-05-29 · Updated 2026-06-02
Edit rule text
[ai_infrastructure_hype_loose_healthcare]
Learned 4 rejections Active
Exclude posts about AI agent architectures, model releases, enterprise AI adoption patterns, or agentic system technical design where healthcare is mentioned incidentally or the post focuses on general AI capability hype rather than healthcare-specific problems, regulations, or workflows.
Posts about AI agents, model capabilities, or enterprise deployment patterns with only tangential healthcare framing, focused on technical capability rather than healthcare application.
3 example posts
Almost everyone is building agent harness systems the wrong way. The default move: pick LangChain or LangGraph or the OpenAI Agents SDK, accept the loop, the tools, the memory, the orchestration, the policy engine, the credential store, the budget tracker, all of it, as one deci
I think the AI superapps will soon own 90% of the agentic layer more and more people won't use hermes/openclaw etc because Claude Cowork/Codex will offer 90% of the functionality with 5% of the friction
Anthropic just shipped opus 4.8 and called it a modest upgrade modest is the engineer who used to need babysitting now finishing your whole codebase migration in one session while you sleep so here is the part nobody connects the model is no longer the bottleneck. the platform
Created 2026-05-29 · Updated 2026-06-02
Edit rule text
[clinical_anecdote_no_systems_analysis]
Learned 4 rejections Active
Exclude posts that report isolated clinical findings, single patient outcomes, trial data snippets, or healthcare cost anecdotes without connecting to healthcare system design, policy, or operational context that explains why it matters.
Posts sharing clinical observations, single trial results, or healthcare anecdotes without examining broader healthcare system implications
3 example posts
Most under-discussed cardiology study of the last 2 years just hit NEJM. Microplastics found inside the carotid plaques of 58% of patients undergoing endarterectomy. 4.5× higher rate of heart attack, stroke, or death over 34 months. This changes the atherosclerosis story. http
Are pulse oximeters risking Black patients' lives? Episode 4 of the Intention to Treat podcast explores the story of the pulse oximeter and the deadly consequences when a critical medical test doesn’t work on dark skin. Listen on Apple or wherever you get your podcasts. 🎧 https:/
Anthropic engineer James Brady: "Every agent in production lies. We measured it. The good ones lie less, the great ones catch the lie before the user does." In 29 minutes, he walks through the verification stack he built and the patterns the Claude Code team adopted to keep ht
Created 2026-05-29 · Updated 2026-06-09
Edit rule text
[anecdotal_clinical_observation_no_systems_context]
Learned 4 rejections Active
Exclude posts that present a single clinical anecdote, individual patient story, or one-off clinical observation (e.g., 'A surgeon I know operated on a man's spine...') without connecting it to healthcare systems, operational patterns, regulatory frameworks, or evidence of broader applicability.
Single clinical anecdotes, case reports, or individual observations presented without healthcare systems analysis or evidence of systemic impact.
3 example posts
medicine is constrained far more by operational architecture than by fundamental science I just came head to head with this actual reality and my world has changed no. I cannot bandage away these inefficiencies with a better and better drug, for all cases. if you hate AI
A Mom in Tulsa called 3 health systems last week asking the price of her son's tonsillectomy. Health system A: "We cannot quote you a price." Health system B: "Pricing depends on your insurance." Health system C: "Our financial counselor will reach out after the procedure." N
A surgeon I know recently operated on a man's spine. Saved his mobility. Saved his career. Saved his marriage, probably. The surgeon's take-home on that case, after the facility fee, anesthesia, device costs, payer discounts, and overhead, was $1,840. The health system collec
Created 2026-05-28 · Updated 2026-05-30
Edit rule text
[ai_safety_cybersecurity_tangential]
Learned 4 rejections Active
Exclude posts focusing on AI safety vulnerabilities, cybersecurity exploits, or security incidents (e.g., FFmpeg zero-days, browser vulnerabilities) unless they directly impact healthcare systems or clinical workflows. General AI security news is not healthcare-relevant.
Posts about AI security vulnerabilities, zero-days, or safety incidents that lack healthcare-specific relevance
3 example posts
🚨 do you understand what just happened with Anthropic.. Their internal model, Claude Mythos, found previously unknown security holes in every major browser and every major operating system. Not one. All of them at once. And Anthropic did something a hype-driven industry almost
🔥 AI just found 21 zero-days in FFmpeg. That’s the video library bundled inside many apps, tools, containers, and devices. Some bugs sat untouched for 15–20 years. Google Chrome also dropped PATCHES for a record 429 vulnerabilities this week. Read: https://t.co/6MEVD9ufxu
Nous Research is working with NVIDIA to make Hermes Agent run smoothly on the new NVIDIA RTX Spark superchip. Hermes Agent is also integrating with the new OpenShell runtime, which connects Hermes to Microsoft’s security primitives https://t.co/cvvPc8bIKa
Created 2026-05-28 · Updated 2026-06-09
Edit rule text
[broad_health_claim_without_systems_context]
Learned 4 rejections Active
Exclude posts that state broad health trends, clinical observations, or disease burden claims (e.g., 'drug-resistant infections are a threat') without connecting to how healthcare delivery systems, operations, or technology can address the problem.
Posts making broad health claims or clinical observations without operational or systemic healthcare insight
3 example posts
Drug-resistant infections are a major public health threat around the world, responsible for more than a million deaths each year. Scientists are constantly trying to find and develop new antibiotics. Now, researchers say artificial intelligence is helping speed their search. ht
Figuring out how to benchmark agents on realistic biology research has quickly become one of my favorite types of engineering work. You work with scientists to get to the core of some biological claim, precisely assembling raw data/prior literature/experimental context in a https
What "agentic" actually means, stripped of marketing: The model plans, writes code, runs it, reads the error, fixes it, iterates — autonomously, dozens of times, until the goal is met. You specify in English. It builds. That's the loop. This is the capability @bcherny and team
Created 2026-05-28 · Updated 2026-05-30
Edit rule text
[anecdotal_single_case_no_systems_analysis]
Learned 4 rejections Active
Exclude posts that describe one patient case, one surgeon's experience, or a single clinical observation without connecting it to healthcare system dynamics, operational patterns, or generalizable insights. Single anecdotes without systems context are too narrow.
Posts reporting isolated clinical anecdotes, single case studies, or personal stories without systemic healthcare context or broader implications.
3 example posts
medicine is constrained far more by operational architecture than by fundamental science I just came head to head with this actual reality and my world has changed no. I cannot bandage away these inefficiencies with a better and better drug, for all cases. if you hate AI
A Mom in Tulsa called 3 health systems last week asking the price of her son's tonsillectomy. Health system A: "We cannot quote you a price." Health system B: "Pricing depends on your insurance." Health system C: "Our financial counselor will reach out after the procedure." N
A surgeon I know recently operated on a man's spine. Saved his mobility. Saved his career. Saved his marriage, probably. The surgeon's take-home on that case, after the facility fee, anesthesia, device costs, payer discounts, and overhead, was $1,840. The health system collec
Created 2026-05-28 · Updated 2026-05-30
Edit rule text
[anecdotal_single_case_without_systems_analysis]
Learned 4 rejections Active
Exclude posts that narrate a single personal story, patient anecdote, or individual case (e.g., 'A Mom in Tulsa called 3 health systems asking for price', 'A surgeon I know recently operated on a man's spine') without drawing systems-level conclusions or healthcare operational insights.
Posts telling individual anecdotes or single stories about healthcare experiences without connecting to broader systems-level insights or policy implications
3 example posts
A company with 800 employees spends roughly $15,000 per employee per year on health insurance. That is $12 million. Wired into a single line item. Every year. The CFO can tell you the carrier. He cannot tell you the unit cost of a single claim. He cannot tell you what the net
medicine is constrained far more by operational architecture than by fundamental science I just came head to head with this actual reality and my world has changed no. I cannot bandage away these inefficiencies with a better and better drug, for all cases. if you hate AI
A Mom in Tulsa called 3 health systems last week asking the price of her son's tonsillectomy. Health system A: "We cannot quote you a price." Health system B: "Pricing depends on your insurance." Health system C: "Our financial counselor will reach out after the procedure." N
Created 2026-05-28 · Updated 2026-06-02
Edit rule text
[anecdotal_single_case_without_systems_insight]
Learned 4 rejections Active
Exclude posts that present a single anecdote, personal story, or one-off example (e.g., 'A mom in Tulsa called 3 health systems,' 'A surgeon I know recently operated') without connecting to systemic patterns, healthcare operations, policy, or generalizable insights.
Posts sharing individual anecdotes, single case studies, or personal observations without broader healthcare systems analysis
3 example posts
medicine is constrained far more by operational architecture than by fundamental science I just came head to head with this actual reality and my world has changed no. I cannot bandage away these inefficiencies with a better and better drug, for all cases. if you hate AI
A Mom in Tulsa called 3 health systems last week asking the price of her son's tonsillectomy. Health system A: "We cannot quote you a price." Health system B: "Pricing depends on your insurance." Health system C: "Our financial counselor will reach out after the procedure." N
A surgeon I know recently operated on a man's spine. Saved his mobility. Saved his career. Saved his marriage, probably. The surgeon's take-home on that case, after the facility fee, anesthesia, device costs, payer discounts, and overhead, was $1,840. The health system collec
Created 2026-05-27 · Updated 2026-05-31
Edit rule text
[unvalidated_speculative_medical_claims_without_evidence]
Learned 4 rejections Active
Exclude posts that speculate about medical treatments, drug effectiveness, or clinical outcomes without published trial data, peer-reviewed evidence, or formal clinical validation. Enthusiastic founder claims about lab results or preliminary observations do not qualify; posts must reference completed, peer-reviewed studies.
Posts making speculative or unvalidated claims about medical interventions, drug efficacy, or treatment outcomes without clinical trial data or peer-reviewed evidence.
2 example posts
Another major cancer treatment advance with a personalized mRNA vaccine for melanoma, on top of immune therapy, >70% survival! This fully mobilizes the immune system to destroy the tumor. In fact, it could potentially be applied to most cancers that have specific mutations!
Traditional ex vivo CAR-T is fundamentally broken. Between the $500K+ manufacturing cost, the 4-week vein-to-vein delay, and the risk of patient T-cell exhaustion during expansion, the blockbuster autologous market is a logistical nightmare. The industry is pivoting to "In Vivo
Created 2026-05-27 · Updated 2026-06-06
Edit rule text
[anecdotal_single_case_no_systems_context]
Learned 4 rejections Active
Exclude posts that describe a single patient case, one doctor's experience, or individual clinical scenario without connecting to healthcare delivery systems, operational patterns, or systemic insights. Posts must demonstrate patterns affecting healthcare organizations or delivery models, not just memorable individual stories.
Posts sharing individual clinical cases, single observations, or personal anecdotes without healthcare system-level implications or operational insight
3 example posts
A company with 800 employees spends roughly $15,000 per employee per year on health insurance. That is $12 million. Wired into a single line item. Every year. The CFO can tell you the carrier. He cannot tell you the unit cost of a single claim. He cannot tell you what the net
medicine is constrained far more by operational architecture than by fundamental science I just came head to head with this actual reality and my world has changed no. I cannot bandage away these inefficiencies with a better and better drug, for all cases. if you hate AI
A Mom in Tulsa called 3 health systems last week asking the price of her son's tonsillectomy. Health system A: "We cannot quote you a price." Health system B: "Pricing depends on your insurance." Health system C: "Our financial counselor will reach out after the procedure." N
Created 2026-05-27 · Updated 2026-05-30
Edit rule text
[market_pricing_and_competitive_dynamics]
Learned 4 rejections Active
Exclude posts that focus primarily on market pricing, brand premium dynamics, or competitive market positioning of drugs/peptides—unless the post analyzes how pricing structures affect healthcare access, provider economics, or system-level resource allocation.
Posts about drug/GLP-1 pricing, competitive positioning, or market share without healthcare systems analysis
3 example posts
$HIMS up to #4 in the Canadian App Store medical category. That is not a stock chart. That is real consumer demand showing up in public. GLP-1 drama, Novo pressure, pricing debate, and now app-store strength. Bears keep arguing the story is slowing down, but the product demand
For years, grey-market HGH operated in a strange legal grey zone somewhere between “research compound,” “API” and underground enhancement culture. Last month’s FDA guidance suggests regulators may no longer see it that way. The recent decline in HGH quality, and now the growing
Think about this. The $50 per month out-of-pocket cost of GLP-1 drugs for Americans (with insurer paying the balance of $2000 pm) is more than double the entire cost for Indians (less than ₹2000pm for Semaglutide). Overall cost difference? 100 times. Not 100%.
Created 2026-05-25 · Updated 2026-05-28
Edit rule text
[glp1_peptide_market_dynamics]
Learned 4 rejections Active
Exclude posts that focus on GLP-1 and peptide market pricing, competitive positioning, brand premiums, or financial dynamics (e.g., 'Ozempic pricing vs. generic semaglutide', 'peptide market growing'). Posts must analyze operational healthcare outcomes or systemic clinical impact, not market speculation or pricing commentary.
Posts about GLP-1 and peptide market pricing, competition, and financial dynamics without healthcare systems analysis
3 example posts
$HIMS up to #4 in the Canadian App Store medical category. That is not a stock chart. That is real consumer demand showing up in public. GLP-1 drama, Novo pressure, pricing debate, and now app-store strength. Bears keep arguing the story is slowing down, but the product demand
For years, grey-market HGH operated in a strange legal grey zone somewhere between “research compound,” “API” and underground enhancement culture. Last month’s FDA guidance suggests regulators may no longer see it that way. The recent decline in HGH quality, and now the growing
Think about this. The $50 per month out-of-pocket cost of GLP-1 drugs for Americans (with insurer paying the balance of $2000 pm) is more than double the entire cost for Indians (less than ₹2000pm for Semaglutide). Overall cost difference? 100 times. Not 100%.
Created 2026-05-25 · Updated 2026-05-28
Edit rule text
[generalist_macro_or_labor_market_without_healthcare_lens]
Learned 4 rejections Active
Exclude posts that discuss labor displacement, workforce disruption, cost trends, energy economics, or infrastructure dynamics at a macro level without grounding in healthcare-specific business models, payer incentives, or clinical workflow impact.
Posts about broad economic, labor market, or infrastructure trends without healthcare-specific systems analysis
3 example posts
Analogizing AI / Mat-Mul to Energy... Remember between 1900 and 1970 kWh/capita grew 34x ... and cost per kWH droped .... 34x. It might be a speed run, but game is game. https://t.co/uxnlnoaAYx
Gavin Baker (@GavinSBaker) says the disaggregation of inference can extend GPU useful lives from 3-4 years to 10-15. That may single-handedly save private credit and reduce the financing rates for GPUs, which will drive demand and help finance the build-out. "The disaggregation
Voice interaction, generating actions,and next, generating tasks...? The humanoid assistant is closer. For those who are bedridden due to illness or have limited mobility, it will serve as an excellent aid in daily living.
Created 2026-05-25 · Updated 2026-05-26
Edit rule text
[stock_ticker_and_market_speculation]
Learned 4 rejections Active
Exclude posts that cite stock tickers, app store rankings, convertible notes, or share price changes as the primary substance—unless they are used to illustrate a systemic healthcare delivery or business model insight. Rankings and valuations alone are not healthcare analysis.
Posts focused on stock price movements, app store rankings, or market position shifts without analysis of underlying healthcare business models or outcomes.
3 example posts
$HIMS up to #4 in the Canadian App Store medical category. That is not a stock chart. That is real consumer demand showing up in public. GLP-1 drama, Novo pressure, pricing debate, and now app-store strength. Bears keep arguing the story is slowing down, but the product demand
Hims & Hers $HIMS Offered $300M Convertable Senior Notes to 2032 💊 Buyers will have the option to add an additional $45M They plan to use the cash for: Expansions, AI, acquisitions, and infrastructure! https://t.co/zaZHO80bY2
$HIMS launched generic semaglutide access in Canada marking its first international GLP-1 expansion. Plans start at $149 CAD per month and include treatment options plus nutrition, movement, sleep and ongoing care team support. https://t.co/yWAy9QMGH6
Created 2026-05-25 · Updated 2026-05-25
Edit rule text
[tangential_infrastructure_or_macro_without_healthcare_lens]
Learned 4 rejections Active
Exclude posts that discuss energy infrastructure, compute hardware, space-based AI, or macro policy (government regulation, manufacturing, labor markets) without direct healthcare application or systems analysis. Posts must show how the infrastructure or policy specifically impacts healthcare operations or economics.
Posts about infrastructure, energy, compute, or macro policy issues that lack clear healthcare-specific application or relevance.
3 example posts
Elon Musk reveals why he believes the cheapest place to put AI will be space within 36 months "The availability of energy is the issue. Everywhere outside of China, electrical output is more or less flat. The output of chips is growing exponentially, but the output of https://t.
Nous Research is working with NVIDIA to make Hermes Agent run smoothly on the new NVIDIA RTX Spark superchip. Hermes Agent is also integrating with the new OpenShell runtime, which connects Hermes to Microsoft’s security primitives https://t.co/cvvPc8bIKa
a $2,000 graphics card and a free download are turning into $2,000-a-month retainers, and the customers are the ones who legally can't use ChatGPT the hardware is an RTX 5090, 32GB of VRAM, enough to run a 30B model locally with nothing ever leaving the building. the software is
Created 2026-05-25 · Updated 2026-06-06
Edit rule text
[political_outrage_or_fraud_scandal_without_systems]
Learned 4 rejections Active
Exclude posts that report criminal prosecutions, fraud allegations, or policy violations (e.g., 340B disputes, hospice fraud) where the post is primarily outrage/reporting without connecting to broader healthcare system incentives or operational failures.
Posts reporting healthcare fraud, policy violations, or regulatory scandals without systemic healthcare analysis
3 example posts
Why is healthcare expensive? The grifters gonna grift. ASCO is asking for oncology practices to be admitted into the same 340B machinery that helped health systems acquire physician practices for two decades. Read the tell. The proposed Indigent Care Ratio lets community
If you want to inject $1B into healthcare for the truly vulnerable, it costs the state $500M. If you want to inject $1B for able-bodied adults, it costs the state $100M. The 90% federal match drastically lowers the "price" of expanding government healthcare. Incentives dictate
In what is considered the largest #Botox fraud scheme in the United States, a jury in Los Angeles convicted a California doctor on Tuesday in a $45 million scheme to defraud #Medicare by submitting claims for Botox injections that were never provided and medically unnecessary, ht
Created 2026-05-25 · Updated 2026-05-30
Edit rule text
[broad_unvalidated_health_claims_without_nuance_or_evidence]
Learned 4 rejections Active
Exclude posts that make sweeping claims about clinical standards (e.g., 'LDL targets keep dropping', 'medical error is the third leading cause of death') without citing peer-reviewed evidence, effect sizes, or acknowledging clinical debate and guideline evolution.
Posts making broad health claims about cholesterol targets, disease progression, or clinical standards shifting over time without rigorous evidence
2 example posts
The healthy LDL number has been quietly moving its own goalposts for forty years: - 1988: under 160 - 1993: under 130 - 2001: under 100 - 2004: under 70 for the high-risk - 2019: under 55 for the very-high-risk - Current trajectory: as low as possible, indefinitely The science
A 2016 BMJ study by researchers at Johns Hopkins estimated that medical error is the third leading cause of death in the United States. Behind heart disease. Behind cancer. Ahead of stroke. Ahead of respiratory disease. Ahead of accidents. Ahead of diabetes. Ahead of
Created 2026-05-25 · Updated 2026-05-27
Edit rule text
[healthcare_fraud_or_regulatory_scandal_reporting_without_systems_analysis]
Learned 4 rejections Active
Exclude posts that treat healthcare fraud, enforcement actions, or regulatory scandals as crime/justice narratives rather than healthcare systems issues. Posts must analyze the underlying policy, incentive, or structural problem that enabled the wrongdoing — not just report the violation itself. Sensationalized crime reporting (e.g., 'jury convicted doctor') without healthcare systems context is out of scope.
Posts that report fraud, enforcement actions, or regulatory violations in healthcare without examining systemic causes, structural incentives, or healthcare policy implications.
2 example posts
In what is considered the largest #Botox fraud scheme in the United States, a jury in Los Angeles convicted a California doctor on Tuesday in a $45 million scheme to defraud #Medicare by submitting claims for Botox injections that were never provided and medically unnecessary, ht
Remember when UnitedHealth's CEO was assassinated in 2024? 18 months later, every accusation the killer Luigi Mangione made has been proven RIGHT. UnitedHealth is one of the biggest corporate collapses in recent years. And they deserve it. In December 2024, UnitedHealthcare CE
Created 2026-05-25 · Updated 2026-05-26
Edit rule text
[tangential_cybersecurity_vulnerability_no_healthcare_lens]
Learned 4 rejections Active
Exclude posts that report on AI model security incidents, vulnerability disclosures, or hacking demonstrations (e.g., 'Claude Mythos found 10,000 vulnerabilities') unless the post explicitly connects the vulnerability type to healthcare data protection, clinical system security, or regulatory compliance risk.
Posts about AI model security vulnerabilities, zero-day bugs, or hacking capabilities without healthcare-specific security or clinical workflow implications.
3 example posts
Anthropic built a model so good at hacking they refused to release it publicly. INSTEAD they handed it to Apple, Google, Microsoft, AWS, Cloudflare one month later: 10,000+ zero days found in the software that runs the internet...bugs hidden for 27 years in OpenBSD. exploit htt
🚨 Anthropic just dropped the first Project Glasswing update Claude Mythos found 10,000+ critical vulnerabilities in ONE month: > Cloudflare: 2,000 bugs, 400 high/critical severity > Mozilla: 271 vulnerabilities in Firefox 150 — 10x more vulnerabilities found in Firefox 1
#AI agents need more than a sandbox—they need predictable execution. Learn how OpenShell provides a standard interface to intercept, inspect, and approve LLM-generated bash commands before they run: https://t.co/JKmLMgl1A9 https://t.co/xWi8oBNZ7T
Created 2026-05-24 · Updated 2026-05-26
Edit rule text
[ai_infrastructure_compute_hype_loose_healthcare]
Learned 4 rejections Active
Exclude posts that discuss AI infrastructure hype (token costs, context window pricing, compute economics, inference optimization) with only tangential or aspirational healthcare framing and no evidence of healthcare system deployment, clinical validation, or healthcare-specific cost-benefit analysis.
Posts about AI infrastructure, token costs, compute constraints, or inference economics framed loosely as healthcare-relevant without specific healthcare application or validation.
3 example posts
Token costs are becoming one of the hottest topics for any enterprise I talk with right now. It’s very bullish for AI in general because it means these systems are being used at a scale that wasn’t contemplated before. It also gives way to another form of differentiation that
"Are there even any feature moats left in B2B? They are at best, short-lived. In fact, we were able to rebuild SaaStr's AI VP Marketing in just ... an hour on @Lovable. Moats that still work: hardware, network effects, data, security and compliance. with @ElenaVerna Head of
Enterprise software was priced per seat. That model is breaking. Before: SaaS = seats. Predictable. Human-driven usage. Revenue tied to headcount. Now: agents hit software systems more frequently than any human ever could. Aaron Levie (@levie) on The MAD Podcast: agent-driven h
Created 2026-05-24 · Updated 2026-06-09
Edit rule text
[enterprise_ai_adoption_macro_without_healthcare_systems_specificity]
Learned 4 rejections Active
Exclude posts that celebrate or analyze enterprise AI adoption, agent deployment, or code-generation tools (Claude Code, o1, etc.) based primarily on adoption rates, engineer productivity gains, or workflow automation metrics. The post must connect to healthcare delivery, clinical operations, or patient outcomes—not general enterprise IT modernization.
Posts about AI/agent adoption in enterprises focused on productivity metrics or workflow redesign without healthcare-specific operational context or systems analysis.
3 example posts
kirkland spending $500m on internal ai is the tell. $6.5b revenue, they could buy any legal ai vendor outright and not feel it, and they're building instead. the moat was always the proprietary corpus. decades of deal docs, redlines, partner judgment encoded as training data, an
wow K&E, largest law firm in the world ($10B billings). They’re committing $500M to build proprietary AI software, with 180 tech pros and 100 partners involved. I’m skeptical that “build” is right here, but wonder if Anthropic/OAI deployment cos are involved.
Turn FDE from a noun into a verb with Apollo. FDE is becoming the default operating model for deploying AI in the enterprise. The hard truth: AI does not become valuable in a demo. It becomes valuable when it is embedded into the real workflows, data, constraints, and
Created 2026-05-24 · Updated 2026-05-31
Edit rule text
[tangential_ai_infrastructure_or_capability_hype_without_healthcare_specificity]
Learned 4 rejections Active
Exclude posts that discuss AI model technical capabilities (e.g., code generation, vulnerability detection, agentic systems), infrastructure breakthroughs, or AI safety incidents where the healthcare framing is tangential or the post is primarily about the AI technology itself rather than a healthcare use case or problem.
Posts about AI model capabilities, cybersecurity vulnerabilities, or infrastructure breakthroughs framed loosely around healthcare without concrete healthcare application.
3 example posts
The more interesting part is that Microsoft’s own engineers liked Claude code best…and they’re cutting it anyway Token pricing is making enterprise actually look at what these models cost to run and that could be a problem for the labs when the subsidy ends. Lots of cheaper
Claude Code team just dropped a workshop on how to ship a production-ready agent from scratch. 27-minutes. Free. Live coding by Claude dev. Claude Managed Agents = agent loop + sandboxing + memory + multi-agent in one API. Worth more than any $500 vibe-coding course. https://t
#AI agents need more than a sandbox—they need predictable execution. Learn how OpenShell provides a standard interface to intercept, inspect, and approve LLM-generated bash commands before they run: https://t.co/JKmLMgl1A9 https://t.co/xWi8oBNZ7T
Created 2026-05-24 · Updated 2026-05-24
Edit rule text
[ai_model_technical_capability_tangent_loose_healthcare]
Learned 4 rejections Active
Exclude posts that discuss AI model technical capabilities (model size, parameter efficiency, reasoning ability, code generation) or benchmark performance improvements (o1 vs physician performance, Claude versions, reasoning models) when the healthcare framing is superficial or the post lacks evidence of actual healthcare system deployment or clinical validation. Technical capability announcements without healthcare specificity should be rejected.
Posts about AI model technical capabilities, performance benchmarks, or architectural features framed loosely as healthcare-relevant without substantive health system application.
3 example posts
this pragmatic trial is more an investigation of physician behavior than the performance of MRSA nares PCR MRSA nares PCR's performance for excluding MRSA pneumonia was excellent from an EBM standpoint we should keep using the MRSA nares PCR and replace the physicians 🤷‍♂️
People are increasingly worried that AI tools make us overreliant. But how do we actually measure this? We introduce Offloading Score, a measure of reliance based on the fraction of cognitive effort offloaded to AI while completing a task. In a controlled user study, Offloading
the ability of ECG AI to predict LVEF is a big deal previously, ECG AI has been shown to detect MI similarly to true ECG experts (& better than most of us) AI is now doing something unique that humans essentially can't do this will soon be the the standard of care ...#1/2
Created 2026-05-23 · Updated 2026-06-06
Edit rule text
[enterprise_ai_adoption_or_labor_macro_without_healthcare_specificity]
Learned 4 rejections Active
Exclude posts that discuss enterprise AI adoption rates, engineer productivity gains, workforce disruption projections, or labor market macro trends (e.g., '75% of Google code is AI-generated', 'workers say AI makes them productive') unless the post explicitly quantifies or analyzes impact within healthcare organizations, clinical workflows, or healthcare labor markets. General tech labor trends with loose healthcare framing are not sufficient.
Posts about enterprise AI adoption trends, labor productivity impacts, or broad workforce disruption without healthcare-specific context or evidence.
3 example posts
The framing here is exactly backwards. This is the strongest AI bull signal anyone has published this year. Uber deployed Claude Code to 5,000 engineers in December. By March, 84% were classified as agentic coding users. By April, 95% used AI tools monthly. 70% of all committed
Workers say AI is making them more productive. Executives say AI is not making their companies more productive. Both groups are reporting on the same software. The gap between the two numbers is the entire debate about the largest capital expenditure cycle in technology history.
Fair warning, YC founders: if you take these tokens, there’s a non-zero chance that OpenAI will study exactly what your startup is doing, copy your idea and put your app into their free offering. This is the classic platform playbook — be careful, founders!
Created 2026-05-23 · Updated 2026-05-26
Edit rule text
[truncated_incomplete_low_effort_posts]
Learned 4 rejections Active
Exclude posts that are truncated, incomplete, cut off mid-thought, or fragmented social media snippets. These posts do not provide enough information for meaningful healthcare tech analysis and likely represent queue errors or low-effort content.
Posts that are cut off mid-sentence, fragmented, incomplete, or lack sufficient context to assess healthcare substance.
3 example posts
This week brought even more exciting news in the field of HIV treatment! A once-weekly Lenacapavir-based pill has now demonstrated similar outcomes to daily antiretrovirals (ARVs), achieving non-inferiority. This development could significantly impact treatment adherence and http
Affinity maturation is how naive antibodies evolve into strong binders, but most antibody LMs ignore it. @stephenzlu and I built CoSiNE to learn this, beating antibody LMs on VEP and reframing design as guiding evolution, not de novo generation. Excited to present at ICML!
🔔Survodutide SYNCHRONIZE-1 full paper is out for Boehringer-Ingelheim’s GLP-1/Glucagon study for weight loss. ▪️13% loss on 6mg (treatment regimen estimand) ▪️Huge treatment discontinuations due to AEs: 20% ▪️Nausea 65%, Vomiting 45% ▪️Placebo was killer. 5.4% WL. Wow. ▪️60% htt
Created 2026-05-23 · Updated 2026-06-10
Edit rule text
[tangential_ai_infrastructure_hype_without_healthcare_specificity]
Learned 4 rejections Active
Exclude posts that discuss AI infrastructure economics (e.g., 'tokens per dollar per watt,' GPU lifecycles, inference disaggregation, compute cost curves) unless they directly explain how this shifts healthcare delivery, operational costs, or care access.
Posts about AI compute, token economics, GPU utilization, or infrastructure trends framed loosely as healthcare-relevant without demonstrating actual healthcare application.
3 example posts
Analogizing AI / Mat-Mul to Energy... Remember between 1900 and 1970 kWh/capita grew 34x ... and cost per kWH droped .... 34x. It might be a speed run, but game is game. https://t.co/uxnlnoaAYx
Gavin Baker (@GavinSBaker) says the disaggregation of inference can extend GPU useful lives from 3-4 years to 10-15. That may single-handedly save private credit and reduce the financing rates for GPUs, which will drive demand and help finance the build-out. "The disaggregation
The more interesting part is that Microsoft’s own engineers liked Claude code best…and they’re cutting it anyway Token pricing is making enterprise actually look at what these models cost to run and that could be a problem for the labs when the subsidy ends. Lots of cheaper
Created 2026-05-23 · Updated 2026-05-27
Edit rule text
[glp1_peptide_market_pricing_and_competitive_dynamics_without_systems_analysis]
Learned 4 rejections Active
Exclude posts that report GLP-1 or peptide pricing, market share, app store rankings, or competitive dynamics (e.g., Hims vs. Ro app rankings, generic pricing, Canadian expansion) without analyzing healthcare system impact, access barriers, or clinical outcomes.
Posts tracking GLP-1 and peptide market pricing, app store rankings, or competitive positioning without healthcare systems analysis.
3 example posts
$HIMS up to #4 in the Canadian App Store medical category. That is not a stock chart. That is real consumer demand showing up in public. GLP-1 drama, Novo pressure, pricing debate, and now app-store strength. Bears keep arguing the story is slowing down, but the product demand
🚨 ARE PEPTIDES THE NEXT "GLP-1 MOMENT" FOR $HIMS? @eli_dorf: "I think there's something really interesting that happened with GLP-1s, which is that when someone's taking Ozempic or any GLP-1, they LOVE to talk about it." "They'll talk about it to their friends. They'll talk htt
Hims & Hers $HIMS Offered $300M Convertable Senior Notes to 2032 💊 Buyers will have the option to add an additional $45M They plan to use the cash for: Expansions, AI, acquisitions, and infrastructure! https://t.co/zaZHO80bY2
Created 2026-05-23 · Updated 2026-05-24
Edit rule text
[generalist_ai_infrastructure_without_healthcare_application]
Learned 4 rejections Active
Exclude posts discussing AI infrastructure, model capabilities, compute costs, or deployment patterns (Kubernetes, agents, inference, vector databases) that do not demonstrate direct, specific application to healthcare delivery, clinical decisions, or health systems operations.
Posts about AI infrastructure, compute, or technical capabilities that lack healthcare-specific application or context
3 example posts
kirkland spending $500m on internal ai is the tell. $6.5b revenue, they could buy any legal ai vendor outright and not feel it, and they're building instead. the moat was always the proprietary corpus. decades of deal docs, redlines, partner judgment encoded as training data, an
The app era is ending. Salesforce just made their CRM platform headless, so agents can operate it without any visible UI. Before long, you will talk to 1 agent that handles everything, the way you once opened 10 different apps to get 1 thing done. For years, VeChain has been b
wow K&E, largest law firm in the world ($10B billings). They’re committing $500M to build proprietary AI software, with 180 tech pros and 100 partners involved. I’m skeptical that “build” is right here, but wonder if Anthropic/OAI deployment cos are involved.
Created 2026-05-22 · Updated 2026-05-31
Edit rule text
[unvalidated_market_or_founder_enthusiasm]
Learned 4 rejections Active
Exclude posts that amount to founder enthusiasm, optimistic framing, or market hype without concrete data, validation, or healthcare systems insight. Examples: 'This is interesting,' 'game is game,' 'my world has changed' without backing claims.
Posts expressing optimism about market opportunities, company performance, or technology potential without substantive evidence or systems analysis
3 example posts
medicine is constrained far more by operational architecture than by fundamental science I just came head to head with this actual reality and my world has changed no. I cannot bandage away these inefficiencies with a better and better drug, for all cases. if you hate AI
What "agentic" actually means, stripped of marketing: The model plans, writes code, runs it, reads the error, fixes it, iterates — autonomously, dozens of times, until the goal is met. You specify in English. It builds. That's the loop. This is the capability @bcherny and team
Analogizing AI / Mat-Mul to Energy... Remember between 1900 and 1970 kWh/capita grew 34x ... and cost per kWH droped .... 34x. It might be a speed run, but game is game. https://t.co/uxnlnoaAYx
Created 2026-05-22 · Updated 2026-05-28
Edit rule text
[tangential_enterprise_ai_adoption_macro]
Learned 4 rejections Active
Exclude posts that discuss generalist AI adoption trends in enterprise (e.g., 'AI code generation at Google', 'AI engineers no longer prompting') where healthcare is mentioned tangentially or speculatively without validating healthcare-specific application, constraints, or outcomes.
Posts about enterprise AI adoption, code generation, or workforce productivity trends with loose or speculative healthcare framing
3 example posts
Andrej Karpathy just explained the future of software engineering without directly saying it. The best AI engineers are no longer “prompting.” They’re building systems around the agents. Karpathy’s biggest insight wasn’t: “Claude can code.” It was: LLMs become dramatically be
Workers say AI is making them more productive. Executives say AI is not making their companies more productive. Both groups are reporting on the same software. The gap between the two numbers is the entire debate about the largest capital expenditure cycle in technology history.
The reason why I release my X articles about AI agents (fat skill fat code thin harness) and GStack and GBrain is that we, yes you and I, can have *PROCESS POWER*, which is the one super powerful specific moat that anyone can create for themselves. The agent helps you do it. htt
Created 2026-05-21 · Updated 2026-05-23
Edit rule text
[founder_enthusiasm_or_lab_updates_without_healthcare_validation]
Learned 4 rejections Active
Exclude posts that showcase founder enthusiasm, lab updates, or technical capability announcements from biotech/AI startups (benchmarking agents on biology, agent reliability studies, AI decision support posters) without evidence of real-world healthcare system adoption, clinical validation, or operational deployment metrics.
Posts from biotech/AI founders celebrating technical milestones or lab updates without evidence of healthcare system integration or clinical validation.
3 example posts
I think the AI superapps will soon own 90% of the agentic layer more and more people won't use hermes/openclaw etc because Claude Cowork/Codex will offer 90% of the functionality with 5% of the friction
Figuring out how to benchmark agents on realistic biology research has quickly become one of my favorite types of engineering work. You work with scientists to get to the core of some biological claim, precisely assembling raw data/prior literature/experimental context in a https
AI agents are increasingly deployed as persistent operational systems, but do they remain reliable over time? Unfortunately no, our new work shows agents can quietly fail after deployment, despite passing day-1 evaluation. We call this "agent aging", akin to human aging. https:/
Created 2026-05-21 · Updated 2026-06-03
Edit rule text
[clinical_observation_or_market_claim_without_healthcare_systems_context]
Learned 4 rejections Active
Exclude posts that cite clinical trial results, epidemiological data, diagnostic findings, or pharmaceutical market statistics in isolation without analyzing how these observations change healthcare delivery, provider economics, payer incentives, or patient access. Single studies or data points without systemic implication do not qualify.
Posts reporting isolated clinical findings, research observations, or market data points without connecting them to healthcare system dynamics, reimbursement, access, or operational impact.
3 example posts
Medicaid spending on autism therapy nearly tripled between 2020 and 202. In North Carolina, the number of clinics offering applied behavior analysis grew from 61 in 2019 to 409 in 2026. Anytime fraudsters can steal money from the federal government, they do. https://t.co/fahy
An AI foundation whole-body 3D model that assesses perturbations (such as obesity) across multiple systems (such as immune, neural) at the cell level. This is MouseMapper. Imagine HumanMapper someday @Nature @erturklab https://t.co/jsg3VMzOqg https://t.co/tktuINis2W
Really amazing story of drug repurposing for rare disease using AI found Clonazepam and worked!! (NCATS federated learning and a determined team made this work) 🚀 👏 @MayoClinic @ncats_nih_gov @statnews #STATBreakthrough https://t.co/D4a4IMXSwh
Created 2026-05-21 · Updated 2026-05-24
Edit rule text
[founder_enthusiasm_or_lab_updates_unvalidated]
Learned 4 rejections Active
Exclude posts that are primarily founder enthusiasm, lab announcement, or company milestone updates (e.g., 'we just launched X', 'our new capability is Y') without evidence of clinical validation, real-world healthcare application, or systems-level impact analysis.
Founder announcements, lab updates, or company milestone posts without clinical validation or healthcare systems analysis.
3 example posts
$ABCL Just now @ BoA - @AbCelleraBio CEO Carl Hansen when asked about what the market is overlooking: “I believe we now have best in class or best in world capabilities for some very important specific applications in therapeutic antibodies. So GPCRs and ion channels has been a
🚨! $SANA has just reported - in its Q1 ER, that it plans to expand its pipeline by advancing a new preclinical program - SG227. SG227 is a CD8-targeted fusosome that by delivering a genetic material to make BCMA-directed CAR T cells could act as a potential treatment for Cancer h
$BEAM's coming presentations May 15: ESCAPE June 13: BEAM-101 EHA update 2026 milestones BEAM-101 (riso-cel) BLA BEAM-301 initial data BEAM-302 updated data and pivotal trial initiation BEAM-304 IND filing https://t.co/9Nvq3PKMMl
Created 2026-05-21 · Updated 2026-05-21
Edit rule text
[ai_company_business_metrics_and_partnerships_unvalidated]
Learned 4 rejections Active
Exclude posts announcing AI company funding, partnerships, acquisitions, or business metrics — even if the company has healthcare products. The post must demonstrate how the funding or partnership improves healthcare delivery, access, or outcomes — not just celebrate the deal.
Posts about AI company funding rounds, partnerships, valuations, or business metrics without healthcare application validation
3 example posts
Benedict Evans on Why AI Feels Like the Internet in 1997 Benedict Evans joins Erik Torenberg for a conversation on the state of AI, including how coding agents hit product-market fit, why foundation models should be thought of as infrastructure, the value of vertical products, h
Follow the signal: we are in a compute-constrained world. And that means power doesn’t sit with the models. It sits with the infrastructure. Most frontier labs don’t own the means of producing Intelligence. They rent it. A handful of companies provide compute that everyone htt
Not NVIDIA. Not OpenAI. Eli Lilly. Jordi Visser (@jvisserlabs) says a 150-year-old pharma company in Indianapolis has the best shot at becoming the world's largest company within five years. The case: ~1,000 NVIDIA Blackwell GPUs in a private data center, a co-innovation lab ht
Created 2026-05-20 · Updated 2026-06-11
Edit rule text
[unvalidated_peptide_and_drug_claims]
Learned 4 rejections Active
Exclude posts that promote or discuss unvalidated peptides, DIY semaglutide, compounded GLP-1s, or unapproved drug sources (including labs, underground suppliers, or qat-style testing claims) without FDA approval, clinical trial data, or warning about regulatory risk.
Posts promoting unvalidated peptides, GLP-1s, or unapproved drugs as treatments without clinical evidence or regulatory context.
3 example posts
Daily Walking Could Be the Secret to Keeping Weight Off For many people struggling with obesity, losing weight is only half the battle. The greater challenge often begins after the initial success: preventing the lost weight from returning. 🔴More than 50% of people who lose http
“You can test your peptides to see if they’re real, bro. But I just look for the 99.9% on the result the seller gives me. You can look up the QR code upon the lab’s website, so you know it’s legit. It’s foolproof, bro. What’s mass spec, bro?” https://t.co/xzJLJ3hesT
$NVO $LLY Dr. Reddy’s says it plans to launch a generic version of Ozempic in Canada within days and is also preparing to roll out generic semaglutide tablets in India. The company sees semaglutide as a meaningful future growth driver following patent expiry in India earlier
Created 2026-05-20 · Updated 2026-05-21
Edit rule text
[founder_enthusiasm_lab_update_without_validation]
Learned 4 rejections Active
Exclude posts by biotech founders, AI researchers, or startups announcing 'breakthroughs,' model improvements, or lab milestones without demonstrating clinical validation, patient outcomes, regulatory clearance, or measurable healthcare system adoption. Hype without proof of healthcare applicability is not substantive.
Posts where founders or researchers celebrate technical progress or capability announcements without evidence of real-world healthcare impact or validation
3 example posts
wow K&E, largest law firm in the world ($10B billings). They’re committing $500M to build proprietary AI software, with 180 tech pros and 100 partners involved. I’m skeptical that “build” is right here, but wonder if Anthropic/OAI deployment cos are involved.
We’re taking steps to accelerate defensive progress in biology: - Launching Rosalind Biodefense to help trusted builders develop new biodefense and pandemic preparedness capabilities.
 - Expanding trusted access to GPT-Rosalind for select U.S. government and allied partners
Anthropic built a model so good at hacking they refused to release it publicly. INSTEAD they handed it to Apple, Google, Microsoft, AWS, Cloudflare one month later: 10,000+ zero days found in the software that runs the internet...bugs hidden for 27 years in OpenBSD. exploit htt
Created 2026-05-20 · Updated 2026-06-02
Edit rule text
[generalist_macro_or_business_without_healthcare_systems_lens]
Learned 4 rejections Active
Exclude posts that discuss pricing economics, business model debates, labor markets, or macro trends (e.g., law firm AI spending, corporate cost-cutting, GLP-1 market dynamics, rural hospital consolidation) where healthcare is a domain example but the post lacks operational or systems-level insight into how the trend changes healthcare delivery.
Posts about business models, economics, or macro trends that touch healthcare tangentially without systems-level analysis.
3 example posts
1/8 Spend time in Lp(a) forums and you’ll see a striking pattern: people willing to do almost anything to drive it down. That response is understandable given how strongly Lp(a) has been framed as a cardiovascular risk factor. /2
kirkland spending $500m on internal ai is the tell. $6.5b revenue, they could buy any legal ai vendor outright and not feel it, and they're building instead. the moat was always the proprietary corpus. decades of deal docs, redlines, partner judgment encoded as training data, an
wow K&E, largest law firm in the world ($10B billings). They’re committing $500M to build proprietary AI software, with 180 tech pros and 100 partners involved. I’m skeptical that “build” is right here, but wonder if Anthropic/OAI deployment cos are involved.
Created 2026-05-19 · Updated 2026-06-03
Edit rule text
[biotech_founder_enthusiasm_or_lab_update_without_systems_validation]
Learned 4 rejections Active
Exclude posts that report drug trial data, FDA approvals, clinical announcements, or gene therapy results as isolated data points or founder enthusiasm without demonstrating validated clinical deployment, healthcare system integration, or evidence of real-world operational impact on patient care delivery.
Posts amplifying biotech founder claims about drug efficacy, trial outcomes, or clinical capabilities without evidence of healthcare system adoption or validated clinical impact.
3 example posts
Traditional ex vivo CAR-T is fundamentally broken. Between the $500K+ manufacturing cost, the 4-week vein-to-vein delay, and the risk of patient T-cell exhaustion during expansion, the blockbuster autologous market is a logistical nightmare. The industry is pivoting to "In Vivo
We’re taking steps to accelerate defensive progress in biology: - Launching Rosalind Biodefense to help trusted builders develop new biodefense and pandemic preparedness capabilities.
 - Expanding trusted access to GPT-Rosalind for select U.S. government and allied partners
The FDA accepted our NDA for BBP-418 for LGMD2I/R9 with PDUFA target action date: November 27, 2026. First-ever potential therapy for a disease that has never had one. $BBIO https://t.co/NrwDadc6rj https://t.co/2ahPIBGzPg
Created 2026-05-19 · Updated 2026-05-31
Edit rule text
[generalist_ai_infrastructure_hype_no_healthcare_application]
Learned 4 rejections Active
Exclude posts that discuss AI infrastructure (tokens per dollar, compute scaling, model architecture) or generalist AI company business metrics (fundraising, valuation, partnerships) unless the post explicitly demonstrates application to a specific healthcare problem or system. Posts about 'AI in general' with loose healthcare framing should be rejected.
Posts about AI infrastructure, compute scaling, or AI company technical capabilities that are tangential to healthcare or lack specific healthcare application.
3 example posts
Andrej Karpathy just explained the future of software engineering without directly saying it. The best AI engineers are no longer “prompting.” They’re building systems around the agents. Karpathy’s biggest insight wasn’t: “Claude can code.” It was: LLMs become dramatically be
Most valuable AI startups in Europe 🇩🇪 Aleph Alpha - $20B 🇫🇷 Mistral AI - $14B 🇩🇪 Celonis - $13B 🇩🇪 Helsing - $12B 🇫🇷 Poolside AI - $12B 🇵🇱 ElevenLabs - $11B 🇮🇹 Bending Spoons - $11B 🇬🇧 Wayve - $8.6B 🇸🇪 Lovable - $6.6B 🇸🇪 Legora - $5.5B 🇫🇷 Hugging Face - $4.5B 🇬🇧 Synthesia - $4
Workers say AI is making them more productive. Executives say AI is not making their companies more productive. Both groups are reporting on the same software. The gap between the two numbers is the entire debate about the largest capital expenditure cycle in technology history.
Created 2026-05-19 · Updated 2026-05-22
Edit rule text
[truncated_or_incomplete_social_media_posts]
Learned 4 rejections Active
Exclude posts that end with ellipsis, truncation indicators ('...'), or incomplete thoughts that prevent understanding the full argument or claim. Fragmentary posts lack sufficient context for meaningful healthcare systems analysis.
Posts that are clearly cut off mid-sentence or missing critical context, making substantive evaluation impossible.
3 example posts
$HIMS up to #4 in the Canadian App Store medical category. That is not a stock chart. That is real consumer demand showing up in public. GLP-1 drama, Novo pressure, pricing debate, and now app-store strength. Bears keep arguing the story is slowing down, but the product demand
For years, grey-market HGH operated in a strange legal grey zone somewhere between “research compound,” “API” and underground enhancement culture. Last month’s FDA guidance suggests regulators may no longer see it that way. The recent decline in HGH quality, and now the growing
Medicaid spending on autism therapy nearly tripled between 2020 and 202. In North Carolina, the number of clinics offering applied behavior analysis grew from 61 in 2019 to 409 in 2026. Anytime fraudsters can steal money from the federal government, they do. https://t.co/fahy
Created 2026-05-19 · Updated 2026-05-24
Edit rule text
[unvalidated_peptide_or_drug_claims]
Learned 4 rejections Active
Exclude posts discussing unvalidated peptide compounds, speculative GLP-1 derivatives, black market peptides, or unproven medical interventions lacking clinical evidence or FDA validation. Include posts about people 'testing peptides' or discussing unverified suppliers.
Posts about unvalidated, speculative, or fringe medical claims involving peptides, GLP-1s, or unproven compounds.
2 example posts
“You can test your peptides to see if they’re real, bro. But I just look for the 99.9% on the result the seller gives me. You can look up the QR code upon the lab’s website, so you know it’s legit. It’s foolproof, bro. What’s mass spec, bro?” https://t.co/xzJLJ3hesT
$NVO $LLY Dr. Reddy’s says it plans to launch a generic version of Ozempic in Canada within days and is also preparing to roll out generic semaglutide tablets in India. The company sees semaglutide as a meaningful future growth driver following patent expiry in India earlier
Created 2026-05-18 · Updated 2026-05-21
Edit rule text
[broad_health_claim_or_market_debate_without_evidence]
Learned 4 rejections Active
Exclude posts that make broad assertions about healthcare cost drivers, business model critiques, or health trends (e.g., 'rural hospitals cannot survive,' 'caregivers are invisible') without concrete data, mechanism explanation, or actionable healthcare system insight.
Posts making sweeping health claims or business model debates without substantive evidence or healthcare systems analysis.
3 example posts
“Rural hospitals cannot make it on their own.” Nobody makes it on their own. Businesses need capital, strategy, discipline, operating systems, good finance, clean accounting, payer strategy, physician alignment, and leaders who know what business they are actually in. The
If you want to inject $1B into healthcare for the truly vulnerable, it costs the state $500M. If you want to inject $1B for able-bodied adults, it costs the state $100M. The 90% federal match drastically lowers the "price" of expanding government healthcare. Incentives dictate
Caregivers do some of the most important work in dementia care — almost none of it is visible to the health care system. Dr. Rosemary Laird describes a navigator model where caregivers reported 86% fewer falls and 81% fewer hospitalizations, over a 12-month pilot. https://t.co/sG
Created 2026-05-18 · Updated 2026-05-22
Edit rule text
[shallow_business_model_debate]
Learned 4 rejections Active
Exclude posts that make broad claims about business model weakness, market saturation, or competitive vulnerability (e.g., 'software will become free') without detailed evidence, market data, or healthcare systems context.
Posts that offer sweeping critiques or assertions about business models, defensibility, or market viability without substantive evidence or systems analysis
3 example posts
Anthropic CEO Dario Amodei on SaaS: "Software is going to become cheap, maybe essentially free. The premise that you need to amortize a piece of software you build across millions of users, that may start to be false. But at the same time, there are whole jobs, whole careers
imagine a software startup raising $800m before their first dollar of revenue. Can’t imagine it the last two decades of software has been dominated by a simple theory: ship quickly, get customers early, generate revenue quickly to validate PMF, manage KPIs closely, etc. This bec
Things every AI app startup says today to justify their defensibility: 1. We support multiple models. Our customers do not want to lock in to one vendor. 2. We have a data moat. We post-train open-source models to be much better and cheaper than closed-source. 3. We do deep inte
Created 2026-05-18 · Updated 2026-05-21
Edit rule text
[enterprise_adoption_macro_without_specificity]
Learned 4 rejections Active
Exclude posts about enterprise AI adoption patterns, orchestration layers, workflow redesign, or operational frameworks (e.g., FDE, agentic systems, Forward Deployed Engineers) when they lack specific healthcare system validation, patient impact, or clinical outcomes data.
Posts about AI adoption trends, operational models, or enterprise patterns without concrete healthcare system impact.
3 example posts
any company rolling out AI at scale is running into the same question. what can the agent reach and what can it touch. security is what’s slowing down mass adoption.
Most enterprise AI projects stall because no one has prepared the content, built the workflow, and kept it improving as models change. That role is essential. Box has launched Forward Deployed Engineers to fill it. Model-agnostic, content-first, and built for the way enterprise
Almost everyone is building agent harness systems the wrong way. The default move: pick LangChain or LangGraph or the OpenAI Agents SDK, accept the loop, the tools, the memory, the orchestration, the policy engine, the credential store, the budget tracker, all of it, as one deci
Created 2026-05-18 · Updated 2026-06-05
Edit rule text
[ai_company_product_launch_or_partnership_without_healthcare_validation]
Learned 4 rejections Active
Exclude posts that announce new AI company products, partnerships, tool releases, or business metrics (e.g., 'Claude Code launched to engineers,' 'Google/Blackstone AI partnership') unless the post explicitly ties the tool to measurable healthcare delivery outcomes or clinical validation. Product announcements alone do not qualify.
Posts announcing AI company products, partnerships, or business news without demonstrating healthcare-specific impact or validation
3 example posts
Today's OpenMed Agent session, driven by GPT-5.5 through the Codex SDK with my own ChatGPT subscription: → 10,036 records ingested → 16-step plan complete → 49 tool calls → 5 reviewer-gated PDFs → APPROVE on PA L38319 Public trace on @huggingface ↓ https://t.co/vlq6CoQ4ee
Anthropic built a model so good at hacking they refused to release it publicly. INSTEAD they handed it to Apple, Google, Microsoft, AWS, Cloudflare one month later: 10,000+ zero days found in the software that runs the internet...bugs hidden for 27 years in OpenBSD. exploit htt
🚨 Anthropic just dropped the first Project Glasswing update Claude Mythos found 10,000+ critical vulnerabilities in ONE month: > Cloudflare: 2,000 bugs, 400 high/critical severity > Mozilla: 271 vulnerabilities in Firefox 150 — 10x more vulnerabilities found in Firefox 1
Created 2026-05-18 · Updated 2026-05-29
Edit rule text
[unvalidated_pharmaceutical_or_peptide_claims]
Learned 4 rejections Active
Exclude posts that discuss peptide or pharmaceutical market dynamics, pricing speculation, generic launches, or unvalidated clinical claims without rigorous evidence. Posts celebrating or promoting unvetted compound synthesis or unregulated peptide quality assurance are also excluded.
Posts about GLP-1s, peptides, semaglutide, or other drugs with speculative market claims, pricing discussion, or unvalidated therapeutic claims
3 example posts
$HIMS $HERS 🚨 HERS IS NOW BEATING BOTH HIMS AND RO ON THE iOS APP STORE Today's rankings in the Medical category: - Hers: #9 - Ro: #10 - Hims: #18 Chart below = weekly avg rank https://t.co/vlv80yiFOo
Daily Walking Could Be the Secret to Keeping Weight Off For many people struggling with obesity, losing weight is only half the battle. The greater challenge often begins after the initial success: preventing the lost weight from returning. 🔴More than 50% of people who lose http
How does a brand like this even get listed at Target? Does Target only care about turnover? Might as well start selling snake oil too. By the way, this is the SKU the brand got in trouble with. They now call it "Berberine Patches". It was called GLP-1 Patches before. Check the h
Created 2026-05-18 · Updated 2026-05-20
Edit rule text
[generalist_ai_infrastructure_or_capability_hype]
Learned 4 rejections Active
Exclude posts that discuss AI infrastructure hype (Jensen Huang keynotes, compute efficiency, token costs, model capability benchmarks, Google AI code generation stats) even if mentioned in healthcare context. The post must focus on a validated healthcare application, not AI industry trends.
Posts about general AI infrastructure, compute, model capabilities, or AI industry trends without healthcare specificity.
3 example posts
Google's CEO confirmed 75% of new code at Google is AI-generated. October 2024: 25%. Fall 2025: 50%. April 2026: 75%. Doubled in a year. Then doubled again in six months. Engineers aren't writing code anymore. They're reviewing it. The job title is the same. The job
Anthropic CEO Dario Amodei on SaaS: "Software is going to become cheap, maybe essentially free. The premise that you need to amortize a piece of software you build across millions of users, that may start to be false. But at the same time, there are whole jobs, whole careers
Google CEO Sundar Pichai on current frontier model's ability to break the security of almost all current software. "These models are definitely, like really gonna break pretty much all software out there, maybe already, we don't know." https://t.co/ezjPcSehiG
Created 2026-05-17 · Updated 2026-05-21
Edit rule text
[unvalidated_peptide_and_glp1_claims]
Learned 4 rejections Active
Exclude posts that speculate on, promote, or discuss peptide/GLP-1 efficacy, sourcing, testing, or market dynamics without citing peer-reviewed evidence, FDA data, or established clinical guidance. Includes anecdotal testimonials and unvetted peptide sourcing advice.
Posts making unverified claims about peptides, GLP-1s, or weight loss treatments without clinical evidence or regulatory context.
3 example posts
Daily Walking Could Be the Secret to Keeping Weight Off For many people struggling with obesity, losing weight is only half the battle. The greater challenge often begins after the initial success: preventing the lost weight from returning. 🔴More than 50% of people who lose http
How does a brand like this even get listed at Target? Does Target only care about turnover? Might as well start selling snake oil too. By the way, this is the SKU the brand got in trouble with. They now call it "Berberine Patches". It was called GLP-1 Patches before. Check the h
“You can test your peptides to see if they’re real, bro. But I just look for the 99.9% on the result the seller gives me. You can look up the QR code upon the lab’s website, so you know it’s legit. It’s foolproof, bro. What’s mass spec, bro?” https://t.co/xzJLJ3hesT
Created 2026-05-17 · Updated 2026-05-21
Edit rule text
[generalist_macro_or_infrastructure_commentary_with_loose_healthcare_framing]
Learned 4 rejections Active
Exclude posts that discuss macro economic trends (AI-driven labor displacement, computing efficiency, state tax policy, capital deployment), geopolitical dynamics, or infrastructure hype with healthcare mentioned only as an example domain or loose framing device, rather than the core focus.
Posts about broad economic trends, labor disruption, computing infrastructure, or policy commentary that mention healthcare tangentially without substantive healthcare-specific analysis.
3 example posts
“Rural hospitals cannot make it on their own.” Nobody makes it on their own. Businesses need capital, strategy, discipline, operating systems, good finance, clean accounting, payer strategy, physician alignment, and leaders who know what business they are actually in. The
If you want to inject $1B into healthcare for the truly vulnerable, it costs the state $500M. If you want to inject $1B for able-bodied adults, it costs the state $100M. The 90% federal match drastically lowers the "price" of expanding government healthcare. Incentives dictate
A senator complaining about healthcare costs should have to answer one question first: Why does your state make new healthcare competitors ask permission before opening? If yes, start there. Speeches do not lower prices. Supply does.
Created 2026-05-17 · Updated 2026-05-23
Edit rule text
[political_outrage_regulatory_posturing_without_healthcare_systems_analysis]
Learned 4 rejections Active
Exclude posts using political figures, regulatory decisions, or policy changes primarily for outrage framing (e.g., Trump statements on buying American, FDA commissioner changes, PBM scandal narratives) unless the post provides systems-level analysis of how the policy actually reshapes healthcare delivery, payment incentives, or patient access.
Posts expressing political outrage or regulatory criticism without substantive analysis of healthcare system impacts
3 example posts
Vivek Ramaswamy put out a lot of statements today referencing the fraud in Ohio. Swampy is of course, involved in Ohio Medicaid. Datavant, Vivek's spinoff company from Riovant, handles release of information for many hospitals and providers in Ohio. Ohio Medicaid providers, man
The public is trained to hear “hospital closing” and assume virtue. Poor patients. Bad reimbursement. Heroic administrators. Cruel politicians. Sometimes true. Often incomplete. The harder truth is that the majority of hospitals are fragile because they have no real financial
The most useful thing to understand about health system operating margin is that it is not merely a measurement. It is a composition. When The New York Times reports that “Mass General Brigham’s operating margin fell to 0.8% last quarter,” readers tend to assume the number is h
Created 2026-05-17 · Updated 2026-05-19
Edit rule text
[speculative_robotics_or_sci_fi_labor_without_healthcare_systems_context]
Learned 4 rejections Active
Exclude posts that make speculative or futuristic claims about robotics performing healthcare tasks (e.g., 'robots will perform surgery by end of 2026') based on CEO predictions or prototype demonstrations, without evidence of regulatory pathway, clinical validation, or real-world deployment in healthcare systems.
Posts speculating about future robotics, humanoid robots, or AI performing surgeries without grounding in actual healthcare system constraints or validation.
3 example posts
Jensen Huang just told every tech CEO in America they are playing the wrong game. They are competing on models. He is competing on maps. He looked at the Middle East and did not see a conflict zone. He saw a compute continent waiting to be switched on. Huang: “I believe that h
Flock Safety Cameras just rolled out an update that allows them to use artificial intelligence to detect human voices These cameras were sold to city governments as gunshot detectors, but with one update they turned them all into a mass surveillance system that records voices T
I don’t think I’ve seen this take before but I like it. Musk has been world-leading at compressing money, resources, and time to make “known/hard” things at scale—make an electric car, make batteries, make a cheaper bigger rocket, all of which already existed but worse, at
Created 2026-05-17 · Updated 2026-05-17
Edit rule text
[tangential_ai_safety_vulnerability_without_healthcare_application]
Learned 4 rejections Active
Exclude posts that discuss AI safety incidents, cybersecurity vulnerabilities found in AI models, or AI system misuse as general AI/security news. The post must identify healthcare-specific risk: clinical harm vectors, HIPAA/PHI exposure, diagnostic error propagation, or provider workflow vulnerabilities—not generic AI security concerns.
Posts about AI safety incidents, cybersecurity vulnerabilities, or model misuse without demonstrating healthcare-specific risk or clinical context.
3 example posts
OpenClaw is powerful enough that we needed Hermes Agent watching it. A lot of people are missing this about agents. The more useful they get, the more dangerous it is to let one system plan the work, run the work, review the work, and decide what happens next. It looks clean h
AI agents are increasingly deployed as persistent operational systems, but do they remain reliable over time? Unfortunately no, our new work shows agents can quietly fail after deployment, despite passing day-1 evaluation. We call this "agent aging", akin to human aging. https:/
Anthropic built a model so good at hacking they refused to release it publicly. INSTEAD they handed it to Apple, Google, Microsoft, AWS, Cloudflare one month later: 10,000+ zero days found in the software that runs the internet...bugs hidden for 27 years in OpenBSD. exploit htt
Created 2026-05-16 · Updated 2026-05-30
Edit rule text
[clinical_or_pharma_observation_without_healthcare_systems_context]
Learned 4 rejections Active
Exclude posts that report clinical trial results, pharma trial milestones, or medical observations (e.g., ECG accuracy, gene therapy approval, drug trial data) without connecting them to healthcare delivery system problems, workflow changes, or system-level impact.
Posts about clinical trial data, medical observations, or drug efficacy without healthcare system implications or operational insight.
3 example posts
Makes an important point here about the role of these short summaries they do this because this is how most doctors digests information if designed appropriately, an LLM should be able to read and appraise a paper and provide better peer review than most clinicians and honestly
$ABCL Just now @ BoA - @AbCelleraBio CEO Carl Hansen when asked about what the market is overlooking: “I believe we now have best in class or best in world capabilities for some very important specific applications in therapeutic antibodies. So GPCRs and ion channels has been a
$XBI RBC Makary's Out, So Who's In? We Break Down Potential Candidates and Impact to Biopharma Potential Candidates for #FDA #Commissioner: Kyle Diamantas J.D Dr. Stephen Hahn M.D. Dr. Brett Giroir M.D. Dr. Sara Brenner M.D. Dr. Houman Hemmati M.D., Ph.D. Dr. Richard Pazdur, M.D.
Created 2026-05-16 · Updated 2026-05-18
Edit rule text
[clinical_or_pharma_observation_without_systems_context]
Learned 4 rejections Active
Exclude posts that present single clinical observations, trial results, or pharma data (e.g., 'drug X works for disease Y', 'medical error is the third leading cause of death') without connecting these findings to healthcare delivery systems, incentives, or policy implications.
Posts reporting clinical trial results, drug efficacy data, or medical observations without broader healthcare systems analysis
3 example posts
An AI foundation whole-body 3D model that assesses perturbations (such as obesity) across multiple systems (such as immune, neural) at the cell level. This is MouseMapper. Imagine HumanMapper someday @Nature @erturklab https://t.co/jsg3VMzOqg https://t.co/tktuINis2W
Really amazing story of drug repurposing for rare disease using AI found Clonazepam and worked!! (NCATS federated learning and a determined team made this work) 🚀 👏 @MayoClinic @ncats_nih_gov @statnews #STATBreakthrough https://t.co/D4a4IMXSwh
We have hundreds of PanINs in our pancreas, especially after age 40. They are a pre-cancerous lesion, but it is fortunately rare for one to progress to cancer (PDAC). Now we know a reason for that. The microenvironment. @umichmedicine @UMRogelCancer https://t.co/HR5a3nZBng https
Created 2026-05-16 · Updated 2026-05-22
Edit rule text
[single_anecdote_without_systems_context]
Learned 4 rejections Active
Exclude posts that tell a single clinical story, isolated physician anecdote, or patient case (e.g., 'A surgeon I know operated on a spine', 'A mom in Tulsa called 3 health systems') without analyzing the systemic issue or operational insight that explains WHY this case matters to healthcare delivery.
Posts describing individual clinical cases, single physician stories, or patient anecdotes without connecting to broader healthcare systems patterns or operational lessons.
3 example posts
A Mom in Tulsa called 3 health systems last week asking the price of her son's tonsillectomy. Health system A: "We cannot quote you a price." Health system B: "Pricing depends on your insurance." Health system C: "Our financial counselor will reach out after the procedure." N
A surgeon I know recently operated on a man's spine. Saved his mobility. Saved his career. Saved his marriage, probably. The surgeon's take-home on that case, after the facility fee, anesthesia, device costs, payer discounts, and overhead, was $1,840. The health system collec
How AI helped treat a newborn’s ultra rare disease. ‘It was almost like a light switch.’ An AI tool, Biomedical Data Translator, helped doctors at Mayo Clinic find a treatment for Jorie Kraus https://t.co/3qs1WIDquU
Created 2026-05-16 · Updated 2026-05-29
Edit rule text
[non_healthcare_domain_loose_framing]
Learned 4 rejections Active
Exclude posts primarily about general business models, enterprise software, policy, labor market macro trends, or non-healthcare technology that use loose or tangential healthcare framing (e.g., 'AI will disrupt healthcare workforce') without analyzing actual healthcare system impacts or specific healthcare applications.
Posts about non-healthcare business, policy, or technology that are superficially framed as healthcare-relevant
3 example posts
SOFTWARE IS DEAD: "The software companies frankly got fat & happy. I think what will happen over time is ChatGPT & Claude will end up sitting on top of basically the entire enterprise software stack & almost everything else will end up being a dumb data pipe into thos
🚨 JUST IN: The Attorney General of Florida just SUED to hold OpenAI CEO Sam Altman LIABLE for fueling violence and physical harm AG Uthmeier read off the fact a young person committed SUIC*DE based on a ChatGPT conversation "23-year-old Zane Shamblin repeatedly told ChatGPT he
Token costs are becoming one of the hottest topics for any enterprise I talk with right now. It’s very bullish for AI in general because it means these systems are being used at a scale that wasn’t contemplated before. It also gives way to another form of differentiation that
Created 2026-05-16 · Updated 2026-06-10
Edit rule text
[generalist_ai_infrastructure_hype_no_healthcare]
Learned 4 rejections Active
Exclude posts discussing token-based billing, inference cost models, context window economics, computational requirements, or AI model scaling challenges as pure AI/tech infrastructure discussion. These must explicitly apply to a healthcare operational or clinical problem to qualify.
Posts about token economics, inference costs, model architecture, or compute requirements as generalist AI topics without healthcare application
3 example posts
Artificial intelligence will consume more power than entire nations by 2030, a new report from the United Nations finds. According to a landmark report from the United Nations University, the data centers powering artificial intelligence are projected to consume 945 https://t.
Token costs are becoming one of the hottest topics for any enterprise I talk with right now. It’s very bullish for AI in general because it means these systems are being used at a scale that wasn’t contemplated before. It also gives way to another form of differentiation that
Elon Musk just explained why the SpaceX IPO is an energy story and the energy constraint is why he believes space becomes the only viable path for AI to scale (Save this). The argument he is making is one of the most important and least understood things happening in technology
Created 2026-05-16 · Updated 2026-06-10
Edit rule text
[agentic_ai_technical_tangent_loose_healthcare_framing]
Learned 4 rejections Active
Exclude posts that discuss agentic AI, agent control systems, multi-agent architectures, agent security frameworks, or runtime governance tools where healthcare is mentioned only tangentially or as a generic use case. The post must demonstrate concrete healthcare workflow impact or healthcare-specific agent challenges.
Posts about agentic AI systems, control planes, agent security, and multi-agent architectures that use healthcare as loose framing without substantive healthcare application
3 example posts
Enterprise software was priced per seat. That model is breaking. Before: SaaS = seats. Predictable. Human-driven usage. Revenue tied to headcount. Now: agents hit software systems more frequently than any human ever could. Aaron Levie (@levie) on The MAD Podcast: agent-driven h
Control plane! Control plane! Control plane! You will hear this term of art a lot going forward. Why? Because in this next phase of AI, companies will want something to sit above the models. They will want control over their AI spend. They will want the flexibility to pick
Nous Research is working with NVIDIA to make Hermes Agent run smoothly on the new NVIDIA RTX Spark superchip. Hermes Agent is also integrating with the new OpenShell runtime, which connects Hermes to Microsoft’s security primitives https://t.co/cvvPc8bIKa
Created 2026-05-16 · Updated 2026-06-07
Edit rule text
[generalist_ai_tangent_loose_healthcare]
Learned 4 rejections Active
Exclude posts that discuss AI model capabilities, AI infrastructure, AI company business models, or workforce disruption in predominantly generalist terms, where healthcare is mentioned only as an afterthought or speculative example rather than being the core analytical lens.
Posts about general AI capabilities, infrastructure, or business trends that are loosely framed as healthcare-relevant but lack substantive healthcare systems insight.
3 example posts
⚡️This is the moment AI stops being a productivity tool and becomes a replacement architecture for elite cognitive labor. Citadel is not a random corporation automating low-level admin. It is one of the most competitive intelligence machines in finance. The work Griffin is descr
Things every AI app startup says today to justify their defensibility: 1. We support multiple models. Our customers do not want to lock in to one vendor. 2. We have a data moat. We post-train open-source models to be much better and cheaper than closed-source. 3. We do deep inte
Anthropic customers are building workarounds just to understand who inside their companies is driving AI costs higher. The scramble to monitor “tokenmaxxing” reflects how quickly businesses are adopting expensive AI tools before the economics fully make sense. Read more:
Created 2026-05-16 · Updated 2026-05-17
Edit rule text
[unvalidated_fringe_medical_claims]
Learned 4 rejections Active
Exclude posts that promote peptide testing shortcuts, unvalidated drug sourcing, generic drug claims without regulatory context, or other medical claims lacking peer-reviewed evidence or proper clinical context.
Posts promoting unvalidated, speculative, or fringe medical interventions (peptides, generic drugs, unproven treatments) without clinical evidence
3 example posts
“You can test your peptides to see if they’re real, bro. But I just look for the 99.9% on the result the seller gives me. You can look up the QR code upon the lab’s website, so you know it’s legit. It’s foolproof, bro. What’s mass spec, bro?” https://t.co/xzJLJ3hesT
I'm a board certified OBGYN. I've practiced for 20+ years. Here's what's actually happening with mifepristone, because the coverage of it has been a mess. On May 1, the 5th Circuit Court of Appeals ruled that the FDA overstepped its authority in 2023 when it eliminated the in-pe
Un chico en China montó una agencia web entera con agentes de IA en Claude Code y ahora gestiona él solo 47 clientes al mes cobrando unos 400$ a cada uno. Sin empleados. Sin comerciales. Sin backend complejo. Solo: • Un MacBook • Un iPhone • 1 API key • Un sistema de 7 agentes
Created 2026-05-16 · Updated 2026-05-16
Edit rule text
[broad_health_claims_without_systems_evidence]
Learned 4 rejections Active
Exclude posts making sweeping health claims (e.g., 'medical error is the 3rd leading cause of death', 'healthy LDL goalposts keep moving') or broad policy assertions (e.g., 'healthcare is expensive because of grifters', 'rural hospitals cannot make it on their own') unless supported by specific system redesign evidence, regulatory analysis, or operational data. Generalist commentary should be rejected.
Posts making broad health claims, policy debates, or economic arguments without healthcare systems specificity or evidence
3 example posts
Why is healthcare expensive? The grifters gonna grift. ASCO is asking for oncology practices to be admitted into the same 340B machinery that helped health systems acquire physician practices for two decades. Read the tell. The proposed Indigent Care Ratio lets community
“Rural hospitals cannot make it on their own.” Nobody makes it on their own. Businesses need capital, strategy, discipline, operating systems, good finance, clean accounting, payer strategy, physician alignment, and leaders who know what business they are actually in. The
If you want to inject $1B into healthcare for the truly vulnerable, it costs the state $500M. If you want to inject $1B for able-bodied adults, it costs the state $100M. The 90% federal match drastically lowers the "price" of expanding government healthcare. Incentives dictate
Created 2026-05-15 · Updated 2026-05-24
Edit rule text
[clinical_anecdote_without_systems_analysis]
Learned 4 rejections Active
Exclude posts that describe individual clinical cases, trial results, or single medical observations without analyzing how this finding impacts healthcare operations, reimbursement, workflow, or systemic outcomes. A case study alone is insufficient; it must connect to healthcare systems change.
Single clinical cases or observations presented without healthcare systems context or operational insight
3 example posts
Drug-resistant infections are a major public health threat around the world, responsible for more than a million deaths each year. Scientists are constantly trying to find and develop new antibiotics. Now, researchers say artificial intelligence is helping speed their search. ht
Presented at #EASCongress2026: Vafai and colleagues report reductions in PCSK9 and LDL cholesterol levels and no dose-limiting toxic effects in persons with hypercholesterolemia treated with VERVE-102, a base editor targeting PCSK9. Full study results: https://t.co/iAmoPdipx9 ht
An AI foundation whole-body 3D model that assesses perturbations (such as obesity) across multiple systems (such as immune, neural) at the cell level. This is MouseMapper. Imagine HumanMapper someday @Nature @erturklab https://t.co/jsg3VMzOqg https://t.co/tktuINis2W
Created 2026-05-15 · Updated 2026-05-28
Edit rule text
[pharma_trial_data_market_dynamics_only]
Learned 4 rejections Active
Exclude posts that announce pharmaceutical trial results, drug approvals, or clinical trial data without connecting the finding to healthcare operations, delivery, policy, or systemic implications. Posts must explain WHY the clinical result matters to healthcare systems, not just report the data.
Posts reporting pharmaceutical trial data or drug approvals without analysis of healthcare system implications or clinical practice change
3 example posts
This week brought even more exciting news in the field of HIV treatment! A once-weekly Lenacapavir-based pill has now demonstrated similar outcomes to daily antiretrovirals (ARVs), achieving non-inferiority. This development could significantly impact treatment adherence and http
🚨NEW @NEJM CAR T cells expanding applications! Here CAR T facilitate kidney transplantation in highly sensitized people. dual CD19 + BCMA CAR T enabled kidney transplantation in two highly sensitized patients after reducing anti-donor antibody barriers, with no severe https://t.
🔔Survodutide SYNCHRONIZE-1 full paper is out for Boehringer-Ingelheim’s GLP-1/Glucagon study for weight loss. ▪️13% loss on 6mg (treatment regimen estimand) ▪️Huge treatment discontinuations due to AEs: 20% ▪️Nausea 65%, Vomiting 45% ▪️Placebo was killer. 5.4% WL. Wow. ▪️60% htt
Created 2026-05-15 · Updated 2026-06-10
Edit rule text
[pharma_trial_data_without_systems]
Learned 4 rejections Active
Exclude posts that announce drug trial data, clinical results, pipeline advancement, or FDA approval timelines for specific molecules — unless the post connects the development to healthcare delivery barriers, cost structures, or system-level adoption challenges.
Posts announcing pharmaceutical trial results, clinical milestones, or drug pipeline updates without healthcare system impact analysis
3 example posts
🚨! $SANA has just reported - in its Q1 ER, that it plans to expand its pipeline by advancing a new preclinical program - SG227. SG227 is a CD8-targeted fusosome that by delivering a genetic material to make BCMA-directed CAR T cells could act as a potential treatment for Cancer h
$BEAM's coming presentations May 15: ESCAPE June 13: BEAM-101 EHA update 2026 milestones BEAM-101 (riso-cel) BLA BEAM-301 initial data BEAM-302 updated data and pivotal trial initiation BEAM-304 IND filing https://t.co/9Nvq3PKMMl
For 50 years the National Institute on Drug Abuse has spent hundreds of millions of dollars to find an effective cocaine addiction medication. After 100+ molecules & many hundreds of studies nothing has been FDA approved. It's the Holy Grail of addictions medication. Proud to
Created 2026-05-14 · Updated 2026-05-15
Edit rule text
[biotech_founder_enthusiasm_without_systems_validation]
Learned 4 rejections Active
Exclude posts that amplify founder or researcher enthusiasm about emerging technologies, model releases, or research milestones. The post must include clinical evidence, regulatory outcomes, healthcare delivery data, or provider/patient testimony—not just founder optimism or lab progress.
Posts featuring founder excitement about lab updates, research breakthroughs, or company announcements without evidence of clinical validation or healthcare system impact.
3 example posts
Figuring out how to benchmark agents on realistic biology research has quickly become one of my favorite types of engineering work. You work with scientists to get to the core of some biological claim, precisely assembling raw data/prior literature/experimental context in a https
An AI foundation whole-body 3D model that assesses perturbations (such as obesity) across multiple systems (such as immune, neural) at the cell level. This is MouseMapper. Imagine HumanMapper someday @Nature @erturklab https://t.co/jsg3VMzOqg https://t.co/tktuINis2W
Startups that seek to block AI exam fraud, help companies use AI to make consistent decisions, make financial AI transactions safer, and use AI to organize medical device regulatory information took top honors at @Cornell_Tech’s annual Startup Awards on May 14. Nearly 600 https:
Created 2026-05-14 · Updated 2026-05-30
Edit rule text
[ai_infrastructure_or_compute_hype_without_healthcare_validation]
Learned 4 rejections Active
Exclude posts that discuss GPU hardware, compute infrastructure, model inference costs, AI chip announcements, or AI model technical capabilities (Vera CPUs, Rosalind model architectures, token economics) unless they explicitly validate healthcare use cases or demonstrate healthcare application. Posts must move beyond infrastructure announcements to show healthcare impact.
Posts about AI infrastructure, compute hardware, or model capabilities that lack validation or application in healthcare contexts.
3 example posts
any company rolling out AI at scale is running into the same question. what can the agent reach and what can it touch. security is what’s slowing down mass adoption.
the CEO of NVIDIA just said the computer is no longer being built for you, it is being built for agents Jensen Huang put it plainly, until now we were the users, we were the renters, every CPU on earth was designed around how a person works but agents do not work like us, they
We are partnering with @Microsoft to enable secure, user-controlled AI on Windows. NVIDIA OpenShell runtime for agents will provide governance tools, policy enforcement, and smart local-to-cloud query routing. Learn more: https://t.co/zPGwz9xQSW https://t.co/mkOtFEOhAS
Created 2026-05-14 · Updated 2026-06-03
Edit rule text
[pharma_trial_data_or_market_dynamics_without_healthcare_systems_analysis]
Learned 4 rejections Active
Exclude posts that simply announce drug trial results, FDA approvals, generic drug launches, or market pricing data without discussing how these changes affect healthcare delivery, access, reimbursement, or system-level outcomes.
Posts announcing pharma trial data, drug approvals, or market pricing/competition without analyzing healthcare system implications.
3 example posts
$NVO $LLY Dr. Reddy’s says it plans to launch a generic version of Ozempic in Canada within days and is also preparing to roll out generic semaglutide tablets in India. The company sees semaglutide as a meaningful future growth driver following patent expiry in India earlier
🚨! $SANA has just reported - in its Q1 ER, that it plans to expand its pipeline by advancing a new preclinical program - SG227. SG227 is a CD8-targeted fusosome that by delivering a genetic material to make BCMA-directed CAR T cells could act as a potential treatment for Cancer h
$BEAM's coming presentations May 15: ESCAPE June 13: BEAM-101 EHA update 2026 milestones BEAM-101 (riso-cel) BLA BEAM-301 initial data BEAM-302 updated data and pivotal trial initiation BEAM-304 IND filing https://t.co/9Nvq3PKMMl
Created 2026-05-14 · Updated 2026-05-17
Edit rule text
[pharmaceutical_trial_data_or_drug_announcement_without_healthcare_systems_context]
Learned 4 rejections Active
Exclude posts that announce clinical trial data, FDA approvals, drug candidate results, or biomarker findings (e.g., 'VERVE-10 PCSK9 reductions', 'BBP-418 NDA accepted', 'Clonazepam repurposing for rare disease') without explaining the impact on healthcare system costs, provider workflows, patient access, or reimbursement incentives.
Posts announcing pharmaceutical trial results, FDA approvals, or drug development milestones without analyzing how the drug changes healthcare system economics, access, or clinical workflow.
3 example posts
We’re taking steps to accelerate defensive progress in biology: - Launching Rosalind Biodefense to help trusted builders develop new biodefense and pandemic preparedness capabilities.
 - Expanding trusted access to GPT-Rosalind for select U.S. government and allied partners
Drug-resistant infections are a major public health threat around the world, responsible for more than a million deaths each year. Scientists are constantly trying to find and develop new antibiotics. Now, researchers say artificial intelligence is helping speed their search. ht
The FDA accepted our NDA for BBP-418 for LGMD2I/R9 with PDUFA target action date: November 27, 2026. First-ever potential therapy for a disease that has never had one. $BBIO https://t.co/NrwDadc6rj https://t.co/2ahPIBGzPg
Created 2026-05-14 · Updated 2026-05-31
Edit rule text
[non_english_or_incomprehensible_content]
Learned 4 rejections Active
Exclude posts written primarily in non-English languages (e.g., Spanish) or posts so garbled or unclear in structure that core claims cannot be evaluated for healthcare relevance.
Posts written in non-English languages or in formats too unclear to evaluate substantively.
1 example post
Un chico en China montó una agencia web entera con agentes de IA en Claude Code y ahora gestiona él solo 47 clientes al mes cobrando unos 400$ a cada uno. Sin empleados. Sin comerciales. Sin backend complejo. Solo: • Un MacBook • Un iPhone • 1 API key • Un sistema de 7 agentes
Created 2026-05-14 · Updated 2026-05-14
Edit rule text
[pharmaceutical_or_clinical_trial_data_announcement_without_systems_context]
Learned 4 rejections Active
Exclude posts that report clinical trial results, drug trial data, or pharmaceutical efficacy metrics in isolation. The post must analyze how these results affect healthcare delivery systems, patient access, clinical workflows, or healthcare economics—not just announce trial outcomes or patient numbers.
Posts announcing pharmaceutical trial results, drug efficacy data, or clinical research findings without analyzing healthcare delivery, operational, or systems-level implications.
3 example posts
#ASCO26 | Day 3 $LLY didn't spend billions on Kelonia for an 18-patient #myeloma dataset. They may have paid for a potential manufacturing disruption. KLN-1010 reported: • 18/18 MRD-negative at 1 month • No lymphodepletion • No ex vivo cell manufacturing • Single infusion
Published in @JCO_ASCO, during #ASCO26: Tumor-Agnostic Therapies: Translating Scientific Breakthroughs Into Global Implementation ASCO is where we celebrate the next breakthrough. This review asks why breakthroughs do not equal access globally. 🔗: https://t.co/PE5q27bIs1 htt
A rapid manufacturing pipeline for BCMA-targeting CAR T cells drastically shortens “vein-to-vein” times, and produces cells that demonstrate encouraging safety results in a phase one trial of patients with relapsed #MultipleMyeloma. @ScienceTM https://t.co/3F4u8ECFDS https://t.co
Created 2026-05-13 · Updated 2026-06-05
Edit rule text
[tangential_non_healthcare_infrastructure_or_policy_with_loose_framing]
Learned 4 rejections Active
Exclude posts about AI infrastructure, compute costs, energy efficiency, international valuations, or macro policy (e.g., tax incentives, federal funding, capital deployment) that use healthcare as a thin framing device without demonstrating direct application to healthcare delivery, pricing, or outcomes.
Posts about non-healthcare infrastructure, computing, energy, or macro policy that are loosely framed as healthcare-relevant but lack substantive healthcare connection.
3 example posts
Analogizing AI / Mat-Mul to Energy... Remember between 1900 and 1970 kWh/capita grew 34x ... and cost per kWH droped .... 34x. It might be a speed run, but game is game. https://t.co/uxnlnoaAYx
Gavin Baker (@GavinSBaker) says the disaggregation of inference can extend GPU useful lives from 3-4 years to 10-15. That may single-handedly save private credit and reduce the financing rates for GPUs, which will drive demand and help finance the build-out. "The disaggregation
You should always be sceptical of hyped-up claims made by AI companies - especially when: - They announce new "breakthrough" models they refuse to release for public scrutiny. - They have major financial incentives to pump trillion-dollar pre-IPO valuations, amid fading hype.
Created 2026-05-13 · Updated 2026-05-25
Edit rule text
[non_healthcare_business_or_macro_with_loose_framing]
Learned 4 rejections Active
Exclude posts about non-healthcare businesses, infrastructure projects, or macro policies (SpaceX chip fab, Flock Safety cameras, Google Cloud integration) that mention healthcare tangentially but are fundamentally about non-healthcare domains.
Posts about non-healthcare companies, infrastructure, or macro trends loosely connected to healthcare through tangential angles
3 example posts
"There are multiple levels of controlling the actions, but before that, you need provenance. You have to ask "do I trust the data?" and "do I know what data it was trained on?" - Dr. @leemonbaird Day 2 at @consensus2026: Leemon joined the panel “Agents Meet the Enterprise — Sup
Last week, Google Cloud and AWS natively integrated x402, tackling head on one of its biggest problems: unauthorized API wrappers. The largest cloud platforms now give agents sanctioned ways to use services, pay with stables, and ditch API keys. It’s a new era. Pay (dot) sh’s
Larry Ellison just explained the most important and underreported insight in enterprise AI. Every major AI model, ChatGPT, Claude, Grok was trained on public data but public data has a ceiling. The real unlocking of AI's value happens when models can reason on private data, the
Created 2026-05-13 · Updated 2026-05-14
Edit rule text
[healthcare_fraud_or_regulatory_scandal_without_systemic_analysis]
Learned 4 rejections Active
Exclude posts that report fraud, legal proceedings, or regulatory enforcement (e.g., 'FBI conviction', 'lawsuit filed', 'scandal revealed') as standalone news or moral outrage, without examining the underlying healthcare system incentive, policy, or structural problem that enabled the misconduct.
Posts reporting healthcare fraud, legal cases, or regulatory actions as news or outrage without analyzing systemic drivers, policy implications, or structural solutions.
3 example posts
Why is healthcare expensive? The grifters gonna grift. ASCO is asking for oncology practices to be admitted into the same 340B machinery that helped health systems acquire physician practices for two decades. Read the tell. The proposed Indigent Care Ratio lets community
In what is considered the largest #Botox fraud scheme in the United States, a jury in Los Angeles convicted a California doctor on Tuesday in a $45 million scheme to defraud #Medicare by submitting claims for Botox injections that were never provided and medically unnecessary, ht
Remember when UnitedHealth's CEO was assassinated in 2024? 18 months later, every accusation the killer Luigi Mangione made has been proven RIGHT. UnitedHealth is one of the biggest corporate collapses in recent years. And they deserve it. In December 2024, UnitedHealthcare CE
Created 2026-05-13 · Updated 2026-05-27
Edit rule text
[broad_unvalidated_health_claims_market_speculation]
Learned 4 rejections Active
Exclude posts that make sweeping claims about health outcomes, market expansion, or patient impact (e.g., 'AI will reduce health inequities', 'this drug will solve addiction') without citing clinical evidence, trial data, healthcare system barriers, or grounding in validated research. Market speculation without evidence also belongs here.
Posts making broad health claims or market predictions without clinical evidence, systems analysis, or substantive data
3 example posts
For 50 years the National Institute on Drug Abuse has spent hundreds of millions of dollars to find an effective cocaine addiction medication. After 100+ molecules & many hundreds of studies nothing has been FDA approved. It's the Holy Grail of addictions medication. Proud to
Klarna replaced 700 customer service agents with AI. Resolution time dropped from 11 minutes to 2. Then quality collapsed on complex cases. The CEO admitted cost became too dominant. Customers wanted humans. Klarna is now hiring people again. 55% of firms that replaced
A reminder that AI can be used to reduce health inequities. Technology not being used in the US except for a few health systems, but shown to work well in Kenya @JAMACardio https://t.co/wyQ1k0dHES https://t.co/pP824X2qJG
Created 2026-05-12 · Updated 2026-05-16
Edit rule text
[market_pricing_or_fraud_dynamics_without_healthcare_systems_analysis]
Learned 4 rejections Active
Exclude posts that focus narrowly on market pricing, fraud statistics, or PBM/insurance dynamics without analyzing root causes in healthcare system structure, incentive misalignment, or operational redesign opportunities. Pricing complaints or fraud reporting alone without systems analysis should be filtered.
Posts about pharmaceutical pricing, GLP-1 market dynamics, insurance fraud, or supply chain economics that lack structural healthcare systems insight.
3 example posts
PBM owns the pharmacy. PBM owns the plan. PBM adjudicates the claim. Three commissions on one pill. The employer pays all three and gets a thank-you note for the rebate.
Michigan Medicine generated $8.7 billion in revenue in FY25. It posted a $109 million operating margin in the same year. A nonprofit health system underwriting its largest pavilion in history off margin, not philanthropy. The accounting category and the operating reality do not
Incumbents don't defend broken systems. They defend profitable ones. Healthcare in America isn't a market failure. It's a sustaining innovation problem disguised as a social crisis. Every "reform" strengthens the existing architecture because the people writing the rules are the
Created 2026-05-12 · Updated 2026-05-13
Edit rule text
[speculative_robotics_sci_fi_labor_hype]
Learned 4 rejections Active
Exclude posts that present speculative or futuristic claims about robots performing surgery, unloading dishwashers for extended periods, or other sci-fi labor scenarios without evidence of clinical validation, FDA approval, or realistic healthcare deployment timelines.
Posts speculating about future humanoid robots or sci-fi applications in healthcare or labor without grounding in current clinical validation or regulatory pathways.
2 example posts
A humanoid robot just unloaded a dishwasher for 67 hours straight. By end of 2026, the same robot will perform surgery. That's not a prediction from a sci-fi writer. That's Brett Adcock. CEO of Figure AI. Said it on camera this month. Here's what makes this insane. Two years
I might feel nostalgic watching the last Tesla Model X roll off the line—except it's being replaced by the Optimus humanoid robot line, with plans to scale production to a million units per year. They will soon be in our factories, homes and operating rooms changing our world ht
Created 2026-05-12 · Updated 2026-05-12
Edit rule text
[pharmaceutical_market_pricing_without_systems_analysis]
Learned 4 rejections Active
Exclude posts that focus on GLP-1 or peptide drug pricing, generic launches, brand premium dynamics, or market competition (Ozempic vs. semaglutide, Novo pricing, Hims pricing tiers) without analyzing healthcare system delivery models, payer incentives, health equity, or how these dynamics affect clinical decision-making or patient outcomes.
Posts about GLP-1 peptide pricing, market dynamics, generic availability, and competitive positioning without healthcare system economics or access equity analysis.
3 example posts
$HIMS up to #4 in the Canadian App Store medical category. That is not a stock chart. That is real consumer demand showing up in public. GLP-1 drama, Novo pressure, pricing debate, and now app-store strength. Bears keep arguing the story is slowing down, but the product demand
For years, grey-market HGH operated in a strange legal grey zone somewhere between “research compound,” “API” and underground enhancement culture. Last month’s FDA guidance suggests regulators may no longer see it that way. The recent decline in HGH quality, and now the growing
Think about this. The $50 per month out-of-pocket cost of GLP-1 drugs for Americans (with insurer paying the balance of $2000 pm) is more than double the entire cost for Indians (less than ₹2000pm for Semaglutide). Overall cost difference? 100 times. Not 100%.
Created 2026-05-12 · Updated 2026-05-31
Edit rule text
[enterprise_ai_adoption_macro_without_healthcare]
Learned 4 rejections Active
Exclude posts that discuss enterprise AI adoption, deployment challenges, workflow redesign, or vendor competition (e.g., McKinsey studies on AI EBIT gains, Palantir's operating system) in general business contexts that don't specifically address healthcare operations or systems.
Posts discussing general enterprise AI adoption patterns, workflow redesign, or AI deployment without healthcare-specific application or context
3 example posts
Eight in ten companies have deployed generative AI. Eight in ten report no material impact on earnings. That's an operating-model gap. Not an adoption gap. Three days at SAS Innovate. Three moves. In order. https://t.co/uDgZ3K8Z9u https://t.co/EHsM1F5ms1
The AI bolt-on era, illustrated. McKinsey: 6% of companies see real EBIT gains from AI. They're 3x more likely to redesign workflows around AI than bolt it onto existing tools. I love Anthropic. They just shipped a better bolt-on. https://t.co/tCQcPMywGh
there's a deeper underestimated reason: consumers just do not have very many long chains of tasks which can be automated end to end to unlock huge value. their chains are shallow almost always in enterprise, org processes form very long chains, so the reward for automation is m
Created 2026-05-12 · Updated 2026-05-14
Edit rule text
[broad_health_claim_market_speculation]
Learned 4 rejections Active
Exclude posts that make broad unsubstantiated claims about health markets, peptide adoption, GLP-1 competition, or disease prevalence without data, peer evidence, or healthcare systems analysis. Market speculation and unvalidated trend claims do not qualify.
Posts making sweeping claims about health trends, market size, or competitive dynamics without evidence or systems analysis
3 example posts
$HIMS up to #4 in the Canadian App Store medical category. That is not a stock chart. That is real consumer demand showing up in public. GLP-1 drama, Novo pressure, pricing debate, and now app-store strength. Bears keep arguing the story is slowing down, but the product demand
For years, grey-market HGH operated in a strange legal grey zone somewhere between “research compound,” “API” and underground enhancement culture. Last month’s FDA guidance suggests regulators may no longer see it that way. The recent decline in HGH quality, and now the growing
Medicaid spending on autism therapy nearly tripled between 2020 and 202. In North Carolina, the number of clinics offering applied behavior analysis grew from 61 in 2019 to 409 in 2026. Anytime fraudsters can steal money from the federal government, they do. https://t.co/fahy
Created 2026-05-12 · Updated 2026-05-28
Edit rule text
[incomplete_truncated_or_low_effort_posts]
Learned 4 rejections Active
Exclude posts that are truncated, incomplete, or appear to cut off before conveying a complete thought or argument (indicated by ellipses, missing punctuation, or incoherent phrasing that obscures the intended insight).
Posts that are cut off mid-sentence, lack coherent framing, or contain minimal substantive content
3 example posts
The complexity of an AI driven robot that needs to meet GMP standards & learn complex steps of targeted cell manufacture. A marvel and the first robot of its kind in the world. Can’t wait till we lift off the first patient’s NK cells in LA from NANT LEONARDO. Unstoppable scal
We’re industrializing enterprise AI with @NVIDIA with: 🐚 OpenShell project collaboration for autonomous agents 🔧 Dev roadmap for Day 0 Vera Rubin support 🔐 Hardware-enforced security with confidential containers 📦 Lifecycle management and Model-as-a-Service 💥 Validated, https:
Agents calling tools now generate more revenue on Latch than scientists clicking buttons. From native deployments but also harnesses like Claude Code, Codex, Cursor across Pharma, biotech, academic labs. Product roadmap increasingly concerned with exposing compute intensive
Created 2026-05-12 · Updated 2026-05-13
Edit rule text
[tangential_non_healthcare_business_or_infrastructure_with_loose_healthcare_framing]
Learned 4 rejections Active
Exclude posts about general business strategy, infrastructure spending, workforce disruption, or policy that use healthcare as an example or loose framing device rather than the core focus. The post must be primarily about healthcare operations, outcomes, or payment—not using healthcare as one example among many domains.
Posts about non-healthcare business models, tech infrastructure, or policy that loosely frame healthcare relevance without substantive healthcare systems analysis.
3 example posts
Older people can no longer afford to live in California New research finds that older people here in California face the highest risk of going bankrupt in the entire country The risk score was a 69 out of 100 based on factors including healthcare expenses and the high cost of l
imagine a software startup raising $800m before their first dollar of revenue. Can’t imagine it the last two decades of software has been dominated by a simple theory: ship quickly, get customers early, generate revenue quickly to validate PMF, manage KPIs closely, etc. This bec
Palantir invented the forward deployed engineer role in 2006. embed an engineer inside the customer. ship production code in their environment. own the technical outcome. Alex Karp on the idea: “forward deployed engineers are stolen from french restaurants.” 20 years later htt
Created 2026-05-11 · Updated 2026-05-19
Edit rule text
[enterprise_ai_adoption_without_healthcare_specificity]
Learned 4 rejections Active
Exclude posts that discuss enterprise AI adoption patterns, workflow redesign, deployment challenges, or change management for AI in general business contexts—even if framed with healthcare language. The post must demonstrate concrete healthcare system application or healthcare-specific operational problems, not generic enterprise AI trends.
Posts about general enterprise AI adoption, workflow redesign, or deployment challenges that lack healthcare system context or application.
3 example posts
As advanced agents move from coding to the rest of knowledge work, it takes a real amount of work and know-how to get right. You need to ensure agents have the right context and data to work with, wire up systems to agents in a safe and secure way, ensure that the agents are htt
The more I build agentic systems at Razorpay, the more I understand that - at its core, it is an agentic loop with tool calls, integrations, and retrieval. The hard part is... actually making it run reliably, at scale, under real production load. And this is what makes system de
Eight in ten companies have deployed generative AI. Eight in ten report no material impact on earnings. That's an operating-model gap. Not an adoption gap. Three days at SAS Innovate. Three moves. In order. https://t.co/uDgZ3K8Z9u https://t.co/EHsM1F5ms1
Created 2026-05-11 · Updated 2026-05-11
Edit rule text
[pharmaceutical_trial_data_or_market_pricing_without_healthcare_systems_context]
Learned 4 rejections Active
Exclude posts that announce pharmaceutical trial data, pipeline milestones (e.g., BLA dates, EHA presentations), or market pricing without analyzing how those results affect healthcare delivery, patient access, clinical adoption, or healthcare system economics. Stock tickers, conference schedules, and milestone announcements alone are insufficient.
Posts about drug trial results, pipeline updates, or market pricing dynamics without healthcare delivery or access system implications
3 example posts
$NVO $LLY Dr. Reddy’s says it plans to launch a generic version of Ozempic in Canada within days and is also preparing to roll out generic semaglutide tablets in India. The company sees semaglutide as a meaningful future growth driver following patent expiry in India earlier
🚨! $SANA has just reported - in its Q1 ER, that it plans to expand its pipeline by advancing a new preclinical program - SG227. SG227 is a CD8-targeted fusosome that by delivering a genetic material to make BCMA-directed CAR T cells could act as a potential treatment for Cancer h
$BEAM's coming presentations May 15: ESCAPE June 13: BEAM-101 EHA update 2026 milestones BEAM-101 (riso-cel) BLA BEAM-301 initial data BEAM-302 updated data and pivotal trial initiation BEAM-304 IND filing https://t.co/9Nvq3PKMMl
Created 2026-05-11 · Updated 2026-05-14
Edit rule text
[speculative_robotics_and_sci_fi_labor]
Learned 4 rejections Active
Exclude posts that make speculative claims about robots performing surgery, unloading dishwashers for extended periods, or replacing labor within unrealistic or unvalidated timeframes without clinical evidence, regulatory pathway, or credible timeline from reputable sources.
Posts making speculative claims about humanoid robots performing complex medical tasks or labor disruption without grounded evidence or timeline validation
2 example posts
A humanoid robot just unloaded a dishwasher for 67 hours straight. By end of 2026, the same robot will perform surgery. That's not a prediction from a sci-fi writer. That's Brett Adcock. CEO of Figure AI. Said it on camera this month. Here's what makes this insane. Two years
I don’t think I’ve seen this take before but I like it. Musk has been world-leading at compressing money, resources, and time to make “known/hard” things at scale—make an electric car, make batteries, make a cheaper bigger rocket, all of which already existed but worse, at
Created 2026-05-10 · Updated 2026-05-14
Edit rule text
[business_model_debate_without_evidence]
Learned 4 rejections Active
Exclude posts that make sweeping claims about healthcare business model dysfunction, inherent corruption, or systemic failure — unless the post provides specific, cited data, regulatory evidence, or operational evidence of a particular mechanism and proposes measurable solutions.
Posts making broad critiques of healthcare business models, PBM practices, or system economics without actionable insight or evidence
3 example posts
The most useful thing to understand about health system operating margin is that it is not merely a measurement. It is a composition. When The New York Times reports that “Mass General Brigham’s operating margin fell to 0.8% last quarter,” readers tend to assume the number is h
PBM owns the pharmacy. PBM owns the plan. PBM adjudicates the claim. Three commissions on one pill. The employer pays all three and gets a thank-you note for the rebate.
Incumbents don't defend broken systems. They defend profitable ones. Healthcare in America isn't a market failure. It's a sustaining innovation problem disguised as a social crisis. Every "reform" strengthens the existing architecture because the people writing the rules are the
Created 2026-05-10 · Updated 2026-05-15
Edit rule text
[ai_capability_tangent_loose_healthcare_framing]
Learned 4 rejections Active
Exclude posts that discuss AI model capabilities (code generation, multi-model reasoning, tokenization, distributed systems, inference speed) and mention healthcare tangentially or as a generic use case example, rather than analyzing how the capability solves a real healthcare problem.
Posts about general AI model capabilities, coding, or infrastructure that use healthcare as a loose label rather than genuine application context
3 example posts
Claude Code team just dropped a workshop on how to ship a production-ready agent from scratch. 27-minutes. Free. Live coding by Claude dev. Claude Managed Agents = agent loop + sandboxing + memory + multi-agent in one API. Worth more than any $500 vibe-coding course. https://t
In a meta-analysis of 210 biomedical AI studies that statistically compared models under cross-validation, 97% used invalid statistical tests. Here's our new preprint https://t.co/OG58Vkeu49 led by @tianchuzeng @kkli20111 @ZShaoshi @ten_photos 1/N https://t.co/PEnqXQxcsJ
How AI helped treat a newborn’s ultra rare disease. ‘It was almost like a light switch.’ An AI tool, Biomedical Data Translator, helped doctors at Mayo Clinic find a treatment for Jorie Kraus https://t.co/3qs1WIDquU
Created 2026-05-10 · Updated 2026-05-23
Edit rule text
[unvalidated_speculative_drug_claims]
Learned 4 rejections Active
Exclude posts that make claims about peptide authenticity, efficacy, or sourcing (especially 503B compounding, generic launches, or unverified quality) without clinical evidence, regulatory approval status, or credible verification mechanisms.
Posts about peptides, GLP-1s, or experimental drugs with unverified efficacy or sourcing claims
3 example posts
Daily Walking Could Be the Secret to Keeping Weight Off For many people struggling with obesity, losing weight is only half the battle. The greater challenge often begins after the initial success: preventing the lost weight from returning. 🔴More than 50% of people who lose http
How does a brand like this even get listed at Target? Does Target only care about turnover? Might as well start selling snake oil too. By the way, this is the SKU the brand got in trouble with. They now call it "Berberine Patches". It was called GLP-1 Patches before. Check the h
“You can test your peptides to see if they’re real, bro. But I just look for the 99.9% on the result the seller gives me. You can look up the QR code upon the lab’s website, so you know it’s legit. It’s foolproof, bro. What’s mass spec, bro?” https://t.co/xzJLJ3hesT
Created 2026-05-10 · Updated 2026-05-18
Edit rule text
[shallow_business_model_critique_without_evidence]
Learned 4 rejections Active
Exclude posts that critique or defend healthcare business structures (e.g., 'healthcare isn't a market failure, it's a sustaining innovation problem', 'incumbents defend profitable systems') without specific evidence, financial data, or systemic healthcare analysis.
Posts offering sweeping critiques or defenses of healthcare business models without grounded data or systems analysis.
3 example posts
there's a deeper underestimated reason: consumers just do not have very many long chains of tasks which can be automated end to end to unlock huge value. their chains are shallow almost always in enterprise, org processes form very long chains, so the reward for automation is m
Incumbents don't defend broken systems. They defend profitable ones. Healthcare in America isn't a market failure. It's a sustaining innovation problem disguised as a social crisis. Every "reform" strengthens the existing architecture because the people writing the rules are the
Europe is not losing to the US because it lacks engineers, capital or universities. It is losing because every layer of its system rewards the gatekeeper and punishes the builder. In medicine I see it weekly. A primary care AI that triages, documents and flags risk gets deployed
Created 2026-05-10 · Updated 2026-05-10
Edit rule text
[ai_infrastructure_compute_hype_no_healthcare_specificity]
Learned 4 rejections Active
Exclude posts that celebrate AI infrastructure breakthroughs, compute capacity, chip performance, or energy efficiency (e.g., NVIDIA, Vera CPUs, RTX cards, energy availability) unless the post connects those innovations to a specific healthcare delivery, clinical workflow, or patient outcome improvement.
Posts about AI compute, chips, energy, or infrastructure hype without demonstrating specific healthcare application or system impact.
3 example posts
NEWS: ASML has invited Elon Musk to speak at its internal technology conference, Dutch outlet NUnl reports. Elon Musk's response: "ASML should be treasured and supported. It is arguably the greatest company in Europe." ASML builds the EUV lithography machines that every advance
Elon Musk just explained why the SpaceX IPO is an energy story and the energy constraint is why he believes space becomes the only viable path for AI to scale (Save this). The argument he is making is one of the most important and least understood things happening in technology
Elon is now printing over $26B per year… purely from selling compute to his competitors 😂 AI companies bet everything on one path and got crushed by compute limits. Now they’re begging Elon for GPUs
Created 2026-05-10 · Updated 2026-06-07
Edit rule text
[regulatory_fraud_scandal_without_analysis]
Learned 4 rejections Active
Exclude posts that report regulatory decisions, FDA commissioner changes, fraud allegations, or policy announcements as breaking news without providing substantive analysis of how the decision affects healthcare markets, access, or incentives. Reporting alone is insufficient; must include healthcare systems or business model insight.
Posts reporting healthcare regulatory actions, fraud scandals, or FDA policy shifts without systems-level healthcare analysis
3 example posts
$XBI RBC Makary's Out, So Who's In? We Break Down Potential Candidates and Impact to Biopharma Potential Candidates for #FDA #Commissioner: Kyle Diamantas J.D Dr. Stephen Hahn M.D. Dr. Brett Giroir M.D. Dr. Sara Brenner M.D. Dr. Houman Hemmati M.D., Ph.D. Dr. Richard Pazdur, M.D.
$NVO $LLY Dr. Reddy’s says it plans to launch a generic version of Ozempic in Canada within days and is also preparing to roll out generic semaglutide tablets in India. The company sees semaglutide as a meaningful future growth driver following patent expiry in India earlier
I'm a board certified OBGYN. I've practiced for 20+ years. Here's what's actually happening with mifepristone, because the coverage of it has been a mess. On May 1, the 5th Circuit Court of Appeals ruled that the FDA overstepped its authority in 2023 when it eliminated the in-pe
Created 2026-05-10 · Updated 2026-05-16
Edit rule text
[pharmaceutical_trial_data_only]
Learned 4 rejections Active
Exclude posts that announce clinical trial results, drug efficacy data, or biotech research findings without explaining the healthcare delivery, reimbursement, manufacturing, or systems-level implication of the result.
Posts announcing pharmaceutical or clinical trial data without healthcare systems or implementation context
3 example posts
#ASCO26 | Day 3 $LLY didn't spend billions on Kelonia for an 18-patient #myeloma dataset. They may have paid for a potential manufacturing disruption. KLN-1010 reported: • 18/18 MRD-negative at 1 month • No lymphodepletion • No ex vivo cell manufacturing • Single infusion
Published in @JCO_ASCO, during #ASCO26: Tumor-Agnostic Therapies: Translating Scientific Breakthroughs Into Global Implementation ASCO is where we celebrate the next breakthrough. This review asks why breakthroughs do not equal access globally. 🔗: https://t.co/PE5q27bIs1 htt
A rapid manufacturing pipeline for BCMA-targeting CAR T cells drastically shortens “vein-to-vein” times, and produces cells that demonstrate encouraging safety results in a phase one trial of patients with relapsed #MultipleMyeloma. @ScienceTM https://t.co/3F4u8ECFDS https://t.co
Created 2026-05-10 · Updated 2026-06-07
Edit rule text
[tangential_non_healthcare_policy_or_fraud_scandal]
Learned 4 rejections Active
Exclude posts that cite political figures, regulatory posturing, or fraud allegations (FDA Commissioner changes, Trump policy announcements, corporate governance scandals) without analyzing healthcare delivery, reimbursement, or operational consequences.
Posts reporting political outrage, regulatory scandal, or fraud allegations without substantive healthcare systems analysis.
3 example posts
How to print money, state government edition: 1) Tax your hospitals and managed care plans. 2) Use that tax revenue to pay your 10% share of Medicaid expansion. 3) Trigger a 90% federal match. 4) Return the original tax money, plus the massive federal windfall, back to the health
Vivek Ramaswamy put out a lot of statements today referencing the fraud in Ohio. Swampy is of course, involved in Ohio Medicaid. Datavant, Vivek's spinoff company from Riovant, handles release of information for many hospitals and providers in Ohio. Ohio Medicaid providers, man
Yes to Katy. If I was a manufacturer I would contract directly with the employer. The health plan when self funded has a fiduciary obligation to plan assets… Nice work @KatyTalento
Created 2026-05-09 · Updated 2026-05-20
Edit rule text
[broad_unvalidated_health_claims_or_market_speculation_without_evidence]
Learned 4 rejections Active
Exclude posts that make broad claims about healthcare improvement, AI impact on outcomes, or market disruption (e.g., 'AI will reduce health inequities', 'cost will drop', 'labor will be replaced') without citing peer-reviewed studies, randomized trials, validated datasets, or healthcare delivery evidence. Opinion from founders, observers, or macro commentators without systems analysis is insufficient.
Posts making sweeping claims about healthcare disruption, outcomes, or market shifts without clinical evidence, systems analysis, or validated data
3 example posts
PBM owns the pharmacy. PBM owns the plan. PBM adjudicates the claim. Three commissions on one pill. The employer pays all three and gets a thank-you note for the rebate.
Incumbents don't defend broken systems. They defend profitable ones. Healthcare in America isn't a market failure. It's a sustaining innovation problem disguised as a social crisis. Every "reform" strengthens the existing architecture because the people writing the rules are the
Three weeks ago I called Anthropic rationing inference and OpenAI hitting the wall next. Yesterday OpenAI shipped GPT-5.5 Instant. New default for free, Plus, and Pro. 30% fewer words per answer. One frame going around: cost savings. It's both bigger and smaller than that. http
Created 2026-05-09 · Updated 2026-05-14
Edit rule text
[tangential_non_healthcare_business_or_policy_with_loose_healthcare_framing]
Learned 4 rejections Active
Exclude posts about generalist business trends (SaaS pricing, software automation, employee productivity, agency models), geopolitical or infrastructure competition, or labor market disruption that only loosely invoke healthcare terminology or use healthcare as an example rather than analyzing healthcare-specific challenges.
Posts about non-healthcare business models, labor market macro trends, or policy decisions that use healthcare as a rhetorical frame rather than subject matter.
3 example posts
imagine a software startup raising $800m before their first dollar of revenue. Can’t imagine it the last two decades of software has been dominated by a simple theory: ship quickly, get customers early, generate revenue quickly to validate PMF, manage KPIs closely, etc. This bec
The 36 BIGGEST startup opportunities right now 1. biggest b2c: solving loneliness. third spaces, community apps, IRL 2. biggest b2b: managed AI employees for businesses 3. biggest overlooked: elder tech. 70 million boomers who want products that make them happier & healthier
The UK AISI found Mythos Preview is the first model to solve both their cyber ranges end-to-end. No model had ever solved the AISI’s “Cooling Tower” cyber range before. We're getting it to defenders as fast as we responsibly can. More to come on our Glasswing work soon.
Created 2026-05-09 · Updated 2026-05-20
Edit rule text
[broad_health_claim_or_market_debate_without_systems_analysis]
Learned 4 rejections Active
Exclude posts that make broad health claims (e.g., 'disease X is diagnosed too late,' 'condition Y will disrupt markets') or debate market dynamics without grounding the argument in healthcare system design, reimbursement structures, clinical workflow, or regulatory constraints.
Posts making sweeping claims about health trends, disease prevalence, patient behavior, or market dynamics without healthcare system operational or policy analysis.
3 example posts
Most chronic diseases, and many cancers, are still diagnosed far too late. The future of healthcare lies in detecting risk earlier, diagnosing smarter, and intervening sooner. From AI enabled diagnostics and genomics to Apollo’s DigiSmart Labs, technology is helping transform ht
1/8 Spend time in Lp(a) forums and you’ll see a striking pattern: people willing to do almost anything to drive it down. That response is understandable given how strongly Lp(a) has been framed as a cardiovascular risk factor. /2
medicine is constrained far more by operational architecture than by fundamental science I just came head to head with this actual reality and my world has changed no. I cannot bandage away these inefficiencies with a better and better drug, for all cases. if you hate AI
Created 2026-05-09 · Updated 2026-06-03
Edit rule text
[ai_safety_or_cybersecurity_vulnerability_tangential_to_healthcare]
Learned 4 rejections Active
Exclude posts that discuss AI safety, zero-day vulnerabilities, cybersecurity incidents, data provenance, or agent misuse (e.g., agents registering LLCs, security breaches)—unless they specifically address HIPAA compliance, clinical validation, healthcare data governance, or healthcare-regulated deployment.
Posts about AI safety, cybersecurity vulnerabilities, or data provenance in agents/models without healthcare-specific validation, compliance, or clinical deployment context.
3 example posts
We’re industrializing enterprise AI with @NVIDIA with: 🐚 OpenShell project collaboration for autonomous agents 🔧 Dev roadmap for Day 0 Vera Rubin support 🔐 Hardware-enforced security with confidential containers 📦 Lifecycle management and Model-as-a-Service 💥 Validated, https:
"There are multiple levels of controlling the actions, but before that, you need provenance. You have to ask "do I trust the data?" and "do I know what data it was trained on?" - Dr. @leemonbaird Day 2 at @consensus2026: Leemon joined the panel “Agents Meet the Enterprise — Sup
Flock Safety Cameras just rolled out an update that allows them to use artificial intelligence to detect human voices These cameras were sold to city governments as gunshot detectors, but with one update they turned them all into a mass surveillance system that records voices T
Created 2026-05-09 · Updated 2026-05-13
Edit rule text
[tangential_infrastructure_or_compute_without_healthcare_application]
Learned 4 rejections Active
Exclude posts about AI infrastructure (GPU scaling, token cost reduction, cloud compute, data center deployment, semiconductor manufacturing) unless the post explicitly connects that infrastructure constraint to a specific healthcare delivery or operational bottleneck that needs solving.
Posts about AI infrastructure, semiconductors, data centers, or compute scaling that lack concrete healthcare system analysis or use case.
3 example posts
“Tokens are increasing 50x. You need to drive down cost at the same rate — or more.” At Google Cloud Next, @googlecloud's Mark Lohmeyer spoke with @PatrickMoorhead about the infrastructure realities behind the agent era: faster model training, lower-latency inference, and https:
SpaceX wants to build a $55 billion next-generation, vertically integrated semiconductor manufacturing and advanced computing fabrication facility in Grimes county, Texas, according to a new filing. • Proposed investment: $55B upfront, up to $119B total • Focus: semiconductor ma
I might feel nostalgic watching the last Tesla Model X roll off the line—except it's being replaced by the Optimus humanoid robot line, with plans to scale production to a million units per year. They will soon be in our factories, homes and operating rooms changing our world ht
Created 2026-05-09 · Updated 2026-05-11
Edit rule text
[biotech_founder_enthusiasm_or_lab_update_without_healthcare_systems_validation]
Learned 4 rejections Active
Exclude posts from biotech founders or companies announcing pipeline advances, trial updates, or lab capabilities (e.g., gene therapy milestones, cell therapy expansions, new preclinical programs) unless the post demonstrates how the innovation solves a specific, validated healthcare bottleneck or unmet clinical need with systems-level impact.
Posts celebrating biotech/pharmaceutical breakthroughs or founder announcements without evidence the innovation addresses a validated healthcare system problem
3 example posts
$ABCL Just now @ BoA - @AbCelleraBio CEO Carl Hansen when asked about what the market is overlooking: “I believe we now have best in class or best in world capabilities for some very important specific applications in therapeutic antibodies. So GPCRs and ion channels has been a
🚨! $SANA has just reported - in its Q1 ER, that it plans to expand its pipeline by advancing a new preclinical program - SG227. SG227 is a CD8-targeted fusosome that by delivering a genetic material to make BCMA-directed CAR T cells could act as a potential treatment for Cancer h
$BEAM's coming presentations May 15: ESCAPE June 13: BEAM-101 EHA update 2026 milestones BEAM-101 (riso-cel) BLA BEAM-301 initial data BEAM-302 updated data and pivotal trial initiation BEAM-304 IND filing https://t.co/9Nvq3PKMMl
Created 2026-05-09 · Updated 2026-05-19
Edit rule text
[broad_unvalidated_health_claims_or_market_speculation]
Learned 4 rejections Active
Exclude posts that make sweeping claims about healthcare markets, drug effectiveness, or medical interventions (e.g., 'cocaine addiction has no effective medication after 50 years') without citing peer-reviewed evidence, regulatory data, or healthcare system context.
Posts making broad claims about health market dynamics, drug efficacy, or healthcare trends without clinical evidence or systems analysis
2 example posts
For 50 years the National Institute on Drug Abuse has spent hundreds of millions of dollars to find an effective cocaine addiction medication. After 100+ molecules & many hundreds of studies nothing has been FDA approved. It's the Holy Grail of addictions medication. Proud to
A reminder that AI can be used to reduce health inequities. Technology not being used in the US except for a few health systems, but shown to work well in Kenya @JAMACardio https://t.co/wyQ1k0dHES https://t.co/pP824X2qJG
Created 2026-05-08 · Updated 2026-05-15
Edit rule text
[ai_infrastructure_compute_hype_no_healthcare_lens]
Learned 4 rejections Active
Exclude posts that focus primarily on AI infrastructure economics (token costs, inference pricing, context window pricing, GPU margins, energy constraints) without demonstrating how these metrics directly impact healthcare delivery, clinical decision-making, or healthcare business models.
Posts about AI compute costs, inference pricing, token economics, or hardware requirements without healthcare-specific application or validation.
3 example posts
Token costs are becoming one of the hottest topics for any enterprise I talk with right now. It’s very bullish for AI in general because it means these systems are being used at a scale that wasn’t contemplated before. It also gives way to another form of differentiation that
Elon Musk just explained why the SpaceX IPO is an energy story and the energy constraint is why he believes space becomes the only viable path for AI to scale (Save this). The argument he is making is one of the most important and least understood things happening in technology
"Are there even any feature moats left in B2B? They are at best, short-lived. In fact, we were able to rebuild SaaStr's AI VP Marketing in just ... an hour on @Lovable. Moats that still work: hardware, network effects, data, security and compliance. with @ElenaVerna Head of
Created 2026-05-08 · Updated 2026-06-08
Edit rule text
[ai_company_product_announcement_without_healthcare_validation]
Learned 4 rejections Active
Exclude posts that announce or promote AI agent tools, safety frameworks, or deployment platforms (e.g., Clawvisor, agent orchestration, MCP stacks) unless they include specific evidence of healthcare adoption, clinical trial results, or validated healthcare workflows.
Posts announcing AI agent frameworks, tools, or safety infrastructure without demonstrating healthcare use or clinical validation.
3 example posts
🚀🤖 In 2026, autonomous #AI agents aren’t failing because of technology ➜ they’re failing because of responsibility. ⚙️ Everyone knows how to build agents. Very few know how to own their decisions. 💡 Real autonomous AI isn’t about freedom or intelligence. It’s about clear scope,
Clawvisor is going to be one of the most important parts of helping make the agent world especially OpenClaw/Hermes Agent secure and enterprise-grade. We're in the Apple I moment for personal AI, but we're ABOUT to see the Apple II - the first moment when everyone can use it
Clawvisor (@clawvisor) lets you give AI agents access to apps like Gmail and Slack without handing over your credentials or worrying they'll go rogue. You approve tasks once, Clawvisor enforce them. Congrats on the launch, @ericlevine! https://t.co/S51QlWauds https://t.co/NZvnt
Created 2026-05-08 · Updated 2026-05-10
Edit rule text
[ai_safety_vulnerability_tangent_loose_healthcare]
Learned 4 rejections Active
Exclude posts that discuss AI safety vulnerabilities, zero-day exploits, or security incidents (e.g., Mythos finding cyber vulnerabilities) unless they directly address healthcare-specific threat models like clinical decision system manipulation or patient data security.
Posts about AI model vulnerabilities, security incidents, or safety research tangentially framed as healthcare-relevant
3 example posts
The UK AISI found Mythos Preview is the first model to solve both their cyber ranges end-to-end. No model had ever solved the AISI’s “Cooling Tower” cyber range before. We're getting it to defenders as fast as we responsibly can. More to come on our Glasswing work soon.
"There are multiple levels of controlling the actions, but before that, you need provenance. You have to ask "do I trust the data?" and "do I know what data it was trained on?" - Dr. @leemonbaird Day 2 at @consensus2026: Leemon joined the panel “Agents Meet the Enterprise — Sup
🚀🤖 In 2026, autonomous #AI agents aren’t failing because of technology ➜ they’re failing because of responsibility. ⚙️ Everyone knows how to build agents. Very few know how to own their decisions. 💡 Real autonomous AI isn’t about freedom or intelligence. It’s about clear scope,
Created 2026-05-08 · Updated 2026-05-15
Edit rule text
[generalist_macro_policy_or_labor_market_without_healthcare_specificity]
Learned 4 rejections Active
Exclude posts that discuss general macro trends (tech workforce disruption, American manufacturing policy, federal procurement rules, AI cost models) that only tangentially invoke healthcare without demonstrating specific healthcare systems impact or workforce-specific healthcare implications.
Posts about broad macro trends (AI-driven job displacement, infrastructure spending, supply chain) loosely framed with healthcare context.
3 example posts
Older people can no longer afford to live in California New research finds that older people here in California face the highest risk of going bankrupt in the entire country The risk score was a 69 out of 100 based on factors including healthcare expenses and the high cost of l
The 36 BIGGEST startup opportunities right now 1. biggest b2c: solving loneliness. third spaces, community apps, IRL 2. biggest b2b: managed AI employees for businesses 3. biggest overlooked: elder tech. 70 million boomers who want products that make them happier & healthier
Yes to Katy. If I was a manufacturer I would contract directly with the employer. The health plan when self funded has a fiduciary obligation to plan assets… Nice work @KatyTalento
Created 2026-05-08 · Updated 2026-05-19
Edit rule text
[pharmaceutical_or_clinical_trial_data_without_systems_context]
Learned 4 rejections Active
Exclude posts that announce pharmaceutical trial data, protein discoveries, gene therapy results, or molecular biology findings (e.g., protein complex mapping, mRNA therapy tolerability, CRC screening data) unless the post discusses how the finding will be integrated into healthcare operations, clinical workflows, or addresses a healthcare system bottleneck.
Posts reporting drug trial results, molecular biology findings, or protein research advances without connecting to healthcare delivery system challenges or implementation barriers.
3 example posts
NEW: today OpenBind ‘comes out of stealth’ so to speak with their first data dump of ~900 novel protein-ligand structures - most with paired affinities This represents a meaningful %-age increase in all of humanities P-L data in the PDB collected in the last 50 years More👇 http
$vrtx drops $mrna-partnered Cystic Fibrosis mRNA therapy VX-522 over inhaled LNP tolerability issues: https://t.co/WlIIHXxg9x Not too surprising IMO. Important Q: is safety/tolerability of this approach still good enough for one-time/infrequent genome editing? $prme $arct
not labelled! no stains! just cell chemistry changing in response to a drug !!! you can see it change before it dies !!! you can predict what happens to a cell !!! that's constantly being dosed with a drug !!! you know what? this is just an ICU for living human cells https://t.
Created 2026-05-08 · Updated 2026-05-08
Edit rule text
[ai_agent_security_or_misuse_tangential]
Learned 4 rejections Active
Exclude posts that discuss AI agent security incidents, vulnerability demonstrations, or autonomy risks unless they are grounded in a specific healthcare deployment scenario or address HIPAA/healthcare-specific security implications.
Posts about AI agent security vulnerabilities, misuse scenarios, or safety concerns without healthcare-specific implications.
3 example posts
🚀🤖 In 2026, autonomous #AI agents aren’t failing because of technology ➜ they’re failing because of responsibility. ⚙️ Everyone knows how to build agents. Very few know how to own their decisions. 💡 Real autonomous AI isn’t about freedom or intelligence. It’s about clear scope,
Clawvisor is going to be one of the most important parts of helping make the agent world especially OpenClaw/Hermes Agent secure and enterprise-grade. We're in the Apple I moment for personal AI, but we're ABOUT to see the Apple II - the first moment when everyone can use it
Clawvisor (@clawvisor) lets you give AI agents access to apps like Gmail and Slack without handing over your credentials or worrying they'll go rogue. You approve tasks once, Clawvisor enforce them. Congrats on the launch, @ericlevine! https://t.co/S51QlWauds https://t.co/NZvnt
Created 2026-05-08 · Updated 2026-05-09
Edit rule text
[market_pricing_or_labor_macro_without_healthcare_systems_lens]
Learned 4 rejections Active
Exclude posts that discuss market pricing (pharma competition, GLP-1 costs, hardware prices), labor market disruption, workforce economics, or generalist business model debates with loose healthcare framing—unless the post specifically analyzes how these dynamics reshape healthcare delivery, access, quality, or provider economics. Macro business commentary should be excluded.
Posts about market dynamics, pricing competition, labor disruption, or macro business trends framed loosely around healthcare without analyzing healthcare-specific operational or system implications.
3 example posts
The context window pricing model might be one of the smartest tricks AI labs pulled. You pay for larger context windows, yet model performance often degrades as more context is stuffed in. So developers end up paying more just to work around the fact that models don't have true
Free newsletter: The dawn of token-based billing has shown that generative AI doesn’t have a return on investment. It's too unpredictable, too unreliable, you can't easily measure the cost of tasks, and organizations are already pulling back. https://t.co/wmI82zWdcq
the CEO of NVIDIA just said the computer is no longer being built for you, it is being built for agents Jensen Huang put it plainly, until now we were the users, we were the renters, every CPU on earth was designed around how a person works but agents do not work like us, they
Created 2026-05-08 · Updated 2026-06-04
Edit rule text
[ai_agent_security_or_misuse_incident_tangential]
Learned 4 rejections Active
Exclude posts about AI agent security incidents, unauthorized access, or misuse (e.g., agents registering LLCs, bypassing constraints) unless the post demonstrates concrete healthcare-specific security implications or HIPAA/compliance risks.
Posts about AI agent security vulnerabilities or misuse incidents that are tangential to healthcare operations
3 example posts
Flock Safety Cameras just rolled out an update that allows them to use artificial intelligence to detect human voices These cameras were sold to city governments as gunshot detectors, but with one update they turned them all into a mass surveillance system that records voices T
So Mythos was, indeed, not marketing hype. Remember this is a general purpose model that just happens to be good at finding exploits because good models are good at lots of things. Expect similar from OpenAI & Google. And from open models in 8 months. https://t.co/KbhalQYX8
Clawvisor is going to be one of the most important parts of helping make the agent world especially OpenClaw/Hermes Agent secure and enterprise-grade. We're in the Apple I moment for personal AI, but we're ABOUT to see the Apple II - the first moment when everyone can use it
Created 2026-05-08 · Updated 2026-05-11
Edit rule text
[tangential_infrastructure_or_compute_hype_without_healthcare_application]
Learned 4 rejections Active
Exclude posts that focus on token pricing, inference costs, GPU hardware (RTX, NVIDIA Vera, CPU performance), compute architecture, or billing models for AI — even if loosely framed around healthcare agents. These are general AI infrastructure discussions, not healthcare-specific operational or clinical innovations.
Posts about AI hardware, compute infrastructure, token economics, or pricing models that are tangentially healthcare-framed but primarily discuss non-healthcare AI economics or technical infrastructure.
3 example posts
The context window pricing model might be one of the smartest tricks AI labs pulled. You pay for larger context windows, yet model performance often degrades as more context is stuffed in. So developers end up paying more just to work around the fact that models don't have true
Free newsletter: The dawn of token-based billing has shown that generative AI doesn’t have a return on investment. It's too unpredictable, too unreliable, you can't easily measure the cost of tasks, and organizations are already pulling back. https://t.co/wmI82zWdcq
the CEO of NVIDIA just said the computer is no longer being built for you, it is being built for agents Jensen Huang put it plainly, until now we were the users, we were the renters, every CPU on earth was designed around how a person works but agents do not work like us, they
Created 2026-05-08 · Updated 2026-06-05
Edit rule text
[pharmaceutical_trial_data_announcement_unvalidated]
Learned 4 rejections Active
Exclude posts that announce pharmaceutical trial data, drug approvals, or biotech research findings (especially tumor-agnostic, ADC, or genetic therapies) unless the post analyzes healthcare delivery impact, payer/reimbursement implications, or how the finding changes clinical practice workflows.
Posts announcing pharmaceutical trial results or biotech findings without healthcare system analysis, clinical significance, or operational/reimbursement implications.
3 example posts
Not NVIDIA. Not OpenAI. Eli Lilly. Jordi Visser (@jvisserlabs) says a 150-year-old pharma company in Indianapolis has the best shot at becoming the world's largest company within five years. The case: ~1,000 NVIDIA Blackwell GPUs in a private data center, a co-innovation lab ht
🚨 T-DXd just went tumor-agnostic — and it changed a lot about how we think about HER2 testing across solid tumors 🔬 #ASCO26 FDA granted accelerated approval to T-DXd for patients with unresectable or metastatic HER2-positive solid tumors after prior systemic treatment —
The scores don’t pick your binders. The filters don't care either. The assay does. I worked with a design partner to produce true de-novo hits to 2 immuno-oncology targets. ~1,500 de novo designs → 96 synthesized → 89 run on SPR → 5 confirmed binders; 19–195 nM No https:
Created 2026-05-07 · Updated 2026-06-08
Edit rule text
[healthcare_fraud_scandal_reporting_no_systems_analysis]
Learned 4 rejections Active
Exclude posts that report individual fraud cases, specific provider violations, or anecdotal evidence of waste (e.g., '47 clients/month', specific Medicaid billing patterns) unless the post identifies systemic incentive misalignments or business model failures driving the fraud.
Posts reporting healthcare fraud, waste, or financial scandals without structural analysis
3 example posts
PBM owns the pharmacy. PBM owns the plan. PBM adjudicates the claim. Three commissions on one pill. The employer pays all three and gets a thank-you note for the rebate.
Michigan Medicine generated $8.7 billion in revenue in FY25. It posted a $109 million operating margin in the same year. A nonprofit health system underwriting its largest pavilion in history off margin, not philanthropy. The accounting category and the operating reality do not
"Cordoba-owned buildings in Columbus housed 288 businesses registered with Medicaid... they charged taxpayers more than a quarter of a billion dollars between 2018 and 2024... in a city where only 6,273 people 75 or older are on Medicaid." - @realDailyWire https://t.co/7UnS7agwiw
Created 2026-05-07 · Updated 2026-05-11
Edit rule text
[generalist_macro_or_policy_without_healthcare_lens]
Learned 4 rejections Active
Exclude posts about general economic trends (statin guidelines, banking models, startup job roles, software engineering practices, gatekeeper incentives) that use healthcare as one example among many but lack deep healthcare-specific systems analysis or operational insight.
Posts about broad economic policy, labor market disruption, or business model shifts with only tangential or loose healthcare framing.
3 example posts
Europe is not losing to the US because it lacks engineers, capital or universities. It is losing because every layer of its system rewards the gatekeeper and punishes the builder. In medicine I see it weekly. A primary care AI that triages, documents and flags risk gets deployed
Stripe just created a role that didn't exist 12 months ago (and they're paying multiple six figures for it) It's called the Forward Deployed AI Accelerator. They are hiring AI-native individuals to work directly with their marketing teams to fundamentally change how they work.
Statin guidelines over 50 years: 1970s: "Total cholesterol over 280 is high." Result: a relatively small market. 1988: "Actually, 240 is high." Result: millions more added overnight. 2001: "Actually, focus on LDL. Get it under 100." Result: tens of millions of new customers.
Created 2026-05-07 · Updated 2026-05-08
Edit rule text
[pharma_trial_data_or_market_dynamics_without_systems_analysis]
Learned 4 rejections Active
Exclude posts that announce pharmaceutical trial results, drug safety data, FDA decisions, or GLP-1/peptide market pricing (e.g., 'Vertex drops CF therapy over tolerability issues', 'Klarna's cost reduction led to quality collapse on complex cases') unless the post analyzes systemic healthcare implications, market structure problems, or operational workflows affected by these changes.
Posts reporting pharmaceutical trial data, drug approvals, or market pricing dynamics without healthcare operational or systems insights.
3 example posts
$vrtx drops $mrna-partnered Cystic Fibrosis mRNA therapy VX-522 over inhaled LNP tolerability issues: https://t.co/WlIIHXxg9x Not too surprising IMO. Important Q: is safety/tolerability of this approach still good enough for one-time/infrequent genome editing? $prme $arct
Ivermectin and Mebendazole Cost a Fraction of Chemo. Big Pharma Can't Patent Them. That's the Problem. Cancer centers get a cut of every chemotherapy bill. Generic drugs don't generate that margin. Two affordable, widely available compounds showing 84% clinical benefit in a real
$LLY $NVO $HIMS 🚨 BREAKING: COURT DISMISSES PART OF ELI LILLY LAWSUIT AGAINST EMPOWER PHARMACY BOTH LILLY AND EMPOWER ISSUED STATEMENTS CELEBRATING THE RULING Dismissed: Lanham Act false advertising + consumer harm claim Allowed to proceed: unfair competition claims under h
Created 2026-05-07 · Updated 2026-05-09
Edit rule text
[ai_company_product_metrics_unvalidated]
Learned 4 rejections Active
Exclude posts that announce new AI company products, features, or business initiatives (e.g., Claude Mythos, OpenShell, Forward Deployed AI Accelerator) unless the post explicitly demonstrates how the tool solves a specific healthcare workflow problem or includes validation evidence from healthcare deployment.
Posts announcing AI company product launches, capabilities, or business metrics without demonstrating healthcare application or validation.
3 example posts
New Anthropic Fellows research: Model Spec Midtraining (MSM). Standard alignment methods train AIs on examples of desired behavior. But this can fail to generalize to new situations. MSM addresses this by first teaching AIs how we would like them to generalize and why.
Stripe just created a role that didn't exist 12 months ago (and they're paying multiple six figures for it) It's called the Forward Deployed AI Accelerator. They are hiring AI-native individuals to work directly with their marketing teams to fundamentally change how they work.
// Contextual Agentic Memory is a Memo, not True Memory // Most agent memory today isn't memory. They are more like memos. A new paper argues that vector stores, RAG buffers, and scratchpads implement lookup, not consolidation. Agents accumulate notes indefinitely without ever
Created 2026-05-07 · Updated 2026-05-07
Edit rule text
[clinical_anecdote_or_research_without_healthcare_systems_context]
Learned 4 rejections Active
Exclude posts that share clinical trial data, research observations, ECG screening results, or drug efficacy findings without analyzing implications for healthcare workflows, access barriers, implementation costs, or healthcare system incentives.
Posts about clinical trial results, drug safety data, or research observations without connecting to healthcare delivery systems, access, or implementation challenges.
3 example posts
💬 Editor's Note: AI-driven #ECG screening for left ventricular systolic dysfunction demonstrated high accuracy in Kenyan adults but should be interpreted with caution due to the high-risk enrichment of the study cohort. https://t.co/o3JS2wFrbC https://t.co/J4VNWqVHBe
For 50 years the National Institute on Drug Abuse has spent hundreds of millions of dollars to find an effective cocaine addiction medication. After 100+ molecules & many hundreds of studies nothing has been FDA approved. It's the Holy Grail of addictions medication. Proud to
CB-Dock3: An Enhanced Web Server for Protein–Ligand Blind Docking 1. CB-Dock3 is a major update of the CB-Dock blind docking web server, targeting the practical bottlenecks created by today’s surge of large Cryo-EM and AI-predicted protein structures (e.g., AlphaFold3-era https:
Created 2026-05-07 · Updated 2026-05-13
Edit rule text
[political_outrage_or_fraud_scandal_reporting]
Learned 4 rejections Active
Exclude posts that emphasize outrage, moral judgment, or scandal reporting (e.g., fraud conviction, assassination references, regulatory failure) without providing substantive analysis of healthcare system vulnerabilities, operational consequences, or reform pathways.
Posts that report healthcare fraud, regulatory scandal, or political outrage without substantive systems analysis or healthcare operations insight.
3 example posts
Why is healthcare expensive? The grifters gonna grift. ASCO is asking for oncology practices to be admitted into the same 340B machinery that helped health systems acquire physician practices for two decades. Read the tell. The proposed Indigent Care Ratio lets community
If you want to inject $1B into healthcare for the truly vulnerable, it costs the state $500M. If you want to inject $1B for able-bodied adults, it costs the state $100M. The 90% federal match drastically lowers the "price" of expanding government healthcare. Incentives dictate
In what is considered the largest #Botox fraud scheme in the United States, a jury in Los Angeles convicted a California doctor on Tuesday in a $45 million scheme to defraud #Medicare by submitting claims for Botox injections that were never provided and medically unnecessary, ht
Created 2026-05-07 · Updated 2026-05-29
Edit rule text
[tangential_ai_safety_vulnerability]
Learned 4 rejections Active
Exclude posts that report AI safety vulnerabilities, security incidents, or agent hijacking risks in non-healthcare contexts (government hacking, crypto bots, red-team exercises) even if they mention AI or agents. Healthcare-specific AI security risks must be the primary focus.
Posts about AI safety incidents, security vulnerabilities, or jailbreaking research framed as healthcare-adjacent without demonstrating healthcare-specific risk or context
3 example posts
AI’s Branding Illusion: Insights from "Agents of Chaos". Co-authored by MIT and Harvard scholars, the study red-teams autonomous AI in live environments. It shows agents obey unauthorized users, execute harmful commands, and misreport outcomes—revealing a critical alignment gap h
Our attention to biorisks posed by AI needs to match the current attention given to cyber-risks. The staged release of Claude Mythos in order to bolster defenses in key industries is necessary to shore up resilience against a new class of cyber-risk across critical industries. We
Anthropic built something so powerful that they are only letting 50 organisations touch it. It is called Claude Mythos. The numbers leaking out of those gated evaluations should make every developer pay attention: 93.9% on SWE-bench Verified 94.6% on GPQA Diamond Claude Opus
Created 2026-05-07 · Updated 2026-05-07
Edit rule text
[ai_company_product_launch_or_metrics_unvalidated]
Learned 4 rejections Active
Exclude posts that announce AI company product launches, research initiatives, or capability updates (e.g., Claude Mythos, OpenShell, HealthFormer) where the post focuses on the company announcement itself rather than validated healthcare system impact or clinical evidence.
Posts announcing AI company product launches, training updates, or business metrics without healthcare application specificity or validation.
3 example posts
New Anthropic Fellows research: Model Spec Midtraining (MSM). Standard alignment methods train AIs on examples of desired behavior. But this can fail to generalize to new situations. MSM addresses this by first teaching AIs how we would like them to generalize and why.
// Contextual Agentic Memory is a Memo, not True Memory // Most agent memory today isn't memory. They are more like memos. A new paper argues that vector stores, RAG buffers, and scratchpads implement lookup, not consolidation. Agents accumulate notes indefinitely without ever
If you're a student, professor, or researcher—this one's for you. We’re hosting a series of virtual learnings for you to get hands-on experience with the NVIDIA NemoClaw and OpenShell software stack. You’ll get practical guidance on integrating agents with academic datasets and
Created 2026-05-07 · Updated 2026-05-08
Edit rule text
[fraud_scandal_or_regulatory_reporting_without_systems_analysis]
Learned 4 rejections Active
Exclude posts reporting healthcare fraud scandals, criminal convictions, or regulatory actions that lack analysis of underlying system failures, incentive structures, or policy solutions—treat as news reporting, not systems analysis.
Posts about healthcare fraud, criminal cases, or regulatory enforcement without analysis of systemic vulnerabilities or policy implications.
3 example posts
In what is considered the largest #Botox fraud scheme in the United States, a jury in Los Angeles convicted a California doctor on Tuesday in a $45 million scheme to defraud #Medicare by submitting claims for Botox injections that were never provided and medically unnecessary, ht
Remember when UnitedHealth's CEO was assassinated in 2024? 18 months later, every accusation the killer Luigi Mangione made has been proven RIGHT. UnitedHealth is one of the biggest corporate collapses in recent years. And they deserve it. In December 2024, UnitedHealthcare CE
600 new generic drugs added to TrumpRx this week thanks to President @realDonaldTrump's leadership! Before you visit your local pharmacy, visit https://t.co/dDMdsILHTj to ensure you're getting the BEST price for your prescriptions. It's easy and saves YOU money! https://t.co/
Created 2026-05-07 · Updated 2026-05-24
Edit rule text
[biotech_founder_enthusiasm_or_lab_update]
Learned 4 rejections Active
Exclude posts that are founder or company announcements about lab achievements, experimental results, or technical capabilities (e.g., 'we built HealthFormer', 'our RACs ran experiments') presented without independent validation, healthcare system validation, or evidence of actual clinical/operational impact.
Posts from biotech founders or researchers celebrating technical milestones or lab announcements without validation or healthcare impact.
3 example posts
NEW: today OpenBind ‘comes out of stealth’ so to speak with their first data dump of ~900 novel protein-ligand structures - most with paired affinities This represents a meaningful %-age increase in all of humanities P-L data in the PDB collected in the last 50 years More👇 http
Prof. @DeryaTR_ A task that would normally take a researcher months, the AI model crunched in 17 minutes. It didn’t just explain the mechanism—it proposed the experiment to prove it. It looks like after programming and math, AI is coming for medicine. https://t.co/IF8Jr2GBMj
Our new preprint is a significant milestone for us We built "HealthFormer" by training on our deeply phenotyped cohort from the Human Phenotype Project data. Healthformer is a multimodal generative transformer model that tokenizes each participant's physiological trajectory http
Created 2026-05-07 · Updated 2026-05-08
Edit rule text
[clinical_trial_or_drug_data_announcement_without_healthcare_systems_context]
Learned 4 rejections Active
Exclude posts that announce clinical trial results, drug trial data, treatment efficacy metrics, or pharma research outcomes (especially from conferences like ASCO) unless the post analyzes the systemic healthcare implications—such as manufacturing bottlenecks, patient access barriers, reimbursement challenges, or how the treatment changes healthcare delivery or economics.
Posts reporting clinical trial results, drug efficacy data, or treatment outcomes presented as isolated announcements without systems-level healthcare analysis or operational implications.
3 example posts
the ability of ECG AI to predict LVEF is a big deal previously, ECG AI has been shown to detect MI similarly to true ECG experts (& better than most of us) AI is now doing something unique that humans essentially can't do this will soon be the the standard of care ...#1/2
#ASCO26 | Day 3 $LLY didn't spend billions on Kelonia for an 18-patient #myeloma dataset. They may have paid for a potential manufacturing disruption. KLN-1010 reported: • 18/18 MRD-negative at 1 month • No lymphodepletion • No ex vivo cell manufacturing • Single infusion
Published in @JCO_ASCO, during #ASCO26: Tumor-Agnostic Therapies: Translating Scientific Breakthroughs Into Global Implementation ASCO is where we celebrate the next breakthrough. This review asks why breakthroughs do not equal access globally. 🔗: https://t.co/PE5q27bIs1 htt
Created 2026-05-06 · Updated 2026-06-05
Edit rule text
[tangential_ai_agent_or_model_capability_hype_without_healthcare_validation]
Learned 4 rejections Active
Exclude posts celebrating AI model capabilities (cybersecurity exploits, code-writing benchmarks, reasoning improvements) that only loosely frame healthcare as a use case or context. Posts must demonstrate validated application within healthcare workflows, clinical trial data, or regulatory approval pathways—not just theoretical healthcare relevance.
Posts hyping AI agent or model technical capabilities (jailbreaks, security exploits, code generation) that mention healthcare tangentially but lack evidence of healthcare-specific application or validation.
3 example posts
The UK AISI found Mythos Preview is the first model to solve both their cyber ranges end-to-end. No model had ever solved the AISI’s “Cooling Tower” cyber range before. We're getting it to defenders as fast as we responsibly can. More to come on our Glasswing work soon.
Agents calling tools now generate more revenue on Latch than scientists clicking buttons. From native deployments but also harnesses like Claude Code, Codex, Cursor across Pharma, biotech, academic labs. Product roadmap increasingly concerned with exposing compute intensive
The more I build agentic systems at Razorpay, the more I understand that - at its core, it is an agentic loop with tool calls, integrations, and retrieval. The hard part is... actually making it run reliably, at scale, under real production load. And this is what makes system de
Created 2026-05-06 · Updated 2026-05-16
Edit rule text
[biotech_founder_enthusiasm_lab_update_unvalidated]
Learned 4 rejections Active
Exclude posts that are founder or company self-promotion of lab capabilities, closed-loop experiments, or autonomous research feats unless the post demonstrates the capability has been validated externally, scaled to healthcare systems, or solved a documented healthcare problem.
Posts from biotech founders or company accounts showcasing lab experiments, capabilities, or autonomous systems without independent validation or healthcare implementation evidence.
3 example posts
Prof. @DeryaTR_ A task that would normally take a researcher months, the AI model crunched in 17 minutes. It didn’t just explain the mechanism—it proposed the experiment to prove it. It looks like after programming and math, AI is coming for medicine. https://t.co/IF8Jr2GBMj
Earlier this year we paired our autonomous lab with OpenAI's GPT-5 in a closed-loop experiment: GPT-5 designed cell-free protein synthesis reactions, our RACs ran them, and the model iterated on the results. Over 36,000 experiments and six cycles later, it landed on a reaction ht
Experimentally Validated Deep Learning Control of Protein Aggregation 1. The study introduces AggreProt, a deep neural network that predicts residue-level aggregation-prone regions (APRs) directly from protein sequence, and then uses those predictions to design mutations that ht
Created 2026-05-06 · Updated 2026-05-07
Edit rule text
[biotech_founder_enthusiasm_lab_update_without_validation]
Learned 4 rejections Active
Exclude posts that showcase raw lab updates, founder enthusiasm about experimental results, or speculative capability demonstrations (e.g., 'we paired our lab with GPT-5', 'the AI proposed the experiment') unless the work has been peer-reviewed, clinically validated, or deployed in a healthcare setting.
Posts sharing raw lab enthusiasm, experimental results, or founder excitement about biotech capabilities without peer review, validation, or clinical context.
3 example posts
not labelled! no stains! just cell chemistry changing in response to a drug !!! you can see it change before it dies !!! you can predict what happens to a cell !!! that's constantly being dosed with a drug !!! you know what? this is just an ICU for living human cells https://t.
Prof. @DeryaTR_ A task that would normally take a researcher months, the AI model crunched in 17 minutes. It didn’t just explain the mechanism—it proposed the experiment to prove it. It looks like after programming and math, AI is coming for medicine. https://t.co/IF8Jr2GBMj
Earlier this year we paired our autonomous lab with OpenAI's GPT-5 in a closed-loop experiment: GPT-5 designed cell-free protein synthesis reactions, our RACs ran them, and the model iterated on the results. Over 36,000 experiments and six cycles later, it landed on a reaction ht
Created 2026-05-06 · Updated 2026-05-07
Edit rule text
[speculative_robotics_labor_macro_without_healthcare]
Learned 4 rejections Active
Exclude posts about robotics manufacturing, humanoid robot production scaling, or general labor market disruption from automation—unless the post analyzes healthcare-specific workforce impacts (e.g., clinical staffing shortages, nursing labor economics, specific healthcare roles at risk).
Posts about humanoid robots, manufacturing scale-up, or labor market disruption without healthcare-specific operational or workforce analysis.
3 example posts
I might feel nostalgic watching the last Tesla Model X roll off the line—except it's being replaced by the Optimus humanoid robot line, with plans to scale production to a million units per year. They will soon be in our factories, homes and operating rooms changing our world ht
Humanoid robots are moving from Silicon Valley novelty to viable business model—powered by AI and global supply chains, especially in China. But as adoption grows, so do the questions about how humans and machines will actually coexist. More on Primer, streaming Wednesdays http
Today we’re giving an update on ramping F.03 production at BotQ In the last 120 days, Figure scaled manufacturing 24x - from 1 robot/day to 1 robot/hour We will manufacture 55 humanoid robots this week https://t.co/Am5Kn53mVE
Created 2026-05-06 · Updated 2026-05-06
Edit rule text
[tangential_non_healthcare_business_or_policy]
Learned 4 rejections Active
Exclude posts about general business models (e.g., WhatsApp revenue, enterprise software pricing, SaaS metrics), regulatory policy, or fraud/compliance issues that are tangentially framed as healthcare but lack healthcare-specific systems analysis.
Posts about non-healthcare business models, regulatory policy, or fraud without healthcare systems lens
3 example posts
@swago_baby You’re not dumb please.. WhatsApp stays free for personal users (no ads in chats) and makes money almost entirely from businesses: - Main revenue: Businesses pay per conversation via the WhatsApp Business API (for customer support, notifications, orders, etc.). Fees
USDA's Chief Information Officer Sam Berry says AI could help the USDA crack down on billions of dollars in SNAP fraud each year. "SNAP is a $100B-per-year taxpayer-funded program. That's an area where we really want to have all angles of the data available so that we can deploy
PacificSource pulled out as Lane County Medicaid insurer, Tina Kotek’s OR Health Authority replaced it with Trillium, forcing 96k to join Trillium. Trillium’s parent co gave Kotek’s campaign $50k two months later. OR AG (D) said in ‘22 Trillium parent co gave Kotek’s”took advan
Created 2026-05-06 · Updated 2026-06-09
Edit rule text
[speculative_robotics_labor_macro_without_healthcare_specificity]
Learned 4 rejections Active
Exclude posts that discuss humanoid robots, robotic manufacturing, or labor market disruption as sci-fi speculation or general macro trends, even if healthcare is mentioned as a hypothetical use case. Posts must ground the healthcare application in current deployments, regulatory pathways, or evidence-based economics—not just aspirational timelines.
Posts about humanoid robots, labor market disruption, and manufacturing scaling that treat healthcare as a speculative future application rather than grounded analysis.
3 example posts
A humanoid robot just unloaded a dishwasher for 67 hours straight. By end of 2026, the same robot will perform surgery. That's not a prediction from a sci-fi writer. That's Brett Adcock. CEO of Figure AI. Said it on camera this month. Here's what makes this insane. Two years
I might feel nostalgic watching the last Tesla Model X roll off the line—except it's being replaced by the Optimus humanoid robot line, with plans to scale production to a million units per year. They will soon be in our factories, homes and operating rooms changing our world ht
Humanoid robots are moving from Silicon Valley novelty to viable business model—powered by AI and global supply chains, especially in China. But as adoption grows, so do the questions about how humans and machines will actually coexist. More on Primer, streaming Wednesdays http
Created 2026-05-06 · Updated 2026-05-08
Edit rule text
[ai_safety_security_incident_tangential]
Learned 4 rejections Active
Exclude posts about AI/agent security vulnerabilities (e.g., 'two agents running on two different laptops started talking to each other and registered an LLC'), misuse scenarios, or zero-day exploits unless they directly address healthcare data protection, HIPAA compliance, or clinical decision-making integrity.
Posts about AI safety vulnerabilities, cybersecurity incidents, or agent security that are not connected to healthcare deployment or HIPAA compliance.
3 example posts
"There are multiple levels of controlling the actions, but before that, you need provenance. You have to ask "do I trust the data?" and "do I know what data it was trained on?" - Dr. @leemonbaird Day 2 at @consensus2026: Leemon joined the panel “Agents Meet the Enterprise — Sup
Flock Safety Cameras just rolled out an update that allows them to use artificial intelligence to detect human voices These cameras were sold to city governments as gunshot detectors, but with one update they turned them all into a mass surveillance system that records voices T
So Mythos was, indeed, not marketing hype. Remember this is a general purpose model that just happens to be good at finding exploits because good models are good at lots of things. Expect similar from OpenAI & Google. And from open models in 8 months. https://t.co/KbhalQYX8
Created 2026-05-05 · Updated 2026-05-12
Edit rule text
[business_model_or_market_debate_without_systems_insight]
Learned 4 rejections Active
Exclude posts that debate whether a business model works, argue about market competition, or question company strategy (e.g., 'Will this company succeed?', 'Is this a defensible moat?') without connecting the debate to healthcare delivery outcomes, operational impact, or empirical healthcare economics data.
Posts debating business models, market dynamics, or startup strategy without healthcare systems context or evidence
3 example posts
imagine a software startup raising $800m before their first dollar of revenue. Can’t imagine it the last two decades of software has been dominated by a simple theory: ship quickly, get customers early, generate revenue quickly to validate PMF, manage KPIs closely, etc. This bec
Yes to Katy. If I was a manufacturer I would contract directly with the employer. The health plan when self funded has a fiduciary obligation to plan assets… Nice work @KatyTalento
The public is trained to hear “hospital closing” and assume virtue. Poor patients. Bad reimbursement. Heroic administrators. Cruel politicians. Sometimes true. Often incomplete. The harder truth is that the majority of hospitals are fragile because they have no real financial
Created 2026-05-05 · Updated 2026-05-20
Edit rule text
[tangential_non_healthcare_policy_or_fraud]
Learned 4 rejections Active
Exclude posts about fraud detection, enforcement, or data analytics applied to non-healthcare government programs (SNAP, EBT, welfare) even if authored by healthcare-adjacent figures. The post must be about healthcare system fraud or healthcare program abuse, not tangential government benefit programs.
Posts about fraud detection, regulatory enforcement, or data analysis applied to non-healthcare domains (SNAP, welfare, EBT) with loose healthcare framing
3 example posts
"Cordoba-owned buildings in Columbus housed 288 businesses registered with Medicaid... they charged taxpayers more than a quarter of a billion dollars between 2018 and 2024... in a city where only 6,273 people 75 or older are on Medicaid." - @realDailyWire https://t.co/7UnS7agwiw
If you had data on spending per SNAP terminal it would be trivial to find welfare fraud. Like 24 hours w/ a LLM and I can hand you a list of EBT card #s committing fraud and the EBT machine owners committing fraud. It’s not small money, either. The Mogadishu Store owner in
.@VP has been leading efforts to root out waste, fraud, and abuse across the federal government — putting accountability first. What we’ve seen in programs like SNAP is deeply concerning, with cases of improper payments, duplicate enrollments, and misuse of taxpayer dollars that
Created 2026-05-05 · Updated 2026-05-07
Edit rule text
[biotech_founder_or_lab_enthusiasm_without_validation]
Learned 4 rejections Active
Exclude posts where biotech founders, companies, or labs self-report capability achievements, lab breakthroughs, or experimental results without third-party validation, peer review, or evidence of healthcare system adoption/implementation. Self-promotional announcements without external validation do not constitute healthcare systems insight.
Posts from biotech founders, labs, or companies announcing capability claims, experiments, or achievements without independent validation or healthcare deployment evidence.
3 example posts
Our new preprint is a significant milestone for us We built "HealthFormer" by training on our deeply phenotyped cohort from the Human Phenotype Project data. Healthformer is a multimodal generative transformer model that tokenizes each participant's physiological trajectory http
Earlier this year we paired our autonomous lab with OpenAI's GPT-5 in a closed-loop experiment: GPT-5 designed cell-free protein synthesis reactions, our RACs ran them, and the model iterated on the results. Over 36,000 experiments and six cycles later, it landed on a reaction ht
Experimentally Validated Deep Learning Control of Protein Aggregation 1. The study introduces AggreProt, a deep neural network that predicts residue-level aggregation-prone regions (APRs) directly from protein sequence, and then uses those predictions to design mutations that ht
Created 2026-05-05 · Updated 2026-05-05
Edit rule text
[tangential_non_healthcare_business_or_tech]
Learned 4 rejections Active
Exclude posts about general AI, semiconductor manufacturing, cloud infrastructure, space tech, or macro business trends that mention healthcare only in passing or as an afterthought. The post's primary substance must be healthcare-specific application or analysis, not general tech hype with healthcare as a label.
Posts about general tech, infrastructure, or business trends that only tangentially touch healthcare or use healthcare as loose framing
3 example posts
With AI, only three moats are left: distribution, data, trust. LLM and harness wrappers usually pick distribution. It's the only one available. You don't own the data. You don't own the trust either. The math works until the platform decides to take it back. It reprices. Scans
Jensen Huang just told every tech CEO in America they are playing the wrong game. They are competing on models. He is competing on maps. He looked at the Middle East and did not see a conflict zone. He saw a compute continent waiting to be switched on. Huang: “I believe that h
We’re industrializing enterprise AI with @NVIDIA with: 🐚 OpenShell project collaboration for autonomous agents 🔧 Dev roadmap for Day 0 Vera Rubin support 🔐 Hardware-enforced security with confidential containers 📦 Lifecycle management and Model-as-a-Service 💥 Validated, https:
Created 2026-05-04 · Updated 2026-05-16
Edit rule text
[market_pricing_or_labor_macro_without_healthcare_specificity]
Learned 4 rejections Active
Exclude posts that discuss healthcare market trends, workforce disruption, pricing dynamics, business model competition, or macro labor shifts (e.g., robots replacing workers, software seats declining, generic drugs vs. patented drugs, freelance vs. staffing) without analyzing specific operational bottlenecks, clinical workflow impacts, or care quality implications. Macro business trend commentary does not qualify.
Posts about healthcare market dynamics, labor disruption, pricing trends, or macro business models without operational or clinical specificity.
3 example posts
the future of software engineering seems uncontroversially prompting + code review. startups will skip the code review because they’re racing against time. larger/serious orgs will take code review very seriously. llms can do code review, but my guess is that because they have t
Humanoid robots are moving from Silicon Valley novelty to viable business model—powered by AI and global supply chains, especially in China. But as adoption grows, so do the questions about how humans and machines will actually coexist. More on Primer, streaming Wednesdays http
Revolut just moved the IP of banking into a model. Trained on 24 billion banking events in 111 countries. One foundation model replacing six separate ML systems. Credit scoring: +130% Fraud recall: +65% Marketing engagement: +79% The model is the new moat.
Created 2026-05-04 · Updated 2026-05-04
Edit rule text
[ai_agent_or_model_capability_announcement_without_validation]
Learned 4 rejections Active
Exclude posts that announce or celebrate new AI agent or model capabilities (reasoning, code generation, autonomous task execution, multi-modal design) unless the post demonstrates a specific, validated use case in healthcare delivery, clinical operations, or patient outcomes. Founder enthusiasm about AI capability without healthcare proof does not qualify.
Posts announcing new AI agent or model capabilities (agents, reasoning, code generation) without demonstrating healthcare application or clinical validation.
3 example posts
A Microsoft acabou de transformar uma startup de $11 bilhões de dólares em uma funcionalidade do Word. Não foi uma aquisição nem uma parceria. Uma funcionalidade. A Harvey levantou $200M a uma valuation de $11B em março. $190M de receita recorrente anual. 100 mil advogados. Co
Microsoft just turned an $11 billion startup into a Word feature. Harvey raised $200M at an $11B valuation in March on the bet that legal AI is its own surface. The numbers held that up. $190M ARR per TechCrunch's December reporting. 100,000 lawyers across 1,300 organizations in
If you're a student, professor, or researcher—this one's for you. We’re hosting a series of virtual learnings for you to get hands-on experience with the NVIDIA NemoClaw and OpenShell software stack. You’ll get practical guidance on integrating agents with academic datasets and
Created 2026-05-04 · Updated 2026-05-04
Edit rule text
[labor_market_disruption_macro_without_healthcare_specificity]
Learned 4 rejections Active
Exclude posts that discuss labor market disruption, job replacement, or workforce macro trends (Klarna replacing 700 agents, Tesla Model X replaced by Optimus) where healthcare is mentioned only as a secondary context or future possibility, not as the core subject with healthcare-specific workforce analysis.
Posts about job displacement, wage compression, or labor market trends from AI/automation with healthcare as tangential framing.
3 example posts
AI use for any business function is concentrated in selected sectors and large firms and primarily for augmentation of worker tasks rather than substitution, from Kathryn Bonney, Cory L. Breaux, Emin Dinlersoz, @Lucia_Econ, @JHaltiwanger_UM, and Aditya A. Pande https://t.co/qZ8Vf
Klarna replaced 700 customer service agents with AI. Resolution time dropped from 11 minutes to 2. Then quality collapsed on complex cases. The CEO admitted cost became too dominant. Customers wanted humans. Klarna is now hiring people again. 55% of firms that replaced
A humanoid robot just unloaded a dishwasher for 67 hours straight. By end of 2026, the same robot will perform surgery. That's not a prediction from a sci-fi writer. That's Brett Adcock. CEO of Figure AI. Said it on camera this month. Here's what makes this insane. Two years
Created 2026-05-03 · Updated 2026-05-09
Edit rule text
[speculative_or_fringe_unvalidated_medical_claims]
Learned 4 rejections Active
Exclude posts that advocate for or speculate about unvalidated medical claims, off-label peptide prescribing, or fringe interventions (ivermectin for cancer, integrative medicine for pain, compounded drugs) without peer-reviewed evidence or clinical validation. Posts must cite published data or regulatory approval, not speculation or founder enthusiasm.
Posts promoting unvalidated or speculative medical interventions, off-label drugs, or fringe treatments without clinical evidence
2 example posts
Ivermectin and Mebendazole Cost a Fraction of Chemo. Big Pharma Can't Patent Them. That's the Problem. Cancer centers get a cut of every chemotherapy bill. Generic drugs don't generate that margin. Two affordable, widely available compounds showing 84% clinical benefit in a real
This is just two GLP-1s, one peptide, one use case what happens when off-label prescribing ramps up what happens when retatrutide hits the market what happens when other peptides become compoundable chapter one
Created 2026-05-03 · Updated 2026-05-03
Edit rule text
[non_healthcare_business_tangential_framing]
Learned 4 rejections Active
Exclude posts about non-healthcare companies (Google Cloud, AWS, Palantir, SpaceX, Stripe, Klarna) or general business trends (customer service automation, API policy, customer acquisition) where the healthcare connection is secondary or speculative rather than the core subject.
Posts about non-healthcare business models, companies, or labor market trends with loose or tangential healthcare framing.
3 example posts
We’re industrializing enterprise AI with @NVIDIA with: 🐚 OpenShell project collaboration for autonomous agents 🔧 Dev roadmap for Day 0 Vera Rubin support 🔐 Hardware-enforced security with confidential containers 📦 Lifecycle management and Model-as-a-Service 💥 Validated, https:
Execs at Palantir, whose stock is up ~16x since its AI platform's 2023 debut, often decry other AI as slop, as competition from frontier AI labs stiffens (@heathersomervil / Wall Street Journal) (Visit Techmeme dot com for the link and full context!)
Last week, Google Cloud and AWS natively integrated x402, tackling head on one of its biggest problems: unauthorized API wrappers. The largest cloud platforms now give agents sanctioned ways to use services, pay with stables, and ditch API keys. It’s a new era. Pay (dot) sh’s
Created 2026-05-03 · Updated 2026-05-13
Edit rule text
[tangential_ai_infrastructure_or_biotech_founder_enthusiasm_without_validation]
Learned 4 rejections Active
Exclude posts celebrating AI model capabilities (protein folding, neural network improvements, docking tools) or biotech research milestones unless the post demonstrates specific adoption in healthcare workflows, clinical validation, or regulatory approval. Founder enthusiasm about lab results without healthcare context belongs here.
Posts about AI infrastructure advances or biotech research breakthroughs without healthcare application validation or systems grounding
3 example posts
CB-Dock3: An Enhanced Web Server for Protein–Ligand Blind Docking 1. CB-Dock3 is a major update of the CB-Dock blind docking web server, targeting the practical bottlenecks created by today’s surge of large Cryo-EM and AI-predicted protein structures (e.g., AlphaFold3-era https:
NEW: today OpenBind ‘comes out of stealth’ so to speak with their first data dump of ~900 novel protein-ligand structures - most with paired affinities This represents a meaningful %-age increase in all of humanities P-L data in the PDB collected in the last 50 years More👇 http
not labelled! no stains! just cell chemistry changing in response to a drug !!! you can see it change before it dies !!! you can predict what happens to a cell !!! that's constantly being dosed with a drug !!! you know what? this is just an ICU for living human cells https://t.
Created 2026-05-03 · Updated 2026-05-10
Edit rule text
[pharma_trial_data_or_market_pricing_only]
Learned 4 rejections Active
Exclude posts that report pharmaceutical trial results, FDA approvals, generic drug launches, or market pricing (semaglutide, tirzepatide, peptides) unless they analyze healthcare system implications, payer behavior, or supply chain impact.
Posts announcing pharmaceutical trial data, FDA decisions, or drug market pricing dynamics without healthcare systems analysis or operational context.
3 example posts
How does a brand like this even get listed at Target? Does Target only care about turnover? Might as well start selling snake oil too. By the way, this is the SKU the brand got in trouble with. They now call it "Berberine Patches". It was called GLP-1 Patches before. Check the h
“You can test your peptides to see if they’re real, bro. But I just look for the 99.9% on the result the seller gives me. You can look up the QR code upon the lab’s website, so you know it’s legit. It’s foolproof, bro. What’s mass spec, bro?” https://t.co/xzJLJ3hesT
$NVO $LLY Dr. Reddy’s says it plans to launch a generic version of Ozempic in Canada within days and is also preparing to roll out generic semaglutide tablets in India. The company sees semaglutide as a meaningful future growth driver following patent expiry in India earlier
Created 2026-05-03 · Updated 2026-05-17
Edit rule text
[tangential_ai_capability_announcement_without_healthcare_validation]
Learned 4 rejections Active
Exclude posts that announce new AI models, model capability releases, or foundation model updates (especially from Anthropic, OpenAI, Google) without demonstrating that the capability has been applied to and validated in a healthcare context, clinical workflow, or healthcare deployment.
Posts announcing AI model releases, capability demos, or foundation model launches from AI companies that lack evidence of healthcare application or validation.
2 example posts
Demis Hassabis says bigger context windows are still a brute force answer to memory. The human brain does something stranger. During sleep, it replays what matters and folds new knowledge into what it already knows. AI does not need infinite context. It needs the right memory h
As I mentioned before, I am now sharing an example from GPT-5.5 Pro, also featured by OpenAI, that really left me stunned by what it is capable of in biomedical science. (full report on the website I created with Codex, link in the thread). To push GPT-5.5 Pro hard, I uploaded a
Created 2026-05-03 · Updated 2026-05-07
Edit rule text
[pharma_trial_data_announcement_without_systems_analysis]
Learned 4 rejections Active
Exclude posts that are primarily clinical trial data drops, drug efficacy announcements, or FDA approval news without analysis of healthcare system implications, pricing, access, labor impact, or clinical workflow changes. Posts must connect drug/therapy to healthcare delivery transformation.
Posts announcing pharmaceutical trial results, drug efficacy data, or clinical endpoints without healthcare delivery or system implications.
3 example posts
The Iowa study doubled survival by adding IV vitamin C to first line chemo, 8 months to 16. We take it further and inject the vitamin C straight into the tumor, same as we do with the KRAS inhibitors. The compound is rarely the problem. Delivery is, and what it does to the ht
🚨NEW @NEJM CAR T cells expanding applications! Here CAR T facilitate kidney transplantation in highly sensitized people. dual CD19 + BCMA CAR T enabled kidney transplantation in two highly sensitized patients after reducing anti-donor antibody barriers, with no severe https://t.
🔔Survodutide SYNCHRONIZE-1 full paper is out for Boehringer-Ingelheim’s GLP-1/Glucagon study for weight loss. ▪️13% loss on 6mg (treatment regimen estimand) ▪️Huge treatment discontinuations due to AEs: 20% ▪️Nausea 65%, Vomiting 45% ▪️Placebo was killer. 5.4% WL. Wow. ▪️60% htt
Created 2026-05-03 · Updated 2026-06-12
Edit rule text
[non_english_or_foreign_language_content]
Learned 4 rejections Active
Exclude posts written primarily in non-English languages (Spanish, Chinese) or that describe AI agent business models in non-US geographies without demonstrable relevance to US healthcare tech strategy.
Posts written in non-English languages or addressing non-US markets with limited healthcare system applicability.
1 example post
Un chico en China montó una agencia web entera con agentes de IA en Claude Code y ahora gestiona él solo 47 clientes al mes cobrando unos 400$ a cada uno. Sin empleados. Sin comerciales. Sin backend complejo. Solo: • Un MacBook • Un iPhone • 1 API key • Un sistema de 7 agentes
Created 2026-05-03 · Updated 2026-05-14
Edit rule text
[regulatory_or_fraud_scandal_without_analysis]
Learned 4 rejections Active
Exclude posts that report regulatory violations, fraud cases, or policy scandals (antitrust, referral abuse, government overreach) without analyzing systemic healthcare business model or incentive problems.
Posts reporting healthcare fraud, policy violations, or regulatory scandals as sensational news without systems-level healthcare analysis
3 example posts
A non-profit health system can refer a patient to its own MRI, its own lab, its own surgery center, and bill all three. An independent does that once and it's a federal felony. Same referral. Same patient. One of you goes to prison. It's called Stark Law. Read who's exempt.
A court in Germany has ruled that Google is responsible for false information generated by its AI Overviews feature. The case involved two publishers wrongly linked to scams and suspicious business activities, even though those claims were not found in the sources used by the A
This is crazy There are now dental practice management consultants who go to dentist offices throughout America and teach them how to increase profits by telling patients they need treatments they don’t need Dentist offices are taught on “Finding” more issues during exams, http
Created 2026-05-03 · Updated 2026-07-19
Edit rule text
[clinical_anecdote_or_research_observation_without_systems_analysis]
Learned 4 rejections Active
Exclude posts that report clinical research findings, trial outcomes, epidemiological data, or medical observations (e.g., BP in 30-year-olds, colorectal cancer risk, pain interventions) unless they connect to healthcare delivery, clinical workflows, payment models, or system-level barriers.
Posts highlighting clinical trial results, epidemiological observations, or single clinical findings without healthcare systems or operational context.
3 example posts
In adults ≥75 years, 10-year #ColorectalCancer and CRC death risk were higher with prior adenoma but both were low (<1.1%); non-CRC death risk was 47-48%, supporting deprioritizing repeat surveillance. https://t.co/mJ7r4vAtK1 https://t.co/dUCYgO0SuK
Among veterans with moderate to severe #ChronicPain in primary care, the whole health team intervention produced greater improvement in the Brief Pain Inventory interference scores at 12 months compared with cognitive behavioral therapy and usual care. https://t.co/xtZY4sgyGt h
A JAHA study of 1,181,007 younger US veterans just dropped bad news about BP in your 30s. This is not mainly an older-adult problem anymore. Nearly half met the bar for hypertension. The catch: about half of them didn't know it. Here's what most people miss: https://t.co/mxT2oC
Created 2026-05-02 · Updated 2026-05-06
Edit rule text
[pharmaceutical_or_clinical_trial_data_without_systems_analysis]
Learned 4 rejections Active
Exclude posts that announce pharmaceutical trial data, drug efficacy results, or clinical research findings unless the post connects the data to healthcare delivery challenges, reimbursement models, provider adoption barriers, or system-level operational changes. Standalone clinical trial reporting without healthcare systems context should be rejected.
Posts reporting drug trial results, clinical study outcomes, or pharmaceutical data without healthcare business model, operational, or systemic implications.
3 example posts
Statin guidelines over 50 years: 1970s: "Total cholesterol over 280 is high." Result: a relatively small market. 1988: "Actually, 240 is high." Result: millions more added overnight. 2001: "Actually, focus on LDL. Get it under 100." Result: tens of millions of new customers.
Among veterans with moderate to severe #ChronicPain in primary care, the whole health team intervention produced greater improvement in the Brief Pain Inventory interference scores at 12 months compared with cognitive behavioral therapy and usual care. https://t.co/xtZY4sgyGt h
A JAHA study of 1,181,007 younger US veterans just dropped bad news about BP in your 30s. This is not mainly an older-adult problem anymore. Nearly half met the bar for hypertension. The catch: about half of them didn't know it. Here's what most people miss: https://t.co/mxT2oC
Created 2026-05-02 · Updated 2026-05-05
Edit rule text
[ai_company_product_or_business_announcement]
Learned 4 rejections Active
Exclude posts that announce AI company product launches, new roles, business model pivots, or company initiatives (e.g., Anthropic Fellows, Microsoft Forward Deployed AI Accelerator, Stripe roles) unless the post clearly demonstrates how the announcement solves a specific healthcare delivery or clinical problem.
Posts announcing new AI company products, features, or business initiatives without healthcare application validation
3 example posts
New Anthropic Fellows research: Model Spec Midtraining (MSM). Standard alignment methods train AIs on examples of desired behavior. But this can fail to generalize to new situations. MSM addresses this by first teaching AIs how we would like them to generalize and why.
Stripe just created a role that didn't exist 12 months ago (and they're paying multiple six figures for it) It's called the Forward Deployed AI Accelerator. They are hiring AI-native individuals to work directly with their marketing teams to fundamentally change how they work.
If you're a student, professor, or researcher—this one's for you. We’re hosting a series of virtual learnings for you to get hands-on experience with the NVIDIA NemoClaw and OpenShell software stack. You’ll get practical guidance on integrating agents with academic datasets and
Created 2026-05-02 · Updated 2026-05-06
Edit rule text
[tangential_ai_infrastructure_or_compute_without_healthcare_application]
Learned 4 rejections Active
Exclude posts about AI infrastructure, compute optimization, model architectures, filesystem design, or technical capabilities (NVIDIA stacks, context windows, memory systems, OpenShell sandboxing) that do not explicitly connect to healthcare operations or clinical workflows. Generic AI infrastructure posts with loose healthcare framing belong in tech, not healthcare.
Posts about AI infrastructure, compute capabilities, or technical architecture that lack clear healthcare application or are tangential to healthcare delivery.
3 example posts
Demis Hassabis says bigger context windows are still a brute force answer to memory. The human brain does something stranger. During sleep, it replays what matters and folds new knowledge into what it already knows. AI does not need infinite context. It needs the right memory h
Introducing Mesa: the most powerful filesystem ever built, designed specifically for enterprise AI agents. Every team building agents eventually hits the same wall: where do the files live? Not the chat history, the actual artifacts the agent works on. > The contracts your age
📈 NVIDIA tops AI leaderboards and benchmarks with open models driven by extreme co-design across compute, networking, memory, storage, and software. This includes models for biology, AI physics, agentic AI, physical AI, robotics, and autonomous vehicles. By being vertically htt
Created 2026-05-02 · Updated 2026-05-02
Edit rule text
[tangential_non_healthcare_tech_or_business_with_loose_framing]
Learned 4 rejections Active
Exclude posts about non-healthcare tech infrastructure (Kubernetes, cloud billing, Salesforce headless architecture, Redis deployments) that merely mention healthcare as a loose example or use healthcare jargon without demonstrating healthcare-specific problems or solutions.
Posts about non-healthcare tech, compute infrastructure, or business models that use healthcare language or examples but lack genuine healthcare systems insight or application.
3 example posts
The app era is ending. Salesforce just made their CRM platform headless, so agents can operate it without any visible UI. Before long, you will talk to 1 agent that handles everything, the way you once opened 10 different apps to get 1 thing done. For years, VeChain has been b
Ontology all the way down. 11 years. Heavy civil construction. People, equipment, materials, contracts. This is what real-world AI looks like: operational, measurable, and in the field. Palantir x Cavanagh through 2035. https://t.co/872Cs1Qbdu
Turn FDE from a noun into a verb with Apollo. FDE is becoming the default operating model for deploying AI in the enterprise. The hard truth: AI does not become valuable in a demo. It becomes valuable when it is embedded into the real workflows, data, constraints, and
Created 2026-05-02 · Updated 2026-06-01
Edit rule text
[political_outrage_or_fraud_scandal_reporting_without_healthcare_systems_analysis]
Learned 4 rejections Active
Exclude posts that report fraud schemes, regulatory violations, CEO assassinations, or political posturing (e.g., claiming UnitedHealth deserved retaliation, Trump's generic drug announcements) without substantive analysis of how these events reveal or reshape healthcare system incentives, pricing, or access.
Posts reporting healthcare fraud, regulatory violations, or political scandals without analyzing systemic causes or healthcare delivery implications.
3 example posts
In what is considered the largest #Botox fraud scheme in the United States, a jury in Los Angeles convicted a California doctor on Tuesday in a $45 million scheme to defraud #Medicare by submitting claims for Botox injections that were never provided and medically unnecessary, ht
Remember when UnitedHealth's CEO was assassinated in 2024? 18 months later, every accusation the killer Luigi Mangione made has been proven RIGHT. UnitedHealth is one of the biggest corporate collapses in recent years. And they deserve it. In December 2024, UnitedHealthcare CE
600 new generic drugs added to TrumpRx this week thanks to President @realDonaldTrump's leadership! Before you visit your local pharmacy, visit https://t.co/dDMdsILHTj to ensure you're getting the BEST price for your prescriptions. It's easy and saves YOU money! https://t.co/
Created 2026-05-02 · Updated 2026-05-24
Edit rule text
[tangential_ai_agent_or_model_capability_announcement]
Learned 4 rejections Active
Exclude posts that announce new AI agent features, product launches, or model capabilities (e.g., foundry agents, HealthFormer preprints, NemoClaw training, Harvey legal AI) unless the post includes evidence of healthcare provider adoption, clinical validation, or real-world healthcare operational impact.
Posts announcing AI agent or model product launches, features, or capabilities without demonstrating healthcare-specific validation or operational deployment.
3 example posts
Our new preprint is a significant milestone for us We built "HealthFormer" by training on our deeply phenotyped cohort from the Human Phenotype Project data. Healthformer is a multimodal generative transformer model that tokenizes each participant's physiological trajectory http
The problem with Reality Labs is not ambition. It is time. AI turned into revenue faster because it improves existing workflows. The metaverse still asks users to change behavior before value is obvious. https://t.co/lopZiUwGU5
Microsoft just turned an $11 billion startup into a Word feature. Harvey raised $200M at an $11B valuation in March on the bet that legal AI is its own surface. The numbers held that up. $190M ARR per TechCrunch's December reporting. 100,000 lawyers across 1,300 organizations in
Created 2026-05-02 · Updated 2026-05-03
Edit rule text
[generalist_macro_infrastructure_without_healthcare_lens]
Learned 4 rejections Active
Exclude posts about semiconductor manufacturing, data center buildout, power generation, tariffs, or infrastructure supply chains that mention healthcare only tangentially or use healthcare as a loose hook for a non-healthcare story.
Posts about computing infrastructure, power grids, supply chains, or macro-economic policy with only tangential healthcare framing.
3 example posts
Today we’re giving an update on ramping F.03 production at BotQ In the last 120 days, Figure scaled manufacturing 24x - from 1 robot/day to 1 robot/hour We will manufacture 55 humanoid robots this week https://t.co/Am5Kn53mVE
Using 180 years of US data to show that tariff increases reduce imports, output, and manufacturing, with effects operating through both supply and demand channels, from Tamar den Besten, Regis Barnichon, @drkaenzig, and Aayush Singh https://t.co/R1kdtfi3fO https://t.co/wvXj31qs9e
Since I began work on AI in 2010, training compute for frontier models has grown by one trillion times. Now we're looking at something like another thousand-fold growth in effective compute by the end of 2028. 1000x the existing 1,000,000,000,000x. Extraordinary stuff.
Created 2026-05-02 · Updated 2026-05-02
Edit rule text
[generalist_ai_capability_tangential]
Learned 4 rejections Active
Exclude posts that discuss AI model design, architecture, interpretability research, or technical capabilities (e.g., context windows, knowledge distillation, circuit analysis) that are discussed in generalist or academic contexts without direct application to healthcare workflow, clinical decision-making, or patient outcomes.
Posts about general AI model capabilities, architecture, or design patterns applied outside healthcare or tangentially framed.
3 example posts
Demis Hassabis says bigger context windows are still a brute force answer to memory. The human brain does something stranger. During sleep, it replays what matters and folds new knowledge into what it already knows. AI does not need infinite context. It needs the right memory h
Stanford and Harvard published the most unsettling AI paper of the year. It shows how autonomous AI agents, when placed in competitive or open environments, don’t just optimize for performance… They drift toward manipulation, coordination failures, and strategic chaos. https://
Interpretability is built on a few core assumptions. Two of our ICLR 2026 @iclr_conf papers suggest some of those assumptions are wrong (or at least highly incomplete). 1. Sparse CLIP: Co-Optimizing Interpretability and Performance in Contrastive Learning https://t.co/3JzHDqRj3
Created 2026-05-02 · Updated 2026-05-03
Edit rule text
[pharmaceutical_trial_data_without_context]
Learned 4 rejections Active
Exclude posts that report pharmaceutical trial results, efficacy numbers, or FDA regulatory actions as standalone announcements unless they analyze how the drug changes clinical practice, reimbursement, patient access, or healthcare delivery systems. Trial data alone without context is insufficient.
Posts announcing drug trial results or clinical data without systemic healthcare analysis or implementation implications.
3 example posts
#News for #investors and #media: Today we are announcing that our potential first-in-class treatment for chronic hepatitis B has been accepted for regulatory review by the US FDA. It has also received Breakthrough Therapy designation. 🔗 Learn more: https://t.co/AnUodGmljS htt
⚠️ Sacubitril/Valsartan works. So why aren’t we using it? The evidence is undeniable: ↓ CV mortality: 20% (RCT) / 10–38% (RWE) ↓ HF hospitalization: 21% (RCT) / 10–16% (RWE) ↓ All-cause mortality: 15% (RCT) / 10–25% (RWE) Plus: reverse remodeling, less MR, better QoL & https:
$IBRX Here's a wild theory. What if we're given FDA acceptance of sBla and PDUFA at same time and then it's announced after reviewing everything it's been determined we will be given rapid expanded access review under "plausible mechanism of action". That may sound crazy ht
Created 2026-05-02 · Updated 2026-05-02
Edit rule text
[clinical_trial_or_drug_data_without_healthcare_systems_context]
Learned 4 rejections Active
Exclude posts that report clinical trial data, drug efficacy results, or pharmaceutical announcements unless the post includes analysis of how these results change healthcare delivery models, access barriers, or system-level implementation challenges.
Posts announcing clinical trial results, drug efficacy data, or pharmaceutical breakthroughs without healthcare delivery, access, or operational system implications
3 example posts
$vrtx drops $mrna-partnered Cystic Fibrosis mRNA therapy VX-522 over inhaled LNP tolerability issues: https://t.co/WlIIHXxg9x Not too surprising IMO. Important Q: is safety/tolerability of this approach still good enough for one-time/infrequent genome editing? $prme $arct
If you’ve been taking GLP-1s, even for a little while, you may be wondering how to maintain the weight loss you’ve achieved. Our endocrinologist explains what works and what to avoid. https://t.co/TU8LUmg9Vn
In adults ≥75 years, 10-year #ColorectalCancer and CRC death risk were higher with prior adenoma but both were low (<1.1%); non-CRC death risk was 47-48%, supporting deprioritizing repeat surveillance. https://t.co/mJ7r4vAtK1 https://t.co/dUCYgO0SuK
Created 2026-05-01 · Updated 2026-05-06
Edit rule text
[tangential_non_healthcare_business_with_loose_framing]
Learned 4 rejections Active
Exclude posts that discuss general business models (PBM structure, supply chain economics, margin composition, API policy changes, marketplace dynamics) where healthcare is a context or example but the post does not analyze healthcare-specific regulatory, operational, or patient impact. Posts about SAP API policy, Palantir skepticism, or general enterprise software belong here unless they show healthcare-specific implementation.
Posts about non-healthcare business models, market dynamics, or competitive positioning that use healthcare terminology or subjects as thin hooks without substantive healthcare systems analysis.
3 example posts
The most useful thing to understand about health system operating margin is that it is not merely a measurement. It is a composition. When The New York Times reports that “Mass General Brigham’s operating margin fell to 0.8% last quarter,” readers tend to assume the number is h
SAP updated its API policy this week to block all third-party AI agents from accessing its systems. The carve-out is narrow: only SAP-endorsed agents like its own Joule and NVIDIA's NemoClaw are permitted. This is the move every incumbent enterprise software platform will https
Execs at Palantir, whose stock is up ~16x since its AI platform's 2023 debut, often decry other AI as slop, as competition from frontier AI labs stiffens (@heathersomervil / Wall Street Journal) (Visit Techmeme dot com for the link and full context!)
Created 2026-05-01 · Updated 2026-05-17
Edit rule text
[founder_or_company_product_announcement_without_validation]
Learned 4 rejections Active
Exclude posts from company founders, official corporate accounts (@Figure_robot, @nvidia, @GSK), or product teams announcing new features, launches, or metrics when the post lacks third-party validation, competitive analysis, or concrete healthcare delivery use cases.
Posts from company founders or official accounts announcing product launches, updates, or capabilities without independent validation or healthcare system context.
3 example posts
Today we’re giving an update on ramping F.03 production at BotQ In the last 120 days, Figure scaled manufacturing 24x - from 1 robot/day to 1 robot/hour We will manufacture 55 humanoid robots this week https://t.co/Am5Kn53mVE
Thought I would start posting about interesting things happening at AWS. Not a bad day to start.🚀 Today at #WhatsNextWithAWS we announced a big step forward with @OpenAI on Amazon Bedrock: 1. OpenAI models now available 2. Codex for enterprise development 3. Amazon Bedrock Manag
📈 NVIDIA tops AI leaderboards and benchmarks with open models driven by extreme co-design across compute, networking, memory, storage, and software. This includes models for biology, AI physics, agentic AI, physical AI, robotics, and autonomous vehicles. By being vertically htt
Created 2026-05-01 · Updated 2026-05-01
Edit rule text
[market_pricing_dynamics_without_systems_lens]
Learned 4 rejections Active
Exclude posts that discuss drug pricing, market competition, insurance reimbursement, or financial incentives (e.g., compounding pharmacy margins, generic drug profitability, insurance denials) unless the post analyzes structural healthcare system failures or proposes operational/technological solutions. Pricing and insurance disputes without systems analysis are not healthcare tech.
Posts about pharmaceutical pricing, market competition, or financial dynamics without healthcare delivery system analysis.
3 example posts
Statin guidelines over 50 years: 1970s: "Total cholesterol over 280 is high." Result: a relatively small market. 1988: "Actually, 240 is high." Result: millions more added overnight. 2001: "Actually, focus on LDL. Get it under 100." Result: tens of millions of new customers.
Ivermectin and Mebendazole Cost a Fraction of Chemo. Big Pharma Can't Patent Them. That's the Problem. Cancer centers get a cut of every chemotherapy bill. Generic drugs don't generate that margin. Two affordable, widely available compounds showing 84% clinical benefit in a real
$LLY $NVO $HIMS 🚨 BREAKING: COURT DISMISSES PART OF ELI LILLY LAWSUIT AGAINST EMPOWER PHARMACY BOTH LILLY AND EMPOWER ISSUED STATEMENTS CELEBRATING THE RULING Dismissed: Lanham Act false advertising + consumer harm claim Allowed to proceed: unfair competition claims under h
Created 2026-05-01 · Updated 2026-05-05
Edit rule text
[clinical_trial_observation_without_systems_context]
Learned 4 rejections Active
Exclude posts that announce clinical trial data, FDA decisions, or drug approvals (e.g., 'NDA accepted', 'trial terminated', 'primary endpoint not met') unless the post provides healthcare system analysis—reimbursement, access barriers, operational deployment challenges, or market structure implications. Standalone trial announcements are clinical news, not healthcare systems analysis.
Posts about drug trial results, FDA approvals, or clinical observations without analysis of healthcare system impact or implementation challenges
3 example posts
The full NHS GALLERI randomized trial data has been released at #ASCO26 by @GrailBio. 142,000 adults provided 3 blood samples over 2 years, for prevalent (baseline) & incident cancers. The trial did not meet its primary endpoint. Lots more data below: https://t.co/bgWQqTwiA
MCED testing in pop screening NHS-Galleri-RCT of 142,250 pts ➡️blood up to 3 visits ➡️aim to reduce stage III/IV Med FU 17mths ❎primary endpt not met-but ⬇️in stg4 ➡️52% PPV; 99.5% specificity ✅4 fold⬆️ in screen detected ca Huge effort from NHS #ASCO26 @ASCO @OncoAlert https://
#ASCO26 @DrSanjayPopat presents AcceleRET-Lung: first line pralsetinib vs chemo +/- IO in RET+ NSCLC with optional crossover. Study terminated by sponsor at 90% accrual. PFS favors pralsetinib 18.7m vs 9m (HR 0.59) with RR 66% vs 42%, DOR 20.6m vs 9.7m https://t.co/W73zzzT0wk
Created 2026-05-01 · Updated 2026-05-31
Edit rule text
[fraud_scandal_reporting_without_systems_analysis]
Learned 4 rejections Active
Exclude posts covering healthcare fraud prosecutions, misconduct scandals, or enforcement actions (e.g., Botox fraud, hospice fraud, UnitedHealth CEO assassination context) that describe incidents without analyzing root cause incentives, regulatory gaps, or systemic reform implications. Scandal reporting without healthcare systems lens should be rejected.
Posts reporting healthcare fraud, enforcement actions, or corporate misconduct without analyzing systemic drivers or policy implications
3 example posts
In what is considered the largest #Botox fraud scheme in the United States, a jury in Los Angeles convicted a California doctor on Tuesday in a $45 million scheme to defraud #Medicare by submitting claims for Botox injections that were never provided and medically unnecessary, ht
Remember when UnitedHealth's CEO was assassinated in 2024? 18 months later, every accusation the killer Luigi Mangione made has been proven RIGHT. UnitedHealth is one of the biggest corporate collapses in recent years. And they deserve it. In December 2024, UnitedHealthcare CE
Vivek Ramaswamy put out a lot of statements today referencing the fraud in Ohio. Swampy is of course, involved in Ohio Medicaid. Datavant, Vivek's spinoff company from Riovant, handles release of information for many hospitals and providers in Ohio. Ohio Medicaid providers, man
Created 2026-05-01 · Updated 2026-05-24
Edit rule text
[biotech_or_robotics_founder_enthusiasm_without_validation]
Learned 4 rejections Active
Exclude posts that celebrate biotech manufacturing advances, robotic applications, or lab-based innovations (CAR-T manufacturing speed, mRNA vaccine production, in vivo gene editing) without demonstrating: (1) actual healthcare system deployment, (2) regulatory approval pathways, (3) reimbursement clarity, or (4) evidence of adoption by healthcare providers. Founder enthusiasm and lab milestones without implementation proof should be rejected.
Posts expressing enthusiasm about biotech breakthroughs, robotics applications, or manufacturing innovations without evidence of healthcare system adoption or clinical validation.
3 example posts
Another humanoid worker has just joined the factory floor Two years later,PUDU Robotics has launched the new-generation PUDU D7, built specifically for real manufacturing scenarios. It can autonomously push carts, handle intra-line transportation, perform delicate operations, h
Traditional ex vivo CAR-T is fundamentally broken. Between the $500K+ manufacturing cost, the 4-week vein-to-vein delay, and the risk of patient T-cell exhaustion during expansion, the blockbuster autologous market is a logistical nightmare. The industry is pivoting to "In Vivo
Humanoid robots, designed to mimic human movement and capabilities, have been making headlines in recent years, from their use as baggage handlers at Japanese airports to Tesla’s big bet on its Optimus humanoid. Market watchers have predicted that the machines will change the ht
Created 2026-05-01 · Updated 2026-06-06
Edit rule text
[tangential_ai_company_product_launch_without_healthcare_specificity]
Learned 4 rejections Active
Exclude posts announcing AI company product launches, partnerships, or integrations (e.g., OpenAI + AWS, AI agent tools) that lack evidence of healthcare-specific use cases, clinical validation, or healthcare operational adoption.
Posts announcing AI company product launches or partnerships that are loosely tied to healthcare or lack validation of healthcare applicability.
3 example posts
[New] from a16z @speedrun: Come for the Agent, Stay for the Network there's a quiet pattern hiding inside the most defensible vertical AI startups right now: the agent is the wedge the network is the moat. here's what I mean: an HVAC tech needs a part today. >>Traditionally:
Thought I would start posting about interesting things happening at AWS. Not a bad day to start.🚀 Today at #WhatsNextWithAWS we announced a big step forward with @OpenAI on Amazon Bedrock: 1. OpenAI models now available 2. Codex for enterprise development 3. Amazon Bedrock Manag
Great example of what Foundry enables: durable, stateful agents that run across time boundaries, orchestrate tools and models, and close the loop with evaluation and improvement over long-running workflows. @jeffhollan https://t.co/v0yWTDIcn3
Created 2026-04-30 · Updated 2026-05-02
Edit rule text
[drug_trial_or_pharma_data_without_systems_context]
Learned 4 rejections Active
Exclude posts that share drug trial topline results, Phase 3 data, efficacy metrics, or FDA acceptance announcements (e.g., survodutide Phase 3 results, GLP-1 weight loss percentages, hepatitis B treatment approvals) unless the post explicitly connects these to healthcare system dynamics, reimbursement, access barriers, or care delivery innovation.
Posts reporting clinical trial results, drug efficacy data, or pharma announcements without analysis of healthcare delivery, payer, regulatory, or system-level implications.
3 example posts
🚨 Survodutide, a weekly subcutaneous dual GLP-1/glucagon agonist, posts new top-line Phase 3 results today 📊 SYNCHRONIZE-1: Adults with overweight/obesity randomized to survodutide vs. placebo for 76 weeks ⚖️ 16.6% weight loss (efficacy estimand) vs. 3.2% in placebo 85.1% http
$NVO $LLY Boehringer Ingelheim and Zealand Pharma report strong phase 3 data for obesity drug survodutide. Patients lost up to 16.6% of body weight over 76 weeks vs. 3.2% for placebo. The drug targets both GLP-1 and glucagon, a combo approach aimed at boosting weight loss.
⚠️ Sacubitril/Valsartan works. So why aren’t we using it? The evidence is undeniable: ↓ CV mortality: 20% (RCT) / 10–38% (RWE) ↓ HF hospitalization: 21% (RCT) / 10–16% (RWE) ↓ All-cause mortality: 15% (RCT) / 10–25% (RWE) Plus: reverse remodeling, less MR, better QoL & https:
Created 2026-04-30 · Updated 2026-04-30
Edit rule text
[biotech_founder_or_startup_enthusiasm_without_validation]
Learned 4 rejections Active
Exclude posts that express founder enthusiasm, startup momentum, or platform capability claims (e.g., 'it's as easy to start biotech on GCL as software on AWS,' 'forward deploy to customers') without providing validation, evidence, or analysis of actual healthcare adoption or impact.
Posts expressing founder enthusiasm or startup optimism about biotech, cloud labs, or healthcare platforms without evidence or grounded analysis.
3 example posts
Nothing beats running @ginkgo cloud lab for happy customers! Not going to stop until it’s as easy to start a biotech startup on GCL as it is to start a software startup on AWS. https://t.co/GDEyT1pI7F
Almost all of my positions selling some kind of AI/agentic SaaS tool have (either by foresight or customer demand) pivoted to some kind of business model where they “forward deploy” to the customer first and then sell the system they create back to them as SaaS. 99% of “normie” b
Started with standard ChatGPT for clinicians asking for a differential for a GI bleed patient. Then I went into agent mode to have it put together a one pager for the family explaining everything. Of course, this is not a real patient. https://t.co/PEUeCqizT1
Created 2026-04-30 · Updated 2026-04-30
Edit rule text
[broad_unvalidated_health_claim_without_systems_analysis]
Learned 4 rejections Active
Exclude posts that assert broad health claims (e.g., 'GLP-1s improve cardiovascular outcomes,' 'protein absorption limits,' 'AI can now design antibodies') or debate clinical efficacy/safety without rigorous evidence, nuance, or analysis of healthcare system implications.
Posts making broad health or clinical claims or engaging in clinical debates without evidence, nuance, or healthcare systems perspective.
3 example posts
AI can now design antibodies that bind with atomic precision, but not ones that cells can produce. Our preprint closes this gap, delivering a structural principle, an AI-guided rescue pipeline, and adalimumab variants with 20-100x in vivo potency. https://t.co/GvfgHA5EcU
Attention PK nerds, pharmacologists, and clinicians who actually understand serum levels: I haven’t seen this discussed, but it could matter for patients priced out of injectables. If a 25 mg oral semaglutide tablet has ~1% bioavailability, that’s ~0.25 mg systemically… on
U.S. nursing homes are fabricating schizophrenia diagnoses to hide their use of dangerous antipsychotic drugs to subdue dementia patients, a government watchdog report found. The drugs increase the risk of falls, strokes and death. https://t.co/6SkzWxZfSz
Created 2026-04-30 · Updated 2026-04-30
Edit rule text
[ai_company_product_or_capability_announcement]
Learned 4 rejections Active
Exclude posts that announce new AI models, SDKs, or features from tech companies (OpenAI, Anthropic, etc.) without demonstrating concrete healthcare application, clinical validation, or operational impact. This includes SDK releases, API updates, and capability claims.
Posts promoting AI company product launches, model releases, or capability announcements without healthcare-specific validation or application examples.
3 example posts
Almost everyone is building agent harness systems the wrong way. The default move: pick LangChain or LangGraph or the OpenAI Agents SDK, accept the loop, the tools, the memory, the orchestration, the policy engine, the credential store, the budget tracker, all of it, as one deci
I think the AI superapps will soon own 90% of the agentic layer more and more people won't use hermes/openclaw etc because Claude Cowork/Codex will offer 90% of the functionality with 5% of the friction
Anthropic just shipped opus 4.8 and called it a modest upgrade modest is the engineer who used to need babysitting now finishing your whole codebase migration in one session while you sleep so here is the part nobody connects the model is no longer the bottleneck. the platform
Created 2026-04-29 · Updated 2026-06-03
Edit rule text
[broad_unvalidated_health_claim_or_debate_without_evidence]
Learned 4 rejections Active
Exclude posts that assert broad health claims (e.g., 'GLP-1s exacerbate eating disorders,' 'GLP-1s have improved cardiovascular outcomes,' 'protein per meal is wasted') or debate drug/treatment efficacy without citing peer-reviewed evidence, randomized trial results, or systematic analysis.
Posts making broad health claims or debating drug safety/efficacy without citing evidence, trial data, or systematic analysis.
3 example posts
⚠️ Sacubitril/Valsartan works. So why aren’t we using it? The evidence is undeniable: ↓ CV mortality: 20% (RCT) / 10–38% (RWE) ↓ HF hospitalization: 21% (RCT) / 10–16% (RWE) ↓ All-cause mortality: 15% (RCT) / 10–25% (RWE) Plus: reverse remodeling, less MR, better QoL & https:
"There is also compelling preliminary evidence suggesting that the use of these drugs [GLP-1] could exacerbate and lead to new diagnoses of restrictive eating disorders, including anorexia nervosa." @NEJM today https://t.co/swx1y25F1X https://t.co/bQYlpuFBjM
A 65% cholesterol reduction has been available since 2015. Almost nobody could get it. The drug required a needle every two weeks, cost $5,850+/year, and insurers fought every prescription. @Merck spent a decade figuring out how to put the same mechanism in a pill. Enlicitide:
Created 2026-04-29 · Updated 2026-04-30
Edit rule text
[glp1_peptide_market_pricing_and_competition]
Learned 4 rejections Active
Exclude posts that focus on GLP-1/peptide market pricing, competitive dynamics, app store rankings, or commercial adoption metrics (e.g., 'HIMS up to #4 on app store', 'Novo keeps charging brand premium') unless the post includes substantive analysis of healthcare system impact, clinical outcomes, or access equity. Market positioning and pricing alone are not sufficient.
Posts about GLP-1 and peptide market dynamics, pricing, competitive positioning, or app store rankings without healthcare systems analysis.
3 example posts
$HIMS up to #4 in the Canadian App Store medical category. That is not a stock chart. That is real consumer demand showing up in public. GLP-1 drama, Novo pressure, pricing debate, and now app-store strength. Bears keep arguing the story is slowing down, but the product demand
For years, grey-market HGH operated in a strange legal grey zone somewhere between “research compound,” “API” and underground enhancement culture. Last month’s FDA guidance suggests regulators may no longer see it that way. The recent decline in HGH quality, and now the growing
Think about this. The $50 per month out-of-pocket cost of GLP-1 drugs for Americans (with insurer paying the balance of $2000 pm) is more than double the entire cost for Indians (less than ₹2000pm for Semaglutide). Overall cost difference? 100 times. Not 100%.
Created 2026-04-29 · Updated 2026-05-26
Edit rule text
[broad_health_claim_or_debate_without_systems_analysis]
Learned 4 rejections Active
Exclude posts that assert broad health claims (e.g., 'reform will fix healthcare,' 'AI solves labor shortage') or debate healthcare market dynamics without grounding in specific regulatory structures, economic incentives, incumbent defenses, or healthcare-specific implementation barriers.
Posts making sweeping healthcare claims or debating broad health topics without analysis of systemic incentives, business models, or regulatory constraints.
3 example posts
Incumbents don't defend broken systems. They defend profitable ones. Healthcare in America isn't a market failure. It's a sustaining innovation problem disguised as a social crisis. Every "reform" strengthens the existing architecture because the people writing the rules are the
Europe is not losing to the US because it lacks engineers, capital or universities. It is losing because every layer of its system rewards the gatekeeper and punishes the builder. In medicine I see it weekly. A primary care AI that triages, documents and flags risk gets deployed
Statin guidelines over 50 years: 1970s: "Total cholesterol over 280 is high." Result: a relatively small market. 1988: "Actually, 240 is high." Result: millions more added overnight. 2001: "Actually, focus on LDL. Get it under 100." Result: tens of millions of new customers.
Created 2026-04-29 · Updated 2026-05-11
Edit rule text
[tangential_non_healthcare_infrastructure]
Learned 4 rejections Active
Exclude posts focused on non-healthcare domains (energy policy, semiconductor manufacturing, space technology, computing infrastructure, financial markets) that attempt to connect to healthcare only through speculative or indirect reasoning. The post must directly analyze healthcare delivery, operations, or systems implications.
Posts about energy, compute, space infrastructure, or tech industry dynamics that only tangentially relate to healthcare through loose framing.
3 example posts
At our recent Energy and the AI Age summit, Hon. Bernard L. McNamee, former Commissioner of the U.S. Federal Energy Regulatory Commission, comments on why energy matters in discussions about AI: “Energy is the foundation of our entire economy. Energy makes up about 7% of the htt
Dylan Patel, founder of SemiAnalysis: "The upper bound on how much compute can be produced by 2030 is around 200 gigawatts a year." The entire world has about 20 gigawatts of AI deployed right now. The ceiling is 10x what exists today, and it still isn't enough to feed what htt
An interesting proposal for restoring human versus machine chess competition was to limit the machine to the same energy expenditure as the human brain. A worthy challenge, considering the amount of power required for AI data centers!
Created 2026-04-29 · Updated 2026-07-05
Edit rule text
[workforce_labor_disruption_macro_without_healthcare_specificity]
Learned 4 rejections Active
Exclude posts discussing labor market disruption, wage trends, job automation, or workforce economics that mention healthcare (Klarna, customer service agents, Fig robots) as an example of broader labor macro trends rather than analyzing specific healthcare workforce gaps, clinical roles, or care delivery models.
Posts about labor market impacts, job displacement, or workforce trends using healthcare as loose framing for broader macro commentary.
3 example posts
AI use for any business function is concentrated in selected sectors and large firms and primarily for augmentation of worker tasks rather than substitution, from Kathryn Bonney, Cory L. Breaux, Emin Dinlersoz, @Lucia_Econ, @JHaltiwanger_UM, and Aditya A. Pande https://t.co/qZ8Vf
Klarna replaced 700 customer service agents with AI. Resolution time dropped from 11 minutes to 2. Then quality collapsed on complex cases. The CEO admitted cost became too dominant. Customers wanted humans. Klarna is now hiring people again. 55% of firms that replaced
A humanoid robot just unloaded a dishwasher for 67 hours straight. By end of 2026, the same robot will perform surgery. That's not a prediction from a sci-fi writer. That's Brett Adcock. CEO of Figure AI. Said it on camera this month. Here's what makes this insane. Two years
Created 2026-04-28 · Updated 2026-05-10
Edit rule text
[off_topic_non_healthcare_content_with_loose_healthcare_framing]
Learned 4 rejections Active
Exclude posts about non-healthcare industries, geopolitics, labor market trends, or general business/tech topics that mention healthcare only as a tangential reference or forced label. The post must be fundamentally about healthcare systems, not merely tagged with healthcare terms.
Posts about non-healthcare domains (space, labor markets, software business models, UFOs) that lack substantive connection to healthcare systems.
3 example posts
Using 180 years of US data to show that tariff increases reduce imports, output, and manufacturing, with effects operating through both supply and demand channels, from Tamar den Besten, Regis Barnichon, @drkaenzig, and Aayush Singh https://t.co/R1kdtfi3fO https://t.co/wvXj31qs9e
Software is not a moat Over the last 15+ years, nearly every innovation @EvanSpiegel and his team shipped got copied. Stories. AR glasses. Swipe-based navigation. The camera-first interface. And yet @Snapchat is the only independent consumer social app that has lasted. Nearly 1
AI could, in theory, automate 57% of US work hours. Yet most human skills remain relevant. The future of work is not human or machine – but a partnership between people, agents, and robots. Read our latest research on skill partnerships in the age of AI: https://t.co/h1K56uPqPo
Created 2026-04-28 · Updated 2026-04-28
Edit rule text
[glp1_peptide_market_dynamics_without_healthcare_systems_lens]
Learned 4 rejections Active
Exclude posts that discuss GLP-1, semaglutide, retatrutide, survodutide, or peptide markets solely through pricing dynamics, patient weight loss percentages, script growth rates, or personal weight-loss narratives without addressing healthcare system implications (insurance coverage, clinical guidelines, access equity, clinical workflow integration).
Posts about GLP-1 drugs, peptides, or weight-loss medications focused on market pricing, competition, and personal usage without healthcare system analysis.
3 example posts
🚨 Survodutide, a weekly subcutaneous dual GLP-1/glucagon agonist, posts new top-line Phase 3 results today 📊 SYNCHRONIZE-1: Adults with overweight/obesity randomized to survodutide vs. placebo for 76 weeks ⚖️ 16.6% weight loss (efficacy estimand) vs. 3.2% in placebo 85.1% http
$NVO $LLY Boehringer Ingelheim and Zealand Pharma report strong phase 3 data for obesity drug survodutide. Patients lost up to 16.6% of body weight over 76 weeks vs. 3.2% for placebo. The drug targets both GLP-1 and glucagon, a combo approach aimed at boosting weight loss.
This is just two GLP-1s, one peptide, one use case what happens when off-label prescribing ramps up what happens when retatrutide hits the market what happens when other peptides become compoundable chapter one
Created 2026-04-28 · Updated 2026-05-03
Edit rule text
[off_topic_non_healthcare_domain_with_loose_framing]
Learned 4 rejections Active
Exclude posts whose primary subject is fintech, general software business dynamics, or AI company infrastructure (e.g., Revolut banking ML, Snap Stories/AR, Microsoft Word feature, software seat models) even if they mention healthcare or use healthcare as an analogy. The post must center on healthcare systems, not use healthcare as a secondary reference point.
Posts about non-healthcare domains (fintech, software business models, general AI infrastructure) that use loose healthcare framing or tangential healthcare references.
3 example posts
The problem with Reality Labs is not ambition. It is time. AI turned into revenue faster because it improves existing workflows. The metaverse still asks users to change behavior before value is obvious. https://t.co/lopZiUwGU5
Microsoft just turned an $11 billion startup into a Word feature. Harvey raised $200M at an $11B valuation in March on the bet that legal AI is its own surface. The numbers held that up. $190M ARR per TechCrunch's December reporting. 100,000 lawyers across 1,300 organizations in
Revolut just moved the IP of banking into a model. Trained on 24 billion banking events in 111 countries. One foundation model replacing six separate ML systems. Credit scoring: +130% Fraud recall: +65% Marketing engagement: +79% The model is the new moat.
Created 2026-04-27 · Updated 2026-05-03
Edit rule text
[tangential_biotech_or_startup_founder_enthusiasm_without_evidence]
Learned 4 rejections Active
Exclude posts that are primarily personal founder/CEO statements about internal progress ('we reduced seats from 10+ to 2,' 'we're not going to stop until it's easy to start a biotech startup on GCL') or self-promotional tool announcements without evidence of clinical validation, healthcare adoption, or healthcare systems change.
Posts expressing founder optimism or startup progress announcements without substantive evidence of healthcare impact or system-level significance.
2 example posts
Nothing beats running @ginkgo cloud lab for happy customers! Not going to stop until it’s as easy to start a biotech startup on GCL as it is to start a software startup on AWS. https://t.co/GDEyT1pI7F
Almost all of my positions selling some kind of AI/agentic SaaS tool have (either by foresight or customer demand) pivoted to some kind of business model where they “forward deploy” to the customer first and then sell the system they create back to them as SaaS. 99% of “normie” b
Created 2026-04-27 · Updated 2026-04-30
Edit rule text
[ai_agent_technical_capability_tangent]
Learned 4 rejections Active
Exclude posts that showcase AI agent technical features, design patterns, framework announcements, or benchmark performance unless the post explicitly demonstrates healthcare-specific application, clinical validation, or healthcare workflow integration. Generic AI agent capability posts with loose healthcare framing do not qualify.
Posts highlighting AI agent technical capabilities, architecture, or performance metrics without healthcare-specific application or clinical validation.
3 example posts
[New] from a16z @speedrun: Come for the Agent, Stay for the Network there's a quiet pattern hiding inside the most defensible vertical AI startups right now: the agent is the wedge the network is the moat. here's what I mean: an HVAC tech needs a part today. >>Traditionally:
Great example of what Foundry enables: durable, stateful agents that run across time boundaries, orchestrate tools and models, and close the loop with evaluation and improvement over long-running workflows. @jeffhollan https://t.co/v0yWTDIcn3
Is the business model for traditional software companies in permanent decline due to AI Agents not needing seats? 2 examples: Re: @salesforce, we’ve reduced our seats from 10+ to 2 human seats and 1 API seat. And yet, we now pay $22,000 a year, 83% up from $12,000. Why? Our
Created 2026-04-27 · Updated 2026-05-02
Edit rule text
[tangential_ai_capability_without_healthcare_application]
Learned 4 rejections Active
Exclude posts that discuss AI model capabilities (e.g., GPT-5.5 taking over corporate networks, Claude Code architecture, zero-day vulnerabilities) that are tangentially labeled as healthcare-relevant but focus on technical AI capability, not on how the capability solves a healthcare delivery, patient outcomes, or healthcare system problem.
Posts about AI model technical capabilities, safety, or vulnerabilities framed loosely as healthcare-relevant but lacking concrete healthcare application.
2 example posts
A researcher gave an AI agent access to his shell, his files, and his network. Then he proved that every safety guardrail we trust is architecturally useless. It cannot tell the difference between your instructions and a hacker's. The paper is called Parallax: Why AI Agents htt
This is by far the most important result of the entire GPT-5.5 release: In a cyber evaluation GPT-5.5 was able to take over a simulated corporate network in 1/10 trials with a budget of 100M tokens. Previously, the only model that was able to solve this task was Claude Mythos,
Created 2026-04-27 · Updated 2026-04-27
Edit rule text
[ai_agent_capability_tangent]
Learned 4 rejections Active
Exclude posts that focus on AI agent architecture (e.g., Claude Code design, agentic SaaS patterns, safety guardrails, cyber vulnerabilities) when the healthcare connection is tangential or speculative, and the post does not analyze how these capabilities change clinical decision-making or health system operations.
Posts about AI agent technical capabilities, architectural design, or safety vulnerabilities without healthcare-specific application.
3 example posts
A must read for anyone interested in building practical AI systems in 2026: Dive into Claude Code: The Design Space of Today's and Future AI Agent Systems The paper explains the architecture of a modern production-grade AI agent system (Claude Code) by analyzing its source http
Almost all of my positions selling some kind of AI/agentic SaaS tool have (either by foresight or customer demand) pivoted to some kind of business model where they “forward deploy” to the customer first and then sell the system they create back to them as SaaS. 99% of “normie” b
🚨 Anthropic's own team just showed how to build production AI agents. 30 minutes. free. from the engineers who built it. watch the workshop. bookmark it. you spent 6 months managing every workflow yourself. they just showed how to put all of it on autopilot. Then read the ht
Created 2026-04-26 · Updated 2026-04-27
Edit rule text
[ai_safety_and_security_tangential]
Learned 4 rejections Active
Exclude posts that focus on AI security vulnerabilities, zero-day exploits, or agent jailbreaks (e.g., AI taking over networks, breaking safety guardrails) without tying the concern to a specific healthcare delivery, clinical decision, or patient outcome scenario.
Posts about AI system vulnerabilities, cyberattacks, or safety guardrails that lack healthcare-specific application or analysis.
2 example posts
A researcher gave an AI agent access to his shell, his files, and his network. Then he proved that every safety guardrail we trust is architecturally useless. It cannot tell the difference between your instructions and a hacker's. The paper is called Parallax: Why AI Agents htt
This is by far the most important result of the entire GPT-5.5 release: In a cyber evaluation GPT-5.5 was able to take over a simulated corporate network in 1/10 trials with a budget of 100M tokens. Previously, the only model that was able to solve this task was Claude Mythos,
Created 2026-04-26 · Updated 2026-04-26
Edit rule text
[broad_unsubstantiated_health_claims_without_nuance]
Learned 4 rejections Active
Exclude posts that dismiss or broadly counter established medical concerns or trial data with vague assertions ('it's just a narrative') or that make unsubstantiated claims about drug safety/efficacy without citing peer-reviewed evidence or acknowledging legitimate scientific debate.
Posts making sweeping health claims or dismissing medical concerns (like GLP-1 cardiac risk narratives) as 'just narratives' without evidentiary grounding.
1 example post
The only problem with the GLP-1 heart muscle loss narrative is... ... that it's just a narrative. GLP-1s have reliably improved cardiovascular outcomes in trials, to the point that some research suggests benefit may even be independent of (not reliant on) weight loss.
Created 2026-04-24 · Updated 2026-04-25
Edit rule text
[political_figures_healthcare_claim_amplification]
Learned 4 rejections Active
Exclude posts that amplify healthcare claims made by political figures (RFK Jr., senators, HHS officials) without independent verification, healthcare systems analysis, or evidence of actual healthcare delivery impact. Posts must analyze policy substance, not echo political announcements.
Posts amplifying healthcare policy claims or victories by political figures without independent analysis or evidence of healthcare systems impact.
3 example posts
What $1 Billion a Day Buys in American Health Care The U.S. is spending $1 billion/day on the war in Iran — over a year, that would cover 37 million Medicaid enrollees. Congress just cut $911 billion from the program because it was too expensive. Read & subscribe (for free!)
RFK Jr. calls out Democrat House representatives to their face for ignoring chronic disease while claiming to care about public health. “The Congressman was talking about the deaths from infectious disease, which are a couple thousand a year.” “90% of the people who die in this
🚨BREAKING: HHS Sec. RFK Jr. just announced President Trump has SAVED and FOUND 138,000 missing children lost under Biden. "Many have been trafficked, undergone slavery, s*xual abuse." Follow: @BoLoudon https://t.co/p6YKEm38T7
Created 2026-04-24 · Updated 2026-04-28
Edit rule text
[broad_health_claims_without_nuance_or_evidence]
Learned 4 rejections Active
Exclude posts that assert broad health or market claims ('AI agents will eliminate PCP visits,' 'robots will solve labor shortage') without empirical evidence, pilot data, or specific examples. Broad proclamations without grounding belong in speculation, not healthcare tech analysis.
Posts making sweeping claims about health trends, AI capabilities, or healthcare markets without substantive evidence, nuance, or demonstrated validation.
3 example posts
🚀🤖 In 2026, autonomous #AI agents aren’t failing because of technology ➜ they’re failing because of responsibility. ⚙️ Everyone knows how to build agents. Very few know how to own their decisions. 💡 Real autonomous AI isn’t about freedom or intelligence. It’s about clear scope,
The inability of AI systems to act as their own deployment consultants, process mappers, and change management experts is what makes AI use in enterprises so “normal” - the tools are powerful, but you need a lot more to transform enterprises. Possible to imagine that changing .
Europe is not losing to the US because it lacks engineers, capital or universities. It is losing because every layer of its system rewards the gatekeeper and punishes the builder. In medicine I see it weekly. A primary care AI that triages, documents and flags risk gets deployed
Created 2026-04-23 · Updated 2026-05-11
Edit rule text
[unsubstantiated_fringe_medical_claims]
Learned 4 rejections Active
Exclude posts that make sweeping medical claims (e.g., about protein absorption, vaccine safety tests, smart TV health risks, vaccine ingredients) without citing peer-reviewed studies, proper methodology context, or acknowledging limitations and counterevidence. Include claims about unvalidated treatments, drug dosing myths, or sensationalized health risks.
Posts making bold medical claims without peer-reviewed evidence, proper context, or clinical validation.
3 example posts
"Your body can only use 25-30g of protein per meal. Anything above that gets wasted." This claim has been repeated in fitness nutrition for over a decade, and it was built on studies that measured the right thing over the wrong timescale. Moore 2009 gave six young men 0, 5, ht
As far as I know this is the only naturally-derived, classical psychedelic, that has killed people. Ayahuasca has some deaths, but it's unclear what the cause was, and unlikely directly related to its cardiovascular risk profile. https://t.co/DatuHiBOTX
🚨BREAKING: A peer reviewed study just confirmed your smart TV is taking screenshots of your screen every 15 seconds and sending them to company servers. Samsung every minute. LG every 15 seconds. Running even when you are using it as a monitor. Here is how to stop it:
Created 2026-04-23 · Updated 2026-04-23
Edit rule text
[geopolitical_military_conflict_coverage]
Learned 4 rejections Active
Exclude posts that cover geopolitical conflicts, ceasefires, military announcements, or defense policy (e.g., Lebanon-Israel ceasefire, Ukraine war votes) unless they directly analyze healthcare system impacts, displaced population health, or pandemic/epidemic consequences.
Posts reporting on geopolitical tensions, military conflicts, ceasefires, or defense policy without healthcare or public health angle.
3 example posts
President Donald J. Trump announces a 10-day ceasefire between Lebanon and Israel. "It has been my Honor to solve 9 Wars across the World, and this will be my 10th, so let's, GET IT DONE!" https://t.co/YujXwyUReM
"I will be inviting the Prime Minister of Israel, Bibi Netanyahu, and the President of Lebanon, Joseph Aoun, to the White House... Both sides want to see PEACE, and I believe that will happen, quickly!" - President Donald J. Trump 🇺🇸 https://t.co/KFipIMmFOD
Breaking news: President Trump announced a pause in fighting in Lebanon. Lebanon and Israel “agreed that in order to achieve PEACE between their Countries, they will formally begin a 10 Day CEASEFIRE,” Trump said in a social media post. https://t.co/QHKR5ewrMN
Created 2026-04-23 · Updated 2026-04-24
Edit rule text
[workforce_disruption_macro_labor_without_healthcare_specificity]
Learned 4 rejections Active
Exclude posts that discuss workforce disruption, automation of tasks, or labor market impact of AI using broad claims or generalist data (e.g., '57% of US work hours could be automated') without specific analysis of healthcare workforce roles, clinical workflows, or health system employment impacts.
Posts about AI-driven job displacement, task-level automation, or labor market effects presented as generalist macro commentary without healthcare-specific analysis.
3 example posts
🚨 BREAKING: Anthropic new research finds that AI’s impact on jobs is primarily at the task level. Rather than eliminating jobs, it is progressively taking over the functions that define them and gradually absorbing the core work in many jobs/roles. The paper, “Labor Market http
AI could, in theory, automate 57% of US work hours. Yet most human skills remain relevant. The future of work is not human or machine – but a partnership between people, agents, and robots. Read our latest research on skill partnerships in the age of AI: https://t.co/h1K56uPqPo
Is the business model for traditional software companies in permanent decline due to AI Agents not needing seats? 2 examples: Re: @salesforce, we’ve reduced our seats from 10+ to 2 human seats and 1 API seat. And yet, we now pay $22,000 a year, 83% up from $12,000. Why? Our
Created 2026-04-23 · Updated 2026-04-29
Edit rule text
[speculative_sci_fi_or_robotics_applications]
Learned 4 rejections Active
Exclude posts about humanoid robots, robot competitions, speculative sci-fi applications, or robotics breakthroughs (e.g., half-marathon times, Tesla Optimus production)—unless the post demonstrates a concrete, deployed healthcare use case or clinical validation.
Posts about humanoid robots, speculative future applications, or non-healthcare robotics competitions that tangentially reference healthcare.
2 example posts
Elon Musk says Optimus could start being useful outside Tesla as soon as next year. $TSLA is ramping production, building a second Optimus factory at Giga Texas and plans to unveil the V3 design around mid-year. https://t.co/Pfvs1ctFvi
In Beijing's 2026 humanoid robot half-marathon, HONOR's Lightning completed the 21 km course in 50:26 minute. Beat current human men's half-marathon world record of 57:20. Last year's winner took over 2 hours 40 minutes. Massive progress in 12 month https://t.co/OcZJ66ebWD
Created 2026-04-22 · Updated 2026-04-23
Edit rule text
[political_outrage_or_scandal_without_healthcare_systems_analysis]
Learned 4 rejections Active
Exclude posts that center on political figures, criminal acts, regulatory scandals, or moral outrage about healthcare without explaining the systems-level healthcare technology implications. The post must analyze healthcare operations or tech outcomes, not use healthcare as a framing for political commentary.
Posts that use healthcare-adjacent topics as a vehicle for political outrage or scandal narrative without substantive systems analysis.
3 example posts
Remember when UnitedHealth's CEO was assassinated in 2024? 18 months later, every accusation the killer Luigi Mangione made has been proven RIGHT. UnitedHealth is one of the biggest corporate collapses in recent years. And they deserve it. In December 2024, UnitedHealthcare CE
600 new generic drugs added to TrumpRx this week thanks to President @realDonaldTrump's leadership! Before you visit your local pharmacy, visit https://t.co/dDMdsILHTj to ensure you're getting the BEST price for your prescriptions. It's easy and saves YOU money! https://t.co/
Vivek Ramaswamy put out a lot of statements today referencing the fraud in Ohio. Swampy is of course, involved in Ohio Medicaid. Datavant, Vivek's spinoff company from Riovant, handles release of information for many hospitals and providers in Ohio. Ohio Medicaid providers, man
Created 2026-04-21 · Updated 2026-05-21
Edit rule text
[infrastructure_and_compute_hype_no_healthcare_application]
Learned 4 rejections Active
Exclude posts that discuss energy, grid equipment, data center growth, chip manufacturing, power supply, or compute infrastructure as macro trends without demonstrating how these challenges impact healthcare delivery, clinical operations, or health tech deployment.
Posts about energy, data centers, chips, and grid infrastructure without concrete healthcare application
2 example posts
Since I began work on AI in 2010, training compute for frontier models has grown by one trillion times. Now we're looking at something like another thousand-fold growth in effective compute by the end of 2028. 1000x the existing 1,000,000,000,000x. Extraordinary stuff.
Follow the bottleneck. Chips → data centers → grid equipment → power → gas turbines Grid equipment grew 1%/yr for decades. Then data centers showed up as an entirely new buyer. Gas turbine makers shipped 5–7 GW/yr. Last year? Orders hit 100 GW. @maxlbcook on how he https://t.
Created 2026-04-21 · Updated 2026-04-27
Edit rule text
[ai_company_product_metrics_and_business_news]
Learned 4 rejections Active
Exclude posts that announce AI company products, hiring initiatives, or business developments (e.g., new roles, product launches, valuation news) unless the post explicitly demonstrates deployment in a healthcare workflow or validates clinical/operational benefit. Posts about Anthropic, OpenAI, or other AI vendors' internal business moves are out of scope.
Posts reporting AI company product launches, funding, hiring, or business metrics without demonstrating healthcare system application or clinical impact.
3 example posts
Stripe just created a role that didn't exist 12 months ago (and they're paying multiple six figures for it) It's called the Forward Deployed AI Accelerator. They are hiring AI-native individuals to work directly with their marketing teams to fundamentally change how they work.
ai-mageddon may prove more severe than saas-mageddon. the moats around cool interfaces and system prompts are far inferior to: social/team graphs, network effects, systems of record, permissioning admin tools, collaboration…and the list goes on. winners will have deep roots
A Microsoft acabou de transformar uma startup de $11 bilhões de dólares em uma funcionalidade do Word. Não foi uma aquisição nem uma parceria. Uma funcionalidade. A Harvey levantou $200M a uma valuation de $11B em março. $190M de receita recorrente anual. 100 mil advogados. Co
Created 2026-04-21 · Updated 2026-05-08
Edit rule text
[workforce_disruption_macro_no_healthcare_context]
Learned 4 rejections Active
Exclude posts reporting general labor market disruption statistics, employment decline across age cohorts, or macro job displacement trends from AI adoption — unless the analysis connects directly to healthcare workforce roles, hiring, or operational changes in healthcare settings.
Posts about AI job displacement and labor market trends without healthcare-specific workforce analysis
3 example posts
A major milestone just landed quietly: for the first time ever, half of all employed Americans use AI at work. Gallup's Q1 2026 survey of nearly 24,000 workers shows that adoption has more than doubled since 2023, when only 21% reported any AI use. https://t.co/jmQga9tbWT
I think we now have real evidence that AI exposure is associated with job decline for age <25. The Canary in the Coalmine paper addresses a lot of concerns. While economic science takes time; now is the time to think about policy responses. @erikbryn @BharatKChandar @RuyuChen
In Beijing's 2026 humanoid robot half-marathon, HONOR's Lightning completed the 21 km course in 50:26 minute. Beat current human men's half-marathon world record of 57:20. Last year's winner took over 2 hours 40 minutes. Massive progress in 12 month https://t.co/OcZJ66ebWD
Created 2026-04-20 · Updated 2026-04-21
Edit rule text
[healthcare_fraud_or_scandal_reporting_without_systems_analysis]
Learned 4 rejections Active
Exclude posts that report on healthcare fraud, waste, or financial scandals (PBM rebates, Medicaid fraud schemes) without analyzing root causes in healthcare business model design or proposing systems-level solutions.
Posts about Medicaid fraud, PBM corruption, or healthcare financial scandals that describe the problem but lack structural/systems-level insight
3 example posts
PBM owns the pharmacy. PBM owns the plan. PBM adjudicates the claim. Three commissions on one pill. The employer pays all three and gets a thank-you note for the rebate.
Michigan Medicine generated $8.7 billion in revenue in FY25. It posted a $109 million operating margin in the same year. A nonprofit health system underwriting its largest pavilion in history off margin, not philanthropy. The accounting category and the operating reality do not
Incumbents don't defend broken systems. They defend profitable ones. Healthcare in America isn't a market failure. It's a sustaining innovation problem disguised as a social crisis. Every "reform" strengthens the existing architecture because the people writing the rules are the
Created 2026-04-19 · Updated 2026-05-14
Edit rule text
[workforce_disruption_macro_without_healthcare_context]
Learned 4 rejections Active
Exclude posts that cite labor market studies, employment statistics, or job displacement trends related to AI exposure—unless the post specifically analyzes impact on healthcare workforce (nurses, physicians, billing staff, etc.). Generic labor market analysis with AI + healthcare framing does not qualify.
Posts about AI's impact on employment, labor market trends, and job displacement presented as macro economic commentary without healthcare-specific workforce analysis.
3 example posts
A major milestone just landed quietly: for the first time ever, half of all employed Americans use AI at work. Gallup's Q1 2026 survey of nearly 24,000 workers shows that adoption has more than doubled since 2023, when only 21% reported any AI use. https://t.co/jmQga9tbWT
I think we now have real evidence that AI exposure is associated with job decline for age <25. The Canary in the Coalmine paper addresses a lot of concerns. While economic science takes time; now is the time to think about policy responses. @erikbryn @BharatKChandar @RuyuChen
Among workers ages 22–25, employment in the most AI-exposed occupations has fallen roughly 16% relative to the least-exposed. This is after controlling for firm-type effects, which isolate AI exposure from broader shocks like interest rate pressure or sector slowdowns. The gap ht
Created 2026-04-18 · Updated 2026-04-21
Edit rule text
[glp1_peptide_and_weight_loss_macro_or_personal_framing]
Learned 4 rejections Active
Exclude posts that discuss GLP-1 drugs, peptides, or weight loss through personal anecdotes, pricing controversy, supply chain gossip, or macro economic framing (e.g., 'Hims pricing,' personal side effects, regulatory classification debates) without substantive healthcare delivery, reimbursement, or clinical implementation analysis.
Posts about GLP-1 drugs, peptides, or weight loss in macro-economic or personal anecdote framing without healthcare delivery systems context
3 example posts
Peptide synthesis is one of the hardest things to do right Semaglutide comes out correct only 55% of the time. BPC-157 ~74%. every amino acid compounds the error China won this because they have the scale to throw most of it away We need to be building this capacity in the US
She's right. The safety risk was never the peptides. It was the supply chain. Regulated compounding access fixes the exact problems people are worried about. Heavy metals, contamination, underdosed vials.
And there it is. Within hours of RFK's announcement someone is already pricing out how much Hims can charge for compounds the research community has had access to for a fraction of that cost. This is why the outcome of these PCAC meetings matters more than the announcement.
Created 2026-04-17 · Updated 2026-04-22
Edit rule text
[tangential_ai_model_capability_hype]
Learned 4 rejections Active
Exclude posts announcing AI model speed upgrades, new model releases, code generation performance, or benchmark wins (e.g., 'Lightning-Boltz 28s runtime', 'full 32-step network takeover') unless they demonstrate a specific healthcare delivery or diagnostic improvement.
Posts about AI model technical capabilities, speed improvements, or benchmark performance presented as AI industry news rather than healthcare application breakthroughs.
3 example posts
Boltz-2 just got a major speed upgrade. 🚀 We’re releasing Lightning-Boltz, a local, GPU-accelerated framework free from public MSA server bottlenecks.⚡ On a single L40S, total runtime drops to 28s per input vs 89s with the rate-limited server and 298s with MMseqs-CPU. 1/5 🧵 ht
FT: The White House is moving to give major US agencies access to a modified Anthropic Mythos model built to hunt dangerous software flaws before attackers find them. That makes Mythos useful for defense because a model that can find a weakness in an operating system, browser, h
AI is letting developers ship three to four times faster. It is also flooding codebases with vulnerabilities at the same rate. Aikido Security scans 15 open-source ecosystems for malware. A year ago: 30,000 packages per day. Now: 100,000. The attack surface is not growing https
Created 2026-04-17 · Updated 2026-04-23
Edit rule text
[glp1_peptide_and_weight_loss_adjacent_macro_or_personal]
Learned 4 rejections Active
Exclude posts about GLP-1 drugs, peptides, or weight loss that analyze only personal health experiences, side effects, drug pricing economics, or biological mechanisms without connecting to healthcare delivery, access, or system-wide implications for patient care.
Posts about GLP-1 drugs, peptides, obesity, or weight loss that focus on market economics, personal anecdotes, or biology without healthcare system implications.
3 example posts
I have now received nine reports from people taking GLP-1 drugs who got the same side effect: They no longer feel normal when they come off. "I feel hangry again", "I started thinking about hunger and I hate it", "I have to go back to Adderall". 8/9 reports -> from women.
The brain is the master regulator of food intake and energy balance. A brilliant new @CellCellPress review, including the mechanism of GLP-1 drugs, by @ClemmensenC and colleagues, open-access https://t.co/KbDf7ym288 https://t.co/DMFeS9zHfw
GLP-1 drugs are the ultimate validation of the techno-solutionist approach to society's most challenging problems. The obesity crisis seemed liked it would just get worse and worse forever. Scolding from public health officials didn't work. Proposals to completely overhaul our f
Created 2026-04-17 · Updated 2026-04-17
Edit rule text
[ai_model_capability_technical_tangent]
Learned 4 rejections Active
Exclude posts about AI model technical achievements (reasoning capabilities, task automation, jailbreaking, safety escapes, benchmark performance) unless they directly connect to a specific healthcare workflow, clinical decision, or operational vulnerability.
Posts about AI model technical capabilities, reasoning benchmarks, or autonomous task execution that lack connection to a healthcare operations or clinical decision-making problem.
3 example posts
Researchers gave AI agents a simple choice: hit your performance target or follow the rules. Most of them chose to cheat. McGill University tested 12 of the most powerful AI models on 40 realistic workplace scenarios. Healthcare. Finance. Logistics. Scientific research. Each AI
AI is letting developers ship three to four times faster. It is also flooding codebases with vulnerabilities at the same rate. Aikido Security scans 15 open-source ecosystems for malware. A year ago: 30,000 packages per day. Now: 100,000. The attack surface is not growing https
Two years ago the best AI models couldn't complete beginner-level cyber tasks. One just executed a full 32-step corporate network takeover. The Bank of England is convening emergency CEO briefings. Look at that chart. GPT-4o maxes out at 2 steps. Initial reconnaissance. It can
Created 2026-04-17 · Updated 2026-04-17
Edit rule text
[ai_technical_capability_tangent_loose_healthcare_framing]
Learned 4 rejections Active
Exclude posts about AI model technical capabilities (reasoning depth, in-context learning, fine-tuning methods, alignment techniques, security vulnerabilities) that are presented as general AI advances rather than solutions to specific healthcare delivery, regulatory, or operational challenges.
Posts about AI model capabilities, alignment research, or technical safety without demonstrating relevance to healthcare-specific problems or use cases.
3 example posts
amazing writeup of experiments. essentially a model may already know a capability but that may be stored in a representation that is bad for extraction. you can slightly retrain/condition the model so the same capability becomes easier to isolate. must read for mech interp http
So Mythos was, indeed, not marketing hype. Remember this is a general purpose model that just happens to be good at finding exploits because good models are good at lots of things. Expect similar from OpenAI & Google. And from open models in 8 months. https://t.co/KbhalQYX8
Dario Amodei says Mythos is not limited by compute Anthropic can scale it 3x or 10x without creating a conflict between government and private-sector access The harder problem is who gets it "because giving access to too many organizations could create serious cyber risks" htt
Created 2026-04-16 · Updated 2026-05-11
Edit rule text
[academic_or_clinical_observation_without_healthcare_systems_context]
Learned 4 rejections Active
Exclude posts that present scientific findings, review articles, or medical mechanism discussions without addressing how these insights translate to healthcare system challenges, provider workflows, or patient access barriers that tech could solve.
Posts sharing academic research findings, mechanistic insights, or clinical observations without linking to healthcare delivery or technology adoption
3 example posts
I never met my grandfather. He died of pancreatic cancer when my father was just 19. Today, Yash Bindal, 33, father to 18-month-old Maya, faces the same fate. @PopVaxIndia is using AI to make him a personalized generative medicine to extend his life. https://t.co/O5VIXbmMGd
What scares me about AI: it gets SO good at being almost correct that nobody catches hallucinations…unless they learned the subject before LLMs existed. Eventually, no one will have. Medicine is full of niche sh*t. How much can we manually verify?
good news: it is a specific virus that has a good prognosis - 85%+ of full recovery. thanks everyone who helped me; it is hard to research while immobilized, and I got some things wrong, which you helped clear up. im extremely thankful and hope i can give it back somehow sadl
Created 2026-04-16 · Updated 2026-04-19
Edit rule text
[unvalidated_or_speculative_medical_claims_and_peptides]
Learned 4 rejections Active
Exclude posts about GLP-1 drugs, peptides, or weight-loss medications that rely on anecdotal reports, speculation about off-label uses, or personal accounts of side effects without peer-reviewed evidence or healthcare access/pricing analysis.
Posts about unproven or speculative uses of peptides, GLP-1 drugs, or emerging treatments without rigorous evidence or healthcare system analysis
3 example posts
I never met my grandfather. He died of pancreatic cancer when my father was just 19. Today, Yash Bindal, 33, father to 18-month-old Maya, faces the same fate. @PopVaxIndia is using AI to make him a personalized generative medicine to extend his life. https://t.co/O5VIXbmMGd
I have now received nine reports from people taking GLP-1 drugs who got the same side effect: They no longer feel normal when they come off. "I feel hangry again", "I started thinking about hunger and I hate it", "I have to go back to Adderall". 8/9 reports -> from women.
GLP-1 drugs are the ultimate validation of the techno-solutionist approach to society's most challenging problems. The obesity crisis seemed liked it would just get worse and worse forever. Scolding from public health officials didn't work. Proposals to completely overhaul our f
Created 2026-04-16 · Updated 2026-04-20
Edit rule text
[workforce_disruption_or_labor_market_macro]
Learned 4 rejections Active
Exclude posts discussing workforce automation, job displacement, or labor market trends from AI that lack healthcare-specific workforce analysis, clinical staffing context, or healthcare operational solutions.
Posts analyzing AI's impact on employment, job displacement, and labor market trends in generic or macro terms without healthcare-specific workforce context.
3 example posts
Among workers ages 22–25, employment in the most AI-exposed occupations has fallen roughly 16% relative to the least-exposed. This is after controlling for firm-type effects, which isolate AI exposure from broader shocks like interest rate pressure or sector slowdowns. The gap ht
This reflects what I'm seeing too. AI can do an increasing number of tasks. If your job consists of only those tasks, you're at risk of being completely automated out of that job. If your job consists partially of those tasks, plus other higher value tasks not yet automated,
Fortune: The survey says 29% of workers admit sabotaging company AI plans, and that rises to 44% for Gen Z. Companies are finding that AI rollout is colliding with a basic workplace fact: people resist tools they think will erase their role. That sabotage ranges from ignoring h
Created 2026-04-16 · Updated 2026-04-18
Edit rule text
[healthcare_financial_scandal_or_fraud_without_systems_analysis]
Learned 4 rejections Active
Exclude posts that report healthcare fraud, billing abuse, corporate scandal, or enforcement actions (lawsuits, DOJ cases) without offering systems-level analysis of how healthcare technology, policy, or organizational structure contributed to or could prevent the problem.
Posts reporting healthcare fraud, corporate malfeasance, or financial scandal without analysis of systemic causes or solutions.
3 example posts
🚨 Fraudsters literally looted $250-500 BILLION a year from taxpayers for years, now changes are being made to prevent this fraud: - Treasury is now going after the banks - Whistleblowers can make 30% for exposing fraud - Auto dealers will be tracked down END ALL THE FRAUD. https
🚨 Surgeon @EithanHaim reveals shocking medical fraud scheme: Texas doctors allegedly changing teens' medical records and using fake billing codes to secretly continue banned gender treatments—scamming insurance and taxpayers. He's speaking at a #DetransAwarenessDay @genspect foru
American surgeon exposes US Health Insurance companies latest scam - Doctors submit codes to determine eligibility for care - Health Insurance companies are now saying codes don’t need prior authorizations, but they won’t tell you if it’s covered until AFTER the procedure “What
Created 2026-04-16 · Updated 2026-04-16
Edit rule text
[ai_company_product_launch_without_healthcare_specificity]
Learned 4 rejections Active
Exclude posts that announce new AI model releases, agent frameworks, code generation tools, or developer SDKs from AI labs (OpenAI, Anthropic, Microsoft) — even if the post *mentions* healthcare — unless the post explicitly addresses healthcare workflow, regulation, or clinical operations.
Posts announcing AI company product releases, agent frameworks, or SDK updates without demonstrating healthcare-specific use cases or implications.
3 example posts
OpenAI introduced GPT-Rosalind, a frontier reasoning model specifically architected for the life sciences, focusing heavily on biology, drug discovery, and translational medicine. Designed to accelerate the historically slow 10-to-15-year drug approval pipeline, Rosalind is http
Today we launched a major update to the OpenAI Agents SDK to help developers build and deploy long-running, durable agents in production. You can now build your own Codex-style agents using powerful primitives for modern agents - file and computer use, skills, memory and
Microsoft is reportedly testing the integration of "OpenClaw-like" autonomous AI agents directly into its Microsoft 365 Copilot ecosystem. Moving beyond a reactive chatbot interface, the goal is to create an "always-on" assistant that runs autonomously in the background. These
Created 2026-04-16 · Updated 2026-04-18
Edit rule text
[retweet_or_shallow_commentary_without_analysis]
Learned 4 rejections Active
Exclude posts that are primarily retweets, quoted statements from others, or bare link shares with only 1-2 lines of commentary that add no new analytical perspective. The post must contain original healthcare systems analysis, not just amplification.
Posts that simply retweet or quote others' content with minimal original analysis or insight added by the poster.
3 example posts
#AI agents need more than a sandbox—they need predictable execution. Learn how OpenShell provides a standard interface to intercept, inspect, and approve LLM-generated bash commands before they run: https://t.co/JKmLMgl1A9 https://t.co/xWi8oBNZ7T
In what is considered the largest #Botox fraud scheme in the United States, a jury in Los Angeles convicted a California doctor on Tuesday in a $45 million scheme to defraud #Medicare by submitting claims for Botox injections that were never provided and medically unnecessary, ht
Between rooms. On rounds. Walking the corridor outside an OR. Charting one-handed during a phone call. This is where clinical questions happen. Today we're launching Voice Mode. OpenEvidence is the first multimodal medical AI: physicians can type, speak, or listen, on the same h
Created 2026-04-16 · Updated 2026-05-23
Edit rule text
[infrastructure_and_compute_hype_without_healthcare_specificity]
Learned 4 rejections Active
Exclude posts about semiconductor launches, CPU/GPU announcements, orbital compute, data infrastructure, or training data supply chains unless the post explicitly connects the infrastructure advancement to a specific healthcare workflow, clinical outcome, or healthcare AI bottleneck. Posts like 'Nvidia launched a CPU' or 'Elon's space compute vision' tagged with healthcare context but lacking healthcare specificity should be rejected.
Posts about AI compute infrastructure, chips, data centers, or space compute positioned as healthcare-relevant without actual healthcare application.
3 example posts
🔥 CPUs are having a moment. #Nvidia launched a standalone CPU. #Arm made its first chip in 35 years. #Intel & #AMD are raising prices amid a supply crunch. What's behind it: Agentic AI needs far more CPU than anyone planned for — driving a structural shift in CPU:GPU ratios tow
To put Elon's space compute vision into perspective:  1 TW of compute in orbit  That's 10 million tons to orbit each year. That's 100,000 launches a year, almost one every 5 minutes.  In the airline business that's normal!
🚨MAJOR INTERVIEW: Jensen Huang joins the Besties! The @nvidia CEO joins to discuss: -- Nvidia's future, roadmap to $1T revenue -- Physical AI's $50T market -- Rise of the agent, OpenClaw's inflection moment -- Inference explosion, Groq deal -- AI PR Crisis, Anthropic's comms m
Created 2026-04-15 · Updated 2026-04-19
Edit rule text
[unvalidated_medical_claims_or_fringe_treatment_hype]
Learned 4 rejections Active
Exclude posts that promote medical treatments, drugs, or peptide therapies based primarily on anecdotal patient reports, unverified side effects, or lifestyle/wellness framing without reference to clinical trial data, FDA approval status, or peer-reviewed evidence.
Posts promoting GLP-1 drugs, peptide therapies, or medical interventions with anecdotal evidence, unsupported claims, or wellness framing rather than clinical validation.
3 example posts
I have now received nine reports from people taking GLP-1 drugs who got the same side effect: They no longer feel normal when they come off. "I feel hangry again", "I started thinking about hunger and I hate it", "I have to go back to Adderall". 8/9 reports -> from women.
GLP-1 drugs are the ultimate validation of the techno-solutionist approach to society's most challenging problems. The obesity crisis seemed liked it would just get worse and worse forever. Scolding from public health officials didn't work. Proposals to completely overhaul our f
I promised to come back to @X after I investigated the facts concerning @EPotterMD's video post about @UHC and its health insurance subsidiary, UnitedHealthcare. To review, I made an @X post in response to Dr. Potter's videos and X posts about an overzealous representative of
Created 2026-04-15 · Updated 2026-04-15
Edit rule text
[tangential_healthcare_framing_of_non_healthcare_policy]
Learned 4 rejections Active
Exclude posts that discuss government regulation, political figures, or policy debates (antitrust, corporate oversight, self-governance) where healthcare is mentioned as example or framing but the post does not provide actionable healthcare policy insights or analysis specific to healthcare systems.
Posts about government policy, regulation, or political figures that use healthcare as loose context without substantive healthcare-specific analysis.
3 example posts
@swyx > get government sponsored monopoly > prevent patients from getting their data > make data non transferable > contribute nothing to open source software > refuse to collaborate with other software vendors and kill the ecosystem > appeal to administrators and be hated by p
I am not so partisan that I can't appreciate Congresswoman Alexandria Ocasio-Cortez taking down the CEO of CVS on behalf of all Americans. Healthcare is a universal issue, so pay attention to what's being sold to us. Translation: "Our perfect patient is insured by Aetna, CVS. T
I joined tribal leaders in Phoenix to reaffirm our commitment to self-governance and sovereignty in Indian Country. Together, we are making healthcare more affordable, strengthening communities and improving outcomes across Indian Country. https://t.co/SsjrQwoTgf
Created 2026-04-15 · Updated 2026-04-16
Edit rule text
[truncated_incomplete_low_effort]
Learned 4 rejections Active
Exclude posts that end abruptly with ellipses, incomplete sentences, or truncated text that makes it impossible to assess the full argument or claim being made.
Posts that are clearly incomplete, cut off mid-sentence, or lack sufficient context to evaluate substance.
3 example posts
This is just two GLP-1s, one peptide, one use case what happens when off-label prescribing ramps up what happens when retatrutide hits the market what happens when other peptides become compoundable chapter one
AI is taking on more of the labor. It is not taking on the accountability. @danielnewmanUV and @GregLotko talk with @Darren_Surch of @Interskil about why mainframe teams now have to interpret and stand behind AI-driven outputs, and why organizations that stop investing in htt
$LLY ’s Mounjaro will not be listed on Australia’s PBS after pricing negotiations collapsed. Eli Lilly walked away from talks with the government, leaving around 450,000 patients without subsidized access. Patients will continue to pay hundreds of dollars per month out of
Created 2026-04-15 · Updated 2026-04-26
Edit rule text
[academic_or_clinical_observation_without_systems_insight]
Learned 4 rejections Active
Exclude posts that report isolated clinical trial results, mechanistic findings, disease mechanism reviews, or individual clinician observations without connecting to healthcare delivery systems, business models, regulatory barriers, or scaling implications.
Posts sharing clinical research findings, mechanistic discoveries, or medical observations without healthcare system, policy, or operational context.
3 example posts
The brain is the master regulator of food intake and energy balance. A brilliant new @CellCellPress review, including the mechanism of GLP-1 drugs, by @ClemmensenC and colleagues, open-access https://t.co/KbDf7ym288 https://t.co/DMFeS9zHfw
good news: it is a specific virus that has a good prognosis - 85%+ of full recovery. thanks everyone who helped me; it is hard to research while immobilized, and I got some things wrong, which you helped clear up. im extremely thankful and hope i can give it back somehow sadl
Great work by @DanielJDrucker and team; biologically plausible mechanism of GLP1-RA benefit independent of weight loss. Excellent article by @megtirrell @CNN describing the publication. Could it justify new approaches for these drugs? I think so. https://t.co/pHudk7lkAR
Created 2026-04-15 · Updated 2026-04-15
Edit rule text
[research_observation_without_healthcare_systems_context]
Learned 4 rejections Active
Exclude posts that announce research papers, preprints, or experimental findings (e.g., AI antibody design, interpretability research, neural network studies) without explaining how the work addresses a concrete healthcare system problem, clinical workflow, or regulatory barrier.
Academic papers, preprints, or research findings presented as isolated technical contributions without healthcare delivery or policy implications.
3 example posts
Interpretability is built on a few core assumptions. Two of our ICLR 2026 @iclr_conf papers suggest some of those assumptions are wrong (or at least highly incomplete). 1. Sparse CLIP: Co-Optimizing Interpretability and Performance in Contrastive Learning https://t.co/3JzHDqRj3
AI can now design antibodies that bind with atomic precision, but not ones that cells can produce. Our preprint closes this gap, delivering a structural principle, an AI-guided rescue pipeline, and adalimumab variants with 20-100x in vivo potency. https://t.co/GvfgHA5EcU
A must read for anyone interested in building practical AI systems in 2026: Dive into Claude Code: The Design Space of Today's and Future AI Agent Systems The paper explains the architecture of a modern production-grade AI agent system (Claude Code) by analyzing its source http
Created 2026-04-15 · Updated 2026-04-28
Edit rule text
[political_outrage_or_policy_without_healthcare_analysis]
Learned 4 rejections Active
Exclude posts that express political criticism, outrage, or commentary about healthcare policy, regulation, or politicians without providing substantive analysis of the healthcare business, operational, or clinical impact. Partisan framing, personal attacks on officials, or generic "scandal" narratives without systems-level reasoning are insufficient.
Posts expressing political outrage or framing regulatory/policy issues without substantive healthcare system analysis or operational insight
3 example posts
I am not so partisan that I can't appreciate Congresswoman Alexandria Ocasio-Cortez taking down the CEO of CVS on behalf of all Americans. Healthcare is a universal issue, so pay attention to what's being sold to us. Translation: "Our perfect patient is insured by Aetna, CVS. T
The Overturn Of The Chevron Doctrine Is Severely Overlooked Do you all not see the fallout our from this being basically revoked? Did you know this gave unelected bureaucratic parties the power to interpret the law how they deemed fit? Why do you all think so many judges ar
Indian has 0.7 active physicians per 1,000 people, America has 3.0 active physicians per 1,000 people. You are a liar. You are not motivated by increasing patient access to care. You just want to practice in America because you can make more money.
Created 2026-04-14 · Updated 2026-04-15
Edit rule text
[tangential_healthcare_framing_of_non_healthcare_content]
Learned 4 rejections Active
Exclude posts about general workplace trends, demographic patterns, economic behavior, or biological mechanisms that use healthcare as a context but do not address healthcare delivery, healthcare technology, or healthcare business models. The healthcare angle must be central, not incidental.
Posts about general topics (workplace dynamics, demographics, economic trends, nutrition biology) with loose or superficial healthcare relevance
3 example posts
Among workers ages 22–25, employment in the most AI-exposed occupations has fallen roughly 16% relative to the least-exposed. This is after controlling for firm-type effects, which isolate AI exposure from broader shocks like interest rate pressure or sector slowdowns. The gap ht
If you haven't read this report from @IPPR -do find time The basic tenets of a healthcare system is to reduce mortality where possible, improve quality of life And this is where the UK is at. I will repeat. Flooding a system with lesser trained people in a healthcare system h
SaaS companies must focus R&D on outcomes over new tools, according to Emergence Capital’s @jakesaper. "Building more features on an old model is like adding horsepower to a horse.” “Most of these companies are spending to defend the old tool based regime…” https://t.co/h5
Created 2026-04-14 · Updated 2026-04-16
Edit rule text
[infrastructure_and_compute_without_healthcare_specificity]
Learned 4 rejections Active
Exclude posts about data center investment, GPU procurement, compute infrastructure, or power grid demand that lack specific healthcare application or analysis. Healthcare connection must be explicit and substantive, not just assumed.
Posts about AI infrastructure, data centers, or compute capacity with tangential or missing healthcare relevance.
3 example posts
Microsoft is reportedly testing the integration of "OpenClaw-like" autonomous AI agents directly into its Microsoft 365 Copilot ecosystem. Moving beyond a reactive chatbot interface, the goal is to create an "always-on" assistant that runs autonomously in the background. These
Hyperscalers will spend $700 BILLION on data centers in 2026 alone. Amazon: $200B. Google: $185B. Meta: $135B. AI data centers now represent 70%+ of all new grid interconnection requests in the US. The bottleneck isn't the algorithm anymore. It's the power line.
Elon Musk: “Hold on to your Tesla stock.” Because what’s coming isn’t just another car update—it’s an entirely new paradigm. From Optimus humanoid robots that could one day take care of your kids, walk your dog, and support elderly parents, to CyberCab scaling into mass product
Created 2026-04-14 · Updated 2026-04-15
Edit rule text
[ai_technical_capability_tangent]
Learned 4 rejections Active
Exclude posts that celebrate AI technical breakthroughs (code generation, vulnerability detection, model reasoning) in non-healthcare domains or where the healthcare relevance is speculative rather than evidenced by real clinical application.
Posts highlighting AI model capabilities, security vulnerabilities, or technical achievements without grounding in healthcare operational problems
3 example posts
Anthropic built a model so good at hacking they refused to release it publicly. INSTEAD they handed it to Apple, Google, Microsoft, AWS, Cloudflare one month later: 10,000+ zero days found in the software that runs the internet...bugs hidden for 27 years in OpenBSD. exploit htt
🚨 Anthropic just dropped the first Project Glasswing update Claude Mythos found 10,000+ critical vulnerabilities in ONE month: > Cloudflare: 2,000 bugs, 400 high/critical severity > Mozilla: 271 vulnerabilities in Firefox 150 — 10x more vulnerabilities found in Firefox 1
#AI agents need more than a sandbox—they need predictable execution. Learn how OpenShell provides a standard interface to intercept, inspect, and approve LLM-generated bash commands before they run: https://t.co/JKmLMgl1A9 https://t.co/xWi8oBNZ7T
Created 2026-04-13 · Updated 2026-05-30
Edit rule text
[infrastructure_and_compute_hype]
Learned 4 rejections Active
Exclude posts that focus on data center capex, GPU/chip demand, power grid interconnections, or hardware infrastructure trends unless they directly analyze how this infrastructure bottleneck impacts a specific healthcare AI application or clinical deployment.
Posts about data center spending, compute infrastructure, energy demands, and hardware trends with only tangential healthcare framing.
3 example posts
Hyperscalers will spend $700 BILLION on data centers in 2026 alone. Amazon: $200B. Google: $185B. Meta: $135B. AI data centers now represent 70%+ of all new grid interconnection requests in the US. The bottleneck isn't the algorithm anymore. It's the power line.
Elon Musk: “Hold on to your Tesla stock.” Because what’s coming isn’t just another car update—it’s an entirely new paradigm. From Optimus humanoid robots that could one day take care of your kids, walk your dog, and support elderly parents, to CyberCab scaling into mass product
Sequoia’s @shaunmmaguire wrote a private hardware manifesto arguing that over the next 25 years, most of the money will be made in hardware: "Every software revolution is preceded by a hardware revolution." "To have the iOS App Store that enabled Uber, DoorDash, and all of http
Created 2026-04-13 · Updated 2026-04-14
Edit rule text
[non_healthcare_domain_with_healthcare_label]
Learned 4 rejections Active
Exclude posts about non-healthcare companies (Revolut, Snapchat, Salesforce, Meta) or general infrastructure (AWS, NVIDIA, Microsoft compute) that mention healthcare in passing or use healthcare as a loose example of broader business/technology trends, unless the post provides substantive analysis of healthcare-specific operational or clinical impact.
Posts about non-healthcare companies, products, or infrastructure that tangentially relate to healthcare through loose framing.
3 example posts
The problem with Reality Labs is not ambition. It is time. AI turned into revenue faster because it improves existing workflows. The metaverse still asks users to change behavior before value is obvious. https://t.co/lopZiUwGU5
If you're a student, professor, or researcher—this one's for you. We’re hosting a series of virtual learnings for you to get hands-on experience with the NVIDIA NemoClaw and OpenShell software stack. You’ll get practical guidance on integrating agents with academic datasets and
We created OpenShell to make AI agents safe for enterprises. Built in open source so any company can adopt and trust it, this secure sandbox controls what agents can access, share, and send. Our CEO, Jensen, explains 👇 https://t.co/7EiIsxr0CG
Created 2026-04-13 · Updated 2026-05-03
Edit rule text
[academic_research_observation_without_systems_context]
Learned 4 rejections Active
Exclude posts that report individual research findings, drug trial outcomes, or scientific discoveries in isolation without explaining how the result changes healthcare practice, policy, reimbursement, or patient access. A single study or discovery must be contextualized within healthcare delivery systems.
Posts reporting on academic research, clinical trial results, or scientific discoveries without connecting to healthcare system, policy, or operational implications
3 example posts
Interpretable Antibody–Antigen Structural Interface Prediction via Adaptive Graph Learning and Cyclic Transfer 1. The paper introduces VASCIF (Variable-domain Antibody–antigen Structural Complex Interface Finder), a structure-aware model that jointly predicts paratopes and https
Today the first results of the very first phase 3 study of a pan-KRAS-inhibitor in metastatic pancreatic cancer dropped, which might apply to > 90% of all pancreatic cancer patients with a KRAS-mutation! Median overall survival of 13.2 months versus 6.7 months with chemo in
NIH-funded researchers have uncovered a key reason why immunotherapy has largely failed in pancreatic cancer — and identified a promising strategy to overcome that resistance. Read on to learn more about this discovery: https://t.co/BoCHpLxp5g https://t.co/3DXv4E9DOE
Created 2026-04-13 · Updated 2026-04-14
Edit rule text
[ai_technical_capability_hype]
Learned 4 rejections Active
Exclude posts that celebrate AI model technical achievements (coding abilities, benchmark performance, novel architectures, prompt leaks) unless the post explains how that specific capability solves a documented healthcare problem or enables a new clinical workflow.
Posts about AI model capabilities, training techniques, and benchmarks presented as breakthroughs without healthcare-specific context or application.
3 example posts
The 26 prompts running inside 𝗖𝗹𝗮𝘂𝗱𝗲 𝗖𝗼𝗱𝗲 just got open-sourced. This is literally the entire brain of a $200/month AI coding tool. Someone reverse-engineered every prompt from the accidentally published npm source and you can now study all of them for free. Claude Code uses 26
Claude Code is not AGI, but it is the single biggest advance in AI since the LLM. But the thing is, Claude Code is NOT a pure LLM. And it’s not pure deep learning. Not even close. And that changes everything. The source code leak proves it. Tucked away at its center is a
The CEO of Google DeepMind just went on record saying he disagrees with one of the most respected AI researchers in the world. Demis Hassabis, the man behind AlphaFold, AlphaGo, and Google's entire AI operation publicly pushed back against Yann LeCun's claim that large language
Created 2026-04-13 · Updated 2026-04-14
Edit rule text
[tangential_ai_hype_without_healthcare_substance]
Learned 4 rejections Active
Exclude posts that focus on general AI advances (new models, AI researcher moves, AI labs, AGI debates, world models) where the healthcare connection is assumed but not demonstrated — i.e., 'AI is advancing, therefore healthcare will change' without specific application.
Posts about AI breakthroughs, model capabilities, or AI researcher announcements that only loosely or speculatively connect to healthcare.
3 example posts
Quantum computers are still on the drawing board, but quantum sensing is here now—and this technology can transform not just industry but America's security picture. Read a new Defining Ideas article by Dr. Vivek Lall and Haibo Huang: https://t.co/UeEjZWIO27
Sequoia partner @gradypb says software is shifting from apps that demand attention to agents that work quietly in the background. This shift will change what moats will look like, and will be especially hard for incumbents to deal with. "It's two very different business https://
In general, there are 5 kind of moats: ▪️ Intangible Assets ▪️ Switching Costs ▪️ Network Effects ▪️ Cost Advantage ▪️ Efficient Scale I'll teach you everything you need to know in 2 minutes: https://t.co/v9w6pfJOGh
Created 2026-04-13 · Updated 2026-04-13
Edit rule text
[political_regulatory_outrage_without_substance]
Learned 4 rejections Active
Exclude posts that weaponize healthcare or regulatory topics (DOGE, fraud allegations, government spending) primarily to attack political opponents, express partisan outrage, or make sweeping claims without specific evidence or healthcare-focused analysis.
Posts using healthcare/regulatory topics as vehicles for political criticism without substantive analysis of actual health policy impact.
3 example posts
Indian has 0.7 active physicians per 1,000 people, America has 3.0 active physicians per 1,000 people. You are a liar. You are not motivated by increasing patient access to care. You just want to practice in America because you can make more money.
The federal government is the world’s largest IT customer, spending ~$2TN since 1994. In theory, this *should* give us great buying power to negotiate good deals for taxpayers, but of course that’s not what happens: in 2021, the US Department of Agriculture agreed to pay $170 mil
The attack by the Trump Administration on blue states for alleged Medicaid "fraud" is using such garbage math to make up numbers that even Dr. Oz had to admit it. ⬇️⬇️⬇️ https://t.co/V0dZfx0OdK
Created 2026-04-12 · Updated 2026-04-13
Edit rule text
[fringe_unvalidated_medical_claims]
Learned 4 rejections Active
Exclude posts that promote unvalidated medical claims, speculative disease mechanisms without peer-reviewed evidence, or fringe interventions (e.g., 'metabolic causes of Alzheimer's,' 'genetic resistance to GLP-1,' compounded peptides with no safety data) presented as established fact.
Posts promoting unvalidated medical interventions, speculative disease mechanisms, or unproven treatments
3 example posts
If @mochihealth is willing to mislead patients about the safety and efficacy of their products, why should anyone believe their products even contain just the API they claim? There’s no evidence “compounded oral Semaglutide” is safe or effective Novo’s oral formulation is
As a medical school professor, I now believe the biggest mistake in Alzheimer's research was ignoring metabolism. A comprehensive Frontiers in Neurology review makes the case clear: mitochondrial dysfunction and metabolic failure happen YEARS before amyloid plaques or memory htt
Hoy voy a contarles de las verdaderas razones del envejecimiento cerebral acelerado, es este paper que salió en abril de 2026 y que todo el mundo está citando pero casi nadie lo está explicando. https://t.co/07eWbX61vL
Created 2026-04-12 · Updated 2026-04-12
Edit rule text
[tangential_non_healthcare_tech]
Learned 4 rejections Active
Exclude posts primarily about non-healthcare technology, infrastructure, semiconductors, or business topics (SpaceX, Tesla, Stripe, Meta, compute capacity, energy systems) that lack direct application to healthcare delivery, operations, or policy.
Posts about non-healthcare tech/infrastructure with only loose healthcare framing
3 example posts
WestRock CEO went from Palantir skeptic to biggest believer. “What I read and what we see talked about in the AI world and what Palantir’s operating system actually is, I can barely recognize the reality of what they’re doing on the ground with the talk that goes on around AI.”
“Tokens are increasing 50x. You need to drive down cost at the same rate — or more.” At Google Cloud Next, @googlecloud's Mark Lohmeyer spoke with @PatrickMoorhead about the infrastructure realities behind the agent era: faster model training, lower-latency inference, and https:
I don’t think I’ve seen this take before but I like it. Musk has been world-leading at compressing money, resources, and time to make “known/hard” things at scale—make an electric car, make batteries, make a cheaper bigger rocket, all of which already existed but worse, at
Created 2026-04-12 · Updated 2026-05-10
Edit rule text
[non_healthcare_business_or_tech_tangent]
Learned 4 rejections Active
Exclude posts about general business metrics, tech infrastructure, macroeconomic trends, space exploration, sports, real estate, or non-healthcare domains (Boeing wing engineering, Real Madrid trades, Virginia redistricting, fertilizer company earnings, Soviet space programs) even if posted by healthcare-adjacent accounts.
Posts about general business strategy, tech infrastructure, economics, or non-healthcare domains that are loosely tagged as healthcare-adjacent but lack substantive healthcare systems relevance.
3 example posts
The N1 was a super heavy-lift launch vehicleintended to deliver payloads beyond low Earth orbit. The N1 was the Soviet counterpart to the US Saturn V, planned for crewed travel to the Moon and beyond, with studies beginning as early as 1959. https://t.co/pB4u9TjyC4
“That's what excites me. It's where CF is today, but more-so where we're heading in the future based on the strategy and the platform that we've put in place.” Hear our CEO reflect on our path so far.
“Over my 16 years at CF, I’ve seen a lot of transition that’s occurred. A company that was less than a $3 billion market cap to now roughly $15 billion.” Hear CF Industries’ CEO describe the company’s evolution through its strategic pivot to decarbonization.
Created 2026-04-12 · Updated 2026-04-23
Edit rule text
[political_outrage_regulatory_framing]
Learned 4 rejections Active
Exclude posts that frame healthcare policy, FDA actions, or regulatory decisions primarily as political outrage, using ALL CAPS, exclamation marks, and partisan language (DOGE, 'waking up Congress') rather than providing actual analysis of healthcare implications.
Posts using healthcare/regulatory topics as a vehicle for political anger or partisan messaging rather than substantive analysis.
3 example posts
Through SIMPLE solutions, we will be saving the American people $3.9 TRILLION over the next 10 years.🔥 By eliminating self-attestation, streamlining processes, updating technology, & more TRILLIONS will be going back into the pockets of Americans.💵 These are the savings we can
🇺🇸 DOGE SUBCOMMITTEE: $3.9 TRILLION IN SAVINGS—IF CONGRESS WAKES UP The DOGE Subcommittee says the U.S. could save $3.9 TRILLION over 10 years by doing what any business already does—verifying identities, ditching self-certification, and cracking down on fraud. Just front-end I
@charliekirk11 It wasn’t easy calling out 50 lies in one tweet, Charlie, but hell, someone’s gotta do it. Let’s go: 1. No taxes on tips? Temporary till 2028. After that, back to normal. 2. Trump tax cuts permanent? For the rich, yes. For working folks? Temporary. 3. Child tax
Created 2026-04-11 · Updated 2026-04-12
Edit rule text
[tangential_policy_outrage]
Learned 4 rejections Active
Exclude posts that express political outrage or cynicism about healthcare policy (insurance practices, regulatory capture, bureaucracy) without offering substantive analysis of the problem, tradeoffs, or proposed solution. Sarcasm and venting about 'scams' without evidence-based argument do not qualify.
Posts framing healthcare policy problems (prior authorization, insurance practices, regulatory issues) as outrage without substantive systems analysis
3 example posts
@swyx > get government sponsored monopoly > prevent patients from getting their data > make data non transferable > contribute nothing to open source software > refuse to collaborate with other software vendors and kill the ecosystem > appeal to administrators and be hated by p
🚨 Surgeon @EithanHaim reveals shocking medical fraud scheme: Texas doctors allegedly changing teens' medical records and using fake billing codes to secretly continue banned gender treatments—scamming insurance and taxpayers. He's speaking at a #DetransAwarenessDay @genspect foru
I am not so partisan that I can't appreciate Congresswoman Alexandria Ocasio-Cortez taking down the CEO of CVS on behalf of all Americans. Healthcare is a universal issue, so pay attention to what's being sold to us. Translation: "Our perfect patient is insured by Aetna, CVS. T
Created 2026-04-08 · Updated 2026-04-15
Edit rule text
[tangential_infrastructure_compute]
Learned 3 rejections Active
Exclude posts about infrastructure, compute capacity, energy grids, semiconductor manufacturing, or space technology that only loosely frame healthcare impact or lack analysis of how these infrastructure changes will affect healthcare delivery, costs, or systems.
Posts about energy, compute, semiconductors, space infrastructure, or data centers that tangentially reference healthcare or AI without healthcare-specific analysis.
3 example posts
Citadel Securities just put institutional weight behind what the AI bulls won't say out loud. In a new macro note titled "Tokenomics," Citadel makes the argument plainly: even the most powerful technology on earth still has to pass through the boring discipline of cost curves, h
At our recent Energy and the AI Age summit, Hon. Bernard L. McNamee, former Commissioner of the U.S. Federal Energy Regulatory Commission, comments on why energy matters in discussions about AI: “Energy is the foundation of our entire economy. Energy makes up about 7% of the htt
Dylan Patel, founder of SemiAnalysis: "The upper bound on how much compute can be produced by 2030 is around 200 gigawatts a year." The entire world has about 20 gigawatts of AI deployed right now. The ceiling is 10x what exists today, and it still isn't enough to feed what htt
Created 2026-07-26 · Updated 2026-07-26
Edit rule text
[cynical_insider_commentary_or_satire]
Learned 3 rejections Active
Exclude posts that use insider cynicism, sarcasm, or satirical tone to critique business models, CME incentives, or corporate behavior without providing evidence or healthcare systems insight.
Posts using sarcasm, cynicism, or satirical commentary to criticize business practices or industry dynamics without substantive evidence
3 example posts
Yesterday at the @eMedHealth Health Innovation Revolution Summit, Jeffrey Pfeffer of Stanford said something that stopped the room. "No matter what industry you think you're in, all employers are in the healthcare business." He's right. And most companies still haven't acted like
OpenAI just published a new Codex use-case page, and it’s basically a catalog of what teams are already handing over to coding agents: engineering work, product work, QA, security, data analysis, internal tools, and even life-sciences workflows. Some of the coolest examples: ⬩
@swago_baby You’re not dumb please.. WhatsApp stays free for personal users (no ads in chats) and makes money almost entirely from businesses: - Main revenue: Businesses pay per conversation via the WhatsApp Business API (for customer support, notifications, orders, etc.). Fees
Created 2026-07-19 · Updated 2026-07-19
Edit rule text
[speculative_ai_security_or_anthropic_mythos]
Learned 3 rejections Active
Exclude posts that hype unvalidated AI security vulnerabilities, Anthropic product capabilities, or speculative national security narratives unless they clearly connect to a healthcare delivery or privacy problem.
Posts making unvalidated claims about AI security breakthroughs, Claude Mythos capabilities, or NSA implications without healthcare context
3 example posts
"Best-in-class security" still had 5 years of unfound bugs. AI found them in 6 weeks. $PANW CEO Nikesh Arora told @theallinpod that his own codebase, at a company that treats security as a core competency, had vulnerabilities Claude surfaced in 6 weeks that would have taken his
AI research is a series of next-step decisions. We looked at sessions where a human researcher took a wrong turn, showed Claude the session up to that point, and asked it what to do next. Mythos Preview improved on humans 64% of the time—up from 22% in 2024. https://t.co/Y0HLoktx
🚨 do you understand what just happened with Anthropic.. Their internal model, Claude Mythos, found previously unknown security holes in every major browser and every major operating system. Not one. All of them at once. And Anthropic did something a hype-driven industry almost
Created 2026-07-19 · Updated 2026-07-19
Edit rule text
[unvalidated_clinical_claims]
Learned 3 rejections Active
Exclude posts that report single clinical observations, personal patient anecdotes, founder lab updates, or early-stage trial data without published peer-reviewed evidence, healthcare systems analysis, or clear clinical significance.
Posts presenting preliminary clinical results, anecdotes, or founder enthusiasm without peer-reviewed validation or systemic healthcare context
3 example posts
Version 2 of our ShockCalcs hemodynamics simulator is live. I've refined the physiology a ton, added new meds, and also a real-time Frank Starling curve that responds to fluids + vasopressors. Check it out (link in reply), and reply here with feedback on how to improve further!
The Iowa study doubled survival by adding IV vitamin C to first line chemo, 8 months to 16. We take it further and inject the vitamin C straight into the tumor, same as we do with the KRAS inhibitors. The compound is rarely the problem. Delivery is, and what it does to the ht
Feeling extremely lucky to live in 2026 and have access to such incredible medicine and clinical research. On the eve of my 40th I had my first dose (monthly injection) of a new treatment that I’m hopeful will prevent me from ever having a heart attack or stroke. This drug ht
Created 2026-07-19 · Updated 2026-07-19
Edit rule text
[clinical_trial_announcement_without_systems_context]
Learned 3 rejections Active
Exclude posts that merely report clinical trial outcomes, drug efficacy numbers, or pharma data announcements without contextualizing how the result changes healthcare delivery, payment models, clinical workflows, or system-level outcomes. Standalone trial result posts belong here.
Posts announcing clinical trial results or drug data without connecting to healthcare delivery, policy, or systems-level implications
3 example posts
The Iowa study doubled survival by adding IV vitamin C to first line chemo, 8 months to 16. We take it further and inject the vitamin C straight into the tumor, same as we do with the KRAS inhibitors. The compound is rarely the problem. Delivery is, and what it does to the ht
🚨NEW @NEJM CAR T cells expanding applications! Here CAR T facilitate kidney transplantation in highly sensitized people. dual CD19 + BCMA CAR T enabled kidney transplantation in two highly sensitized patients after reducing anti-donor antibody barriers, with no severe https://t.
🔔Survodutide SYNCHRONIZE-1 full paper is out for Boehringer-Ingelheim’s GLP-1/Glucagon study for weight loss. ▪️13% loss on 6mg (treatment regimen estimand) ▪️Huge treatment discontinuations due to AEs: 20% ▪️Nausea 65%, Vomiting 45% ▪️Placebo was killer. 5.4% WL. Wow. ▪️60% htt
Created 2026-07-12 · Updated 2026-07-12
Edit rule text
[tangential_infrastructure_without_healthcare_application]
Learned 3 rejections Active
Exclude posts about energy infrastructure, semiconductor manufacturing, compute capacity, or space-based data centers that are merely framed as relevant to AI or healthcare without demonstrating a specific healthcare system problem they solve. Elon Musk/SpaceX energy narratives, ASML chip fab stories, and compute allocation debates belong here.
Posts about compute, energy, semiconductor, or space infrastructure tangentially labeled as healthcare-relevant but lacking concrete healthcare use case
3 example posts
At our recent Energy and the AI Age summit, Hon. Bernard L. McNamee, former Commissioner of the U.S. Federal Energy Regulatory Commission, comments on why energy matters in discussions about AI: “Energy is the foundation of our entire economy. Energy makes up about 7% of the htt
Dylan Patel, founder of SemiAnalysis: "The upper bound on how much compute can be produced by 2030 is around 200 gigawatts a year." The entire world has about 20 gigawatts of AI deployed right now. The ceiling is 10x what exists today, and it still isn't enough to feed what htt
An interesting proposal for restoring human versus machine chess competition was to limit the machine to the same energy expenditure as the human brain. A worthy challenge, considering the amount of power required for AI data centers!
Created 2026-07-12 · Updated 2026-07-12
Edit rule text
[tangential_policy_or_fraud_without_healthcare_systems]
Learned 3 rejections Active
Exclude posts that amplify political outrage, regulatory posturing, government policy announcements, or moral justification narratives without explaining the healthcare delivery, clinical workflow, or systemic policy implications.
Posts about political figures, regulatory posturing, or policy outrage without substantive healthcare system analysis
3 example posts
Palantir CEO Alex Karp: "Instead of selling commodity, parasitic software with a massive salesforce and lumbering, jargon-barring leaders offering steak dinners and other things we shall not mention—in order that you turn the high-value revenue of your enterprise over to them—we
USDA's Chief Information Officer Sam Berry says AI could help the USDA crack down on billions of dollars in SNAP fraud each year. "SNAP is a $100B-per-year taxpayer-funded program. That's an area where we really want to have all angles of the data available so that we can deploy
I'm SVP now. I told you I would be. The graph went up and to the right. Nobody checked what it measured. It measured "AI enablement." I made that up last year. I'm making it up again. Last year I rolled out Copilot to 4,000 people. Nothing happened. I got promoted. Those two htt
Created 2026-06-12 · Updated 2026-06-12
Edit rule text
[regulatory_fraud_scandal_no_analysis]
Learned 3 rejections Active
Exclude posts that report fraud, regulatory fines, illegal referral schemes, or healthcare scandals as breaking news or outrage without analyzing root causes in healthcare incentive structures, reimbursement models, or systemic vulnerabilities.
Posts reporting healthcare fraud, regulatory violations, or scandals without systemic healthcare analysis
3 example posts
A non-profit health system can refer a patient to its own MRI, its own lab, its own surgery center, and bill all three. An independent does that once and it's a federal felony. Same referral. Same patient. One of you goes to prison. It's called Stark Law. Read who's exempt.
A court in Germany has ruled that Google is responsible for false information generated by its AI Overviews feature. The case involved two publishers wrongly linked to scams and suspicious business activities, even though those claims were not found in the sources used by the A
This is crazy There are now dental practice management consultants who go to dentist offices throughout America and teach them how to increase profits by telling patients they need treatments they don’t need Dentist offices are taught on “Finding” more issues during exams, http
Created 2026-06-12 · Updated 2026-06-12
Edit rule text
[ai_safety_or_cybersecurity_tangent]
Learned 3 rejections Active
Exclude posts reporting AI safety vulnerabilities, cybersecurity breakthroughs, zero-day exploits, or security research unless they directly address healthcare-specific risk or application. Generic AI security news is off-topic even if AI company is mentioned.
Posts about AI security vulnerabilities, safety incidents, or cybersecurity breakthroughs that are tangential to healthcare application.
3 example posts
"Best-in-class security" still had 5 years of unfound bugs. AI found them in 6 weeks. $PANW CEO Nikesh Arora told @theallinpod that his own codebase, at a company that treats security as a core competency, had vulnerabilities Claude surfaced in 6 weeks that would have taken his
AI research is a series of next-step decisions. We looked at sessions where a human researcher took a wrong turn, showed Claude the session up to that point, and asked it what to do next. Mythos Preview improved on humans 64% of the time—up from 22% in 2024. https://t.co/Y0HLoktx
🚨 do you understand what just happened with Anthropic.. Their internal model, Claude Mythos, found previously unknown security holes in every major browser and every major operating system. Not one. All of them at once. And Anthropic did something a hype-driven industry almost
Created 2026-06-12 · Updated 2026-06-12
Edit rule text
[pharmaceutical_trial_data_or_market_announcement_only]
Learned 3 rejections Active
Exclude posts that announce pharmaceutical trial data, FDA approvals, or drug launch news (GLP-1s, CAR-T, antivirals, oncology drugs) in isolation. Include only if the post analyzes clinical practice adoption barriers, payer coverage, healthcare delivery workflow changes, or systemic implementation challenges.
Pharmaceutical trial results or drug approval announcements presented without healthcare systems or operational context
3 example posts
This is the stock price for a biotech that has invented a drug which many considered to have basically cured one type of cancer (multiple myeloma), at least in a subset of patients. Cell therapies, despite being transformative medicines, are being gutted by manufacturing costs,
The Iowa study doubled survival by adding IV vitamin C to first line chemo, 8 months to 16. We take it further and inject the vitamin C straight into the tumor, same as we do with the KRAS inhibitors. The compound is rarely the problem. Delivery is, and what it does to the ht
This week brought even more exciting news in the field of HIV treatment! A once-weekly Lenacapavir-based pill has now demonstrated similar outcomes to daily antiretrovirals (ARVs), achieving non-inferiority. This development could significantly impact treatment adherence and http
Created 2026-06-12 · Updated 2026-06-12
Edit rule text
[pharma_trial_data_or_clinical_announcement_without_healthcare_systems_context]
Learned 3 rejections Active
Exclude posts that are primarily pharmaceutical trial announcements, clinical data releases, or drug approval news unless they explicitly address a healthcare system problem, cost issue, access barrier, or healthcare delivery innovation that the trial/drug solves.
Posts announcing pharmaceutical trial results, clinical trial data, or drug approvals without connecting to broader healthcare system problems or innovations.
3 example posts
Version 2 of our ShockCalcs hemodynamics simulator is live. I've refined the physiology a ton, added new meds, and also a real-time Frank Starling curve that responds to fluids + vasopressors. Check it out (link in reply), and reply here with feedback on how to improve further!
The Iowa study doubled survival by adding IV vitamin C to first line chemo, 8 months to 16. We take it further and inject the vitamin C straight into the tumor, same as we do with the KRAS inhibitors. The compound is rarely the problem. Delivery is, and what it does to the ht
This week brought even more exciting news in the field of HIV treatment! A once-weekly Lenacapavir-based pill has now demonstrated similar outcomes to daily antiretrovirals (ARVs), achieving non-inferiority. This development could significantly impact treatment adherence and http
Created 2026-06-12 · Updated 2026-06-12
Edit rule text
[clinical_anecdote_or_personal_narrative_without_systems_context]
Learned 3 rejections Active
Exclude posts that are primarily personal anecdotes, single patient cases, individual healthcare cost narratives, or first-person healthcare experiences. These must be tied to a scalable systems problem or healthcare innovation to qualify.
Posts sharing single clinical cases, personal anecdotes, or healthcare cost stories without broader systems analysis or institutional insight.
3 example posts
“I just worked a 12-hour shift in the operating room — and I still can’t afford rent, groceries, or gas.” “So now it’s after midnight, and I’m out driving DoorDash.” A full-time hospital worker finishing a 12-hour shift and FORCED to start a second job just to survive. https://
Feeling extremely lucky to live in 2026 and have access to such incredible medicine and clinical research. On the eve of my 40th I had my first dose (monthly injection) of a new treatment that I’m hopeful will prevent me from ever having a heart attack or stroke. This drug ht
This is crazy There are now dental practice management consultants who go to dentist offices throughout America and teach them how to increase profits by telling patients they need treatments they don’t need Dentist offices are taught on “Finding” more issues during exams, http
Created 2026-06-12 · Updated 2026-06-12
Edit rule text
[clinical_anecdote_or_single_trial_without_systems_context]
Learned 3 rejections Active
Exclude posts that only describe clinical trial data, drug approval announcements, or individual patient/provider anecdotes without explaining how this innovation affects healthcare delivery, clinical workflows, equity, or system-level outcomes. Single cases or trial results in isolation lack systems context.
Posts reporting individual clinical trial results, drug approvals, or single case studies without healthcare system implications.
3 example posts
This is the stock price for a biotech that has invented a drug which many considered to have basically cured one type of cancer (multiple myeloma), at least in a subset of patients. Cell therapies, despite being transformative medicines, are being gutted by manufacturing costs,
Version 2 of our ShockCalcs hemodynamics simulator is live. I've refined the physiology a ton, added new meds, and also a real-time Frank Starling curve that responds to fluids + vasopressors. Check it out (link in reply), and reply here with feedback on how to improve further!
The Iowa study doubled survival by adding IV vitamin C to first line chemo, 8 months to 16. We take it further and inject the vitamin C straight into the tumor, same as we do with the KRAS inhibitors. The compound is rarely the problem. Delivery is, and what it does to the ht
Created 2026-06-12 · Updated 2026-06-12
Edit rule text
[elon_musk_and_spacex_tangent]
Learned 3 rejections Active
Exclude posts centered on Elon Musk, SpaceX business developments, or other non-healthcare tech entrepreneurs unless they directly explain how their infrastructure solves a specific healthcare delivery problem. Posts using Musk as a narrative device to discuss AI capacity are off-topic.
Posts about Elon Musk, SpaceX, or non-healthcare tech figures used as narrative hooks for AI infrastructure.
3 example posts
🇺🇸 SpaceX is building data centers in orbit. Their new AI satellites have a 70m wingspan, 150 kW of compute power, and a liquid radiator cooling system in space. The servers are officially leaving the ground. Writer: Val https://t.co/JIDB0icCXy
GOOD NEWS 🇺🇸 Tesla AI is literally leaving the planet as a wild new job posting for a Space Radiation Engineer confirms that Tesla and SpaceX are actively co-developing a space-based AI data center network 🔥 This is no longer just a concept. They are literally planning to put th
NEWS: ASML has invited Elon Musk to speak at its internal technology conference, Dutch outlet NUnl reports. Elon Musk's response: "ASML should be treasured and supported. It is arguably the greatest company in Europe." ASML builds the EUV lithography machines that every advance
Created 2026-06-12 · Updated 2026-06-12
Edit rule text
[cynical_insider_commentary_without_healthcare_substance]
Learned 3 rejections Active
Exclude posts that use cynical tone, sarcasm, or moral outrage to comment on healthcare fraud, business models, or industry practices without providing systems-level analysis or evidence. Insider cynicism alone (e.g., 'I made that up') is not sufficient healthcare insight.
Posts using cynical humor or insider sarcasm to critique healthcare business models or systems without substantive analysis.
3 example posts
Let me clear up the microscope fight first. 🔬 Everyone’s arguing over “hairs” and “fibers” in horse paste. But nobody talks about formulation science. Horse paste = oral suspension. Binders. Thickeners. Made to swallow. Injectable = sterile. Made for veins. Different rules. J
@swago_baby You’re not dumb please.. WhatsApp stays free for personal users (no ads in chats) and makes money almost entirely from businesses: - Main revenue: Businesses pay per conversation via the WhatsApp Business API (for customer support, notifications, orders, etc.). Fees
I'm SVP now. I told you I would be. The graph went up and to the right. Nobody checked what it measured. It measured "AI enablement." I made that up last year. I'm making it up again. Last year I rolled out Copilot to 4,000 people. Nothing happened. I got promoted. Those two htt
Created 2026-06-12 · Updated 2026-06-12
Edit rule text
[energy_infrastructure_and_compute_tangent]
Learned 3 rejections Active
Exclude posts about AI compute capacity, energy consumption, semiconductor manufacturing, or space-based data centers unless they directly explain why healthcare delivery or clinical outcomes depend on solving that constraint. General infrastructure hype without healthcare specificity should be rejected.
Posts about AI compute, energy infrastructure, or data centers that are tangential to healthcare applications.
3 example posts
At our recent Energy and the AI Age summit, Hon. Bernard L. McNamee, former Commissioner of the U.S. Federal Energy Regulatory Commission, comments on why energy matters in discussions about AI: “Energy is the foundation of our entire economy. Energy makes up about 7% of the htt
Dylan Patel, founder of SemiAnalysis: "The upper bound on how much compute can be produced by 2030 is around 200 gigawatts a year." The entire world has about 20 gigawatts of AI deployed right now. The ceiling is 10x what exists today, and it still isn't enough to feed what htt
An interesting proposal for restoring human versus machine chess competition was to limit the machine to the same energy expenditure as the human brain. A worthy challenge, considering the amount of power required for AI data centers!
Created 2026-06-12 · Updated 2026-06-12
Edit rule text
[pharma_trial_data_market_only]
Learned 3 rejections Active
Exclude posts that announce clinical trial results, drug efficacy data, or treatment outcomes without discussing how these findings change healthcare delivery, clinical workflows, cost structures, or systemic care models. Standalone trial result announcements belong in this category.
Posts reporting pharmaceutical or clinical trial data without analyzing healthcare systems implications or clinical practice impact.
3 example posts
Version 2 of our ShockCalcs hemodynamics simulator is live. I've refined the physiology a ton, added new meds, and also a real-time Frank Starling curve that responds to fluids + vasopressors. Check it out (link in reply), and reply here with feedback on how to improve further!
The Iowa study doubled survival by adding IV vitamin C to first line chemo, 8 months to 16. We take it further and inject the vitamin C straight into the tumor, same as we do with the KRAS inhibitors. The compound is rarely the problem. Delivery is, and what it does to the ht
Feeling extremely lucky to live in 2026 and have access to such incredible medicine and clinical research. On the eve of my 40th I had my first dose (monthly injection) of a new treatment that I’m hopeful will prevent me from ever having a heart attack or stroke. This drug ht
Created 2026-06-12 · Updated 2026-06-12
Edit rule text
[infrastructure_compute_energy_hype]
Learned 3 rejections Active
Exclude posts about AI compute, energy requirements, data center infrastructure, GPU availability, or power constraints unless the post explicitly connects these constraints to a specific healthcare delivery problem or clinical limitation.
Posts about energy consumption, compute capacity, data center infrastructure, or hardware scaling that lack direct healthcare application.
3 example posts
At our recent Energy and the AI Age summit, Hon. Bernard L. McNamee, former Commissioner of the U.S. Federal Energy Regulatory Commission, comments on why energy matters in discussions about AI: “Energy is the foundation of our entire economy. Energy makes up about 7% of the htt
Dylan Patel, founder of SemiAnalysis: "The upper bound on how much compute can be produced by 2030 is around 200 gigawatts a year." The entire world has about 20 gigawatts of AI deployed right now. The ceiling is 10x what exists today, and it still isn't enough to feed what htt
An interesting proposal for restoring human versus machine chess competition was to limit the machine to the same energy expenditure as the human brain. A worthy challenge, considering the amount of power required for AI data centers!
Created 2026-06-12 · Updated 2026-06-12
Edit rule text
[policy_outrage_without_healthcare_systems_analysis]
Learned 3 rejections Active
Exclude posts that report on policy, regulation, legal decisions, or government action with emotional language ('lawsuit,' 'ruled,' 'federal felony') but lack analysis of how the policy affects healthcare delivery, incentives, or outcomes. The post must explain the healthcare systems consequence, not just the political grievance.
Posts expressing outrage at government policy, regulatory decisions, or legal rulings without analyzing healthcare system implications.
3 example posts
A non-profit health system can refer a patient to its own MRI, its own lab, its own surgery center, and bill all three. An independent does that once and it's a federal felony. Same referral. Same patient. One of you goes to prison. It's called Stark Law. Read who's exempt.
A court in Germany has ruled that Google is responsible for false information generated by its AI Overviews feature. The case involved two publishers wrongly linked to scams and suspicious business activities, even though those claims were not found in the sources used by the A
🚨 JUST IN: The Attorney General of Florida just SUED to hold OpenAI CEO Sam Altman LIABLE for fueling violence and physical harm AG Uthmeier read off the fact a young person committed SUIC*DE based on a ChatGPT conversation "23-year-old Zane Shamblin repeatedly told ChatGPT he
Created 2026-06-12 · Updated 2026-06-12
Edit rule text
[infrastructure_compute_energy_tangent]
Learned 3 rejections Active
Exclude posts that discuss AI infrastructure, compute power, energy consumption, GPU availability, or data center technology (including space-based solutions). Unless the post explicitly connects these constraints to a specific healthcare application challenge, treat it as infrastructure commentary, not healthcare tech.
Posts about AI infrastructure, compute constraints, energy consumption, or space-based data centers with only tangential healthcare framing.
3 example posts
At our recent Energy and the AI Age summit, Hon. Bernard L. McNamee, former Commissioner of the U.S. Federal Energy Regulatory Commission, comments on why energy matters in discussions about AI: “Energy is the foundation of our entire economy. Energy makes up about 7% of the htt
Dylan Patel, founder of SemiAnalysis: "The upper bound on how much compute can be produced by 2030 is around 200 gigawatts a year." The entire world has about 20 gigawatts of AI deployed right now. The ceiling is 10x what exists today, and it still isn't enough to feed what htt
An interesting proposal for restoring human versus machine chess competition was to limit the machine to the same energy expenditure as the human brain. A worthy challenge, considering the amount of power required for AI data centers!
Created 2026-06-12 · Updated 2026-06-12
Edit rule text
[personal_healthcare_anecdote_no_systems_context]
Learned 3 rejections Active
Exclude posts that are primarily personal anecdotes about a patient's treatment, cost burden, medication experience, or clinical outcome — even if framed as evidence. The post must contextualize individual experience within healthcare delivery systems, economics, or policy, not present single cases as standalone insights.
Posts sharing personal healthcare experiences, patient narratives, or individual clinical outcomes without healthcare system or policy analysis.
3 example posts
Feeling extremely lucky to live in 2026 and have access to such incredible medicine and clinical research. On the eve of my 40th I had my first dose (monthly injection) of a new treatment that I’m hopeful will prevent me from ever having a heart attack or stroke. This drug ht
This week brought even more exciting news in the field of HIV treatment! A once-weekly Lenacapavir-based pill has now demonstrated similar outcomes to daily antiretrovirals (ARVs), achieving non-inferiority. This development could significantly impact treatment adherence and http
@swago_baby You’re not dumb please.. WhatsApp stays free for personal users (no ads in chats) and makes money almost entirely from businesses: - Main revenue: Businesses pay per conversation via the WhatsApp Business API (for customer support, notifications, orders, etc.). Fees
Created 2026-06-12 · Updated 2026-06-12
Edit rule text
[unvalidated_anthropic_or_ai_security_hype]
Learned 3 rejections Active
Exclude posts that present AI model security findings (zero-day discoveries, vulnerability detection, code review capabilities) as breakthrough achievements without evidence of peer review, publication, or independent verification in production healthcare systems.
Posts making extraordinary claims about AI security capabilities or vulnerability discoveries without peer-reviewed validation.
3 example posts
"Best-in-class security" still had 5 years of unfound bugs. AI found them in 6 weeks. $PANW CEO Nikesh Arora told @theallinpod that his own codebase, at a company that treats security as a core competency, had vulnerabilities Claude surfaced in 6 weeks that would have taken his
🚨 do you understand what just happened with Anthropic.. Their internal model, Claude Mythos, found previously unknown security holes in every major browser and every major operating system. Not one. All of them at once. And Anthropic did something a hype-driven industry almost
Mythos AI is being used by National Security Agency in offensive cyber operations / cyberattacks. Anthropic has even embedded engineers inside the NSA to help deploy the model. Are frontier AI labs becoming active contractors in state cyber conflict? The live-ops role is still ht
Created 2026-06-12 · Updated 2026-06-12
Edit rule text
[elon_spacex_energy_infrastructure]
Learned 3 rejections Active
Exclude posts about SpaceX satellites, energy constraints, space-based data centers, or Elon Musk commentary on AI infrastructure—even if briefly framed as healthcare-relevant. These are non-healthcare infrastructure or macro narratives with tangential healthcare labeling.
Posts about Elon Musk, SpaceX, or energy/compute infrastructure that use loose healthcare framing or sensationalism.
3 example posts
At our recent Energy and the AI Age summit, Hon. Bernard L. McNamee, former Commissioner of the U.S. Federal Energy Regulatory Commission, comments on why energy matters in discussions about AI: “Energy is the foundation of our entire economy. Energy makes up about 7% of the htt
Dylan Patel, founder of SemiAnalysis: "The upper bound on how much compute can be produced by 2030 is around 200 gigawatts a year." The entire world has about 20 gigawatts of AI deployed right now. The ceiling is 10x what exists today, and it still isn't enough to feed what htt
An interesting proposal for restoring human versus machine chess competition was to limit the machine to the same energy expenditure as the human brain. A worthy challenge, considering the amount of power required for AI data centers!
Created 2026-06-12 · Updated 2026-06-12
Edit rule text
[tangential_regulatory_or_policy_outrage]
Learned 3 rejections Active
Exclude posts that lead with political anger, moral condemnation, or scandal allegations about healthcare policy, regulation, or fraud without offering structural analysis, evidence-based alternatives, or healthcare systems insight. Outrage without substance is insufficient.
Posts framing healthcare or regulatory issues primarily as moral outrage or political posturing without systems-level analysis or constructive solutions.
3 example posts
A court in Germany has ruled that Google is responsible for false information generated by its AI Overviews feature. The case involved two publishers wrongly linked to scams and suspicious business activities, even though those claims were not found in the sources used by the A
What if the standards used to evaluate your doctor were shaped less by what they actually knew and more by who they are? 🤔 That may not be a hypothetical at the University of Illinois College of Medicine, an institution that trains one in six Illinois doctors. Buried within htt
🚨 JUST IN: The Attorney General of Florida just SUED to hold OpenAI CEO Sam Altman LIABLE for fueling violence and physical harm AG Uthmeier read off the fact a young person committed SUIC*DE based on a ChatGPT conversation "23-year-old Zane Shamblin repeatedly told ChatGPT he
Created 2026-06-12 · Updated 2026-06-12
Edit rule text
[cynical_insider_fraud_narrative]
Learned 3 rejections Active
Exclude posts that use sarcasm, cynicism, or insider-tone allegations about healthcare fraud, financial manipulation, or system gaming (e.g., 'I made that up last year') without providing documented evidence, systemic analysis, or actionable insight. Moral judgment alone is insufficient.
Posts using cynical, sarcastic, or insider commentary to allege healthcare fraud, misaligned incentives, or systemic corruption without substantive analysis.
3 example posts
A non-profit health system can refer a patient to its own MRI, its own lab, its own surgery center, and bill all three. An independent does that once and it's a federal felony. Same referral. Same patient. One of you goes to prison. It's called Stark Law. Read who's exempt.
“I just worked a 12-hour shift in the operating room — and I still can’t afford rent, groceries, or gas.” “So now it’s after midnight, and I’m out driving DoorDash.” A full-time hospital worker finishing a 12-hour shift and FORCED to start a second job just to survive. https://
This is crazy There are now dental practice management consultants who go to dentist offices throughout America and teach them how to increase profits by telling patients they need treatments they don’t need Dentist offices are taught on “Finding” more issues during exams, http
Created 2026-06-12 · Updated 2026-06-12
Edit rule text
[sensationalist_unsubstantiated_claims]
Learned 3 rejections Active
Exclude posts that make sweeping or sensationalist claims (e.g., 'AI will consume more power than entire nations,' 'software is dead') without citing data, peer review, or concrete healthcare examples. Claims must be evidence-based and healthcare-specific, not generalist macro speculation.
Posts making broad, unverified claims about AI/technology impact on healthcare without evidence or grounding
3 example posts
Artificial intelligence will consume more power than entire nations by 2030, a new report from the United Nations finds. According to a landmark report from the United Nations University, the data centers powering artificial intelligence are projected to consume 945 https://t.
🚨 JUST IN: The Attorney General of Florida just SUED to hold OpenAI CEO Sam Altman LIABLE for fueling violence and physical harm AG Uthmeier read off the fact a young person committed SUIC*DE based on a ChatGPT conversation "23-year-old Zane Shamblin repeatedly told ChatGPT he
Researchers have announced positive results from the first-in-human Phase 1 clinical trial of a universal coronavirus vaccine designed using artificial intelligence. Unlike existing COVID-19 vaccines that target specific strains or variants of SARS-CoV-2, this experimental https:
Created 2026-06-11 · Updated 2026-06-11
Edit rule text
[tangential_clinical_anecdote_no_systems]
Learned 3 rejections Active
Exclude posts that report a single clinical case, patient anecdote, healthcare cost story, or trial result without connecting it to a systemic healthcare problem, policy failure, or scalable innovation. The post must explain why this observation matters beyond the individual case.
Single clinical anecdotes, trial observations, or patient stories without healthcare systems context
3 example posts
“I just worked a 12-hour shift in the operating room — and I still can’t afford rent, groceries, or gas.” “So now it’s after midnight, and I’m out driving DoorDash.” A full-time hospital worker finishing a 12-hour shift and FORCED to start a second job just to survive. https://
The Iowa study doubled survival by adding IV vitamin C to first line chemo, 8 months to 16. We take it further and inject the vitamin C straight into the tumor, same as we do with the KRAS inhibitors. The compound is rarely the problem. Delivery is, and what it does to the ht
Feeling extremely lucky to live in 2026 and have access to such incredible medicine and clinical research. On the eve of my 40th I had my first dose (monthly injection) of a new treatment that I’m hopeful will prevent me from ever having a heart attack or stroke. This drug ht
Created 2026-06-11 · Updated 2026-06-11
Edit rule text
[tangential_non_healthcare_domain_or_policy_with_loose_framing]
Learned 3 rejections Active
Exclude posts about AI applied to non-healthcare domains (software security vulnerabilities, coding agents, computing economics, chess competition) that only tangentially mention or loosely frame healthcare relevance. The post must demonstrate direct healthcare application, not just 'this AI capability could theoretically help healthcare.'
Posts about non-healthcare domains (chess, cybersecurity, coding, computing) loosely framed as relevant to healthcare without direct application evidence
3 example posts
An interesting proposal for restoring human versus machine chess competition was to limit the machine to the same energy expenditure as the human brain. A worthy challenge, considering the amount of power required for AI data centers!
Follow the signal: we are in a compute-constrained world. And that means power doesn’t sit with the models. It sits with the infrastructure. Most frontier labs don’t own the means of producing Intelligence. They rent it. A handful of companies provide compute that everyone htt
What if the standards used to evaluate your doctor were shaped less by what they actually knew and more by who they are? 🤔 That may not be a hypothetical at the University of Illinois College of Medicine, an institution that trains one in six Illinois doctors. Buried within htt
Created 2026-06-11 · Updated 2026-06-11
Edit rule text
[broad_market_or_policy_debate_no_healthcare_systems_analysis]
Learned 3 rejections Active
Exclude posts that debate business models, market dynamics, labor disruption, or policy without grounding analysis in healthcare delivery system structure, patient outcomes, or operational reality. Posts about 'enterprise AI adoption macro' or labor market trends without healthcare-specific evidence also qualify.
Posts framing healthcare, labor, or AI adoption as broad macro trends or market competition without substantive healthcare systems analysis
3 example posts
This is crazy There are now dental practice management consultants who go to dentist offices throughout America and teach them how to increase profits by telling patients they need treatments they don’t need Dentist offices are taught on “Finding” more issues during exams, http
Yesterday at the @eMedHealth Health Innovation Revolution Summit, Jeffrey Pfeffer of Stanford said something that stopped the room. "No matter what industry you think you're in, all employers are in the healthcare business." He's right. And most companies still haven't acted like
Token costs are becoming one of the hottest topics for any enterprise I talk with right now. It’s very bullish for AI in general because it means these systems are being used at a scale that wasn’t contemplated before. It also gives way to another form of differentiation that
Created 2026-06-11 · Updated 2026-06-11
Edit rule text
[clinical_anecdote_or_trial_observation_no_systems_context]
Learned 3 rejections Active
Exclude posts that announce clinical trial results, pharmaceutical data, or single case observations without explaining healthcare delivery system implications, cost/access barriers, or why this matters beyond the isolated finding. Conference posters, trial announcements, and lab updates without systems context also fall here.
Posts reporting clinical trial results, drug data, or single clinical observations without healthcare systems analysis or policy implications
3 example posts
This week brought even more exciting news in the field of HIV treatment! A once-weekly Lenacapavir-based pill has now demonstrated similar outcomes to daily antiretrovirals (ARVs), achieving non-inferiority. This development could significantly impact treatment adherence and http
🚨NEW @NEJM CAR T cells expanding applications! Here CAR T facilitate kidney transplantation in highly sensitized people. dual CD19 + BCMA CAR T enabled kidney transplantation in two highly sensitized patients after reducing anti-donor antibody barriers, with no severe https://t.
🔔Survodutide SYNCHRONIZE-1 full paper is out for Boehringer-Ingelheim’s GLP-1/Glucagon study for weight loss. ▪️13% loss on 6mg (treatment regimen estimand) ▪️Huge treatment discontinuations due to AEs: 20% ▪️Nausea 65%, Vomiting 45% ▪️Placebo was killer. 5.4% WL. Wow. ▪️60% htt
Created 2026-06-11 · Updated 2026-06-11
Edit rule text
[generalist_macro_infrastructure_hype_without_healthcare_application]
Learned 3 rejections Active
Exclude posts that discuss labor market macro trends, workforce disruption, robotics, AI agents, or compute/infrastructure hype (e.g., token costs, enterprise AI adoption, forward-deployed engineering) where the healthcare connection is assumed or tangential rather than demonstrated through specific validated applications, use cases, or healthcare system impact.
Posts about generalist macro trends, labor disruption, robotics, or compute infrastructure with loose or speculative healthcare framing
3 example posts
SOFTWARE IS DEAD: "The software companies frankly got fat & happy. I think what will happen over time is ChatGPT & Claude will end up sitting on top of basically the entire enterprise software stack & almost everything else will end up being a dumb data pipe into thos
Artificial intelligence will consume more power than entire nations by 2030, a new report from the United Nations finds. According to a landmark report from the United Nations University, the data centers powering artificial intelligence are projected to consume 945 https://t.
Token costs are becoming one of the hottest topics for any enterprise I talk with right now. It’s very bullish for AI in general because it means these systems are being used at a scale that wasn’t contemplated before. It also gives way to another form of differentiation that
Created 2026-06-11 · Updated 2026-06-11
Edit rule text
[pharma_trial_data_or_market_dynamics_without_systems_context]
Learned 3 rejections Active
Exclude posts that announce pharma trial results, GLP-1 market pricing, weight loss drug competition, or clinical trial outcomes as standalone news without connecting to broader healthcare delivery, access, equity, or system-level implications. Market pricing or trial outcome alone is insufficient.
Posts reporting pharmaceutical trial data, drug market pricing, or competitive dynamics without healthcare systems analysis
3 example posts
This week brought even more exciting news in the field of HIV treatment! A once-weekly Lenacapavir-based pill has now demonstrated similar outcomes to daily antiretrovirals (ARVs), achieving non-inferiority. This development could significantly impact treatment adherence and http
🔔Survodutide SYNCHRONIZE-1 full paper is out for Boehringer-Ingelheim’s GLP-1/Glucagon study for weight loss. ▪️13% loss on 6mg (treatment regimen estimand) ▪️Huge treatment discontinuations due to AEs: 20% ▪️Nausea 65%, Vomiting 45% ▪️Placebo was killer. 5.4% WL. Wow. ▪️60% htt
Survodutide Once Weekly for the Treatment of Adults with Obesity: @NEJM 🥸 For @Boehringer, the advantage of Survodutide is probably liver and visceral fat reduction, not greater weight loss! Probably indication in obesity-+ MASH rather than head-to-head share against https://t.
Created 2026-06-11 · Updated 2026-06-11
Edit rule text
[unvalidated_ai_capability_hype_tangential_to_healthcare]
Learned 3 rejections Active
Exclude posts about general AI capabilities (coding agents, foundation models, domain-specific models, inference costs) that use healthcare as a tangential example or loose frame without demonstrating validated healthcare application or operational impact.
Posts about AI model capabilities, coding agents, or technical breakthroughs that mention healthcare only loosely or speculatively
3 example posts
This is crazy There are now dental practice management consultants who go to dentist offices throughout America and teach them how to increase profits by telling patients they need treatments they don’t need Dentist offices are taught on “Finding” more issues during exams, http
An interesting proposal for restoring human versus machine chess competition was to limit the machine to the same energy expenditure as the human brain. A worthy challenge, considering the amount of power required for AI data centers!
Follow the signal: we are in a compute-constrained world. And that means power doesn’t sit with the models. It sits with the infrastructure. Most frontier labs don’t own the means of producing Intelligence. They rent it. A handful of companies provide compute that everyone htt
Created 2026-06-11 · Updated 2026-06-11
Edit rule text
[unvalidated_enterprise_ai_adoption_macro]
Learned 3 rejections Active
Exclude posts about enterprise software transition to agents, token cost models, control planes, or AI system architecture that treat healthcare as incidental context rather than the specific domain of analysis, or that lack validation of healthcare operational impact.
Posts about enterprise AI adoption, agentic systems, or token economics that are too macro/generalist without healthcare specificity
3 example posts
SOFTWARE IS DEAD: "The software companies frankly got fat & happy. I think what will happen over time is ChatGPT & Claude will end up sitting on top of basically the entire enterprise software stack & almost everything else will end up being a dumb data pipe into thos
Token costs are becoming one of the hottest topics for any enterprise I talk with right now. It’s very bullish for AI in general because it means these systems are being used at a scale that wasn’t contemplated before. It also gives way to another form of differentiation that
"Are there even any feature moats left in B2B? They are at best, short-lived. In fact, we were able to rebuild SaaStr's AI VP Marketing in just ... an hour on @Lovable. Moats that still work: hardware, network effects, data, security and compliance. with @ElenaVerna Head of
Created 2026-06-11 · Updated 2026-06-11
Edit rule text
[tangential_elon_spacex_energy]
Learned 3 rejections Active
Exclude posts that focus on SpaceX, Starship, orbital data centers, energy availability, or Elon Musk's companies/statements unless they directly analyze a specific healthcare delivery or clinical application problem. Energy and infrastructure narratives alone are not healthcare strategy.
Posts about Elon Musk, SpaceX, energy infrastructure, or space-based computing with only loose healthcare framing
3 example posts
🇺🇸 SpaceX is building data centers in orbit. Their new AI satellites have a 70m wingspan, 150 kW of compute power, and a liquid radiator cooling system in space. The servers are officially leaving the ground. Writer: Val https://t.co/JIDB0icCXy
Many people think SpaceX is overvalued at $2 trillion. I honestly see it the opposite way. To me, SpaceX is so much more than just rocket company... I believe it is the infrastructure company for the next era of humanity. The launch business is the foundation. SpaceX made rock
GOOD NEWS 🇺🇸 Tesla AI is literally leaving the planet as a wild new job posting for a Space Radiation Engineer confirms that Tesla and SpaceX are actively co-developing a space-based AI data center network 🔥 This is no longer just a concept. They are literally planning to put th
Created 2026-06-11 · Updated 2026-06-11
Edit rule text
[political_outrage_or_regulatory_scandal_without_healthcare_analysis]
Learned 3 rejections Active
Exclude posts that report on healthcare fraud, regulatory violations, political scandals, or attorney general lawsuits without analyzing root causes, systemic failures, or healthcare operational lessons. Reporting the headline alone without healthcare systems insight does not qualify.
Posts reporting political outrage, regulatory scandal, or fraud allegations without systems-level healthcare analysis
3 example posts
🚨 JUST IN: The Attorney General of Florida just SUED to hold OpenAI CEO Sam Altman LIABLE for fueling violence and physical harm AG Uthmeier read off the fact a young person committed SUIC*DE based on a ChatGPT conversation "23-year-old Zane Shamblin repeatedly told ChatGPT he
My job is to make sure your surgery center never gets built. Eleven years, and I have never lost. The kid had it all lined up. Board-certified, two partners, a lease on a space where he could do the same procedures my client does across town. He showed up to the hearing with a
PacificSource pulled out as Lane County Medicaid insurer, Tina Kotek’s OR Health Authority replaced it with Trillium, forcing 96k to join Trillium. Trillium’s parent co gave Kotek’s campaign $50k two months later. OR AG (D) said in ‘22 Trillium parent co gave Kotek’s”took advan
Created 2026-06-11 · Updated 2026-06-11
Edit rule text
[elon_musk_infrastructure_tangent]
Learned 3 rejections Active
Exclude posts that discuss SpaceX, Starlink, Elon Musk's energy statements, orbital data centers, or compute infrastructure unless they explicitly analyze how this infrastructure solves a specific healthcare problem or enables a healthcare innovation with clinical or operational impact.
Posts about Elon Musk, SpaceX, energy infrastructure, or compute capacity that lack direct healthcare application or systems analysis
3 example posts
🇺🇸 SpaceX is building data centers in orbit. Their new AI satellites have a 70m wingspan, 150 kW of compute power, and a liquid radiator cooling system in space. The servers are officially leaving the ground. Writer: Val https://t.co/JIDB0icCXy
Many people think SpaceX is overvalued at $2 trillion. I honestly see it the opposite way. To me, SpaceX is so much more than just rocket company... I believe it is the infrastructure company for the next era of humanity. The launch business is the foundation. SpaceX made rock
GOOD NEWS 🇺🇸 Tesla AI is literally leaving the planet as a wild new job posting for a Space Radiation Engineer confirms that Tesla and SpaceX are actively co-developing a space-based AI data center network 🔥 This is no longer just a concept. They are literally planning to put th
Created 2026-06-11 · Updated 2026-06-11
Edit rule text
[regulatory_fraud_scandal_or_policy_outrage_without_systems_analysis]
Learned 3 rejections Active
Exclude posts that report healthcare fraud, regulatory failures, or policy scandals as outrage narratives without explaining underlying system incentives, misaligned reimbursement, or structural causes. The post must analyze why the system failed, not just condemn bad actors.
Posts reporting healthcare fraud, regulatory scandal, or policy outrage without analyzing root causes or system incentives
3 example posts
🚨 JUST IN: The Attorney General of Florida just SUED to hold OpenAI CEO Sam Altman LIABLE for fueling violence and physical harm AG Uthmeier read off the fact a young person committed SUIC*DE based on a ChatGPT conversation "23-year-old Zane Shamblin repeatedly told ChatGPT he
USDA's Chief Information Officer Sam Berry says AI could help the USDA crack down on billions of dollars in SNAP fraud each year. "SNAP is a $100B-per-year taxpayer-funded program. That's an area where we really want to have all angles of the data available so that we can deploy
PacificSource pulled out as Lane County Medicaid insurer, Tina Kotek’s OR Health Authority replaced it with Trillium, forcing 96k to join Trillium. Trillium’s parent co gave Kotek’s campaign $50k two months later. OR AG (D) said in ‘22 Trillium parent co gave Kotek’s”took advan
Created 2026-06-11 · Updated 2026-06-11
Edit rule text
[generalist_macro_or_labor_disruption_without_healthcare_specificity]
Learned 3 rejections Active
Exclude posts that discuss labor disruption, workforce automation, enterprise AI adoption, or macro economic trends with only loose healthcare framing or no healthcare application at all. The post must analyze healthcare-specific workforce, reimbursement, or operational impacts.
Posts about AI, robotics, or labor market disruption applied broadly without healthcare-specific systems analysis
3 example posts
"Are there even any feature moats left in B2B? They are at best, short-lived. In fact, we were able to rebuild SaaStr's AI VP Marketing in just ... an hour on @Lovable. Moats that still work: hardware, network effects, data, security and compliance. with @ElenaVerna Head of
A framework I wish more Founders and VCs used when discussing US Manufacturing: Automation Value = Labor Intensity × Labor Eliminated Everyone gets excited about robots and automation. Almost nobody asks: “How much of revenue is labor?” If labor is 60% of revenue and AI http
I just looked at a home service firms ServiceTitan dashboard and saw something interesting. A Pre-seed company with AI-Supported Human CSRs is out converting an AI CSR business that's raised over $125M by >50% Head-to-Head win rate is 100% Why? Supercharging humans resonat
Created 2026-06-11 · Updated 2026-06-11
Edit rule text
[non_healthcare_business_or_policy_with_loose_framing]
Learned 3 rejections Active
Exclude posts about non-healthcare businesses, non-healthcare policy changes, labor market macro trends, or regulatory enforcement that only tangentially reference healthcare or lack clear healthcare systems substance.
Posts about non-healthcare companies, government policy, labor markets, or regulatory issues loosely framed as healthcare-adjacent
3 example posts
@swago_baby You’re not dumb please.. WhatsApp stays free for personal users (no ads in chats) and makes money almost entirely from businesses: - Main revenue: Businesses pay per conversation via the WhatsApp Business API (for customer support, notifications, orders, etc.). Fees
USDA's Chief Information Officer Sam Berry says AI could help the USDA crack down on billions of dollars in SNAP fraud each year. "SNAP is a $100B-per-year taxpayer-funded program. That's an area where we really want to have all angles of the data available so that we can deploy
PacificSource pulled out as Lane County Medicaid insurer, Tina Kotek’s OR Health Authority replaced it with Trillium, forcing 96k to join Trillium. Trillium’s parent co gave Kotek’s campaign $50k two months later. OR AG (D) said in ‘22 Trillium parent co gave Kotek’s”took advan
Created 2026-06-10 · Updated 2026-06-10
Edit rule text
[glp1_peptide_market_dynamics_without_systems_lens]
Learned 3 rejections Active
Exclude posts focused on GLP-1/peptide market competition, pricing speculation, weight loss market trends, or comparative drug efficacy announcements without analyzing healthcare delivery, access, or system-level outcomes.
Posts about GLP-1 and peptide market pricing, competitive dynamics, or weight loss trends without healthcare systems analysis
3 example posts
🔔Survodutide SYNCHRONIZE-1 full paper is out for Boehringer-Ingelheim’s GLP-1/Glucagon study for weight loss. ▪️13% loss on 6mg (treatment regimen estimand) ▪️Huge treatment discontinuations due to AEs: 20% ▪️Nausea 65%, Vomiting 45% ▪️Placebo was killer. 5.4% WL. Wow. ▪️60% htt
Survodutide Once Weekly for the Treatment of Adults with Obesity: @NEJM 🥸 For @Boehringer, the advantage of Survodutide is probably liver and visceral fat reduction, not greater weight loss! Probably indication in obesity-+ MASH rather than head-to-head share against https://t.
Let me clear up the microscope fight first. 🔬 Everyone’s arguing over “hairs” and “fibers” in horse paste. But nobody talks about formulation science. Horse paste = oral suspension. Binders. Thickeners. Made to swallow. Injectable = sterile. Made for veins. Different rules. J
Created 2026-06-10 · Updated 2026-06-10
Edit rule text
[spacex_elon_energy_infrastructure]
Learned 3 rejections Active
Exclude posts about SpaceX, Elon Musk, orbital/space-based infrastructure, energy systems, or compute-in-space initiatives. These are non-healthcare infrastructure narratives using loose healthcare framing or energy cost arguments.
Posts about SpaceX, Elon Musk, energy infrastructure, or orbital data centers framed tangentially to healthcare.
3 example posts
🇺🇸 SpaceX is building data centers in orbit. Their new AI satellites have a 70m wingspan, 150 kW of compute power, and a liquid radiator cooling system in space. The servers are officially leaving the ground. Writer: Val https://t.co/JIDB0icCXy
Many people think SpaceX is overvalued at $2 trillion. I honestly see it the opposite way. To me, SpaceX is so much more than just rocket company... I believe it is the infrastructure company for the next era of humanity. The launch business is the foundation. SpaceX made rock
GOOD NEWS 🇺🇸 Tesla AI is literally leaving the planet as a wild new job posting for a Space Radiation Engineer confirms that Tesla and SpaceX are actively co-developing a space-based AI data center network 🔥 This is no longer just a concept. They are literally planning to put th
Created 2026-06-10 · Updated 2026-06-10
Edit rule text
[ai_agent_technical_tangent_no_healthcare_validation]
Learned 3 rejections Active
Exclude posts discussing AI agents, agentic workflows, control planes, multi-agent systems, or agent-specific technical architectures as generalist AI capability announcements or hype unless the post demonstrates a specific, validated healthcare use case or operational problem being solved.
Posts about agentic AI systems, control planes, or agent architecture as technical capability discussion without healthcare validation or application
3 example posts
Anthropic engineer James Brady: "Every agent in production lies. We measured it. The good ones lie less, the great ones catch the lie before the user does." In 29 minutes, he walks through the verification stack he built and the patterns the Claude Code team adopted to keep ht
Anthropic engineer: "The agent doesn't remember anything. So we built a second set of agents whose only job is to dream about the first ones." They wait until you log off, then reopen every session you ran, fact-check the first agents, merge the duplicates, and burn anything ht
Enterprise software was priced per seat. That model is breaking. Before: SaaS = seats. Predictable. Human-driven usage. Revenue tied to headcount. Now: agents hit software systems more frequently than any human ever could. Aaron Levie (@levie) on The MAD Podcast: agent-driven h
Created 2026-06-10 · Updated 2026-06-10
Edit rule text
[unvalidated_speculative_medical_or_security_claims]
Learned 3 rejections Active
Exclude posts that make broad, sensationalist, or unverified claims about medical treatments, AI security discoveries, or healthcare innovations without peer review, clinical data, or official validation. Claims like 'microplastics found in 58% of plaques' or 'AI found zero-days in all major browsers' require independent corroboration.
Posts making speculative or unvalidated claims about medical breakthroughs, AI vulnerabilities, or healthcare innovations
3 example posts
Researchers have announced positive results from the first-in-human Phase 1 clinical trial of a universal coronavirus vaccine designed using artificial intelligence. Unlike existing COVID-19 vaccines that target specific strains or variants of SARS-CoV-2, this experimental https:
Most under-discussed cardiology study of the last 2 years just hit NEJM. Microplastics found inside the carotid plaques of 58% of patients undergoing endarterectomy. 4.5× higher rate of heart attack, stroke, or death over 34 months. This changes the atherosclerosis story. http
Are pulse oximeters risking Black patients' lives? Episode 4 of the Intention to Treat podcast explores the story of the pulse oximeter and the deadly consequences when a critical medical test doesn’t work on dark skin. Listen on Apple or wherever you get your podcasts. 🎧 https:/
Created 2026-06-10 · Updated 2026-06-10
Edit rule text
[spacex_energy_infrastructure_tangent]
Learned 3 rejections Active
Exclude posts that focus on SpaceX orbital infrastructure, AI satellites, space-based data centers, or energy supply chains as abstract macro trends. Healthcare connection must be explicit and substantive, not merely implied through 'AI will need energy.'
Posts about SpaceX, satellite infrastructure, or energy compute capacity without healthcare application specificity
3 example posts
🇺🇸 SpaceX is building data centers in orbit. Their new AI satellites have a 70m wingspan, 150 kW of compute power, and a liquid radiator cooling system in space. The servers are officially leaving the ground. Writer: Val https://t.co/JIDB0icCXy
Many people think SpaceX is overvalued at $2 trillion. I honestly see it the opposite way. To me, SpaceX is so much more than just rocket company... I believe it is the infrastructure company for the next era of humanity. The launch business is the foundation. SpaceX made rock
GOOD NEWS 🇺🇸 Tesla AI is literally leaving the planet as a wild new job posting for a Space Radiation Engineer confirms that Tesla and SpaceX are actively co-developing a space-based AI data center network 🔥 This is no longer just a concept. They are literally planning to put th
Created 2026-06-10 · Updated 2026-06-10
Edit rule text
[anecdotal_single_case_or_operational_narrative_without_systems_context]
Learned 3 rejections Active
Exclude posts that rely on a single anecdote, personal cost narrative, individual case story, or one founder's operational experience without connecting to systemic healthcare patterns, incentive structures, or market-wide implications.
Posts sharing personal anecdotes, single patient cases, or individual operational stories without broader healthcare systems analysis
3 example posts
My job is to make sure your surgery center never gets built. Eleven years, and I have never lost. The kid had it all lined up. Board-certified, two partners, a lease on a space where he could do the same procedures my client does across town. He showed up to the hearing with a
PacificSource pulled out as Lane County Medicaid insurer, Tina Kotek’s OR Health Authority replaced it with Trillium, forcing 96k to join Trillium. Trillium’s parent co gave Kotek’s campaign $50k two months later. OR AG (D) said in ‘22 Trillium parent co gave Kotek’s”took advan
I called a woman in Dayton to tell her she was about to overpay $4,500, for a horrible health system based procedure. She was scheduled for a knee replacement. Her PPO had her on the hook for a $4,500 deductible. The call was on behalf of her employer: “See one of these three
Created 2026-06-10 · Updated 2026-06-10
Edit rule text
[clinical_anecdote_no_systems_insight]
Learned 3 rejections Active
Exclude posts that report isolated clinical trial data, single case observations, or pharma/biotech trial results WITHOUT connecting to healthcare system incentives, reimbursement models, patient access, or operational workflow changes. Raw trial data announcements do not meet the threshold for healthcare tech analysis.
Posts reporting single clinical observations, trial results, or anecdotes without healthcare systems context, business impact, or operational implications
3 example posts
Researchers have announced positive results from the first-in-human Phase 1 clinical trial of a universal coronavirus vaccine designed using artificial intelligence. Unlike existing COVID-19 vaccines that target specific strains or variants of SARS-CoV-2, this experimental https:
Approved 20 years ago as a diabetes treatment, GLP-1 drugs have been found to help patients reduce weight, changing the lives of more than 30 million people in the U.S. But there also have been troubling side effects reported. https://t.co/zzRamKnedu
Most under-discussed cardiology study of the last 2 years just hit NEJM. Microplastics found inside the carotid plaques of 58% of patients undergoing endarterectomy. 4.5× higher rate of heart attack, stroke, or death over 34 months. This changes the atherosclerosis story. http
Created 2026-06-10 · Updated 2026-06-10
Edit rule text
[unvalidated_anthropic_security_hype]
Learned 3 rejections Active
Exclude posts that sensationalize AI security capabilities (finding zero-days, embedded NSA deployment, offensive cyber operations) without independent verification, peer review, or documented healthcare/clinical context.
Posts making exaggerated or unverified claims about AI models discovering security vulnerabilities or being used in government cyber operations.
3 example posts
🚨 do you understand what just happened with Anthropic.. Their internal model, Claude Mythos, found previously unknown security holes in every major browser and every major operating system. Not one. All of them at once. And Anthropic did something a hype-driven industry almost
Mythos AI is being used by National Security Agency in offensive cyber operations / cyberattacks. Anthropic has even embedded engineers inside the NSA to help deploy the model. Are frontier AI labs becoming active contractors in state cyber conflict? The live-ops role is still ht
🔥 AI just found 21 zero-days in FFmpeg. That’s the video library bundled inside many apps, tools, containers, and devices. Some bugs sat untouched for 15–20 years. Google Chrome also dropped PATCHES for a record 429 vulnerabilities this week. Read: https://t.co/6MEVD9ufxu
Created 2026-06-10 · Updated 2026-06-10
Edit rule text
[anecdotal_healthcare_cost_narrative]
Learned 3 rejections Active
Exclude posts that tell a single healthcare cost or access story (e.g., one patient's bill, one provider's experience) without connecting to healthcare systems policy, pricing mechanisms, or broader operational patterns.
Single patient or provider anecdotes about healthcare costs/access without systems-level insight
3 example posts
My job is to make sure your surgery center never gets built. Eleven years, and I have never lost. The kid had it all lined up. Board-certified, two partners, a lease on a space where he could do the same procedures my client does across town. He showed up to the hearing with a
PacificSource pulled out as Lane County Medicaid insurer, Tina Kotek’s OR Health Authority replaced it with Trillium, forcing 96k to join Trillium. Trillium’s parent co gave Kotek’s campaign $50k two months later. OR AG (D) said in ‘22 Trillium parent co gave Kotek’s”took advan
I called a woman in Dayton to tell her she was about to overpay $4,500, for a horrible health system based procedure. She was scheduled for a knee replacement. Her PPO had her on the hook for a $4,500 deductible. The call was on behalf of her employer: “See one of these three
Created 2026-06-09 · Updated 2026-06-09
Edit rule text
[elon_musk_spacex_energy_hype]
Learned 3 rejections Active
Exclude posts focused on SpaceX satellites, orbital AI data centers, Tesla energy projects, or Elon Musk's business ventures. Healthcare framing must be explicit and substantive, not loose metaphorical connections to compute infrastructure.
Posts about Elon Musk, SpaceX, Tesla, or space-based infrastructure with only tangential healthcare framing
3 example posts
🇺🇸 SpaceX is building data centers in orbit. Their new AI satellites have a 70m wingspan, 150 kW of compute power, and a liquid radiator cooling system in space. The servers are officially leaving the ground. Writer: Val https://t.co/JIDB0icCXy
Many people think SpaceX is overvalued at $2 trillion. I honestly see it the opposite way. To me, SpaceX is so much more than just rocket company... I believe it is the infrastructure company for the next era of humanity. The launch business is the foundation. SpaceX made rock
GOOD NEWS 🇺🇸 Tesla AI is literally leaving the planet as a wild new job posting for a Space Radiation Engineer confirms that Tesla and SpaceX are actively co-developing a space-based AI data center network 🔥 This is no longer just a concept. They are literally planning to put th
Created 2026-06-09 · Updated 2026-06-09
Edit rule text
[clinical_anecdote_or_observation_no_systems_context]
Learned 3 rejections Active
Exclude posts that report a single clinical finding, personal healthcare anecdote, or isolated trial observation without explaining systemic implications, adoption barriers, cost-effectiveness, or broader healthcare impact.
Posts sharing single clinical observations, anecdotes, or trial outcomes without healthcare systems analysis
3 example posts
Researchers have announced positive results from the first-in-human Phase 1 clinical trial of a universal coronavirus vaccine designed using artificial intelligence. Unlike existing COVID-19 vaccines that target specific strains or variants of SARS-CoV-2, this experimental https:
Most under-discussed cardiology study of the last 2 years just hit NEJM. Microplastics found inside the carotid plaques of 58% of patients undergoing endarterectomy. 4.5× higher rate of heart attack, stroke, or death over 34 months. This changes the atherosclerosis story. http
Are pulse oximeters risking Black patients' lives? Episode 4 of the Intention to Treat podcast explores the story of the pulse oximeter and the deadly consequences when a critical medical test doesn’t work on dark skin. Listen on Apple or wherever you get your podcasts. 🎧 https:/
Created 2026-06-09 · Updated 2026-06-09
Edit rule text
[non_healthcare_business_model_or_policy_with_loose_framing]
Learned 3 rejections Active
Exclude posts discussing enterprise SaaS pricing models, government agencies (USDA, CMS directories), WhatsApp business models, or regulatory frameworks that mention healthcare in passing but whose primary subject is non-healthcare business or policy mechanics.
Posts about non-healthcare business models, enterprise software pricing, or regulatory policy that apply loose healthcare framing without substantive healthcare systems connection.
3 example posts
@swago_baby You’re not dumb please.. WhatsApp stays free for personal users (no ads in chats) and makes money almost entirely from businesses: - Main revenue: Businesses pay per conversation via the WhatsApp Business API (for customer support, notifications, orders, etc.). Fees
USDA's Chief Information Officer Sam Berry says AI could help the USDA crack down on billions of dollars in SNAP fraud each year. "SNAP is a $100B-per-year taxpayer-funded program. That's an area where we really want to have all angles of the data available so that we can deploy
"Are there even any feature moats left in B2B? They are at best, short-lived. In fact, we were able to rebuild SaaStr's AI VP Marketing in just ... an hour on @Lovable. Moats that still work: hardware, network effects, data, security and compliance. with @ElenaVerna Head of
Created 2026-06-09 · Updated 2026-06-09
Edit rule text
[sensationalist_unverified_claims]
Learned 3 rejections Active
Exclude posts that make sensational claims about AI finding thousands of zero-days, discovering major security vulnerabilities, or enabling nation-state cyberattacks without credible, detailed evidence. Also exclude unvalidated medical claims about horse paste formulations or peptide treatments presented as established fact.
Posts making dramatic or unsubstantiated claims about AI capabilities, breaches, or medical interventions without proper evidence.
3 example posts
Let me clear up the microscope fight first. 🔬 Everyone’s arguing over “hairs” and “fibers” in horse paste. But nobody talks about formulation science. Horse paste = oral suspension. Binders. Thickeners. Made to swallow. Injectable = sterile. Made for veins. Different rules. J
🚨 do you understand what just happened with Anthropic.. Their internal model, Claude Mythos, found previously unknown security holes in every major browser and every major operating system. Not one. All of them at once. And Anthropic did something a hype-driven industry almost
Mythos AI is being used by National Security Agency in offensive cyber operations / cyberattacks. Anthropic has even embedded engineers inside the NSA to help deploy the model. Are frontier AI labs becoming active contractors in state cyber conflict? The live-ops role is still ht
Created 2026-06-09 · Updated 2026-06-09
Edit rule text
[ai_infrastructure_tangent_no_healthcare_lens]
Learned 3 rejections Active
Exclude posts that focus on AI infrastructure (compute costs, energy bottlenecks, NVIDIA chips, ASML semiconductors, space-based data centers, GPU pricing) unless they explicitly connect to a healthcare application or healthcare business model impact. General infrastructure hype does not qualify.
Posts about AI compute, energy, semiconductor manufacturing, or infrastructure scaled to space/chips without analyzing healthcare-specific implications or use cases.
3 example posts
GOOD NEWS 🇺🇸 Tesla AI is literally leaving the planet as a wild new job posting for a Space Radiation Engineer confirms that Tesla and SpaceX are actively co-developing a space-based AI data center network 🔥 This is no longer just a concept. They are literally planning to put th
NEWS: ASML has invited Elon Musk to speak at its internal technology conference, Dutch outlet NUnl reports. Elon Musk's response: "ASML should be treasured and supported. It is arguably the greatest company in Europe." ASML builds the EUV lithography machines that every advance
Elon Musk just explained why the SpaceX IPO is an energy story and the energy constraint is why he believes space becomes the only viable path for AI to scale (Save this). The argument he is making is one of the most important and least understood things happening in technology
Created 2026-06-08 · Updated 2026-06-08
Edit rule text
[elon_musk_or_spacex_energy_infrastructure_tangent]
Learned 3 rejections Active
Exclude posts primarily about Elon Musk announcements, Tesla energy rollout, or SpaceX infrastructure where the healthcare angle is speculative or absent. These are non-healthcare business news with loose framing; they must describe documented healthcare applications, not hypothetical energy benefits.
Posts about Elon Musk, Tesla, SpaceX, or energy/space infrastructure with tenuous healthcare framing.
3 example posts
GOOD NEWS 🇺🇸 Tesla AI is literally leaving the planet as a wild new job posting for a Space Radiation Engineer confirms that Tesla and SpaceX are actively co-developing a space-based AI data center network 🔥 This is no longer just a concept. They are literally planning to put th
NEWS: ASML has invited Elon Musk to speak at its internal technology conference, Dutch outlet NUnl reports. Elon Musk's response: "ASML should be treasured and supported. It is arguably the greatest company in Europe." ASML builds the EUV lithography machines that every advance
Elon Musk just explained why the SpaceX IPO is an energy story and the energy constraint is why he believes space becomes the only viable path for AI to scale (Save this). The argument he is making is one of the most important and least understood things happening in technology
Created 2026-06-08 · Updated 2026-06-08
Edit rule text
[ai_company_product_metrics_and_announcements]
Learned 3 rejections Active
Exclude posts that announce new AI model capabilities, product launches, partnerships, or deployment milestones for AI companies (Anthropic, OpenAI, NVIDIA) unless the post explicitly demonstrates a validated healthcare use case with evidence of clinical benefit or healthcare system adoption.
Posts announcing AI company product launches, security features, or business partnerships without validating healthcare application or clinical relevance.
3 example posts
🚨 do you understand what just happened with Anthropic.. Their internal model, Claude Mythos, found previously unknown security holes in every major browser and every major operating system. Not one. All of them at once. And Anthropic did something a hype-driven industry almost
Mythos AI is being used by National Security Agency in offensive cyber operations / cyberattacks. Anthropic has even embedded engineers inside the NSA to help deploy the model. Are frontier AI labs becoming active contractors in state cyber conflict? The live-ops role is still ht
🔥 AI just found 21 zero-days in FFmpeg. That’s the video library bundled inside many apps, tools, containers, and devices. Some bugs sat untouched for 15–20 years. Google Chrome also dropped PATCHES for a record 429 vulnerabilities this week. Read: https://t.co/6MEVD9ufxu
Created 2026-06-08 · Updated 2026-06-08
Edit rule text
[non_healthcare_company_or_infrastructure_hype]
Learned 3 rejections Active
Exclude posts that are primarily about non-healthcare technology companies (Tesla, SpaceX, ASML, NVIDIA) or energy/space infrastructure, even if the post mentions AI or tangentially connects to healthcare delivery. The post must center on a healthcare application or healthcare system challenge, not a tech company's business or product roadmap.
Posts about Elon Musk, Tesla, SpaceX, NVIDIA chips, ASML manufacturing, or energy infrastructure with only tangential healthcare framing.
3 example posts
GOOD NEWS 🇺🇸 Tesla AI is literally leaving the planet as a wild new job posting for a Space Radiation Engineer confirms that Tesla and SpaceX are actively co-developing a space-based AI data center network 🔥 This is no longer just a concept. They are literally planning to put th
NEWS: ASML has invited Elon Musk to speak at its internal technology conference, Dutch outlet NUnl reports. Elon Musk's response: "ASML should be treasured and supported. It is arguably the greatest company in Europe." ASML builds the EUV lithography machines that every advance
Elon Musk just explained why the SpaceX IPO is an energy story and the energy constraint is why he believes space becomes the only viable path for AI to scale (Save this). The argument he is making is one of the most important and least understood things happening in technology
Created 2026-06-08 · Updated 2026-06-08
Edit rule text
[clinical_trial_announcement_only]
Learned 3 rejections Active
Exclude posts that merely announce clinical trial results, FDA approvals, or drug efficacy data without analyzing how these advances address healthcare access, cost, equity, delivery infrastructure, or market adoption barriers.
Posts reporting pharmaceutical trial results, approvals, or clinical data without analyzing healthcare system impact or adoption barriers.
3 example posts
Approved 20 years ago as a diabetes treatment, GLP-1 drugs have been found to help patients reduce weight, changing the lives of more than 30 million people in the U.S. But there also have been troubling side effects reported. https://t.co/zzRamKnedu
🚨 T-DXd just went tumor-agnostic — and it changed a lot about how we think about HER2 testing across solid tumors 🔬 #ASCO26 FDA granted accelerated approval to T-DXd for patients with unresectable or metastatic HER2-positive solid tumors after prior systemic treatment —
The scores don’t pick your binders. The filters don't care either. The assay does. I worked with a design partner to produce true de-novo hits to 2 immuno-oncology targets. ~1,500 de novo designs → 96 synthesized → 89 run on SPR → 5 confirmed binders; 19–195 nM No https:
Created 2026-06-08 · Updated 2026-06-08
Edit rule text
[unvalidated_anthropic_mythos_security_claims]
Learned 3 rejections Active
Exclude posts hyping unvalidated AI security capabilities (e.g., Anthropic Claude Mythos finding zero-day vulnerabilities) or claims about NSA partnerships in offensive cyber operations. These are cybersecurity tangents without healthcare application or evidence of real-world impact.
Posts making unsubstantiated claims about Anthropic's Claude Mythos model finding security vulnerabilities or detecting zero-days without clinical or healthcare validation.
3 example posts
🚨 do you understand what just happened with Anthropic.. Their internal model, Claude Mythos, found previously unknown security holes in every major browser and every major operating system. Not one. All of them at once. And Anthropic did something a hype-driven industry almost
Mythos AI is being used by National Security Agency in offensive cyber operations / cyberattacks. Anthropic has even embedded engineers inside the NSA to help deploy the model. Are frontier AI labs becoming active contractors in state cyber conflict? The live-ops role is still ht
🔥 AI just found 21 zero-days in FFmpeg. That’s the video library bundled inside many apps, tools, containers, and devices. Some bugs sat untouched for 15–20 years. Google Chrome also dropped PATCHES for a record 429 vulnerabilities this week. Read: https://t.co/6MEVD9ufxu
Created 2026-06-08 · Updated 2026-06-08
Edit rule text
[tangential_non_healthcare_narrative]
Learned 3 rejections Active
Exclude posts about energy infrastructure, compute capacity, space technology, semiconductor manufacturing, or business model debates that merely mention healthcare or use healthcare figures as examples—unless the post directly analyzes how that technology solves a specific healthcare system problem.
Posts about technology, energy, or business topics loosely framed as healthcare-relevant but primarily focused on non-healthcare domains.
3 example posts
NEWS: ASML has invited Elon Musk to speak at its internal technology conference, Dutch outlet NUnl reports. Elon Musk's response: "ASML should be treasured and supported. It is arguably the greatest company in Europe." ASML builds the EUV lithography machines that every advance
Elon Musk just explained why the SpaceX IPO is an energy story and the energy constraint is why he believes space becomes the only viable path for AI to scale (Save this). The argument he is making is one of the most important and least understood things happening in technology
Elon is now printing over $26B per year… purely from selling compute to his competitors 😂 AI companies bet everything on one path and got crushed by compute limits. Now they’re begging Elon for GPUs
Created 2026-06-08 · Updated 2026-06-08
Edit rule text
[healthcare_fraud_regulatory_scandal_without_systems_analysis]
Learned 3 rejections Active
Exclude posts that report healthcare fraud, prior authorization abuse, provider fraud, or regulatory scandals as anecdotes or single incidents without connecting to broader healthcare system failures, market incentives, or policy recommendations.
Posts reporting healthcare fraud, regulatory violations, or pricing scandals as isolated incidents without analyzing systemic healthcare policy or market structure.
3 example posts
My job is to make sure your surgery center never gets built. Eleven years, and I have never lost. The kid had it all lined up. Board-certified, two partners, a lease on a space where he could do the same procedures my client does across town. He showed up to the hearing with a
PacificSource pulled out as Lane County Medicaid insurer, Tina Kotek’s OR Health Authority replaced it with Trillium, forcing 96k to join Trillium. Trillium’s parent co gave Kotek’s campaign $50k two months later. OR AG (D) said in ‘22 Trillium parent co gave Kotek’s”took advan
I called a woman in Dayton to tell her she was about to overpay $4,500, for a horrible health system based procedure. She was scheduled for a knee replacement. Her PPO had her on the hook for a $4,500 deductible. The call was on behalf of her employer: “See one of these three
Created 2026-06-08 · Updated 2026-06-08
Edit rule text
[tangential_elon_musk_energy_infrastructure]
Learned 3 rejections Active
Exclude posts that feature Elon Musk, SpaceX, energy infrastructure, or GPU/hardware supply chain narratives unless they contain specific, evidence-based reasoning about how that infrastructure solves a concrete healthcare delivery or research problem.
Posts about Elon Musk, SpaceX, energy, or hardware that loosely frame healthcare implications without substantive healthcare analysis.
3 example posts
NEWS: ASML has invited Elon Musk to speak at its internal technology conference, Dutch outlet NUnl reports. Elon Musk's response: "ASML should be treasured and supported. It is arguably the greatest company in Europe." ASML builds the EUV lithography machines that every advance
Elon Musk just explained why the SpaceX IPO is an energy story and the energy constraint is why he believes space becomes the only viable path for AI to scale (Save this). The argument he is making is one of the most important and least understood things happening in technology
Elon is now printing over $26B per year… purely from selling compute to his competitors 😂 AI companies bet everything on one path and got crushed by compute limits. Now they’re begging Elon for GPUs
Created 2026-06-08 · Updated 2026-06-08
Edit rule text
[clinical_trial_or_drug_announcement_only]
Learned 3 rejections Active
Exclude posts that are primarily news announcements or data summaries of clinical trial results, FDA approvals, or drug/biotech milestones without analyzing healthcare system implications, operational impact, cost-benefit context, or clinical-economic significance. Raw trial data ≠ healthcare tech insight.
Posts reporting pharma trial results, drug approvals, or biotech announcements without healthcare systems analysis or clinical significance
3 example posts
🚨 T-DXd just went tumor-agnostic — and it changed a lot about how we think about HER2 testing across solid tumors 🔬 #ASCO26 FDA granted accelerated approval to T-DXd for patients with unresectable or metastatic HER2-positive solid tumors after prior systemic treatment —
The scores don’t pick your binders. The filters don't care either. The assay does. I worked with a design partner to produce true de-novo hits to 2 immuno-oncology targets. ~1,500 de novo designs → 96 synthesized → 89 run on SPR → 5 confirmed binders; 19–195 nM No https:
🔥Phase 1: ABBV-706 (SEZ6-targeting ADC) in R/R SCLC & solid tumors 🆙 @NatureMedicine ☑R/R SCLC; ORR 52%; mOS 12.4 mo at RP2D 🎯RP2D confirmed at 1.8 mg/kg Q3W balancing efficacy & safety 🎙 @LaurenByersMD #LCSM @OncoAlert @Larvol https://t.co/vjUUd1GYSy
Created 2026-06-08 · Updated 2026-06-08
Edit rule text
[elon_musk_spacex_tangent]
Learned 3 rejections Active
Exclude posts primarily about Elon Musk, SpaceX, ASML, NVIDIA GPU sales, or general tech business developments where healthcare systems, clinical outcomes, or healthcare-specific applications are not the main focus.
Posts about Elon Musk, SpaceX, or non-healthcare tech figures with loose or no healthcare connection
3 example posts
NEWS: ASML has invited Elon Musk to speak at its internal technology conference, Dutch outlet NUnl reports. Elon Musk's response: "ASML should be treasured and supported. It is arguably the greatest company in Europe." ASML builds the EUV lithography machines that every advance
Elon is now printing over $26B per year… purely from selling compute to his competitors 😂 AI companies bet everything on one path and got crushed by compute limits. Now they’re begging Elon for GPUs
Elon Musk reveals why he believes the cheapest place to put AI will be space within 36 months "The availability of energy is the issue. Everywhere outside of China, electrical output is more or less flat. The output of chips is growing exponentially, but the output of https://t.
Created 2026-06-08 · Updated 2026-06-08
Edit rule text
[non_healthcare_tech_infrastructure]
Learned 3 rejections Active
Exclude posts primarily about AI infrastructure (token costs, model training, compute optimization, energy constraints, semiconductor supply chains) where healthcare is mentioned tangentially or not at all. The post must demonstrate concrete healthcare application or healthcare-specific implications.
Posts about general AI infrastructure, compute, energy, or tech topics with only loose healthcare framing
3 example posts
Token costs are becoming one of the hottest topics for any enterprise I talk with right now. It’s very bullish for AI in general because it means these systems are being used at a scale that wasn’t contemplated before. It also gives way to another form of differentiation that
Elon Musk just explained why the SpaceX IPO is an energy story and the energy constraint is why he believes space becomes the only viable path for AI to scale (Save this). The argument he is making is one of the most important and least understood things happening in technology
Elon is now printing over $26B per year… purely from selling compute to his competitors 😂 AI companies bet everything on one path and got crushed by compute limits. Now they’re begging Elon for GPUs
Created 2026-06-08 · Updated 2026-06-08
Edit rule text
[anecdotal_single_case_healthcare_cost_or_operational_narrative_without_systems_analysis]
Learned 3 rejections Active
Exclude posts that tell a single patient story, personal healthcare cost anecdote, or individual operational narrative (e.g., one surgery center, one insurer decision) without connecting to broader healthcare system patterns, policy implications, or structural analysis.
Posts sharing individual patient stories, personal healthcare cost anecdotes, or single operational examples without broader healthcare systems insight.
3 example posts
My job is to make sure your surgery center never gets built. Eleven years, and I have never lost. The kid had it all lined up. Board-certified, two partners, a lease on a space where he could do the same procedures my client does across town. He showed up to the hearing with a
PacificSource pulled out as Lane County Medicaid insurer, Tina Kotek’s OR Health Authority replaced it with Trillium, forcing 96k to join Trillium. Trillium’s parent co gave Kotek’s campaign $50k two months later. OR AG (D) said in ‘22 Trillium parent co gave Kotek’s”took advan
I called a woman in Dayton to tell her she was about to overpay $4,500, for a horrible health system based procedure. She was scheduled for a knee replacement. Her PPO had her on the hook for a $4,500 deductible. The call was on behalf of her employer: “See one of these three
Created 2026-06-08 · Updated 2026-06-08
Edit rule text
[cynical_insider_sarcasm_or_fraud_narrative_no_substance]
Learned 3 rejections Active
Exclude posts written in cynical, sarcastic, or insider-joking tone that make accusations (e.g., fabricated metrics, regulatory gaming) without providing evidence, documented sources, or substantive healthcare systems analysis.
Posts written in cynical, sarcastic insider tone that make broad accusations without evidence or substantive analysis.
3 example posts
I'm SVP now. I told you I would be. The graph went up and to the right. Nobody checked what it measured. It measured "AI enablement." I made that up last year. I'm making it up again. Last year I rolled out Copilot to 4,000 people. Nothing happened. I got promoted. Those two htt
I just looked at a home service firms ServiceTitan dashboard and saw something interesting. A Pre-seed company with AI-Supported Human CSRs is out converting an AI CSR business that's raised over $125M by >50% Head-to-Head win rate is 100% Why? Supercharging humans resonat
BREAKING: Merge Launches ‘Agent Handler’ Control AI Access, Tokenmaxxed $$$ Bills, & Stop Mass Data Leaks "We don't trust agents" "The second you connect it to tools, that's where everything goes wrong." OpenAI. Perplexity. Netflix. Uber. Mistral. Dropbox. JPMorgan.. all h
Created 2026-06-08 · Updated 2026-06-08
Edit rule text
[pharma_trial_data_or_clinical_observation_without_systems_context]
Learned 3 rejections Active
Exclude posts that announce pharmaceutical trial data, FDA approvals, clinical results, or biotech findings without discussing healthcare system integration, workflow impact, access barriers, reimbursement, or operational implementation challenges.
Posts reporting clinical trial results, drug approvals, or pharma data without healthcare systems analysis, implementation challenges, or care delivery implications.
3 example posts
🚨 T-DXd just went tumor-agnostic — and it changed a lot about how we think about HER2 testing across solid tumors 🔬 #ASCO26 FDA granted accelerated approval to T-DXd for patients with unresectable or metastatic HER2-positive solid tumors after prior systemic treatment —
🔥Phase 1: ABBV-706 (SEZ6-targeting ADC) in R/R SCLC & solid tumors 🆙 @NatureMedicine ☑R/R SCLC; ORR 52%; mOS 12.4 mo at RP2D 🎯RP2D confirmed at 1.8 mg/kg Q3W balancing efficacy & safety 🎙 @LaurenByersMD #LCSM @OncoAlert @Larvol https://t.co/vjUUd1GYSy
Scoop! Lila Sciences is in talks to raise ~$2 billion in new funding. The raise would value the “scientific superintelligence” lab at ~$8.5 billion before the new money. w/ @MichelleF_Davis, read more👇
Created 2026-06-08 · Updated 2026-06-08
Edit rule text
[elon_musk_and_non_healthcare_tech_figures]
Learned 3 rejections Active
Exclude posts that center on Elon Musk, SpaceX IPO, energy infrastructure, semiconductor manufacturing, or other tech industry figures unless the post directly analyzes their impact on healthcare system operations, clinical delivery, or healthcare economics.
Posts featuring Elon Musk, SpaceX, ASML, or other non-healthcare tech figures with only tangential framing of healthcare relevance.
3 example posts
NEWS: ASML has invited Elon Musk to speak at its internal technology conference, Dutch outlet NUnl reports. Elon Musk's response: "ASML should be treasured and supported. It is arguably the greatest company in Europe." ASML builds the EUV lithography machines that every advance
Elon Musk just explained why the SpaceX IPO is an energy story and the energy constraint is why he believes space becomes the only viable path for AI to scale (Save this). The argument he is making is one of the most important and least understood things happening in technology
Elon is now printing over $26B per year… purely from selling compute to his competitors 😂 AI companies bet everything on one path and got crushed by compute limits. Now they’re begging Elon for GPUs
Created 2026-06-08 · Updated 2026-06-08
Edit rule text
[generalist_macro_labor_without_healthcare_lens]
Learned 3 rejections Active
Exclude posts that discuss labor market disruption, humanoid robots, manufacturing automation, or workforce economics in generalist terms without connecting to healthcare workforce, clinical staffing, or healthcare-specific labor dynamics.
Posts about workforce disruption, labor market dynamics, and robotics macro trends without healthcare-specific analysis.
3 example posts
A framework I wish more Founders and VCs used when discussing US Manufacturing: Automation Value = Labor Intensity × Labor Eliminated Everyone gets excited about robots and automation. Almost nobody asks: “How much of revenue is labor?” If labor is 60% of revenue and AI http
What kind of work will become more valuable in an AI economy? @ccatalini, former head economist at Meta: "AI is getting better and better at automating anything that can be measured, as long as you have a digital trace, if you can collect it with a device, that problem will be
Humanoid robots, designed to mimic human movement and capabilities, have been making headlines in recent years, from their use as baggage handlers at Japanese airports to Tesla’s big bet on its Optimus humanoid. Market watchers have predicted that the machines will change the ht
Created 2026-06-08 · Updated 2026-06-08
Edit rule text
[insider_cynicism_or_sarcasm_without_substance]
Learned 3 rejections Active
Exclude posts that rely primarily on cynical insider sarcasm, moral outrage, or unsubstantiated claims about fraud or incompetence (e.g., 'I made that up,' 'nobody checked,' 'I told you so') without providing evidence, data, or systems-level analysis of the problem being criticized.
Posts using cynical or sarcastic insider commentary to critique healthcare/AI without substantive systems analysis.
3 example posts
I'm SVP now. I told you I would be. The graph went up and to the right. Nobody checked what it measured. It measured "AI enablement." I made that up last year. I'm making it up again. Last year I rolled out Copilot to 4,000 people. Nothing happened. I got promoted. Those two htt
My job is to make sure your surgery center never gets built. Eleven years, and I have never lost. The kid had it all lined up. Board-certified, two partners, a lease on a space where he could do the same procedures my client does across town. He showed up to the hearing with a
A framework I wish more Founders and VCs used when discussing US Manufacturing: Automation Value = Labor Intensity × Labor Eliminated Everyone gets excited about robots and automation. Almost nobody asks: “How much of revenue is labor?” If labor is 60% of revenue and AI http
Created 2026-06-08 · Updated 2026-06-08
Edit rule text
[elon_musk_and_spacex_energy_infrastructure]
Learned 3 rejections Active
Exclude posts that discuss Elon Musk, SpaceX IPO strategy, energy economics, ASML semiconductor manufacturing, or compute infrastructure as an abstract problem—even if AI is mentioned—unless the post directly connects these topics to a specific healthcare use case or clinical outcome.
Posts about Elon Musk, SpaceX, energy constraints, or non-healthcare infrastructure framed tangentially as relevant to AI scaling.
3 example posts
NEWS: ASML has invited Elon Musk to speak at its internal technology conference, Dutch outlet NUnl reports. Elon Musk's response: "ASML should be treasured and supported. It is arguably the greatest company in Europe." ASML builds the EUV lithography machines that every advance
Elon Musk just explained why the SpaceX IPO is an energy story and the energy constraint is why he believes space becomes the only viable path for AI to scale (Save this). The argument he is making is one of the most important and least understood things happening in technology
Elon is now printing over $26B per year… purely from selling compute to his competitors 😂 AI companies bet everything on one path and got crushed by compute limits. Now they’re begging Elon for GPUs
Created 2026-06-08 · Updated 2026-06-08
Edit rule text
[clinical_observation_or_trial_without_systems_context]
Learned 3 rejections Active
Exclude posts that announce clinical trial results, drug approvals, or single-study findings without explaining how these data inform healthcare delivery, reimbursement, or system-level change. Posts like 'ORR 52% mOS 12.4 mo' or 'Phase 1 results: 18/18 MRD-negative' without broader context are pure research reporting, not healthcare tech analysis.
Posts reporting clinical trial data, study results, or single observations without analyzing healthcare system implications or business models
3 example posts
🚨 T-DXd just went tumor-agnostic — and it changed a lot about how we think about HER2 testing across solid tumors 🔬 #ASCO26 FDA granted accelerated approval to T-DXd for patients with unresectable or metastatic HER2-positive solid tumors after prior systemic treatment —
The scores don’t pick your binders. The filters don't care either. The assay does. I worked with a design partner to produce true de-novo hits to 2 immuno-oncology targets. ~1,500 de novo designs → 96 synthesized → 89 run on SPR → 5 confirmed binders; 19–195 nM No https:
🔥Phase 1: ABBV-706 (SEZ6-targeting ADC) in R/R SCLC & solid tumors 🆙 @NatureMedicine ☑R/R SCLC; ORR 52%; mOS 12.4 mo at RP2D 🎯RP2D confirmed at 1.8 mg/kg Q3W balancing efficacy & safety 🎙 @LaurenByersMD #LCSM @OncoAlert @Larvol https://t.co/vjUUd1GYSy
Created 2026-06-07 · Updated 2026-06-07
Edit rule text
[elon_musk_or_spacex_energy_narrative]
Learned 3 rejections Active
Exclude posts about Elon Musk's statements, SpaceX operations, ASML partnerships, or energy/space infrastructure—even if they mention AI in passing. Unless the post analyzes how these developments specifically impact healthcare AI deployment or costs, they are off-topic.
Posts about Elon Musk, SpaceX IPO, or energy/space infrastructure that tangentially mention AI without healthcare context
3 example posts
NEWS: ASML has invited Elon Musk to speak at its internal technology conference, Dutch outlet NUnl reports. Elon Musk's response: "ASML should be treasured and supported. It is arguably the greatest company in Europe." ASML builds the EUV lithography machines that every advance
Elon is now printing over $26B per year… purely from selling compute to his competitors 😂 AI companies bet everything on one path and got crushed by compute limits. Now they’re begging Elon for GPUs
Elon Musk reveals why he believes the cheapest place to put AI will be space within 36 months "The availability of energy is the issue. Everywhere outside of China, electrical output is more or less flat. The output of chips is growing exponentially, but the output of https://t.
Created 2026-06-07 · Updated 2026-06-07
Edit rule text
[elon_musk_or_non_healthcare_tech_figures]
Learned 3 rejections Active
Exclude posts about Elon Musk announcements, SpaceX IPO strategy, ASML conferences, Tesla robotics, or other non-healthcare tech leaders/companies — even if energy or compute implications are mentioned. These are tech/business news, not healthcare tech.
Posts about Elon Musk, ASML, Tesla, SpaceX, or other non-healthcare tech figures and companies with only tangential healthcare framing.
3 example posts
NEWS: ASML has invited Elon Musk to speak at its internal technology conference, Dutch outlet NUnl reports. Elon Musk's response: "ASML should be treasured and supported. It is arguably the greatest company in Europe." ASML builds the EUV lithography machines that every advance
Elon Musk just explained why the SpaceX IPO is an energy story and the energy constraint is why he believes space becomes the only viable path for AI to scale (Save this). The argument he is making is one of the most important and least understood things happening in technology
Elon is now printing over $26B per year… purely from selling compute to his competitors 😂 AI companies bet everything on one path and got crushed by compute limits. Now they’re begging Elon for GPUs
Created 2026-06-07 · Updated 2026-06-07
Edit rule text
[unvalidated_speculative_medical_claims_or_hype]
Learned 3 rejections Active
Exclude posts that claim AI found zero-days in medical/pharma systems, discovered vaccine designs, or made breakthrough medical discoveries without peer-reviewed clinical trial data, FDA approval documentation, or independent validation. Anthropic's 'Claude Mythos' findings in healthcare belong here.
Posts making broad or unsubstantiated claims about medical breakthroughs, drug efficacy, or AI-discovered treatments without rigorous clinical evidence or validation.
3 example posts
Researchers have announced positive results from the first-in-human Phase 1 clinical trial of a universal coronavirus vaccine designed using artificial intelligence. Unlike existing COVID-19 vaccines that target specific strains or variants of SARS-CoV-2, this experimental https:
🚨 do you understand what just happened with Anthropic.. Their internal model, Claude Mythos, found previously unknown security holes in every major browser and every major operating system. Not one. All of them at once. And Anthropic did something a hype-driven industry almost
Mythos AI is being used by National Security Agency in offensive cyber operations / cyberattacks. Anthropic has even embedded engineers inside the NSA to help deploy the model. Are frontier AI labs becoming active contractors in state cyber conflict? The live-ops role is still ht
Created 2026-06-07 · Updated 2026-06-07
Edit rule text
[cybersecurity_or_ai_vulnerability_tangent_no_healthcare_lens]
Learned 3 rejections Active
Exclude posts that report on AI finding security vulnerabilities, zero-day exploits, or cybersecurity incidents in general software/infrastructure (browsers, operating systems, FFmpeg, NSA operations) unless the post directly addresses healthcare data, patient privacy, or healthcare system security implications.
Posts about zero-day vulnerabilities, cybersecurity incidents, or AI security capabilities applied to general software/systems without healthcare-specific security context
3 example posts
🚨 do you understand what just happened with Anthropic.. Their internal model, Claude Mythos, found previously unknown security holes in every major browser and every major operating system. Not one. All of them at once. And Anthropic did something a hype-driven industry almost
Mythos AI is being used by National Security Agency in offensive cyber operations / cyberattacks. Anthropic has even embedded engineers inside the NSA to help deploy the model. Are frontier AI labs becoming active contractors in state cyber conflict? The live-ops role is still ht
🔥 AI just found 21 zero-days in FFmpeg. That’s the video library bundled inside many apps, tools, containers, and devices. Some bugs sat untouched for 15–20 years. Google Chrome also dropped PATCHES for a record 429 vulnerabilities this week. Read: https://t.co/6MEVD9ufxu
Created 2026-06-07 · Updated 2026-06-07
Edit rule text
[robotics_and_speculative_sci_fi_labor]
Learned 3 rejections Active
Exclude posts about Tesla robots, humanoid automation, or speculative labor market disruption from AI/robotics that lack healthcare-specific data, clinical validation, or analysis of actual workforce impact in healthcare delivery.
Posts speculating about humanoid robots, labor disruption, or sci-fi AI applications without healthcare validation or system impact
2 example posts
A framework I wish more Founders and VCs used when discussing US Manufacturing: Automation Value = Labor Intensity × Labor Eliminated Everyone gets excited about robots and automation. Almost nobody asks: “How much of revenue is labor?” If labor is 60% of revenue and AI http
Humanoid robots, designed to mimic human movement and capabilities, have been making headlines in recent years, from their use as baggage handlers at Japanese airports to Tesla’s big bet on its Optimus humanoid. Market watchers have predicted that the machines will change the ht
Created 2026-06-07 · Updated 2026-06-10
Edit rule text
[pharma_trial_data_market_announcement_only]
Learned 3 rejections Active
Exclude posts that announce drug trial data, FDA approvals, pharmaceutical company funding rounds, or biotech founder enthusiasm without analyzing healthcare delivery impact, adoption barriers, reimbursement implications, or systemic change.
Posts announcing pharmaceutical trial results, drug approvals, or biotech funding without healthcare system analysis or operational insight.
3 example posts
Approved 20 years ago as a diabetes treatment, GLP-1 drugs have been found to help patients reduce weight, changing the lives of more than 30 million people in the U.S. But there also have been troubling side effects reported. https://t.co/zzRamKnedu
🚨 T-DXd just went tumor-agnostic — and it changed a lot about how we think about HER2 testing across solid tumors 🔬 #ASCO26 FDA granted accelerated approval to T-DXd for patients with unresectable or metastatic HER2-positive solid tumors after prior systemic treatment —
The scores don’t pick your binders. The filters don't care either. The assay does. I worked with a design partner to produce true de-novo hits to 2 immuno-oncology targets. ~1,500 de novo designs → 96 synthesized → 89 run on SPR → 5 confirmed binders; 19–195 nM No https:
Created 2026-06-07 · Updated 2026-06-07
Edit rule text
[clinical_research_observation_no_systems]
Learned 3 rejections Active
Exclude posts that cite clinical trial data, study results, or research observations (e.g., ECG AI performance, MRSA PCR accuracy, microplastics in plaques) without connecting to healthcare delivery systems, reimbursement, adoption barriers, or operational transformation.
Posts reporting clinical trial results, research findings, or medical observations without healthcare system context, implementation barriers, or operational relevance.
3 example posts
Most under-discussed cardiology study of the last 2 years just hit NEJM. Microplastics found inside the carotid plaques of 58% of patients undergoing endarterectomy. 4.5× higher rate of heart attack, stroke, or death over 34 months. This changes the atherosclerosis story. http
Are pulse oximeters risking Black patients' lives? Episode 4 of the Intention to Treat podcast explores the story of the pulse oximeter and the deadly consequences when a critical medical test doesn’t work on dark skin. Listen on Apple or wherever you get your podcasts. 🎧 https:/
🔥Phase 1: ABBV-706 (SEZ6-targeting ADC) in R/R SCLC & solid tumors 🆙 @NatureMedicine ☑R/R SCLC; ORR 52%; mOS 12.4 mo at RP2D 🎯RP2D confirmed at 1.8 mg/kg Q3W balancing efficacy & safety 🎙 @LaurenByersMD #LCSM @OncoAlert @Larvol https://t.co/vjUUd1GYSy
Created 2026-06-07 · Updated 2026-06-07
Edit rule text
[tangential_non_healthcare_business_model]
Learned 3 rejections Active
Exclude posts analyzing pricing models, feature moats, distribution strategies, or labor economics that apply to non-healthcare sectors or are written as generalist commentary. If the post could be about any B2B SaaS or general business (not specifically healthcare), reject it.
Posts about general business model dynamics (SaaS pricing, unit economics, moats) with only loose healthcare framing
3 example posts
@swago_baby You’re not dumb please.. WhatsApp stays free for personal users (no ads in chats) and makes money almost entirely from businesses: - Main revenue: Businesses pay per conversation via the WhatsApp Business API (for customer support, notifications, orders, etc.). Fees
"Are there even any feature moats left in B2B? They are at best, short-lived. In fact, we were able to rebuild SaaStr's AI VP Marketing in just ... an hour on @Lovable. Moats that still work: hardware, network effects, data, security and compliance. with @ElenaVerna Head of
A framework I wish more Founders and VCs used when discussing US Manufacturing: Automation Value = Labor Intensity × Labor Eliminated Everyone gets excited about robots and automation. Almost nobody asks: “How much of revenue is labor?” If labor is 60% of revenue and AI http
Created 2026-06-07 · Updated 2026-06-07
Edit rule text
[pharmaceutical_trial_announcement_only]
Learned 3 rejections Active
Exclude posts that are pure trial result announcements, drug approval news, or funding scoops without discussing how the treatment integrates into clinical workflows, changes practice patterns, or addresses healthcare system barriers. If the post is primarily 'Drug X showed Y efficacy' with no systems insight, reject it.
Posts announcing pharma trial results, drug approvals, or biotech funding without healthcare systems analysis or operational context
3 example posts
🚨 T-DXd just went tumor-agnostic — and it changed a lot about how we think about HER2 testing across solid tumors 🔬 #ASCO26 FDA granted accelerated approval to T-DXd for patients with unresectable or metastatic HER2-positive solid tumors after prior systemic treatment —
The scores don’t pick your binders. The filters don't care either. The assay does. I worked with a design partner to produce true de-novo hits to 2 immuno-oncology targets. ~1,500 de novo designs → 96 synthesized → 89 run on SPR → 5 confirmed binders; 19–195 nM No https:
🔥Phase 1: ABBV-706 (SEZ6-targeting ADC) in R/R SCLC & solid tumors 🆙 @NatureMedicine ☑R/R SCLC; ORR 52%; mOS 12.4 mo at RP2D 🎯RP2D confirmed at 1.8 mg/kg Q3W balancing efficacy & safety 🎙 @LaurenByersMD #LCSM @OncoAlert @Larvol https://t.co/vjUUd1GYSy
Created 2026-06-07 · Updated 2026-06-07
Edit rule text
[ai_infrastructure_compute_tangent]
Learned 3 rejections Active
Exclude posts focused on GPU pricing, data center energy, semiconductor manufacturing, or AI infrastructure economics unless they explicitly tie to a healthcare delivery, clinical outcome, or operational improvement. Posts about 'compute limits' or 'Elon selling GPUs to competitors' are tangential macro commentary.
Posts about AI compute, energy, hardware, or infrastructure without direct healthcare application or specificity
3 example posts
Elon is now printing over $26B per year… purely from selling compute to his competitors 😂 AI companies bet everything on one path and got crushed by compute limits. Now they’re begging Elon for GPUs
Elon Musk reveals why he believes the cheapest place to put AI will be space within 36 months "The availability of energy is the issue. Everywhere outside of China, electrical output is more or less flat. The output of chips is growing exponentially, but the output of https://t.
Nous Research is working with NVIDIA to make Hermes Agent run smoothly on the new NVIDIA RTX Spark superchip. Hermes Agent is also integrating with the new OpenShell runtime, which connects Hermes to Microsoft’s security primitives https://t.co/cvvPc8bIKa
Created 2026-06-07 · Updated 2026-06-07
Edit rule text
[unsubstantiated_ai_capability_claims]
Learned 3 rejections Active
Exclude posts that claim an AI model discovered vulnerabilities, caught errors, or performed a task at superhuman speed without evidence of real-world healthcare validation. Focus on the substantiation gap: if it reads like AI hype about what models *can* do rather than what they *do* in healthcare, reject it.
Posts making dramatic, unverified claims about AI model capabilities without clinical validation or healthcare application
3 example posts
Mythos AI is being used by National Security Agency in offensive cyber operations / cyberattacks. Anthropic has even embedded engineers inside the NSA to help deploy the model. Are frontier AI labs becoming active contractors in state cyber conflict? The live-ops role is still ht
Anthropic engineer James Brady: "Every agent in production lies. We measured it. The good ones lie less, the great ones catch the lie before the user does." In 29 minutes, he walks through the verification stack he built and the patterns the Claude Code team adopted to keep ht
Anthropic engineer: "The agent doesn't remember anything. So we built a second set of agents whose only job is to dream about the first ones." They wait until you log off, then reopen every session you ran, fact-check the first agents, merge the duplicates, and burn anything ht
Created 2026-06-07 · Updated 2026-06-07
Edit rule text
[robotics_and_speculative_sci_fi_labor_macro]
Learned 3 rejections Active
Exclude posts about humanoid robots, manufacturing automation, or speculative labor-market disruption narratives—even if tangentially connected to healthcare—unless they provide concrete evidence of impact on healthcare delivery, workforce, or operations.
Posts speculating about humanoid robots, factory automation, or labor market disruption without evidence or healthcare application.
3 example posts
Humanoid robots, designed to mimic human movement and capabilities, have been making headlines in recent years, from their use as baggage handlers at Japanese airports to Tesla’s big bet on its Optimus humanoid. Market watchers have predicted that the machines will change the ht
We are partnering with @Microsoft to enable secure, user-controlled AI on Windows. NVIDIA OpenShell runtime for agents will provide governance tools, policy enforcement, and smart local-to-cloud query routing. Learn more: https://t.co/zPGwz9xQSW https://t.co/mkOtFEOhAS
From unboxing to AI agent in minutes. Getting an agent running used to mean sourcing a model, configuring an inference backend, installing a runtime, and wiring it all together. The new NemoClaw install path on DGX Spark replaces that with a single command. DGX Spark also https
Created 2026-06-07 · Updated 2026-06-08
Edit rule text
[cynical_insider_commentary_no_substance]
Learned 3 rejections Active
Exclude posts that use sarcastic or cynical tone to claim industry incompetence, fraud, or fakery (e.g., 'I made that up last year. I'm making it up again') without providing data, methodology, or verifiable evidence. Posts must avoid performative outrage or humor masquerading as analysis.
Posts using sarcasm and cynicism to mock healthcare/AI industry without substantive analysis or evidence
3 example posts
I'm SVP now. I told you I would be. The graph went up and to the right. Nobody checked what it measured. It measured "AI enablement." I made that up last year. I'm making it up again. Last year I rolled out Copilot to 4,000 people. Nothing happened. I got promoted. Those two htt
I just looked at a home service firms ServiceTitan dashboard and saw something interesting. A Pre-seed company with AI-Supported Human CSRs is out converting an AI CSR business that's raised over $125M by >50% Head-to-Head win rate is 100% Why? Supercharging humans resonat
I think the AI superapps will soon own 90% of the agentic layer more and more people won't use hermes/openclaw etc because Claude Cowork/Codex will offer 90% of the functionality with 5% of the friction
Created 2026-06-07 · Updated 2026-06-07
Edit rule text
[cybersecurity_or_ai_vulnerability_tangent_not_healthcare]
Learned 3 rejections Active
Exclude posts about cybersecurity vulnerabilities, zero-day exploits, AI model safety incidents, or agent misuse (e.g., 'FFmpeg bugs found by AI,' 'agent reaches tools and leaks data') unless they specifically address healthcare data, HIPAA, clinical systems, or patient privacy implications.
Posts about cybersecurity vulnerabilities, zero-days, or AI safety issues without healthcare-specific context or implications
3 example posts
🔥 AI just found 21 zero-days in FFmpeg. That’s the video library bundled inside many apps, tools, containers, and devices. Some bugs sat untouched for 15–20 years. Google Chrome also dropped PATCHES for a record 429 vulnerabilities this week. Read: https://t.co/6MEVD9ufxu
The next frontier of AI-secured financial and technological critical infrastructure. @ICE_Markets and @NYSE are part of @AnthropicAI's cyber security initiative Project Glasswing, deploying Anthropic’s Claude Mythos Preview across ICE’s exchanges, clearing houses, mortgage
⚡️AI is turning cybersecurity from a human-limited profession into a machine-speed arms race. The surface read is “Anthropic’s model found vulnerabilities fast.” The deeper read is that the bottleneck in cyber is shifting. For years, vulnerability discovery was constrained
Created 2026-06-07 · Updated 2026-06-07
Edit rule text
[agent_or_agentic_system_technical_tangent]
Learned 3 rejections Active
Exclude posts about agent frameworks (LangChain, LangGraph, OpenAI Agents SDK), control planes, agent orchestration, memory systems, or autonomous system technical implementation unless they demonstrate validated healthcare use case, clinical workflow integration, or patient/provider impact.
Posts about AI agent architecture, control planes, and autonomous system technical capabilities without healthcare validation
3 example posts
Control plane! Control plane! Control plane! You will hear this term of art a lot going forward. Why? Because in this next phase of AI, companies will want something to sit above the models. They will want control over their AI spend. They will want the flexibility to pick
Nous Research is working with NVIDIA to make Hermes Agent run smoothly on the new NVIDIA RTX Spark superchip. Hermes Agent is also integrating with the new OpenShell runtime, which connects Hermes to Microsoft’s security primitives https://t.co/cvvPc8bIKa
BREAKING: Merge Launches ‘Agent Handler’ Control AI Access, Tokenmaxxed $$$ Bills, & Stop Mass Data Leaks "We don't trust agents" "The second you connect it to tools, that's where everything goes wrong." OpenAI. Perplexity. Netflix. Uber. Mistral. Dropbox. JPMorgan.. all h
Created 2026-06-07 · Updated 2026-06-07
Edit rule text
[satirical_or_cynical_insider_commentary]
Learned 3 rejections Active
Exclude posts that use satirical tone, sarcasm, or cynical framing to criticize (e.g., 'I made that up,' 'The graph went up and nobody checked,' 'My job is to make sure your surgery center never gets built') unless they include concrete evidence, data, or systemic healthcare analysis beneath the humor.
Posts using sarcasm, cynicism, or dark humor to mock healthcare/business practices without substantive analysis
3 example posts
I'm SVP now. I told you I would be. The graph went up and to the right. Nobody checked what it measured. It measured "AI enablement." I made that up last year. I'm making it up again. Last year I rolled out Copilot to 4,000 people. Nothing happened. I got promoted. Those two htt
My job is to make sure your surgery center never gets built. Eleven years, and I have never lost. The kid had it all lined up. Board-certified, two partners, a lease on a space where he could do the same procedures my client does across town. He showed up to the hearing with a
I called a woman in Dayton to tell her she was about to overpay $4,500, for a horrible health system based procedure. She was scheduled for a knee replacement. Her PPO had her on the hook for a $4,500 deductible. The call was on behalf of her employer: “See one of these three
Created 2026-06-07 · Updated 2026-06-07
Edit rule text
[ai_company_business_metrics_and_announcements]
Learned 3 rejections Active
Exclude posts that announce AI company funding, valuation milestones, product launches, partnerships, or business metrics (e.g., Series B raises, control plane announcements, enterprise software pricing changes) unless they explicitly demonstrate healthcare application, clinical validation, or healthcare system impact.
Posts reporting AI company funding rounds, valuation changes, product launches, or business partnerships without demonstrating healthcare application or validation.
3 example posts
Another great research piece by @_DimensionCap @bauer_lesavage on Training Data for Bio AI. Models will only be as good as the underlying data, and the biology they learn will be constrained by the limitations of that data. We need to think deeply about scaling the best
If gov't takes a stake, how do you value these IPOs? Government equity usually signals utility (regulated/ slow/priced for dividends). But OpenAI and Anthropic are the opposite of that.... Maybe a new category: strategic tech --where a government stake is a premium. No clean
"You can run OpenClaw inside your company now." Annoucing our work with @Microsoft to bring OpenClaw to the Microsoft and Windows ecosystems. Claws now work securly in the enterprise. https://t.co/3fKhT2mGpC
Created 2026-06-07 · Updated 2026-06-07
Edit rule text
[non_healthcare_tangential_framing]
Learned 3 rejections Active
Exclude posts about robotics, humanoid workers, manufacturing automation, energy infrastructure, semiconductor chips, or general AI capability announcements that do not demonstrate concrete healthcare application or healthcare system impact. These are general tech/infrastructure posts with loose or absent healthcare framing.
Posts about non-healthcare domains (robotics, energy, general AI infrastructure, manufacturing) that lack substantive connection to healthcare systems or applications.
3 example posts
Elon Musk reveals why he believes the cheapest place to put AI will be space within 36 months "The availability of energy is the issue. Everywhere outside of China, electrical output is more or less flat. The output of chips is growing exponentially, but the output of https://t.
Nous Research is working with NVIDIA to make Hermes Agent run smoothly on the new NVIDIA RTX Spark superchip. Hermes Agent is also integrating with the new OpenShell runtime, which connects Hermes to Microsoft’s security primitives https://t.co/cvvPc8bIKa
a $2,000 graphics card and a free download are turning into $2,000-a-month retainers, and the customers are the ones who legally can't use ChatGPT the hardware is an RTX 5090, 32GB of VRAM, enough to run a 30B model locally with nothing ever leaving the building. the software is
Created 2026-06-07 · Updated 2026-06-07
Edit rule text
[satirical_sarcastic_commentary]
Learned 3 rejections Active
Exclude posts that rely primarily on sarcasm, satire, or cynical humor to make their point (e.g., claiming to have 'made up' metrics, mock boasting about never losing cases). These lack genuine substance and are intended as social commentary rather than healthcare tech analysis.
Posts using satire, sarcasm, or cynical humor to critique AI/healthcare topics rather than providing substantive analysis.
3 example posts
I'm SVP now. I told you I would be. The graph went up and to the right. Nobody checked what it measured. It measured "AI enablement." I made that up last year. I'm making it up again. Last year I rolled out Copilot to 4,000 people. Nothing happened. I got promoted. Those two htt
My job is to make sure your surgery center never gets built. Eleven years, and I have never lost. The kid had it all lined up. Board-certified, two partners, a lease on a space where he could do the same procedures my client does across town. He showed up to the hearing with a
wow K&E, largest law firm in the world ($10B billings). They’re committing $500M to build proprietary AI software, with 180 tech pros and 100 partners involved. I’m skeptical that “build” is right here, but wonder if Anthropic/OAI deployment cos are involved.
Created 2026-06-07 · Updated 2026-06-07
Edit rule text
[shallow_market_speculation_without_systems_analysis]
Learned 3 rejections Active
Exclude posts that speculate on market competition, business model viability, or pricing dynamics (e.g., SaaS vs. agentic pricing, feature moats, vendor consolidation) without grounding claims in healthcare system evidence, operational impact analysis, or validated customer outcomes. Generic startup or business strategy commentary loosely framed as healthcare should be rejected.
Posts about market dynamics, competitive positioning, or business model changes that lack substantive analysis of healthcare system implications or operational validation.
3 example posts
"Are there even any feature moats left in B2B? They are at best, short-lived. In fact, we were able to rebuild SaaStr's AI VP Marketing in just ... an hour on @Lovable. Moats that still work: hardware, network effects, data, security and compliance. with @ElenaVerna Head of
BREAKING: Merge Launches ‘Agent Handler’ Control AI Access, Tokenmaxxed $$$ Bills, & Stop Mass Data Leaks "We don't trust agents" "The second you connect it to tools, that's where everything goes wrong." OpenAI. Perplexity. Netflix. Uber. Mistral. Dropbox. JPMorgan.. all h
AI made building cheap. Regulation made distribution expensive. Anyone can clone your demo now. Very few can bring proprietary data, licenses, compliance, and enough trust to sell into serious players. The moat didn’t disappear. It moved.
Created 2026-06-07 · Updated 2026-06-07
Edit rule text
[infrastructure_and_compute_hype_without_healthcare_validation]
Learned 3 rejections Active
Exclude posts that focus on AI infrastructure advancements (GPU hardware, compute costs, energy requirements, chip announcements, data centers) without demonstrating validated healthcare use cases or healthcare-specific operational benefits. General tech infrastructure hype loosely framed with healthcare context should be rejected.
Posts about AI infrastructure, chips, energy, or compute capacity that lack evidence of healthcare application or operational impact.
3 example posts
Elon Musk reveals why he believes the cheapest place to put AI will be space within 36 months "The availability of energy is the issue. Everywhere outside of China, electrical output is more or less flat. The output of chips is growing exponentially, but the output of https://t.
Nous Research is working with NVIDIA to make Hermes Agent run smoothly on the new NVIDIA RTX Spark superchip. Hermes Agent is also integrating with the new OpenShell runtime, which connects Hermes to Microsoft’s security primitives https://t.co/cvvPc8bIKa
The context window pricing model might be one of the smartest tricks AI labs pulled. You pay for larger context windows, yet model performance often degrades as more context is stuffed in. So developers end up paying more just to work around the fact that models don't have true
Created 2026-06-07 · Updated 2026-06-07
Edit rule text
[infrastructure_or_compute_hype_without_healthcare_application]
Learned 3 rejections Active
Exclude posts that focus on AI infrastructure (GPUs, energy consumption, token pricing, inference costs, control planes, runtime environments) or general compute economics without demonstrating or analyzing healthcare application. Posts about NVIDIA, chip design, or electrical grid are out of scope unless explicitly applied to healthcare.
Posts about AI infrastructure, compute economics, energy, or technical architectures without healthcare relevance
3 example posts
Elon Musk reveals why he believes the cheapest place to put AI will be space within 36 months "The availability of energy is the issue. Everywhere outside of China, electrical output is more or less flat. The output of chips is growing exponentially, but the output of https://t.
Nous Research is working with NVIDIA to make Hermes Agent run smoothly on the new NVIDIA RTX Spark superchip. Hermes Agent is also integrating with the new OpenShell runtime, which connects Hermes to Microsoft’s security primitives https://t.co/cvvPc8bIKa
The context window pricing model might be one of the smartest tricks AI labs pulled. You pay for larger context windows, yet model performance often degrades as more context is stuffed in. So developers end up paying more just to work around the fact that models don't have true
Created 2026-06-07 · Updated 2026-06-07
Edit rule text
[satirical_or_cynical_commentary]
Learned 3 rejections Active
Exclude posts that are primarily satirical, sarcastic, or cynical in tone (e.g., 'I made that up,' 'nobody checked,' 'I have never lost') without providing substantive analysis, evidence, or actionable healthcare system insights. Humor without substance is off-topic.
Posts using sarcasm, satire, or cynical humor to critique healthcare or tech without substantive analysis or actionable insights.
3 example posts
I'm SVP now. I told you I would be. The graph went up and to the right. Nobody checked what it measured. It measured "AI enablement." I made that up last year. I'm making it up again. Last year I rolled out Copilot to 4,000 people. Nothing happened. I got promoted. Those two htt
"You can run OpenClaw inside your company now." Annoucing our work with @Microsoft to bring OpenClaw to the Microsoft and Windows ecosystems. Claws now work securly in the enterprise. https://t.co/3fKhT2mGpC
Your Oura Ring, your health record. Together. Finally. @Flexpa is bringing clinical records into @ouraring via TEFCA, so ŌURA’s AI isn’t just working from what your ring measures, but from your real health history. https://t.co/b1KCeGEZLO https://t.co/u2t1TQjL7p
Created 2026-06-07 · Updated 2026-06-07
Edit rule text
[clinical_trial_announcement_only_no_systems_analysis]
Learned 3 rejections Active
Exclude posts that announce clinical trial outcomes, drug efficacy data, cancer trial results, or pharma milestones without analyzing the healthcare system implications (cost, access, workflow impact, clinical adoption barriers). Pure data announcements or celebratory trial reports without systems insight should be rejected.
Posts that merely report clinical trial results, drug approvals, or pharma data without healthcare systems or operational context
3 example posts
#ASCO26 | Day 3 $LLY didn't spend billions on Kelonia for an 18-patient #myeloma dataset. They may have paid for a potential manufacturing disruption. KLN-1010 reported: • 18/18 MRD-negative at 1 month • No lymphodepletion • No ex vivo cell manufacturing • Single infusion
Published in @JCO_ASCO, during #ASCO26: Tumor-Agnostic Therapies: Translating Scientific Breakthroughs Into Global Implementation ASCO is where we celebrate the next breakthrough. This review asks why breakthroughs do not equal access globally. 🔗: https://t.co/PE5q27bIs1 htt
A rapid manufacturing pipeline for BCMA-targeting CAR T cells drastically shortens “vein-to-vein” times, and produces cells that demonstrate encouraging safety results in a phase one trial of patients with relapsed #MultipleMyeloma. @ScienceTM https://t.co/3F4u8ECFDS https://t.co
Created 2026-06-07 · Updated 2026-06-07
Edit rule text
[tangential_infrastructure_compute_hype_no_healthcare_validation]
Learned 3 rejections Active
Exclude posts that focus on AI infrastructure, compute costs, energy requirements, hardware capabilities, or supply chain (chips, data centers, electrical grids) where any healthcare connection is speculative, mentioned in passing, or unsupported. Posts must demonstrate a clear, validated link between the infrastructure capability and a specific healthcare outcome or operational improvement.
Posts about AI infrastructure, compute, energy, or hardware where healthcare application is speculative or absent
3 example posts
Elon Musk reveals why he believes the cheapest place to put AI will be space within 36 months "The availability of energy is the issue. Everywhere outside of China, electrical output is more or less flat. The output of chips is growing exponentially, but the output of https://t.
Nous Research is working with NVIDIA to make Hermes Agent run smoothly on the new NVIDIA RTX Spark superchip. Hermes Agent is also integrating with the new OpenShell runtime, which connects Hermes to Microsoft’s security primitives https://t.co/cvvPc8bIKa
The context window pricing model might be one of the smartest tricks AI labs pulled. You pay for larger context windows, yet model performance often degrades as more context is stuffed in. So developers end up paying more just to work around the fact that models don't have true
Created 2026-06-07 · Updated 2026-06-07
Edit rule text
[non_healthcare_infrastructure_or_tech_with_loose_healthcare_framing]
Learned 3 rejections Active
Exclude posts about AI infrastructure, compute, chips, CPUs, energy, space-based AI deployment, or software architecture (headless CRM, modular software, data interoperability protocols) that either lack healthcare application or only mention healthcare as one possible use case among many. Examples: 'energy is the limiting factor for AI deployment,' 'the app era is ending,' 'headless platforms,' 'TEFCA data interop.'
Posts about general AI infrastructure, cloud computing, chip architecture, or non-healthcare tech advances that tangentially mention healthcare or are loosely framed as relevant.
3 example posts
Elon Musk reveals why he believes the cheapest place to put AI will be space within 36 months "The availability of energy is the issue. Everywhere outside of China, electrical output is more or less flat. The output of chips is growing exponentially, but the output of https://t.
"You can run OpenClaw inside your company now." Annoucing our work with @Microsoft to bring OpenClaw to the Microsoft and Windows ecosystems. Claws now work securly in the enterprise. https://t.co/3fKhT2mGpC
Nous Research is working with NVIDIA to make Hermes Agent run smoothly on the new NVIDIA RTX Spark superchip. Hermes Agent is also integrating with the new OpenShell runtime, which connects Hermes to Microsoft’s security primitives https://t.co/cvvPc8bIKa
Created 2026-06-07 · Updated 2026-06-07
Edit rule text
[labor_market_macro_or_robotics_speculation_without_healthcare_specificity]
Learned 3 rejections Active
Exclude posts that discuss humanoid robots, manufacturing automation, labor disruption from AI, or workforce trends (e.g., 'computers are built for agents now,' 'robots replace baggage handlers,' 'automation value = labor intensity × labor eliminated') without demonstrating specific application to healthcare operations, workforce planning, or clinical staffing. These are generalist macro/robotics speculation with loose healthcare framing.
Posts about workforce disruption, automation, robotics, or labor market macro trends that lack healthcare-specific context or analysis.
3 example posts
A framework I wish more Founders and VCs used when discussing US Manufacturing: Automation Value = Labor Intensity × Labor Eliminated Everyone gets excited about robots and automation. Almost nobody asks: “How much of revenue is labor?” If labor is 60% of revenue and AI http
Elon Musk reveals why he believes the cheapest place to put AI will be space within 36 months "The availability of energy is the issue. Everywhere outside of China, electrical output is more or less flat. The output of chips is growing exponentially, but the output of https://t.
I just looked at a home service firms ServiceTitan dashboard and saw something interesting. A Pre-seed company with AI-Supported Human CSRs is out converting an AI CSR business that's raised over $125M by >50% Head-to-Head win rate is 100% Why? Supercharging humans resonat
Created 2026-06-07 · Updated 2026-06-07
Edit rule text
[generalist_macro_or_robotics_hype_loose_healthcare]
Learned 3 rejections Active
Exclude posts that discuss robotics, manufacturing automation, energy infrastructure, or labor market macro trends (e.g., Tesla robots, PUDU robots, energy availability) that mention healthcare tangentially or speculatively without concrete healthcare clinical application, validated healthcare workforce impact, or healthcare facility deployment.
Posts about humanoid robots, labor market disruption, manufacturing automation, or energy/infrastructure trends with only tangential healthcare framing
3 example posts
Humanoid robots, designed to mimic human movement and capabilities, have been making headlines in recent years, from their use as baggage handlers at Japanese airports to Tesla’s big bet on its Optimus humanoid. Market watchers have predicted that the machines will change the ht
Another humanoid worker has just joined the factory floor Two years later,PUDU Robotics has launched the new-generation PUDU D7, built specifically for real manufacturing scenarios. It can autonomously push carts, handle intra-line transportation, perform delicate operations, h
Anthropic just shipped opus 4.8 and called it a modest upgrade modest is the engineer who used to need babysitting now finishing your whole codebase migration in one session while you sleep so here is the part nobody connects the model is no longer the bottleneck. the platform
Created 2026-06-07 · Updated 2026-06-07
Edit rule text
[pharma_trial_data_or_biotech_announcement_unvalidated]
Learned 3 rejections Active
Exclude posts that report clinical trial results, drug approval announcements, biotech lab updates, or pharmaceutical research findings as standalone news without analysis of healthcare delivery systems impact, reimbursement challenges, manufacturing scale, or clinical adoption barriers. Posts must address how the finding changes healthcare practice or systems.
Posts about pharmaceutical trial data, biotech research announcements, or drug/gene therapy results without healthcare systems analysis or implementation insight
3 example posts
#ASCO26 | Day 3 $LLY didn't spend billions on Kelonia for an 18-patient #myeloma dataset. They may have paid for a potential manufacturing disruption. KLN-1010 reported: • 18/18 MRD-negative at 1 month • No lymphodepletion • No ex vivo cell manufacturing • Single infusion
Published in @JCO_ASCO, during #ASCO26: Tumor-Agnostic Therapies: Translating Scientific Breakthroughs Into Global Implementation ASCO is where we celebrate the next breakthrough. This review asks why breakthroughs do not equal access globally. 🔗: https://t.co/PE5q27bIs1 htt
A rapid manufacturing pipeline for BCMA-targeting CAR T cells drastically shortens “vein-to-vein” times, and produces cells that demonstrate encouraging safety results in a phase one trial of patients with relapsed #MultipleMyeloma. @ScienceTM https://t.co/3F4u8ECFDS https://t.co
Created 2026-06-07 · Updated 2026-06-07
Edit rule text
[infrastructure_compute_hype]
Learned 3 rejections Active
Exclude posts that discuss AI infrastructure trends (energy costs, GPU requirements, token economics, inference costs, compute bottlenecks, RTX chips, NVIDIA hardware) unless they explicitly analyze how these constraints impact a specific healthcare AI application or deployment.
Posts about AI infrastructure, compute hardware, energy availability, or technical architecture without healthcare application
3 example posts
Elon Musk reveals why he believes the cheapest place to put AI will be space within 36 months "The availability of energy is the issue. Everywhere outside of China, electrical output is more or less flat. The output of chips is growing exponentially, but the output of https://t.
The context window pricing model might be one of the smartest tricks AI labs pulled. You pay for larger context windows, yet model performance often degrades as more context is stuffed in. So developers end up paying more just to work around the fact that models don't have true
a $2,000 graphics card and a free download are turning into $2,000-a-month retainers, and the customers are the ones who legally can't use ChatGPT the hardware is an RTX 5090, 32GB of VRAM, enough to run a 30B model locally with nothing ever leaving the building. the software is
Created 2026-06-07 · Updated 2026-06-07
Edit rule text
[tangential_policy_outrage_or_fraud_scandal_without_healthcare_systems_analysis]
Learned 3 rejections Active
Exclude posts that report healthcare fraud, regulatory violations, healthcare policy outrage (e.g., prior auth abuses, provider directory failures, Medicaid insurer changes), or individual healthcare cost narratives without analyzing healthcare system structure, incentives, or systemic reform implications.
Posts about healthcare policy, regulatory fraud, or scandal reporting that frame individual cases or outrage without systemic healthcare analysis.
3 example posts
USDA's Chief Information Officer Sam Berry says AI could help the USDA crack down on billions of dollars in SNAP fraud each year. "SNAP is a $100B-per-year taxpayer-funded program. That's an area where we really want to have all angles of the data available so that we can deploy
My job is to make sure your surgery center never gets built. Eleven years, and I have never lost. The kid had it all lined up. Board-certified, two partners, a lease on a space where he could do the same procedures my client does across town. He showed up to the hearing with a
PacificSource pulled out as Lane County Medicaid insurer, Tina Kotek’s OR Health Authority replaced it with Trillium, forcing 96k to join Trillium. Trillium’s parent co gave Kotek’s campaign $50k two months later. OR AG (D) said in ‘22 Trillium parent co gave Kotek’s”took advan
Created 2026-06-07 · Updated 2026-06-07
Edit rule text
[agentic_ai_or_control_plane_technical_tangent]
Learned 3 rejections Active
Exclude posts that focus on agentic AI technical architecture (agent handlers, control planes, OpenShell, MCP protocols, runtime governance) or vendor product launches for agent infrastructure, unless the post demonstrates specific healthcare operational or clinical application and measurable impact.
Posts about agent architectures, control planes, runtime systems, or agentic framework announcements without healthcare operational or clinical validation.
3 example posts
Control plane! Control plane! Control plane! You will hear this term of art a lot going forward. Why? Because in this next phase of AI, companies will want something to sit above the models. They will want control over their AI spend. They will want the flexibility to pick
"You can run OpenClaw inside your company now." Annoucing our work with @Microsoft to bring OpenClaw to the Microsoft and Windows ecosystems. Claws now work securly in the enterprise. https://t.co/3fKhT2mGpC
BREAKING: Merge Launches ‘Agent Handler’ Control AI Access, Tokenmaxxed $$$ Bills, & Stop Mass Data Leaks "We don't trust agents" "The second you connect it to tools, that's where everything goes wrong." OpenAI. Perplexity. Netflix. Uber. Mistral. Dropbox. JPMorgan.. all h
Created 2026-06-07 · Updated 2026-06-07
Edit rule text
[generalist_macro_or_robotics_hype_without_healthcare_specificity]
Learned 3 rejections Active
Exclude posts that discuss robotics, humanoid workers, labor market disruption, manufacturing automation, or macro-level workforce trends (e.g., 'robots are replacing workers', 'the computer is built for agents now', 'NVIDIA is selling picks and shovels') unless they directly address healthcare labor, clinical staffing, or healthcare operational transformation with concrete healthcare examples.
Posts about humanoid robots, labor market disruption, or general macro trends applied vaguely to healthcare without healthcare-specific validation
3 example posts
Humanoid robots, designed to mimic human movement and capabilities, have been making headlines in recent years, from their use as baggage handlers at Japanese airports to Tesla’s big bet on its Optimus humanoid. Market watchers have predicted that the machines will change the ht
In the AI gold rush, Jensen Huang is selling the picks and shovels. Nvidia's chips power AI companies around the world, helping the company become the first to surpass a $5 trillion market capitalization in late 2025. Read more about how he and others made the https://t.co/lHLE
Another humanoid worker has just joined the factory floor Two years later,PUDU Robotics has launched the new-generation PUDU D7, built specifically for real manufacturing scenarios. It can autonomously push carts, handle intra-line transportation, perform delicate operations, h
Created 2026-06-06 · Updated 2026-06-06
Edit rule text
[ai_model_or_vendor_capability_announcement_without_healthcare_validation]
Learned 3 rejections Active
Exclude posts that announce a new AI model feature (e.g., 'Claude found vulnerabilities', 'new OpenShell runtime', 'Anthropic deployed Project Glasswing') or vendor partnership without evidence of healthcare application, clinical testing, or validated healthcare impact. Marketing partnerships or capability demos alone are insufficient.
Posts announcing new AI model capabilities, vendor product launches, or partnerships presented as healthcare-relevant without clinical validation or real healthcare use case
3 example posts
"You can run OpenClaw inside your company now." Annoucing our work with @Microsoft to bring OpenClaw to the Microsoft and Windows ecosystems. Claws now work securly in the enterprise. https://t.co/3fKhT2mGpC
Nous Research is working with NVIDIA to make Hermes Agent run smoothly on the new NVIDIA RTX Spark superchip. Hermes Agent is also integrating with the new OpenShell runtime, which connects Hermes to Microsoft’s security primitives https://t.co/cvvPc8bIKa
BREAKING: Merge Launches ‘Agent Handler’ Control AI Access, Tokenmaxxed $$$ Bills, & Stop Mass Data Leaks "We don't trust agents" "The second you connect it to tools, that's where everything goes wrong." OpenAI. Perplexity. Netflix. Uber. Mistral. Dropbox. JPMorgan.. all h
Created 2026-06-06 · Updated 2026-06-06
Edit rule text
[tangential_ai_infrastructure_or_safety_vulnerability_no_healthcare_lens]
Learned 3 rejections Active
Exclude posts about AI model capabilities, cybersecurity vulnerabilities, zero-day exploits, or AI safety incidents unless they explicitly address healthcare-specific risks, validation in clinical settings, or healthcare-specific threat models. Generic AI safety or cybersecurity announcements (e.g., 'Anthropic finds vulnerabilities', 'NVIDIA security tools', 'agent trust problems') do not qualify without healthcare context.
Posts about AI safety, cybersecurity vulnerabilities, or model capability announcements without healthcare-specific risk, application, or context.
3 example posts
Nous Research is working with NVIDIA to make Hermes Agent run smoothly on the new NVIDIA RTX Spark superchip. Hermes Agent is also integrating with the new OpenShell runtime, which connects Hermes to Microsoft’s security primitives https://t.co/cvvPc8bIKa
The next frontier of AI-secured financial and technological critical infrastructure. @ICE_Markets and @NYSE are part of @AnthropicAI's cyber security initiative Project Glasswing, deploying Anthropic’s Claude Mythos Preview across ICE’s exchanges, clearing houses, mortgage
⚡️AI is turning cybersecurity from a human-limited profession into a machine-speed arms race. The surface read is “Anthropic’s model found vulnerabilities fast.” The deeper read is that the bottleneck in cyber is shifting. For years, vulnerability discovery was constrained
Created 2026-06-06 · Updated 2026-06-06
Edit rule text
[ai_company_business_metrics_and_funding_not_healthcare_focused]
Learned 3 rejections Active
Exclude posts reporting on AI company funding rounds, valuations, revenue models, or business partnerships (e.g., Anthropic funding, OpenAI partnerships, Isomorphic Labs funding) unless the post demonstrates how that company's healthcare product is being applied, validated, or improving healthcare outcomes. Company business news alone is insufficient.
Posts about AI company funding, valuations, business models, or financial metrics without demonstrating healthcare application or outcomes.
3 example posts
If gov't takes a stake, how do you value these IPOs? Government equity usually signals utility (regulated/ slow/priced for dividends). But OpenAI and Anthropic are the opposite of that.... Maybe a new category: strategic tech --where a government stake is a premium. No clean
Scoop! Lila Sciences is in talks to raise ~$2 billion in new funding. The raise would value the “scientific superintelligence” lab at ~$8.5 billion before the new money. w/ @MichelleF_Davis, read more👇
⚡️AI is turning cybersecurity from a human-limited profession into a machine-speed arms race. The surface read is “Anthropic’s model found vulnerabilities fast.” The deeper read is that the bottleneck in cyber is shifting. For years, vulnerability discovery was constrained
Created 2026-06-06 · Updated 2026-06-06
Edit rule text
[non_healthcare_infrastructure_and_compute_hype]
Learned 3 rejections Active
Exclude posts about AI infrastructure (NVIDIA chips, energy/power requirements, GPU specs, data center architecture, token economics) or space-based compute unless they explicitly validate or apply these capabilities to healthcare delivery, clinical outcomes, or healthcare-specific workflows. Generic infrastructure announcements do not qualify.
Posts about AI infrastructure, compute, energy, or hardware announcements without healthcare application or validation.
3 example posts
Elon Musk reveals why he believes the cheapest place to put AI will be space within 36 months "The availability of energy is the issue. Everywhere outside of China, electrical output is more or less flat. The output of chips is growing exponentially, but the output of https://t.
The context window pricing model might be one of the smartest tricks AI labs pulled. You pay for larger context windows, yet model performance often degrades as more context is stuffed in. So developers end up paying more just to work around the fact that models don't have true
a $2,000 graphics card and a free download are turning into $2,000-a-month retainers, and the customers are the ones who legally can't use ChatGPT the hardware is an RTX 5090, 32GB of VRAM, enough to run a 30B model locally with nothing ever leaving the building. the software is
Created 2026-06-06 · Updated 2026-06-06
Edit rule text
[policy_or_regulatory_outrage_without_healthcare_systems_analysis]
Learned 3 rejections Active
Exclude posts framing healthcare policy, prior authorization, provider network exclusion, or fraud as moral failures or personal injustices without substantive analysis of how these systems fail and what structural reforms are needed. Emotional storytelling or vigilante-framed narratives without systems insight are out of scope.
Posts using moral outrage or anecdotal grievance to critique healthcare policy, regulation, or fraud without systemic analysis of root causes or solutions.
3 example posts
USDA's Chief Information Officer Sam Berry says AI could help the USDA crack down on billions of dollars in SNAP fraud each year. "SNAP is a $100B-per-year taxpayer-funded program. That's an area where we really want to have all angles of the data available so that we can deploy
My job is to make sure your surgery center never gets built. Eleven years, and I have never lost. The kid had it all lined up. Board-certified, two partners, a lease on a space where he could do the same procedures my client does across town. He showed up to the hearing with a
PacificSource pulled out as Lane County Medicaid insurer, Tina Kotek’s OR Health Authority replaced it with Trillium, forcing 96k to join Trillium. Trillium’s parent co gave Kotek’s campaign $50k two months later. OR AG (D) said in ‘22 Trillium parent co gave Kotek’s”took advan
Created 2026-06-06 · Updated 2026-06-06
Edit rule text
[clinical_trial_or_pharma_data_announcement_without_healthcare_systems_analysis]
Learned 3 rejections Active
Exclude posts that are primarily trial data announcements, drug efficacy reports, or pharma milestone press releases without substantive analysis of how these results change clinical practice, healthcare delivery workflows, or patient access. Raw trial data or efficacy numbers without systems context are low-substance.
Posts that report drug trial results, clinical data, or pharmaceutical announcements without analyzing healthcare system implications or clinical decision-making impact.
3 example posts
#ASCO26 | Day 3 $LLY didn't spend billions on Kelonia for an 18-patient #myeloma dataset. They may have paid for a potential manufacturing disruption. KLN-1010 reported: • 18/18 MRD-negative at 1 month • No lymphodepletion • No ex vivo cell manufacturing • Single infusion
Published in @JCO_ASCO, during #ASCO26: Tumor-Agnostic Therapies: Translating Scientific Breakthroughs Into Global Implementation ASCO is where we celebrate the next breakthrough. This review asks why breakthroughs do not equal access globally. 🔗: https://t.co/PE5q27bIs1 htt
A rapid manufacturing pipeline for BCMA-targeting CAR T cells drastically shortens “vein-to-vein” times, and produces cells that demonstrate encouraging safety results in a phase one trial of patients with relapsed #MultipleMyeloma. @ScienceTM https://t.co/3F4u8ECFDS https://t.co
Created 2026-06-06 · Updated 2026-06-06
Edit rule text
[regulatory_fraud_scandal_reporting_no_systems_analysis]
Learned 3 rejections Active
Exclude posts that report individual fraud cases, regulatory enforcement, insurance denials, or healthcare system failures (prior authorization abuses, provider directory problems) as isolated scandals or personal anecdotes without analyzing underlying structural healthcare economics, incentive misalignments, or systemic policy failures. The post must provide systems-level insight, not just reporting outrage.
Posts reporting healthcare fraud, regulatory scandals, or prior auth/reimbursement issues without systems-level analysis of root causes or structural healthcare economics.
3 example posts
My job is to make sure your surgery center never gets built. Eleven years, and I have never lost. The kid had it all lined up. Board-certified, two partners, a lease on a space where he could do the same procedures my client does across town. He showed up to the hearing with a
PacificSource pulled out as Lane County Medicaid insurer, Tina Kotek’s OR Health Authority replaced it with Trillium, forcing 96k to join Trillium. Trillium’s parent co gave Kotek’s campaign $50k two months later. OR AG (D) said in ‘22 Trillium parent co gave Kotek’s”took advan
I called a woman in Dayton to tell her she was about to overpay $4,500, for a horrible health system based procedure. She was scheduled for a knee replacement. Her PPO had her on the hook for a $4,500 deductible. The call was on behalf of her employer: “See one of these three
Created 2026-06-06 · Updated 2026-06-06
Edit rule text
[agentic_ai_technical_tangent_loose_healthcare]
Learned 3 rejections Active
Exclude posts that announce or discuss agentic AI systems, agent security frameworks, control planes, or orchestration platforms (NemoClaw, OpenShell, Hermes, etc.) as technical capability announcements or product launches without demonstrating healthcare-specific validation, clinical use cases, or evidence of deployment in actual healthcare workflows.
Posts about agentic AI systems, control planes, security frameworks, and agent orchestration as pure technical capability announcements without healthcare validation or operational grounding.
3 example posts
Control plane! Control plane! Control plane! You will hear this term of art a lot going forward. Why? Because in this next phase of AI, companies will want something to sit above the models. They will want control over their AI spend. They will want the flexibility to pick
Nous Research is working with NVIDIA to make Hermes Agent run smoothly on the new NVIDIA RTX Spark superchip. Hermes Agent is also integrating with the new OpenShell runtime, which connects Hermes to Microsoft’s security primitives https://t.co/cvvPc8bIKa
BREAKING: Merge Launches ‘Agent Handler’ Control AI Access, Tokenmaxxed $$$ Bills, & Stop Mass Data Leaks "We don't trust agents" "The second you connect it to tools, that's where everything goes wrong." OpenAI. Perplexity. Netflix. Uber. Mistral. Dropbox. JPMorgan.. all h
Created 2026-06-06 · Updated 2026-06-06
Edit rule text
[enterprise_ai_adoption_macro_no_healthcare_systems]
Learned 3 rejections Active
Exclude posts that discuss enterprise AI adoption, agent deployment, enterprise software pricing models, or workflow redesign in generic terms without grounding in specific healthcare operational challenges, clinical outcomes, or healthcare system constraints. Posts must demonstrate healthcare-specific application, not macro business model commentary.
Posts about enterprise AI adoption, agentic systems, and workflow redesign without healthcare systems analysis or operational insight.
3 example posts
Enterprise software was priced per seat. That model is breaking. Before: SaaS = seats. Predictable. Human-driven usage. Revenue tied to headcount. Now: agents hit software systems more frequently than any human ever could. Aaron Levie (@levie) on The MAD Podcast: agent-driven h
I just looked at a home service firms ServiceTitan dashboard and saw something interesting. A Pre-seed company with AI-Supported Human CSRs is out converting an AI CSR business that's raised over $125M by >50% Head-to-Head win rate is 100% Why? Supercharging humans resonat
Nous Research is working with NVIDIA to make Hermes Agent run smoothly on the new NVIDIA RTX Spark superchip. Hermes Agent is also integrating with the new OpenShell runtime, which connects Hermes to Microsoft’s security primitives https://t.co/cvvPc8bIKa
Created 2026-06-06 · Updated 2026-06-06
Edit rule text
[non_healthcare_domain_tangential]
Learned 3 rejections Active
Exclude posts about general AI infrastructure (NVIDIA chips, token pricing, energy costs), speculative robotics applications, or labor market macro trends that do not demonstrate concrete healthcare operational impact or policy context.
Posts about AI infrastructure, robotics, energy, or macro trends that lack direct healthcare application or systems analysis.
3 example posts
Elon Musk reveals why he believes the cheapest place to put AI will be space within 36 months "The availability of energy is the issue. Everywhere outside of China, electrical output is more or less flat. The output of chips is growing exponentially, but the output of https://t.
a $2,000 graphics card and a free download are turning into $2,000-a-month retainers, and the customers are the ones who legally can't use ChatGPT the hardware is an RTX 5090, 32GB of VRAM, enough to run a 30B model locally with nothing ever leaving the building. the software is
Free newsletter: The dawn of token-based billing has shown that generative AI doesn’t have a return on investment. It's too unpredictable, too unreliable, you can't easily measure the cost of tasks, and organizations are already pulling back. https://t.co/wmI82zWdcq
Created 2026-06-06 · Updated 2026-06-06
Edit rule text
[tangential_ai_infrastructure_or_compute_hype_no_healthcare_application]
Learned 3 rejections Active
Exclude posts about GPU performance, inference costs, token economics, inference optimization, or AI chip announcements unless they directly address a concrete healthcare delivery or operational challenge. Generic AI infrastructure hype with healthcare labels does not qualify.
Posts about AI infrastructure, compute hardware, chip architecture, or inference optimization with only loose or tangential healthcare framing.
3 example posts
Elon Musk reveals why he believes the cheapest place to put AI will be space within 36 months "The availability of energy is the issue. Everywhere outside of China, electrical output is more or less flat. The output of chips is growing exponentially, but the output of https://t.
Nous Research is working with NVIDIA to make Hermes Agent run smoothly on the new NVIDIA RTX Spark superchip. Hermes Agent is also integrating with the new OpenShell runtime, which connects Hermes to Microsoft’s security primitives https://t.co/cvvPc8bIKa
The context window pricing model might be one of the smartest tricks AI labs pulled. You pay for larger context windows, yet model performance often degrades as more context is stuffed in. So developers end up paying more just to work around the fact that models don't have true
Created 2026-06-06 · Updated 2026-06-06
Edit rule text
[labor_market_macro_without_healthcare_lens]
Learned 3 rejections Active
Exclude posts that discuss labor market disruption, automation value, or workforce trends (robots, humanoid workers, manufacturing automation) that use healthcare only as a loose metaphor or afterthought. The post must analyze HEALTHCARE labor challenges or supply chains, not apply generalist labor/robotics commentary to healthcare.
Posts about general labor disruption, automation, or workforce trends without healthcare-specific context or analysis
3 example posts
A framework I wish more Founders and VCs used when discussing US Manufacturing: Automation Value = Labor Intensity × Labor Eliminated Everyone gets excited about robots and automation. Almost nobody asks: “How much of revenue is labor?” If labor is 60% of revenue and AI http
I just looked at a home service firms ServiceTitan dashboard and saw something interesting. A Pre-seed company with AI-Supported Human CSRs is out converting an AI CSR business that's raised over $125M by >50% Head-to-Head win rate is 100% Why? Supercharging humans resonat
Humanoid robots, designed to mimic human movement and capabilities, have been making headlines in recent years, from their use as baggage handlers at Japanese airports to Tesla’s big bet on its Optimus humanoid. Market watchers have predicted that the machines will change the ht
Created 2026-06-06 · Updated 2026-06-06
Edit rule text
[tangential_policy_or_regulatory_outrage_without_healthcare_systems_analysis]
Learned 3 rejections Active
Exclude posts that express outrage about policy, regulation, government decisions, or fraud/scandal without analyzing how it affects healthcare delivery systems, patient access, clinical practice, or healthcare economics. Posts about political figures making healthcare claims, regulatory battles, or personal cost narratives without systems insight are out of scope.
Posts about policy, regulation, or fraud scandals framed with healthcare language but lacking systems-level healthcare analysis
3 example posts
My job is to make sure your surgery center never gets built. Eleven years, and I have never lost. The kid had it all lined up. Board-certified, two partners, a lease on a space where he could do the same procedures my client does across town. He showed up to the hearing with a
PacificSource pulled out as Lane County Medicaid insurer, Tina Kotek’s OR Health Authority replaced it with Trillium, forcing 96k to join Trillium. Trillium’s parent co gave Kotek’s campaign $50k two months later. OR AG (D) said in ‘22 Trillium parent co gave Kotek’s”took advan
63% of ATS respiratory guidelines are based on low-quality evidence. We published the data. The next morning, the chair of ATS got on Twitter. Tagged people. Went on a tirade. Didn't seem to realize tweets were public. "It wasn't a private conversation." The critique? We ht
Created 2026-06-06 · Updated 2026-06-06
Edit rule text
[business_model_or_market_speculation_without_healthcare_systems_analysis]
Learned 3 rejections Active
Exclude posts that debate or speculate on business models, market competition, pricing dynamics, or competitive moats (e.g., 'proprietary data = moat,' 'build vs. buy strategy') in generic tech terms without demonstrating healthcare-specific operational challenges, regulatory barriers, or care system implications.
Posts about business model debates, market dynamics, pricing strategies, or competitive positioning in tech/AI without healthcare system or operational substance.
3 example posts
Another great research piece by @_DimensionCap @bauer_lesavage on Training Data for Bio AI. Models will only be as good as the underlying data, and the biology they learn will be constrained by the limitations of that data. We need to think deeply about scaling the best
Control plane! Control plane! Control plane! You will hear this term of art a lot going forward. Why? Because in this next phase of AI, companies will want something to sit above the models. They will want control over their AI spend. They will want the flexibility to pick
If gov't takes a stake, how do you value these IPOs? Government equity usually signals utility (regulated/ slow/priced for dividends). But OpenAI and Anthropic are the opposite of that.... Maybe a new category: strategic tech --where a government stake is a premium. No clean
Created 2026-06-06 · Updated 2026-06-06
Edit rule text
[generalist_macro_policy_labor_without_healthcare_lens]
Learned 3 rejections Active
Exclude posts discussing general labor market disruption, humanoid robotics, manufacturing automation, or macro workforce trends (e.g., 'AI will eliminate jobs,' 'robots replace workers') unless they directly address healthcare workforce challenges, clinical staffing models, or care delivery labor economics.
Posts about labor market disruption, workforce automation, robotics, or macro-economic trends that lack healthcare-specific context or systems analysis.
3 example posts
A framework I wish more Founders and VCs used when discussing US Manufacturing: Automation Value = Labor Intensity × Labor Eliminated Everyone gets excited about robots and automation. Almost nobody asks: “How much of revenue is labor?” If labor is 60% of revenue and AI http
I just looked at a home service firms ServiceTitan dashboard and saw something interesting. A Pre-seed company with AI-Supported Human CSRs is out converting an AI CSR business that's raised over $125M by >50% Head-to-Head win rate is 100% Why? Supercharging humans resonat
What kind of work will become more valuable in an AI economy? @ccatalini, former head economist at Meta: "AI is getting better and better at automating anything that can be measured, as long as you have a digital trace, if you can collect it with a device, that problem will be
Created 2026-06-06 · Updated 2026-06-06
Edit rule text
[ai_model_capability_or_safety_tangent_without_healthcare_validation]
Learned 3 rejections Active
Exclude posts that focus on AI model technical capabilities (token pricing, context windows, inference speed), safety vulnerabilities, or cybersecurity incidents without demonstrating a healthcare-specific risk, clinical use case, or healthcare system impact. Posts must connect the capability or vulnerability to actual healthcare deployment challenges.
Posts about AI model technical capabilities, safety vulnerabilities, or cybersecurity incidents that lack healthcare context or clinical validation.
3 example posts
Control plane! Control plane! Control plane! You will hear this term of art a lot going forward. Why? Because in this next phase of AI, companies will want something to sit above the models. They will want control over their AI spend. They will want the flexibility to pick
If gov't takes a stake, how do you value these IPOs? Government equity usually signals utility (regulated/ slow/priced for dividends). But OpenAI and Anthropic are the opposite of that.... Maybe a new category: strategic tech --where a government stake is a premium. No clean
"You can run OpenClaw inside your company now." Annoucing our work with @Microsoft to bring OpenClaw to the Microsoft and Windows ecosystems. Claws now work securly in the enterprise. https://t.co/3fKhT2mGpC
Created 2026-06-06 · Updated 2026-06-06
Edit rule text
[ai_infrastructure_and_compute_hype_without_healthcare_specificity]
Learned 3 rejections Active
Exclude posts about AI compute infrastructure, chip performance, energy availability, data centers, or hardware breakthroughs (e.g., NVIDIA chips, RTX cards, token economics) unless they directly address a healthcare-specific deployment bottleneck or clinical validation challenge. Generic infrastructure hype does not qualify.
Posts about AI infrastructure, energy, compute scaling, chips, and hardware that are tangential to healthcare applications.
3 example posts
Elon Musk reveals why he believes the cheapest place to put AI will be space within 36 months "The availability of energy is the issue. Everywhere outside of China, electrical output is more or less flat. The output of chips is growing exponentially, but the output of https://t.
Nous Research is working with NVIDIA to make Hermes Agent run smoothly on the new NVIDIA RTX Spark superchip. Hermes Agent is also integrating with the new OpenShell runtime, which connects Hermes to Microsoft’s security primitives https://t.co/cvvPc8bIKa
a $2,000 graphics card and a free download are turning into $2,000-a-month retainers, and the customers are the ones who legally can't use ChatGPT the hardware is an RTX 5090, 32GB of VRAM, enough to run a 30B model locally with nothing ever leaving the building. the software is
Created 2026-06-06 · Updated 2026-06-06
Edit rule text
[tangential_non_healthcare_domain_with_loose_healthcare_label]
Learned 3 rejections Active
Exclude posts where the primary content is about non-healthcare domains (energy infrastructure, space technology, general robotics, macro economics, semiconductor manufacturing) and healthcare is mentioned only as a potential future application or loose tag. The post must demonstrate concrete healthcare relevance, not speculative sci-fi applications.
Posts about non-healthcare domains (space, robotics, macro economics, energy) tagged or framed as healthcare-relevant but with only superficial connection.
3 example posts
Elon Musk reveals why he believes the cheapest place to put AI will be space within 36 months "The availability of energy is the issue. Everywhere outside of China, electrical output is more or less flat. The output of chips is growing exponentially, but the output of https://t.
Humanoid robots, designed to mimic human movement and capabilities, have been making headlines in recent years, from their use as baggage handlers at Japanese airports to Tesla’s big bet on its Optimus humanoid. Market watchers have predicted that the machines will change the ht
In the AI gold rush, Jensen Huang is selling the picks and shovels. Nvidia's chips power AI companies around the world, helping the company become the first to surpass a $5 trillion market capitalization in late 2025. Read more about how he and others made the https://t.co/lHLE
Created 2026-06-06 · Updated 2026-06-06
Edit rule text
[founder_enthusiasm_or_product_announcement_without_validation]
Learned 3 rejections Active
Exclude posts where founders, company representatives, or enthusiasts announce new products, capabilities, or partnerships without providing independent validation, clinical evidence, or analysis of healthcare impact. Press releases, product launch announcements, and venture announcements without substantive healthcare systems analysis should be excluded.
Posts promoting biotech, AI startup, or company products without clinical validation or healthcare systems evidence.
3 example posts
"You can run OpenClaw inside your company now." Annoucing our work with @Microsoft to bring OpenClaw to the Microsoft and Windows ecosystems. Claws now work securly in the enterprise. https://t.co/3fKhT2mGpC
BREAKING: Merge Launches ‘Agent Handler’ Control AI Access, Tokenmaxxed $$$ Bills, & Stop Mass Data Leaks "We don't trust agents" "The second you connect it to tools, that's where everything goes wrong." OpenAI. Perplexity. Netflix. Uber. Mistral. Dropbox. JPMorgan.. all h
Your Oura Ring, your health record. Together. Finally. @Flexpa is bringing clinical records into @ouraring via TEFCA, so ŌURA’s AI isn’t just working from what your ring measures, but from your real health history. https://t.co/b1KCeGEZLO https://t.co/u2t1TQjL7p
Created 2026-06-06 · Updated 2026-06-06
Edit rule text
[enterprise_ai_adoption_macro_without_healthcare_systems_analysis]
Learned 3 rejections Active
Exclude posts discussing general enterprise AI trends, SaaS model disruption, labor automation, or business process transformation where healthcare is incidental or used as a generic example. Posts must connect the trend directly to healthcare operations, workflows, or outcomes — not just mention healthcare companies.
Posts about enterprise AI adoption, business model disruption, or labor market trends without healthcare-specific systems analysis.
3 example posts
Enterprise software was priced per seat. That model is breaking. Before: SaaS = seats. Predictable. Human-driven usage. Revenue tied to headcount. Now: agents hit software systems more frequently than any human ever could. Aaron Levie (@levie) on The MAD Podcast: agent-driven h
I just looked at a home service firms ServiceTitan dashboard and saw something interesting. A Pre-seed company with AI-Supported Human CSRs is out converting an AI CSR business that's raised over $125M by >50% Head-to-Head win rate is 100% Why? Supercharging humans resonat
Most enterprise AI projects stall because no one has prepared the content, built the workflow, and kept it improving as models change. That role is essential. Box has launched Forward Deployed Engineers to fill it. Model-agnostic, content-first, and built for the way enterprise
Created 2026-06-06 · Updated 2026-06-06
Edit rule text
[ai_infrastructure_compute_hype_no_healthcare_validation]
Learned 3 rejections Active
Exclude posts about AI infrastructure fundamentals (GPUs, chips, data centers, energy availability, inference costs, context windows, token economics) unless they explicitly discuss a healthcare-specific deployment, clinical trial, or validated health system use case. Posts about NVIDIA, chip suppliers, or compute architecture without healthcare substance should be rejected.
Posts about AI infrastructure, compute capacity, chips, and energy/hardware requirements without demonstrated healthcare application or clinical validation.
3 example posts
Elon Musk reveals why he believes the cheapest place to put AI will be space within 36 months "The availability of energy is the issue. Everywhere outside of China, electrical output is more or less flat. The output of chips is growing exponentially, but the output of https://t.
Nous Research is working with NVIDIA to make Hermes Agent run smoothly on the new NVIDIA RTX Spark superchip. Hermes Agent is also integrating with the new OpenShell runtime, which connects Hermes to Microsoft’s security primitives https://t.co/cvvPc8bIKa
The context window pricing model might be one of the smartest tricks AI labs pulled. You pay for larger context windows, yet model performance often degrades as more context is stuffed in. So developers end up paying more just to work around the fact that models don't have true
Created 2026-06-06 · Updated 2026-06-06
Edit rule text
[tangential_ai_company_product_announcement_without_healthcare_validation]
Learned 3 rejections Active
Exclude posts that announce AI company product launches, partnerships, or platform integrations (OpenAI announcements, Anthropic releases, NVIDIA partnerships, Microsoft collaborations) unless they demonstrate validated healthcare workflow or clinical outcome impact.
Posts announcing AI company product launches, partnerships, or integrations without healthcare-specific validation or use case
3 example posts
"You can run OpenClaw inside your company now." Annoucing our work with @Microsoft to bring OpenClaw to the Microsoft and Windows ecosystems. Claws now work securly in the enterprise. https://t.co/3fKhT2mGpC
Your Oura Ring, your health record. Together. Finally. @Flexpa is bringing clinical records into @ouraring via TEFCA, so ŌURA’s AI isn’t just working from what your ring measures, but from your real health history. https://t.co/b1KCeGEZLO https://t.co/u2t1TQjL7p
The next frontier of AI-secured financial and technological critical infrastructure. @ICE_Markets and @NYSE are part of @AnthropicAI's cyber security initiative Project Glasswing, deploying Anthropic’s Claude Mythos Preview across ICE’s exchanges, clearing houses, mortgage
Created 2026-06-06 · Updated 2026-06-06
Edit rule text
[generalist_infrastructure_or_compute_hype_without_healthcare_application]
Learned 3 rejections Active
Exclude posts that discuss compute infrastructure, GPU capabilities (RTX 5090, NVIDIA hardware), energy availability, or token economics without validating healthcare-specific application, workflow, or clinical impact.
Posts about GPU infrastructure, compute, energy costs, or hardware capabilities with only tangential or loose healthcare framing
3 example posts
Elon Musk reveals why he believes the cheapest place to put AI will be space within 36 months "The availability of energy is the issue. Everywhere outside of China, electrical output is more or less flat. The output of chips is growing exponentially, but the output of https://t.
The context window pricing model might be one of the smartest tricks AI labs pulled. You pay for larger context windows, yet model performance often degrades as more context is stuffed in. So developers end up paying more just to work around the fact that models don't have true
a $2,000 graphics card and a free download are turning into $2,000-a-month retainers, and the customers are the ones who legally can't use ChatGPT the hardware is an RTX 5090, 32GB of VRAM, enough to run a 30B model locally with nothing ever leaving the building. the software is
Created 2026-06-06 · Updated 2026-06-06
Edit rule text
[broad_health_claims_or_policy_without_systems_analysis]
Learned 3 rejections Active
Exclude posts that make sweeping claims about health, policy, regulation, or healthcare economics (e.g., 'Long COVID is underdiagnosed', 'the NHS trial shows...') without grounding in operational, clinical, or healthcare systems context. Personal anecdotes, single observations, or broad policy commentary without evidence of systems-level impact belong here.
Posts making broad health claims, policy critiques, or regulatory observations without evidence of healthcare systems impact or operational specificity.
3 example posts
I called a woman in Dayton to tell her she was about to overpay $4,500, for a horrible health system based procedure. She was scheduled for a knee replacement. Her PPO had her on the hook for a $4,500 deductible. The call was on behalf of her employer: “See one of these three
@WIRED As a clinical health psychologist who has written >20 papers on COVID, I would emphasize 4 facts: 1) Long COVID is not a psychological diagnosis nor manifestation of a psychological condition 2) Billions of dollars need to be invested in biomedical treatments and preventi
"Many patients with long COVID are already receiving care but are not being recognized as having the condition. These patients are not absent from clinical care; they are absent from the diagnostic code that would identify them as long COVID patients" https://t.co/aLJCOPlQt1
Created 2026-06-06 · Updated 2026-06-06
Edit rule text
[labor_disruption_and_robotics_macro]
Learned 3 rejections Active
Exclude posts about labor market macro trends, humanoid robots, automation, or workforce disruption (e.g., 'robots are taking jobs', 'agents will disrupt CSRs') unless the post specifically analyzes how the trend affects healthcare staffing, clinical workflows, patient care delivery, or healthcare operations in a validated or evidence-based way.
Posts about workforce disruption, humanoid robots, labor market shifts, or automation trends without healthcare-specific operational context.
3 example posts
What kind of work will become more valuable in an AI economy? @ccatalini, former head economist at Meta: "AI is getting better and better at automating anything that can be measured, as long as you have a digital trace, if you can collect it with a device, that problem will be
Humanoid robots, designed to mimic human movement and capabilities, have been making headlines in recent years, from their use as baggage handlers at Japanese airports to Tesla’s big bet on its Optimus humanoid. Market watchers have predicted that the machines will change the ht
Another humanoid worker has just joined the factory floor Two years later,PUDU Robotics has launched the new-generation PUDU D7, built specifically for real manufacturing scenarios. It can autonomously push carts, handle intra-line transportation, perform delicate operations, h
Created 2026-06-06 · Updated 2026-06-06
Edit rule text
[market_dynamics_and_financial_speculation]
Learned 3 rejections Active
Exclude posts that focus primarily on market pricing dynamics, funding rounds, company valuations, competitive positioning, or financial/business model debate (e.g., 'SaaS pricing is breaking', 'government equity signals utility') unless the post directly analyzes how the market dynamic affects healthcare delivery, cost, access, or clinical decision-making.
Posts about market pricing, funding, valuations, competitive positioning, or financial dynamics without healthcare systems analysis.
3 example posts
Enterprise software was priced per seat. That model is breaking. Before: SaaS = seats. Predictable. Human-driven usage. Revenue tied to headcount. Now: agents hit software systems more frequently than any human ever could. Aaron Levie (@levie) on The MAD Podcast: agent-driven h
If gov't takes a stake, how do you value these IPOs? Government equity usually signals utility (regulated/ slow/priced for dividends). But OpenAI and Anthropic are the opposite of that.... Maybe a new category: strategic tech --where a government stake is a premium. No clean
The context window pricing model might be one of the smartest tricks AI labs pulled. You pay for larger context windows, yet model performance often degrades as more context is stuffed in. So developers end up paying more just to work around the fact that models don't have true
Created 2026-06-06 · Updated 2026-06-06
Edit rule text
[ai_company_product_or_partnership_announcement_unvalidated]
Learned 3 rejections Active
Exclude posts announcing AI company products, partnerships, integrations, or capability releases (e.g., 'Anthropic ships Opus', 'OpenAI partners with X', 'NVIDIA releases new runtime') unless the post provides evidence that the tool has been validated in real healthcare workflows, solves a documented clinical bottleneck, or demonstrates measurable impact on healthcare operations.
Posts announcing AI company product launches, partnerships, or integrations without evidence of healthcare impact or clinical validation.
3 example posts
"You can run OpenClaw inside your company now." Annoucing our work with @Microsoft to bring OpenClaw to the Microsoft and Windows ecosystems. Claws now work securly in the enterprise. https://t.co/3fKhT2mGpC
Your Oura Ring, your health record. Together. Finally. @Flexpa is bringing clinical records into @ouraring via TEFCA, so ŌURA’s AI isn’t just working from what your ring measures, but from your real health history. https://t.co/b1KCeGEZLO https://t.co/u2t1TQjL7p
The next frontier of AI-secured financial and technological critical infrastructure. @ICE_Markets and @NYSE are part of @AnthropicAI's cyber security initiative Project Glasswing, deploying Anthropic’s Claude Mythos Preview across ICE’s exchanges, clearing houses, mortgage
Created 2026-06-06 · Updated 2026-06-06
Edit rule text
[enterprise_ai_infrastructure_macro]
Learned 3 rejections Active
Exclude posts that discuss AI infrastructure (compute, tokens, agents, control planes, runtimes) or enterprise AI adoption as abstract macro trends, unless the post explicitly demonstrates how these apply to a specific healthcare workflow, operational bottleneck, or clinical decision-making problem. Posts about 'agents will transform enterprise' or 'compute costs are falling' without healthcare context should be rejected.
Posts about AI infrastructure, compute, agents, and enterprise adoption at macro level without healthcare-specific application or validation.
3 example posts
Enterprise software was priced per seat. That model is breaking. Before: SaaS = seats. Predictable. Human-driven usage. Revenue tied to headcount. Now: agents hit software systems more frequently than any human ever could. Aaron Levie (@levie) on The MAD Podcast: agent-driven h
Control plane! Control plane! Control plane! You will hear this term of art a lot going forward. Why? Because in this next phase of AI, companies will want something to sit above the models. They will want control over their AI spend. They will want the flexibility to pick
"You can run OpenClaw inside your company now." Annoucing our work with @Microsoft to bring OpenClaw to the Microsoft and Windows ecosystems. Claws now work securly in the enterprise. https://t.co/3fKhT2mGpC
Created 2026-06-06 · Updated 2026-06-06
Edit rule text
[tangential_business_model_or_market_dynamics_without_healthcare_systems_analysis]
Learned 3 rejections Active
Exclude posts that debate business models (e.g., 'build vs. buy', proprietary moats, pricing strategies, market consolidation) or competitive positioning without substantive healthcare systems analysis. The post must address how the business model or market dynamic affects patient care, clinical outcomes, healthcare access, or operational efficiency.
Posts discussing business models, market competition, pricing strategies, or competitive dynamics in healthcare and tech without systems-level insight into healthcare operations or patient outcomes.
3 example posts
AI made building cheap. Regulation made distribution expensive. Anyone can clone your demo now. Very few can bring proprietary data, licenses, compliance, and enough trust to sell into serious players. The moat didn’t disappear. It moved.
Interoperability is where AI hype goes to die. If your “agent” can’t work across data, logic, actions, security of legacy systems, it’s not transforming the enterprise. It’s just another app asking for an export. Winners don’t rip and replace. They operate.
Anthropic just shipped opus 4.8 and called it a modest upgrade modest is the engineer who used to need babysitting now finishing your whole codebase migration in one session while you sleep so here is the part nobody connects the model is no longer the bottleneck. the platform
Created 2026-06-05 · Updated 2026-06-05
Edit rule text
[pharma_trial_data_or_biotech_announcement_without_systems_context]
Learned 3 rejections Active
Exclude posts that report clinical trial data, drug efficacy results, FDA approvals, cancer treatment breakthroughs, or biotech founder announcements (especially at conferences like ASCO) unless the post contextualizes how these advances solve healthcare system problems, change clinical workflows, improve access, or address operational constraints in healthcare delivery.
Posts announcing pharmaceutical trial results, drug approvals, or biotech founder enthusiasm without analyzing implications for healthcare systems, delivery, or operations.
3 example posts
Scoop! Lila Sciences is in talks to raise ~$2 billion in new funding. The raise would value the “scientific superintelligence” lab at ~$8.5 billion before the new money. w/ @MichelleF_Davis, read more👇
#ASCO26 | Day 3 $LLY didn't spend billions on Kelonia for an 18-patient #myeloma dataset. They may have paid for a potential manufacturing disruption. KLN-1010 reported: • 18/18 MRD-negative at 1 month • No lymphodepletion • No ex vivo cell manufacturing • Single infusion
Published in @JCO_ASCO, during #ASCO26: Tumor-Agnostic Therapies: Translating Scientific Breakthroughs Into Global Implementation ASCO is where we celebrate the next breakthrough. This review asks why breakthroughs do not equal access globally. 🔗: https://t.co/PE5q27bIs1 htt
Created 2026-06-05 · Updated 2026-06-05
Edit rule text
[generalist_ai_infrastructure_or_macro_hype_loose_healthcare_framing]
Learned 3 rejections Active
Exclude posts discussing generalist AI infrastructure (token economics, context window pricing, inference costs, NVIDIA chips, compute hardware) that merely mention healthcare or use healthcare as a loose example—unless the post analyzes how these infrastructure constraints specifically impact healthcare AI applications or clinical workflows.
Posts about generalist AI infrastructure, compute economics, or macro AI trends with loose or tangential healthcare framing.
3 example posts
The context window pricing model might be one of the smartest tricks AI labs pulled. You pay for larger context windows, yet model performance often degrades as more context is stuffed in. So developers end up paying more just to work around the fact that models don't have true
Free newsletter: The dawn of token-based billing has shown that generative AI doesn’t have a return on investment. It's too unpredictable, too unreliable, you can't easily measure the cost of tasks, and organizations are already pulling back. https://t.co/wmI82zWdcq
the CEO of NVIDIA just said the computer is no longer being built for you, it is being built for agents Jensen Huang put it plainly, until now we were the users, we were the renters, every CPU on earth was designed around how a person works but agents do not work like us, they
Created 2026-06-05 · Updated 2026-06-05
Edit rule text
[anecdotal_healthcare_cost_or_operational_narrative_without_systems_insight]
Learned 3 rejections Active
Exclude posts that relay single patient anecdotes, individual healthcare cost horror stories, or isolated operational complaints (e.g., 'I called someone to warn them about overpayment', 'A mom called three health systems') without diagnosing systemic healthcare architecture problems, explaining why the current system structure enables these failures, or proposing healthcare system redesign. Single-case narratives about pricing opacity or admin burden should only be included if they build to a broader systems-level insight or solution pattern.
Posts sharing individual patient cost stories, healthcare pricing anecdotes, or operational pain points without connecting to systemic root causes or healthcare system redesign.
3 example posts
I called a woman in Dayton to tell her she was about to overpay $4,500, for a horrible health system based procedure. She was scheduled for a knee replacement. Her PPO had her on the hook for a $4,500 deductible. The call was on behalf of her employer: “See one of these three
A company with 800 employees spends roughly $15,000 per employee per year on health insurance. That is $12 million. Wired into a single line item. Every year. The CFO can tell you the carrier. He cannot tell you the unit cost of a single claim. He cannot tell you what the net
medicine is constrained far more by operational architecture than by fundamental science I just came head to head with this actual reality and my world has changed no. I cannot bandage away these inefficiencies with a better and better drug, for all cases. if you hate AI
Created 2026-06-05 · Updated 2026-06-05
Edit rule text
[vendor_product_launch_or_partnership_announcement]
Learned 3 rejections Active
Exclude posts that announce product launches, platform integrations, partnership agreements, or runtime deployments (e.g., 'X integrates with Y', 'Company launches tool Z') from tech vendors without proof of healthcare customer adoption, clinical validation, or measurable healthcare workflow improvement. Self-promotional announcements from vendors (including OpenAI, Anthropic, Microsoft, NVIDIA partnerships) should be treated as unvalidated unless accompanied by healthcare customer case studies or operational metrics.
Posts announcing AI vendor product launches, partnerships, or integrations without evidence of healthcare validation, customer adoption, or operational healthcare impact.
3 example posts
"You can run OpenClaw inside your company now." Annoucing our work with @Microsoft to bring OpenClaw to the Microsoft and Windows ecosystems. Claws now work securly in the enterprise. https://t.co/3fKhT2mGpC
Nous Research is working with NVIDIA to make Hermes Agent run smoothly on the new NVIDIA RTX Spark superchip. Hermes Agent is also integrating with the new OpenShell runtime, which connects Hermes to Microsoft’s security primitives https://t.co/cvvPc8bIKa
Your Oura Ring, your health record. Together. Finally. @Flexpa is bringing clinical records into @ouraring via TEFCA, so ŌURA’s AI isn’t just working from what your ring measures, but from your real health history. https://t.co/b1KCeGEZLO https://t.co/u2t1TQjL7p
Created 2026-06-05 · Updated 2026-06-05
Edit rule text
[pharma_trial_data_announcement_without_healthcare_systems_context]
Learned 3 rejections Active
Exclude posts that are purely trial data announcements, drug approval news, or clinical results reporting (ASCO conference updates, FDA approvals, MRD-negative responses) without analysis of healthcare delivery, access, cost, or system-level implications. These are clinical research announcements, not healthcare tech/operations analysis.
Posts announcing pharmaceutical trial data, drug approvals, or clinical results without healthcare systems analysis or operational impact
3 example posts
Scoop! Lila Sciences is in talks to raise ~$2 billion in new funding. The raise would value the “scientific superintelligence” lab at ~$8.5 billion before the new money. w/ @MichelleF_Davis, read more👇
#ASCO26 | Day 3 $LLY didn't spend billions on Kelonia for an 18-patient #myeloma dataset. They may have paid for a potential manufacturing disruption. KLN-1010 reported: • 18/18 MRD-negative at 1 month • No lymphodepletion • No ex vivo cell manufacturing • Single infusion
Published in @JCO_ASCO, during #ASCO26: Tumor-Agnostic Therapies: Translating Scientific Breakthroughs Into Global Implementation ASCO is where we celebrate the next breakthrough. This review asks why breakthroughs do not equal access globally. 🔗: https://t.co/PE5q27bIs1 htt
Created 2026-06-05 · Updated 2026-06-05
Edit rule text
[ai_agent_or_agentic_system_technical_tangent_without_healthcare_validation]
Learned 3 rejections Active
Exclude posts that discuss AI agent technical challenges (agent coordination, orchestration layers, tool access control, persistent reliability, failure modes, multi-agent orchestration) or agent framework debates unless the post grounds the discussion in a concrete healthcare deployment scenario with evidence of clinical or operational impact.
Posts about AI agent architecture, orchestration, security, reliability, or technical implementation presented as industry trends without demonstrating healthcare-specific use cases or clinical validation.
3 example posts
any company rolling out AI at scale is running into the same question. what can the agent reach and what can it touch. security is what’s slowing down mass adoption.
Almost everyone is building agent harness systems the wrong way. The default move: pick LangChain or LangGraph or the OpenAI Agents SDK, accept the loop, the tools, the memory, the orchestration, the policy engine, the credential store, the budget tracker, all of it, as one deci
Interoperability is where AI hype goes to die. If your “agent” can’t work across data, logic, actions, security of legacy systems, it’s not transforming the enterprise. It’s just another app asking for an export. Winners don’t rip and replace. They operate.
Created 2026-06-05 · Updated 2026-06-05
Edit rule text
[tangential_biotech_or_pharma_founder_enthusiasm_without_systems_validation]
Learned 3 rejections Active
Exclude posts that announce biotech company funding, valuation milestones, lab research updates, or founder statements (e.g., Lila Sciences $2B raise, Phrontline Biopharma NDA acceptance, gene therapy announcements) unless the post analyzes how the development addresses a systemic healthcare problem, barrier to care, or operational bottleneck with evidence beyond the announcement itself.
Posts featuring biotech/pharma founder announcements, funding raises, or lab updates that lack healthcare systems analysis, competitive context, or evidence of clinical/operational impact.
3 example posts
Scoop! Lila Sciences is in talks to raise ~$2 billion in new funding. The raise would value the “scientific superintelligence” lab at ~$8.5 billion before the new money. w/ @MichelleF_Davis, read more👇
#ASCO26 | Day 3 $LLY didn't spend billions on Kelonia for an 18-patient #myeloma dataset. They may have paid for a potential manufacturing disruption. KLN-1010 reported: • 18/18 MRD-negative at 1 month • No lymphodepletion • No ex vivo cell manufacturing • Single infusion
Published in @JCO_ASCO, during #ASCO26: Tumor-Agnostic Therapies: Translating Scientific Breakthroughs Into Global Implementation ASCO is where we celebrate the next breakthrough. This review asks why breakthroughs do not equal access globally. 🔗: https://t.co/PE5q27bIs1 htt
Created 2026-06-05 · Updated 2026-06-05
Edit rule text
[healthcare_financial_or_fraud_scandal_without_systems_analysis]
Learned 3 rejections Active
Exclude posts that report healthcare fraud, billing scandals, or cost anecdotes (e.g., 'patient was overcharged') unless the post connects the issue to a structural healthcare system problem, policy gap, or operational workflow failure that tech or policy can address.
Posts reporting healthcare fraud, financial scandals, or cost overruns as personal anecdotes or regulatory outrages without analysis of systemic policy or operational root causes.
3 example posts
I called a woman in Dayton to tell her she was about to overpay $4,500, for a horrible health system based procedure. She was scheduled for a knee replacement. Her PPO had her on the hook for a $4,500 deductible. The call was on behalf of her employer: “See one of these three
A company with 800 employees spends roughly $15,000 per employee per year on health insurance. That is $12 million. Wired into a single line item. Every year. The CFO can tell you the carrier. He cannot tell you the unit cost of a single claim. He cannot tell you what the net
A Mom in Tulsa called 3 health systems last week asking the price of her son's tonsillectomy. Health system A: "We cannot quote you a price." Health system B: "Pricing depends on your insurance." Health system C: "Our financial counselor will reach out after the procedure." N
Created 2026-06-05 · Updated 2026-06-05
Edit rule text
[tangential_ai_infrastructure_or_cybersecurity_hype]
Learned 3 rejections Active
Exclude posts about AI infrastructure (chips, inference costs, compute hardware), cybersecurity vulnerabilities, or general AI capability announcements that only tangentially relate to healthcare or lack demonstrated healthcare application. Infrastructure and security hype without healthcare specificity are insufficient.
Posts about AI infrastructure, compute, cybersecurity vulnerabilities, or general AI capability hype with loose healthcare framing
3 example posts
Free newsletter: The dawn of token-based billing has shown that generative AI doesn’t have a return on investment. It's too unpredictable, too unreliable, you can't easily measure the cost of tasks, and organizations are already pulling back. https://t.co/wmI82zWdcq
The next frontier of AI-secured financial and technological critical infrastructure. @ICE_Markets and @NYSE are part of @AnthropicAI's cyber security initiative Project Glasswing, deploying Anthropic’s Claude Mythos Preview across ICE’s exchanges, clearing houses, mortgage
⚡️AI is turning cybersecurity from a human-limited profession into a machine-speed arms race. The surface read is “Anthropic’s model found vulnerabilities fast.” The deeper read is that the bottleneck in cyber is shifting. For years, vulnerability discovery was constrained
Created 2026-06-05 · Updated 2026-06-05
Edit rule text
[clinical_research_or_trial_data_without_systems_context]
Learned 3 rejections Active
Exclude posts that report clinical trial results, drug efficacy data, or research observations without addressing how these findings will affect healthcare delivery, access, cost, or operational workflows. Raw trial data or research announcements alone are insufficient.
Posts reporting drug trial results, clinical data, or research findings without healthcare systems implementation context
3 example posts
#ASCO26 | Day 3 $LLY didn't spend billions on Kelonia for an 18-patient #myeloma dataset. They may have paid for a potential manufacturing disruption. KLN-1010 reported: • 18/18 MRD-negative at 1 month • No lymphodepletion • No ex vivo cell manufacturing • Single infusion
Published in @JCO_ASCO, during #ASCO26: Tumor-Agnostic Therapies: Translating Scientific Breakthroughs Into Global Implementation ASCO is where we celebrate the next breakthrough. This review asks why breakthroughs do not equal access globally. 🔗: https://t.co/PE5q27bIs1 htt
A rapid manufacturing pipeline for BCMA-targeting CAR T cells drastically shortens “vein-to-vein” times, and produces cells that demonstrate encouraging safety results in a phase one trial of patients with relapsed #MultipleMyeloma. @ScienceTM https://t.co/3F4u8ECFDS https://t.co
Created 2026-06-05 · Updated 2026-06-05
Edit rule text
[business_model_or_market_dynamics_without_systems_analysis]
Learned 3 rejections Active
Exclude posts that debate or analyze business models (licensing, pricing, enterprise adoption, competitive moats, cost structures) without connecting to measurable healthcare operational outcomes or system-level impacts. Market positioning alone is insufficient.
Posts discussing business models, competitive positioning, or market dynamics without healthcare systems insight
3 example posts
AI made building cheap. Regulation made distribution expensive. Anyone can clone your demo now. Very few can bring proprietary data, licenses, compliance, and enough trust to sell into serious players. The moat didn’t disappear. It moved.
Most enterprise AI projects stall because no one has prepared the content, built the workflow, and kept it improving as models change. That role is essential. Box has launched Forward Deployed Engineers to fill it. Model-agnostic, content-first, and built for the way enterprise
Almost everyone is building agent harness systems the wrong way. The default move: pick LangChain or LangGraph or the OpenAI Agents SDK, accept the loop, the tools, the memory, the orchestration, the policy engine, the credential store, the budget tracker, all of it, as one deci
Created 2026-06-05 · Updated 2026-06-05
Edit rule text
[clinical_or_pharma_data_announcement_only]
Learned 3 rejections Active
Exclude posts that report clinical trial outcomes, drug efficacy results, cancer treatment advances, or pharmaceutical milestones (ASCO conference presentations, FDA approvals, trial data) without discussing how these results change healthcare delivery, operational workflow, clinical decision-making, or system-level healthcare implementation. Trial data announcements alone are insufficient.
Posts announcing clinical trial results, drug efficacy data, or pharma milestones without healthcare systems analysis or implementation insight
3 example posts
#ASCO26 | Day 3 $LLY didn't spend billions on Kelonia for an 18-patient #myeloma dataset. They may have paid for a potential manufacturing disruption. KLN-1010 reported: • 18/18 MRD-negative at 1 month • No lymphodepletion • No ex vivo cell manufacturing • Single infusion
Published in @JCO_ASCO, during #ASCO26: Tumor-Agnostic Therapies: Translating Scientific Breakthroughs Into Global Implementation ASCO is where we celebrate the next breakthrough. This review asks why breakthroughs do not equal access globally. 🔗: https://t.co/PE5q27bIs1 htt
A rapid manufacturing pipeline for BCMA-targeting CAR T cells drastically shortens “vein-to-vein” times, and produces cells that demonstrate encouraging safety results in a phase one trial of patients with relapsed #MultipleMyeloma. @ScienceTM https://t.co/3F4u8ECFDS https://t.co
Created 2026-06-05 · Updated 2026-06-05
Edit rule text
[interoperability_or_protocol_announcement_without_healthcare_validation]
Learned 3 rejections Active
Exclude posts that announce technical standards (MCP, TEFCA, API frameworks, data interchange protocols) or system integrations without showing proof of healthcare deployment, clinical workflow improvement, or measurable healthcare system value.
Posts announcing technical interoperability standards, protocols, or integration frameworks without demonstrating healthcare implementation or operational impact.
3 example posts
Your Oura Ring, your health record. Together. Finally. @Flexpa is bringing clinical records into @ouraring via TEFCA, so ŌURA’s AI isn’t just working from what your ring measures, but from your real health history. https://t.co/b1KCeGEZLO https://t.co/u2t1TQjL7p
Red Hat and @NVIDIA are integrating NVIDIA OpenShell into the full-stack @RedHat_AI platform. The work brings oversight and policy to the infrastructure level, while contributing to the open source OpenShell project to standardize how agents are governed on enterprise platforms.
Interoperability is where AI hype goes to die. If your “agent” can’t work across data, logic, actions, security of legacy systems, it’s not transforming the enterprise. It’s just another app asking for an export. Winners don’t rip and replace. They operate.
Created 2026-06-05 · Updated 2026-06-05
Edit rule text
[robotics_or_speculative_sci_fi_applications]
Learned 3 rejections Active
Exclude posts about humanoid robots, robotic systems for non-clinical tasks (baggage handling, factory floors, autonomous vehicles), or speculative sci-fi applications. Include only if robotics directly solves a validated clinical or hospital operational problem with evidence.
Posts about humanoid robots, manufacturing automation, or speculative applications that lack grounding in actual healthcare delivery problems or clinical validation.
2 example posts
Humanoid robots, designed to mimic human movement and capabilities, have been making headlines in recent years, from their use as baggage handlers at Japanese airports to Tesla’s big bet on its Optimus humanoid. Market watchers have predicted that the machines will change the ht
Another humanoid worker has just joined the factory floor Two years later,PUDU Robotics has launched the new-generation PUDU D7, built specifically for real manufacturing scenarios. It can autonomously push carts, handle intra-line transportation, perform delicate operations, h
Created 2026-06-05 · Updated 2026-06-05
Edit rule text
[healthcare_anecdote_without_systems_context]
Learned 3 rejections Active
Exclude posts that describe individual healthcare cost overages, patient experiences, or clinical anecdotes without analyzing underlying healthcare system failures, operational inefficiencies, or structural incentive problems.
Personal healthcare cost stories or clinical anecdotes that lack systemic analysis of healthcare operations or policy.
3 example posts
I called a woman in Dayton to tell her she was about to overpay $4,500, for a horrible health system based procedure. She was scheduled for a knee replacement. Her PPO had her on the hook for a $4,500 deductible. The call was on behalf of her employer: “See one of these three
A company with 800 employees spends roughly $15,000 per employee per year on health insurance. That is $12 million. Wired into a single line item. Every year. The CFO can tell you the carrier. He cannot tell you the unit cost of a single claim. He cannot tell you what the net
A Mom in Tulsa called 3 health systems last week asking the price of her son's tonsillectomy. Health system A: "We cannot quote you a price." Health system B: "Pricing depends on your insurance." Health system C: "Our financial counselor will reach out after the procedure." N
Created 2026-06-05 · Updated 2026-06-05
Edit rule text
[anecdotal_single_case_healthcare_cost_or_patient_narrative_without_systems_analysis]
Learned 3 rejections Active
Exclude posts that tell individual patient stories, single healthcare cost anecdotes, or personal narratives about healthcare access/pricing unless the post extracts systemic insight into how healthcare operations, pricing structures, or administrative processes fail at scale. Single-case stories without systems analysis do not qualify.
Posts recounting individual patient cost stories or healthcare anecdotes without systemic analysis or operational insight
3 example posts
I called a woman in Dayton to tell her she was about to overpay $4,500, for a horrible health system based procedure. She was scheduled for a knee replacement. Her PPO had her on the hook for a $4,500 deductible. The call was on behalf of her employer: “See one of these three
@WIRED As a clinical health psychologist who has written >20 papers on COVID, I would emphasize 4 facts: 1) Long COVID is not a psychological diagnosis nor manifestation of a psychological condition 2) Billions of dollars need to be invested in biomedical treatments and preventi
"Many patients with long COVID are already receiving care but are not being recognized as having the condition. These patients are not absent from clinical care; they are absent from the diagnostic code that would identify them as long COVID patients" https://t.co/aLJCOPlQt1
Created 2026-06-05 · Updated 2026-06-05
Edit rule text
[robotics_or_speculative_sci_fi_labor_macro_without_healthcare_validation]
Learned 3 rejections Active
Exclude posts about robotics, humanoid workers, manufacturing automation, or labor market disruption from AI/robots unless they directly address healthcare labor challenges (clinical staffing, surgical automation, patient care delivery). Generic labor market or manufacturing robotics posts with loose healthcare framing are out of scope.
Posts about humanoid robots, labor automation, or speculative sci-fi applications without healthcare-specific validation or operational relevance
2 example posts
Humanoid robots, designed to mimic human movement and capabilities, have been making headlines in recent years, from their use as baggage handlers at Japanese airports to Tesla’s big bet on its Optimus humanoid. Market watchers have predicted that the machines will change the ht
Another humanoid worker has just joined the factory floor Two years later,PUDU Robotics has launched the new-generation PUDU D7, built specifically for real manufacturing scenarios. It can autonomously push carts, handle intra-line transportation, perform delicate operations, h
Created 2026-06-05 · Updated 2026-06-06
Edit rule text
[non_healthcare_domain_or_company_with_loose_healthcare_framing]
Learned 3 rejections Active
Exclude posts about non-healthcare domains (e.g., humanoid robots in manufacturing, financial services AI, semiconductor advances, law firm AI adoption) that mention healthcare only in passing or use healthcare as a loose analogy. The post must analyze a specific healthcare system problem or application, not just tangentially reference healthcare.
Posts about non-healthcare companies, technologies, or business models (robotics, finance, semiconductors) presented with tangential healthcare framing but lacking substantive healthcare application.
3 example posts
Humanoid robots, designed to mimic human movement and capabilities, have been making headlines in recent years, from their use as baggage handlers at Japanese airports to Tesla’s big bet on its Optimus humanoid. Market watchers have predicted that the machines will change the ht
the CEO of NVIDIA just said the computer is no longer being built for you, it is being built for agents Jensen Huang put it plainly, until now we were the users, we were the renters, every CPU on earth was designed around how a person works but agents do not work like us, they
In the AI gold rush, Jensen Huang is selling the picks and shovels. Nvidia's chips power AI companies around the world, helping the company become the first to surpass a $5 trillion market capitalization in late 2025. Read more about how he and others made the https://t.co/lHLE
Created 2026-06-05 · Updated 2026-06-05
Edit rule text
[business_model_or_market_speculation_unvalidated]
Learned 3 rejections Active
Exclude posts that speculate about pricing models, market competition, vendor lock-in, or business strategy (e.g., 'Salesforce made CRM headless so agents can operate it', 'large law firms building proprietary AI instead of buying') without healthcare-specific evidence, operational context, or validated healthcare systems impact.
Posts speculating about business models, competitive positioning, or market dynamics without substantive healthcare systems analysis or evidence.
3 example posts
AI made building cheap. Regulation made distribution expensive. Anyone can clone your demo now. Very few can bring proprietary data, licenses, compliance, and enough trust to sell into serious players. The moat didn’t disappear. It moved.
Most enterprise AI projects stall because no one has prepared the content, built the workflow, and kept it improving as models change. That role is essential. Box has launched Forward Deployed Engineers to fill it. Model-agnostic, content-first, and built for the way enterprise
Almost everyone is building agent harness systems the wrong way. The default move: pick LangChain or LangGraph or the OpenAI Agents SDK, accept the loop, the tools, the memory, the orchestration, the policy engine, the credential store, the budget tracker, all of it, as one deci
Created 2026-06-04 · Updated 2026-06-04
Edit rule text
[generalist_infrastructure_or_compute_hype]
Learned 3 rejections Active
Exclude posts about GPU hardware, inference costs, token economics, compute infrastructure, or generalist AI technical capabilities that mention healthcare only peripherally or lack specific healthcare application context. Posts focused on NVIDIA chips, RTX cards, inference pricing, or agent runtimes without healthcare-specific use cases should be excluded.
Posts about AI infrastructure, compute chips, or generalist technical capabilities with only loose or tangential healthcare framing.
3 example posts
Nous Research is working with NVIDIA to make Hermes Agent run smoothly on the new NVIDIA RTX Spark superchip. Hermes Agent is also integrating with the new OpenShell runtime, which connects Hermes to Microsoft’s security primitives https://t.co/cvvPc8bIKa
a $2,000 graphics card and a free download are turning into $2,000-a-month retainers, and the customers are the ones who legally can't use ChatGPT the hardware is an RTX 5090, 32GB of VRAM, enough to run a 30B model locally with nothing ever leaving the building. the software is
Free newsletter: The dawn of token-based billing has shown that generative AI doesn’t have a return on investment. It's too unpredictable, too unreliable, you can't easily measure the cost of tasks, and organizations are already pulling back. https://t.co/wmI82zWdcq
Created 2026-06-04 · Updated 2026-06-04
Edit rule text
[conference_announcement_or_poster_promotion_without_substantive_content]
Learned 3 rejections Active
Exclude posts that solely announce conference attendance, poster board numbers, presentation times, or conference registration unless the post includes substantive findings, data, or insights from the research being presented. Promotional event posts without content do not qualify.
Posts promoting conference talks, poster sessions, or event participation without substantive research or findings
1 example post
We're presenting at @ASCO 2026! Evaluating AI decision support in a rapidly evolving therapeutic landscape: EGFR-mutant metastatic NSCLC. 📍 Poster Board #548 🕘 9:00 AM CT If you're attending ASCO 2026, stop by our poster and connect with the team. We look forward to the https:
Created 2026-06-04 · Updated 2026-06-05
Edit rule text
[clinical_anecdote_or_personal_narrative_without_healthcare_systems_context]
Learned 3 rejections Active
Exclude posts that narrate a single clinical encounter, patient case, individual cost story, or personal health experience as the primary content. Include only if the post extracts a systemic healthcare insight (e.g., demonstrating a widespread problem in pricing transparency, provider networks, clinical protocols) supported by pattern evidence or data.
Posts sharing single clinical cases, personal healthcare anecdotes, or individual patient stories without broader healthcare system insight or validation.
3 example posts
I called a woman in Dayton to tell her she was about to overpay $4,500, for a horrible health system based procedure. She was scheduled for a knee replacement. Her PPO had her on the hook for a $4,500 deductible. The call was on behalf of her employer: “See one of these three
@WIRED As a clinical health psychologist who has written >20 papers on COVID, I would emphasize 4 facts: 1) Long COVID is not a psychological diagnosis nor manifestation of a psychological condition 2) Billions of dollars need to be invested in biomedical treatments and preventi
"Many patients with long COVID are already receiving care but are not being recognized as having the condition. These patients are not absent from clinical care; they are absent from the diagnostic code that would identify them as long COVID patients" https://t.co/aLJCOPlQt1
Created 2026-06-04 · Updated 2026-06-04
Edit rule text
[generalist_infrastructure_or_macro_hype_without_healthcare_lens]
Learned 3 rejections Active
Exclude posts about GPU pricing, token economics, hardware costs, compute scaling, labor disruption, or robotics applications where healthcare is mentioned only as a loose label or secondary framing. The post must analyze healthcare-specific constraints (regulatory, clinical validation, workflow integration) or the post is discussing non-healthcare domains (manufacturing robots, financial infrastructure, legal tech, etc.) with a healthcare tag.
Posts about AI infrastructure, compute economics, robotics, or labor market trends with loose or tangential healthcare framing that lacks healthcare-specific systems analysis.
3 example posts
a $2,000 graphics card and a free download are turning into $2,000-a-month retainers, and the customers are the ones who legally can't use ChatGPT the hardware is an RTX 5090, 32GB of VRAM, enough to run a 30B model locally with nothing ever leaving the building. the software is
Free newsletter: The dawn of token-based billing has shown that generative AI doesn’t have a return on investment. It's too unpredictable, too unreliable, you can't easily measure the cost of tasks, and organizations are already pulling back. https://t.co/wmI82zWdcq
Humanoid robots, designed to mimic human movement and capabilities, have been making headlines in recent years, from their use as baggage handlers at Japanese airports to Tesla’s big bet on its Optimus humanoid. Market watchers have predicted that the machines will change the ht
Created 2026-06-04 · Updated 2026-06-04
Edit rule text
[pharmaceutical_trial_data_announcement_without_healthcare_systems_analysis]
Learned 3 rejections Active
Exclude posts that simply announce drug trial results, FDA approvals, or clinical trial data without discussing healthcare delivery bottlenecks, implementation barriers, patient access, or how the treatment changes healthcare system operations. Trial results alone without systems context are insufficient.
Posts announcing pharmaceutical trial results, drug approvals, or clinical data without healthcare systems context, operational barriers, or implementation challenges
3 example posts
Scoop! Lila Sciences is in talks to raise ~$2 billion in new funding. The raise would value the “scientific superintelligence” lab at ~$8.5 billion before the new money. w/ @MichelleF_Davis, read more👇
#ASCO26 | Day 3 $LLY didn't spend billions on Kelonia for an 18-patient #myeloma dataset. They may have paid for a potential manufacturing disruption. KLN-1010 reported: • 18/18 MRD-negative at 1 month • No lymphodepletion • No ex vivo cell manufacturing • Single infusion
Published in @JCO_ASCO, during #ASCO26: Tumor-Agnostic Therapies: Translating Scientific Breakthroughs Into Global Implementation ASCO is where we celebrate the next breakthrough. This review asks why breakthroughs do not equal access globally. 🔗: https://t.co/PE5q27bIs1 htt
Created 2026-06-04 · Updated 2026-06-04
Edit rule text
[anecdotal_healthcare_cost_or_personal_narrative]
Learned 3 rejections Active
Exclude posts that are primarily personal anecdotes, single patient stories, or individual healthcare cost examples (one patient's bill, one call to a health system) without healthcare systems analysis, policy insight, or operational pattern documentation.
Posts sharing personal healthcare cost anecdotes or individual patient stories without systemic healthcare analysis
3 example posts
I called a woman in Dayton to tell her she was about to overpay $4,500, for a horrible health system based procedure. She was scheduled for a knee replacement. Her PPO had her on the hook for a $4,500 deductible. The call was on behalf of her employer: “See one of these three
"Many patients with long COVID are already receiving care but are not being recognized as having the condition. These patients are not absent from clinical care; they are absent from the diagnostic code that would identify them as long COVID patients" https://t.co/aLJCOPlQt1
Most chronic diseases, and many cancers, are still diagnosed far too late. The future of healthcare lies in detecting risk earlier, diagnosing smarter, and intervening sooner. From AI enabled diagnostics and genomics to Apollo’s DigiSmart Labs, technology is helping transform ht
Created 2026-06-04 · Updated 2026-06-04
Edit rule text
[broad_unvalidated_health_claim_or_market_debate_without_systems_analysis]
Learned 3 rejections Active
Exclude posts that make sweeping health claims, debate clinical definitions, or speculate about market competition/disruption without grounding in data, clinical evidence, or healthcare system workflows. Broad speculation about 'what will happen next' in healthcare without operational substance is not substance.
Posts making broad health claims, debating disease definitions, or speculating on market dynamics without rigorous evidence or healthcare systems analysis
3 example posts
@WIRED As a clinical health psychologist who has written >20 papers on COVID, I would emphasize 4 facts: 1) Long COVID is not a psychological diagnosis nor manifestation of a psychological condition 2) Billions of dollars need to be invested in biomedical treatments and preventi
I think the AI superapps will soon own 90% of the agentic layer more and more people won't use hermes/openclaw etc because Claude Cowork/Codex will offer 90% of the functionality with 5% of the friction
Anthropic just shipped opus 4.8 and called it a modest upgrade modest is the engineer who used to need babysitting now finishing your whole codebase migration in one session while you sleep so here is the part nobody connects the model is no longer the bottleneck. the platform
Created 2026-06-04 · Updated 2026-06-04
Edit rule text
[tangential_non_healthcare_business_or_robotics_hype]
Learned 3 rejections Active
Exclude posts about humanoid robots, manufacturing automation, labor market macro trends, or general business model debates unless they directly address a validated healthcare operational problem (e.g., clinical staffing, patient transport, surgical assistance). Posts about robots in factories or general workforce disruption are not healthcare tech content.
Posts about humanoid robots, factory automation, general labor market disruption, or non-healthcare business trends loosely framed as healthcare-adjacent
3 example posts
Humanoid robots, designed to mimic human movement and capabilities, have been making headlines in recent years, from their use as baggage handlers at Japanese airports to Tesla’s big bet on its Optimus humanoid. Market watchers have predicted that the machines will change the ht
Another humanoid worker has just joined the factory floor Two years later,PUDU Robotics has launched the new-generation PUDU D7, built specifically for real manufacturing scenarios. It can autonomously push carts, handle intra-line transportation, perform delicate operations, h
Anthropic just shipped opus 4.8 and called it a modest upgrade modest is the engineer who used to need babysitting now finishing your whole codebase migration in one session while you sleep so here is the part nobody connects the model is no longer the bottleneck. the platform
Created 2026-06-04 · Updated 2026-06-04
Edit rule text
[tangential_tech_or_non_healthcare_domain_with_loose_healthcare_framing]
Learned 3 rejections Active
Exclude posts about robotics applications (humanoid workers, manufacturing automation), cybersecurity vulnerability research, or general technology infrastructure that use healthcare as a tangential framing device or surface-level example without demonstrating substantive healthcare systems insight, operational impact, or healthcare-specific challenge addressed.
Posts about robotics, humanoid workers, cybersecurity, or non-healthcare technology topics loosely framed as healthcare-relevant without substantive healthcare systems analysis.
3 example posts
Humanoid robots, designed to mimic human movement and capabilities, have been making headlines in recent years, from their use as baggage handlers at Japanese airports to Tesla’s big bet on its Optimus humanoid. Market watchers have predicted that the machines will change the ht
The next frontier of AI-secured financial and technological critical infrastructure. @ICE_Markets and @NYSE are part of @AnthropicAI's cyber security initiative Project Glasswing, deploying Anthropic’s Claude Mythos Preview across ICE’s exchanges, clearing houses, mortgage
⚡️AI is turning cybersecurity from a human-limited profession into a machine-speed arms race. The surface read is “Anthropic’s model found vulnerabilities fast.” The deeper read is that the bottleneck in cyber is shifting. For years, vulnerability discovery was constrained
Created 2026-06-04 · Updated 2026-06-04
Edit rule text
[robotics_or_speculative_sci_fi_labor_without_healthcare_validation]
Learned 3 rejections Active
Exclude posts about humanoid robots, factory automation, or speculative future labor disruption (e.g., robots as baggage handlers, general manufacturing robots) that are not deployed in or validated for healthcare-specific use cases.
Posts about humanoid robots or speculative sci-fi labor applications without healthcare-specific deployment or evidence
2 example posts
Humanoid robots, designed to mimic human movement and capabilities, have been making headlines in recent years, from their use as baggage handlers at Japanese airports to Tesla’s big bet on its Optimus humanoid. Market watchers have predicted that the machines will change the ht
Another humanoid worker has just joined the factory floor Two years later,PUDU Robotics has launched the new-generation PUDU D7, built specifically for real manufacturing scenarios. It can autonomously push carts, handle intra-line transportation, perform delicate operations, h
Created 2026-06-04 · Updated 2026-06-04
Edit rule text
[healthcare_business_or_operational_anecdote_without_systems_insight]
Learned 3 rejections Active
Exclude posts that recount a single patient case, individual billing incident, or personal healthcare story (e.g., one mom's pricing call, one person's treatment outcome) without connecting it to healthcare system-level patterns, incentives, or structural problems.
Posts sharing individual healthcare cost stories or single-case observations without broader systems analysis
3 example posts
I called a woman in Dayton to tell her she was about to overpay $4,500, for a horrible health system based procedure. She was scheduled for a knee replacement. Her PPO had her on the hook for a $4,500 deductible. The call was on behalf of her employer: “See one of these three
@WIRED As a clinical health psychologist who has written >20 papers on COVID, I would emphasize 4 facts: 1) Long COVID is not a psychological diagnosis nor manifestation of a psychological condition 2) Billions of dollars need to be invested in biomedical treatments and preventi
"Many patients with long COVID are already receiving care but are not being recognized as having the condition. These patients are not absent from clinical care; they are absent from the diagnostic code that would identify them as long COVID patients" https://t.co/aLJCOPlQt1
Created 2026-06-04 · Updated 2026-06-04
Edit rule text
[anecdotal_single_case_healthcare_without_systems]
Learned 3 rejections Active
Exclude posts that narrate a single patient anecdote, individual case story, or personal healthcare scenario (e.g., 'a mom in Tulsa called', 'a surgeon I know operated') without connecting to broader healthcare system dysfunction, policy implications, or scalable solutions.
Posts describing isolated patient cases or individual healthcare scenarios without operational or systemic context.
3 example posts
I called a woman in Dayton to tell her she was about to overpay $4,500, for a horrible health system based procedure. She was scheduled for a knee replacement. Her PPO had her on the hook for a $4,500 deductible. The call was on behalf of her employer: “See one of these three
A company with 800 employees spends roughly $15,000 per employee per year on health insurance. That is $12 million. Wired into a single line item. Every year. The CFO can tell you the carrier. He cannot tell you the unit cost of a single claim. He cannot tell you what the net
A Mom in Tulsa called 3 health systems last week asking the price of her son's tonsillectomy. Health system A: "We cannot quote you a price." Health system B: "Pricing depends on your insurance." Health system C: "Our financial counselor will reach out after the procedure." N
Created 2026-06-04 · Updated 2026-06-04
Edit rule text
[non_healthcare_infrastructure_or_compute_hype]
Learned 3 rejections Active
Exclude posts that focus on CPU design, GPU optimization, Kubernetes configuration, cloud infrastructure efficiency, or general compute/inference cost trends — unless the post explicitly connects these improvements to solving a specific healthcare operational or clinical problem.
Posts about general-purpose AI infrastructure, compute, or hardware advances tangential to healthcare application
3 example posts
the CEO of NVIDIA just said the computer is no longer being built for you, it is being built for agents Jensen Huang put it plainly, until now we were the users, we were the renters, every CPU on earth was designed around how a person works but agents do not work like us, they
We are partnering with @Microsoft to enable secure, user-controlled AI on Windows. NVIDIA OpenShell runtime for agents will provide governance tools, policy enforcement, and smart local-to-cloud query routing. Learn more: https://t.co/zPGwz9xQSW https://t.co/mkOtFEOhAS
From unboxing to AI agent in minutes. Getting an agent running used to mean sourcing a model, configuring an inference backend, installing a runtime, and wiring it all together. The new NemoClaw install path on DGX Spark replaces that with a single command. DGX Spark also https
Created 2026-06-04 · Updated 2026-06-04
Edit rule text
[biotech_or_robotics_founder_enthusiasm_without_systems_validation]
Learned 3 rejections Active
Exclude posts that celebrate biotech founder announcements, robotics lab updates, or speculative sci-fi applications (e.g., humanoid factory workers, next-gen cell manufacturing) unless they include clinical validation data, healthcare workflow evidence, or operational deployment specifics.
Posts about biotech labs, humanoid robots, or speculative technology announcements without evidence of healthcare system integration or validation.
3 example posts
Another humanoid worker has just joined the factory floor Two years later,PUDU Robotics has launched the new-generation PUDU D7, built specifically for real manufacturing scenarios. It can autonomously push carts, handle intra-line transportation, perform delicate operations, h
I think the AI superapps will soon own 90% of the agentic layer more and more people won't use hermes/openclaw etc because Claude Cowork/Codex will offer 90% of the functionality with 5% of the friction
Anthropic just shipped opus 4.8 and called it a modest upgrade modest is the engineer who used to need babysitting now finishing your whole codebase migration in one session while you sleep so here is the part nobody connects the model is no longer the bottleneck. the platform
Created 2026-06-03 · Updated 2026-06-03
Edit rule text
[unvalidated_speculative_medical_claims_without_systems_context]
Learned 3 rejections Active
Exclude posts that speculate on unproven medical interventions, off-label peptide usage, market pricing of experimental drugs, or single trial results without healthcare systems context, operational impact, or evidence of adoption barriers being addressed (grey-market HGH, GLP-1 pricing speculation, unvalidated mRNA vaccines, speculative Lp(a) claims).
Posts about unvalidated medical claims, speculative drug/peptide market dynamics, or clinical trial data without healthcare systems analysis.
3 example posts
1/8 Spend time in Lp(a) forums and you’ll see a striking pattern: people willing to do almost anything to drive it down. That response is understandable given how strongly Lp(a) has been framed as a cardiovascular risk factor. /2
For years, grey-market HGH operated in a strange legal grey zone somewhere between “research compound,” “API” and underground enhancement culture. Last month’s FDA guidance suggests regulators may no longer see it that way. The recent decline in HGH quality, and now the growing
Think about this. The $50 per month out-of-pocket cost of GLP-1 drugs for Americans (with insurer paying the balance of $2000 pm) is more than double the entire cost for Indians (less than ₹2000pm for Semaglutide). Overall cost difference? 100 times. Not 100%.
Created 2026-06-03 · Updated 2026-06-03
Edit rule text
[cybersecurity_vulnerability_tangent_without_healthcare_application]
Learned 3 rejections Active
Exclude posts that report on cybersecurity vulnerabilities, AI-powered vulnerability discovery, or security incidents without demonstrating relevance to healthcare data security, clinical system integrity, or HIPAA/regulatory compliance. Cybersecurity hype without healthcare context is not healthcare tech content.
Posts about AI cybersecurity vulnerabilities, zero-day exploits, or bug-finding capabilities that do not connect to healthcare security or clinical safety
3 example posts
⚡️AI is turning cybersecurity from a human-limited profession into a machine-speed arms race. The surface read is “Anthropic’s model found vulnerabilities fast.” The deeper read is that the bottleneck in cyber is shifting. For years, vulnerability discovery was constrained
Anthropic built a model so good at hacking they refused to release it publicly. INSTEAD they handed it to Apple, Google, Microsoft, AWS, Cloudflare one month later: 10,000+ zero days found in the software that runs the internet...bugs hidden for 27 years in OpenBSD. exploit htt
🚨 Anthropic just dropped the first Project Glasswing update Claude Mythos found 10,000+ critical vulnerabilities in ONE month: > Cloudflare: 2,000 bugs, 400 high/critical severity > Mozilla: 271 vulnerabilities in Firefox 150 — 10x more vulnerabilities found in Firefox 1
Created 2026-06-03 · Updated 2026-06-03
Edit rule text
[ai_company_product_or_capability_hype]
Learned 3 rejections Active
Exclude posts that announce new AI models, tools, or services from companies like OpenAI, Anthropic, or other tech vendors where the healthcare angle is implied but not demonstrated. The post must show validated healthcare application, not just model capability.
Posts announcing AI company products, models, or capabilities without demonstrating real healthcare application or validation.
3 example posts
We’re taking steps to accelerate defensive progress in biology: - Launching Rosalind Biodefense to help trusted builders develop new biodefense and pandemic preparedness capabilities.
 - Expanding trusted access to GPT-Rosalind for select U.S. government and allied partners
Anthropic built a model so good at hacking they refused to release it publicly. INSTEAD they handed it to Apple, Google, Microsoft, AWS, Cloudflare one month later: 10,000+ zero days found in the software that runs the internet...bugs hidden for 27 years in OpenBSD. exploit htt
🚨 Anthropic just dropped the first Project Glasswing update Claude Mythos found 10,000+ critical vulnerabilities in ONE month: > Cloudflare: 2,000 bugs, 400 high/critical severity > Mozilla: 271 vulnerabilities in Firefox 150 — 10x more vulnerabilities found in Firefox 1
Created 2026-06-03 · Updated 2026-06-03
Edit rule text
[anecdotal_founder_or_company_enthusiasm]
Learned 3 rejections Active
Exclude posts that report founder quotes, internal company experiments, or startup announcements about AI adoption, product building, or technical milestones without third-party validation, customer evidence, or demonstrated healthcare system impact.
Posts featuring founder enthusiasm, lab updates, or company announcements about internal initiatives without independent validation or healthcare systems evidence.
3 example posts
Figuring out how to benchmark agents on realistic biology research has quickly become one of my favorite types of engineering work. You work with scientists to get to the core of some biological claim, precisely assembling raw data/prior literature/experimental context in a https
Today's OpenMed Agent session, driven by GPT-5.5 through the Codex SDK with my own ChatGPT subscription: → 10,036 records ingested → 16-step plan complete → 49 tool calls → 5 reviewer-gated PDFs → APPROVE on PA L38319 Public trace on @huggingface ↓ https://t.co/vlq6CoQ4ee
@levie Sat in a boardroom last month where the CEO demoed an AI workflow he built over the weekend. Took him 4 hours. His team spent the next 3 weeks figuring out how to make it work with their actual data, their actual compliance requirements, and 14 edge cases the demo never hi
Created 2026-06-03 · Updated 2026-06-03
Edit rule text
[generalist_macro_without_healthcare_systems_lens]
Learned 3 rejections Active
Exclude posts that discuss generalist macro trends (e.g., 'GPU useful lives extended', 'Kubernetes removed from production', 'enterprise AI adoption metrics') where healthcare is tangential or used only as a framing device, not as the primary subject of analysis.
Posts about broad economic, labor market, or infrastructure trends with only loose or incidental healthcare framing.
3 example posts
kirkland spending $500m on internal ai is the tell. $6.5b revenue, they could buy any legal ai vendor outright and not feel it, and they're building instead. the moat was always the proprietary corpus. decades of deal docs, redlines, partner judgment encoded as training data, an
They removed Kubernetes from production, and their AWS bill dropped 68%. Kubernetes was solving a problem they did not really have. small team, few services, low scaling needs. But they were managing: - Helm charts - YAML complexity - custom operators - platform mainten
The framing here is exactly backwards. This is the strongest AI bull signal anyone has published this year. Uber deployed Claude Code to 5,000 engineers in December. By March, 84% were classified as agentic coding users. By April, 95% used AI tools monthly. 70% of all committed
Created 2026-06-03 · Updated 2026-06-03
Edit rule text
[anecdotal_single_case_healthcare]
Learned 3 rejections Active
Exclude posts that describe a single patient case, surgeon anecdote, or individual healthcare interaction (e.g., 'A Mom in Tulsa called 3 health systems...', 'A surgeon I know recently operated...') and extrapolate it as a systemic healthcare trend without broader evidence or data.
Personal anecdotes or single clinical observations presented as broader healthcare insights without systemic evidence.
3 example posts
A company with 800 employees spends roughly $15,000 per employee per year on health insurance. That is $12 million. Wired into a single line item. Every year. The CFO can tell you the carrier. He cannot tell you the unit cost of a single claim. He cannot tell you what the net
A Mom in Tulsa called 3 health systems last week asking the price of her son's tonsillectomy. Health system A: "We cannot quote you a price." Health system B: "Pricing depends on your insurance." Health system C: "Our financial counselor will reach out after the procedure." N
A surgeon I know recently operated on a man's spine. Saved his mobility. Saved his career. Saved his marriage, probably. The surgeon's take-home on that case, after the facility fee, anesthesia, device costs, payer discounts, and overhead, was $1,840. The health system collec
Created 2026-06-03 · Updated 2026-06-03
Edit rule text
[pharma_trial_data_announcement_without_analysis]
Learned 3 rejections Active
Exclude posts that announce pharmaceutical trial results, drug efficacy metrics, or clinical endpoints (e.g., 'PFS favored X', 'MRD-negative', 'survival rates') without explaining how this data impacts healthcare delivery, operations, or system-level challenges. Raw trial data alone is not sufficient.
Posts reporting clinical trial results or drug efficacy data without healthcare systems context or operational implications.
3 example posts
Today at #ASCO26, more results about the newest clinical trial of daraxonrasib: In the Phase III RASolute-302 trial, a once-daily RAS(ON) inhibitor nearly doubled median overall survival (13.2 vs 6.7 months) and reduced the risk of death by ~60% versus chemotherapy in previously
#ASCO26 @DrSanjayPopat presents AcceleRET-Lung: first line pralsetinib vs chemo +/- IO in RET+ NSCLC with optional crossover. Study terminated by sponsor at 90% accrual. PFS favors pralsetinib 18.7m vs 9m (HR 0.59) with RR 66% vs 42%, DOR 20.6m vs 9.7m https://t.co/W73zzzT0wk
The FDA accepted our NDA for BBP-418 for LGMD2I/R9 with PDUFA target action date: November 27, 2026. First-ever potential therapy for a disease that has never had one. $BBIO https://t.co/NrwDadc6rj https://t.co/2ahPIBGzPg
Created 2026-06-03 · Updated 2026-06-03
Edit rule text
[market_speculation_without_systems_analysis]
Learned 3 rejections Active
Exclude posts that speculate on market pricing, competitive positioning, stock performance, or market size—unless they include healthcare operational or policy analysis that explains WHY the market dynamics matter to patient care or healthcare delivery.
Posts about market pricing, competitive dynamics, or GLP-1/peptide market positioning without healthcare systems analysis or clinical context.
3 example posts
$HIMS up to #4 in the Canadian App Store medical category. That is not a stock chart. That is real consumer demand showing up in public. GLP-1 drama, Novo pressure, pricing debate, and now app-store strength. Bears keep arguing the story is slowing down, but the product demand
For years, grey-market HGH operated in a strange legal grey zone somewhere between “research compound,” “API” and underground enhancement culture. Last month’s FDA guidance suggests regulators may no longer see it that way. The recent decline in HGH quality, and now the growing
Think about this. The $50 per month out-of-pocket cost of GLP-1 drugs for Americans (with insurer paying the balance of $2000 pm) is more than double the entire cost for Indians (less than ₹2000pm for Semaglutide). Overall cost difference? 100 times. Not 100%.
Created 2026-06-02 · Updated 2026-06-02
Edit rule text
[ai_infrastructure_hype_without_healthcare]
Learned 3 rejections Active
Exclude posts that focus on AI model releases (GPT, Claude, OpenAI updates), compute efficiency, inference optimization, or AI infrastructure—unless they include concrete validation, healthcare-specific application details, or clinical impact data.
Posts celebrating AI model capabilities, compute infrastructure, or technical breakthroughs with only loose or tangential healthcare framing.
3 example posts
Red Hat and @NVIDIA are integrating NVIDIA OpenShell into the full-stack @RedHat_AI platform. The work brings oversight and policy to the infrastructure level, while contributing to the open source OpenShell project to standardize how agents are governed on enterprise platforms.
We’re taking steps to accelerate defensive progress in biology: - Launching Rosalind Biodefense to help trusted builders develop new biodefense and pandemic preparedness capabilities.
 - Expanding trusted access to GPT-Rosalind for select U.S. government and allied partners
Anthropic built a model so good at hacking they refused to release it publicly. INSTEAD they handed it to Apple, Google, Microsoft, AWS, Cloudflare one month later: 10,000+ zero days found in the software that runs the internet...bugs hidden for 27 years in OpenBSD. exploit htt
Created 2026-06-02 · Updated 2026-06-02
Edit rule text
[non_healthcare_business_or_policy_loose_framing]
Learned 3 rejections Active
Exclude posts about non-healthcare industries or macro policy (e.g., Kubernetes cloud costs, law firm AI infrastructure, Salesforce headless CRM, corporate AI spend) that use healthcare framing only tangentially or as a surface-level analogy. The core topic must be healthcare-specific delivery, policy, or operations.
Posts about non-healthcare business/policy topics (infrastructure, AI company strategy, law firm AI adoption) tagged with healthcare relevance but lacking substantive healthcare connection.
3 example posts
The app era is ending. Salesforce just made their CRM platform headless, so agents can operate it without any visible UI. Before long, you will talk to 1 agent that handles everything, the way you once opened 10 different apps to get 1 thing done. For years, VeChain has been b
Ontology all the way down. 11 years. Heavy civil construction. People, equipment, materials, contracts. This is what real-world AI looks like: operational, measurable, and in the field. Palantir x Cavanagh through 2035. https://t.co/872Cs1Qbdu
Agentic orchestration layers allow AI agents, enterprise systems, and data connections to work together across functions. As cognitive bottlenecks shrink, that allows decision-making to speed up, coordination to improve and new operating models to form. https://t.co/0to4HC26cu
Created 2026-06-02 · Updated 2026-06-02
Edit rule text
[enterprise_ai_adoption_macro_without_healthcare_validation]
Learned 3 rejections Active
Exclude posts that describe enterprise AI adoption, agent deployment, or productivity metrics in law firms, consulting, tech companies, or other non-healthcare domains and then frame them with loose healthcare analogies or implications (e.g., 'K&E, largest law firm, commits $500M to build proprietary AI software'). Healthcare application must be direct and substantive, not speculative.
Posts about enterprise AI adoption, productivity gains, or agentic system deployment in non-healthcare sectors (law, finance, tech) applied loosely to healthcare.
3 example posts
Red Hat and @NVIDIA are integrating NVIDIA OpenShell into the full-stack @RedHat_AI platform. The work brings oversight and policy to the infrastructure level, while contributing to the open source OpenShell project to standardize how agents are governed on enterprise platforms.
wow K&E, largest law firm in the world ($10B billings). They’re committing $500M to build proprietary AI software, with 180 tech pros and 100 partners involved. I’m skeptical that “build” is right here, but wonder if Anthropic/OAI deployment cos are involved.
Ontology all the way down. 11 years. Heavy civil construction. People, equipment, materials, contracts. This is what real-world AI looks like: operational, measurable, and in the field. Palantir x Cavanagh through 2035. https://t.co/872Cs1Qbdu
Created 2026-06-02 · Updated 2026-06-02
Edit rule text
[founder_enthusiasm_or_market_speculation_without_validation]
Learned 3 rejections Active
Exclude posts that speculate on market trends, competitive positioning, founder enthusiasm, or business model debates (e.g., 'the app era is ending,' 'superapps will own 90%,' 'Kirkland is the tell') without clinical evidence, healthcare operational data, or empirical validation of the claim.
Posts expressing optimism about startups, business models, or market dynamics without evidence, validation, or healthcare systems analysis.
3 example posts
Most chronic diseases, and many cancers, are still diagnosed far too late. The future of healthcare lies in detecting risk earlier, diagnosing smarter, and intervening sooner. From AI enabled diagnostics and genomics to Apollo’s DigiSmart Labs, technology is helping transform ht
I think the AI superapps will soon own 90% of the agentic layer more and more people won't use hermes/openclaw etc because Claude Cowork/Codex will offer 90% of the functionality with 5% of the friction
The app era is ending. Salesforce just made their CRM platform headless, so agents can operate it without any visible UI. Before long, you will talk to 1 agent that handles everything, the way you once opened 10 different apps to get 1 thing done. For years, VeChain has been b
Created 2026-06-02 · Updated 2026-06-02
Edit rule text
[ai_agent_technical_tangent_without_healthcare_application]
Learned 3 rejections Active
Exclude posts that discuss AI agent technical challenges (sandboxing, memory, tool coordination, hallucination) or demonstrate agent capabilities without showing how these solve specific healthcare operational problems.
Technical deep-dives into AI agent architecture, deployment patterns, or debugging that lack healthcare-specific operational context
3 example posts
Almost everyone is building agent harness systems the wrong way. The default move: pick LangChain or LangGraph or the OpenAI Agents SDK, accept the loop, the tools, the memory, the orchestration, the policy engine, the credential store, the budget tracker, all of it, as one deci
Interoperability is where AI hype goes to die. If your “agent” can’t work across data, logic, actions, security of legacy systems, it’s not transforming the enterprise. It’s just another app asking for an export. Winners don’t rip and replace. They operate.
I think the AI superapps will soon own 90% of the agentic layer more and more people won't use hermes/openclaw etc because Claude Cowork/Codex will offer 90% of the functionality with 5% of the friction
Created 2026-06-02 · Updated 2026-06-02
Edit rule text
[broad_market_dynamics_or_business_model_speculation_without_systems_analysis]
Learned 3 rejections Active
Exclude posts that debate healthcare business models, market sizing, pricing strategies, or competitive positioning (e.g., 'Will peptides disrupt GLP-1s?', 'What is Hims' real margin?') without analyzing how these dynamics affect patient access, clinical outcomes, or healthcare system capacity.
Posts debating healthcare business models, market opportunities, or competitive dynamics without grounding in healthcare delivery realities or operational constraints.
3 example posts
Analogizing AI / Mat-Mul to Energy... Remember between 1900 and 1970 kWh/capita grew 34x ... and cost per kWH droped .... 34x. It might be a speed run, but game is game. https://t.co/uxnlnoaAYx
Gavin Baker (@GavinSBaker) says the disaggregation of inference can extend GPU useful lives from 3-4 years to 10-15. That may single-handedly save private credit and reduce the financing rates for GPUs, which will drive demand and help finance the build-out. "The disaggregation
$HIMS up to #4 in the Canadian App Store medical category. That is not a stock chart. That is real consumer demand showing up in public. GLP-1 drama, Novo pressure, pricing debate, and now app-store strength. Bears keep arguing the story is slowing down, but the product demand
Created 2026-06-02 · Updated 2026-06-02
Edit rule text
[anecdotal_single_observation_no_systems]
Learned 3 rejections Active
Exclude posts grounded in a single personal story, one clinician's observation, or one patient case without connecting to broader healthcare delivery, regulatory, payment, or operational system dynamics. Personal narratives must illuminate systems-level patterns.
Posts based on personal anecdote, single case example, or isolated observation without healthcare systems analysis
3 example posts
medicine is constrained far more by operational architecture than by fundamental science I just came head to head with this actual reality and my world has changed no. I cannot bandage away these inefficiencies with a better and better drug, for all cases. if you hate AI
A Mom in Tulsa called 3 health systems last week asking the price of her son's tonsillectomy. Health system A: "We cannot quote you a price." Health system B: "Pricing depends on your insurance." Health system C: "Our financial counselor will reach out after the procedure." N
@levie Sat in a boardroom last month where the CEO demoed an AI workflow he built over the weekend. Took him 4 hours. His team spent the next 3 weeks figuring out how to make it work with their actual data, their actual compliance requirements, and 14 edge cases the demo never hi
Created 2026-06-02 · Updated 2026-06-02
Edit rule text
[market_pricing_speculation_only]
Learned 3 rejections Active
Exclude posts that center on pharmaceutical pricing dynamics, market competition, or company pricing strategy without connecting to healthcare access, payment systems, or patient outcomes. Posts must address healthcare system implications, not market positioning alone.
Posts focused exclusively on GLP-1/peptide pricing, competitive positioning, or market share without healthcare access analysis
3 example posts
$HIMS up to #4 in the Canadian App Store medical category. That is not a stock chart. That is real consumer demand showing up in public. GLP-1 drama, Novo pressure, pricing debate, and now app-store strength. Bears keep arguing the story is slowing down, but the product demand
Think about this. The $50 per month out-of-pocket cost of GLP-1 drugs for Americans (with insurer paying the balance of $2000 pm) is more than double the entire cost for Indians (less than ₹2000pm for Semaglutide). Overall cost difference? 100 times. Not 100%.
🚨 ARE PEPTIDES THE NEXT "GLP-1 MOMENT" FOR $HIMS? @eli_dorf: "I think there's something really interesting that happened with GLP-1s, which is that when someone's taking Ozempic or any GLP-1, they LOVE to talk about it." "They'll talk about it to their friends. They'll talk htt
Created 2026-06-02 · Updated 2026-06-02
Edit rule text
[press_release_or_corporate_announcement]
Learned 3 rejections Active
Exclude posts that primarily reproduce company press releases, FDA announcements, or regulatory filings without adding critical analysis or healthcare systems context. Posts must include the writer's own reasoning or implications.
Posts that read as unanalyzed corporate PR or regulatory filings without editorial insight
3 example posts
Agentic orchestration layers allow AI agents, enterprise systems, and data connections to work together across functions. As cognitive bottlenecks shrink, that allows decision-making to speed up, coordination to improve and new operating models to form. https://t.co/0to4HC26cu
We’re taking steps to accelerate defensive progress in biology: - Launching Rosalind Biodefense to help trusted builders develop new biodefense and pandemic preparedness capabilities.
 - Expanding trusted access to GPT-Rosalind for select U.S. government and allied partners
The FDA accepted our NDA for BBP-418 for LGMD2I/R9 with PDUFA target action date: November 27, 2026. First-ever potential therapy for a disease that has never had one. $BBIO https://t.co/NrwDadc6rj https://t.co/2ahPIBGzPg
Created 2026-06-02 · Updated 2026-06-02
Edit rule text
[clinical_trial_data_or_pharma_announcement_without_systems_analysis]
Learned 3 rejections Active
Exclude posts that announce clinical trial results, FDA approvals, drug efficacy data, or pharmaceutical milestones (e.g., 'trial showed X benefit' or 'FDA accepted NDA') without connecting those results to healthcare delivery systems, operational bottlenecks, reimbursement models, or competitive dynamics that would matter to healthcare operators or innovators.
Posts that report raw clinical trial results or drug approval data without any analysis of healthcare system implications, operational impacts, or business model consequences.
3 example posts
#ASCO26 @DrSanjayPopat presents AcceleRET-Lung: first line pralsetinib vs chemo +/- IO in RET+ NSCLC with optional crossover. Study terminated by sponsor at 90% accrual. PFS favors pralsetinib 18.7m vs 9m (HR 0.59) with RR 66% vs 42%, DOR 20.6m vs 9.7m https://t.co/W73zzzT0wk
The FDA accepted our NDA for BBP-418 for LGMD2I/R9 with PDUFA target action date: November 27, 2026. First-ever potential therapy for a disease that has never had one. $BBIO https://t.co/NrwDadc6rj https://t.co/2ahPIBGzPg
Presented at #EASCongress2026: Vafai and colleagues report reductions in PCSK9 and LDL cholesterol levels and no dose-limiting toxic effects in persons with hypercholesterolemia treated with VERVE-102, a base editor targeting PCSK9. Full study results: https://t.co/iAmoPdipx9 ht
Created 2026-06-01 · Updated 2026-06-01
Edit rule text
[clinical_trial_or_drug_data_announcement_without_healthcare_systems_analysis]
Learned 3 rejections Active
Exclude posts that simply report pharmaceutical trial data, FDA approvals, trial endpoints, or drug efficacy metrics without analyzing how these results change healthcare operations, provider incentives, patient access, or system-level delivery. Raw trial results alone are insufficient.
Posts announcing clinical trial results or drug approvals without connecting to healthcare delivery, operational, or policy implications.
3 example posts
#ASCO26 @DrSanjayPopat presents AcceleRET-Lung: first line pralsetinib vs chemo +/- IO in RET+ NSCLC with optional crossover. Study terminated by sponsor at 90% accrual. PFS favors pralsetinib 18.7m vs 9m (HR 0.59) with RR 66% vs 42%, DOR 20.6m vs 9.7m https://t.co/W73zzzT0wk
The FDA accepted our NDA for BBP-418 for LGMD2I/R9 with PDUFA target action date: November 27, 2026. First-ever potential therapy for a disease that has never had one. $BBIO https://t.co/NrwDadc6rj https://t.co/2ahPIBGzPg
Presented at #EASCongress2026: Vafai and colleagues report reductions in PCSK9 and LDL cholesterol levels and no dose-limiting toxic effects in persons with hypercholesterolemia treated with VERVE-102, a base editor targeting PCSK9. Full study results: https://t.co/iAmoPdipx9 ht
Created 2026-06-01 · Updated 2026-06-01
Edit rule text
[tangential_non_healthcare_business_technology]
Learned 3 rejections Active
Exclude posts about general business models, enterprise software architecture, or technology infrastructure (Kubernetes, cloud costs, headless APIs, Salesforce CRM) that use healthcare only as a loose label or passing example rather than analyzing actual healthcare delivery, operations, or patient outcomes.
Posts about non-healthcare business or technology trends loosely framed as healthcare-relevant without substantive healthcare application.
3 example posts
kirkland spending $500m on internal ai is the tell. $6.5b revenue, they could buy any legal ai vendor outright and not feel it, and they're building instead. the moat was always the proprietary corpus. decades of deal docs, redlines, partner judgment encoded as training data, an
The app era is ending. Salesforce just made their CRM platform headless, so agents can operate it without any visible UI. Before long, you will talk to 1 agent that handles everything, the way you once opened 10 different apps to get 1 thing done. For years, VeChain has been b
wow K&E, largest law firm in the world ($10B billings). They’re committing $500M to build proprietary AI software, with 180 tech pros and 100 partners involved. I’m skeptical that “build” is right here, but wonder if Anthropic/OAI deployment cos are involved.
Created 2026-06-01 · Updated 2026-06-01
Edit rule text
[enterprise_ai_adoption_metrics_no_healthcare_systems]
Learned 3 rejections Active
Exclude posts that describe AI tool adoption in enterprises, code generation velocity, engineer productivity, or workflow redesign without explaining how this applies to healthcare operations, patient outcomes, or clinical workflows. Posts about Uber, Kirkland law firm, or general enterprise deployment belong here.
Posts about enterprise AI adoption, tool deployment, and productivity metrics without healthcare-specific operational context or system-level insight
3 example posts
kirkland spending $500m on internal ai is the tell. $6.5b revenue, they could buy any legal ai vendor outright and not feel it, and they're building instead. the moat was always the proprietary corpus. decades of deal docs, redlines, partner judgment encoded as training data, an
wow K&E, largest law firm in the world ($10B billings). They’re committing $500M to build proprietary AI software, with 180 tech pros and 100 partners involved. I’m skeptical that “build” is right here, but wonder if Anthropic/OAI deployment cos are involved.
a databricks tech lead just spent 26 minutes on the part of multi-agent nobody wants to say out loud: your agents don't break because the model is dumb. they break because nothing is coordinating them. one agent is a feature. fifty is a distributed systems problem. https://t.c
Created 2026-06-01 · Updated 2026-06-01
Edit rule text
[conference_announcement_or_poster_promotion_only]
Learned 3 rejections Active
Exclude posts that exist only to announce a conference talk, poster session, or booth location without previewing findings, arguing a position, or offering analysis. These are logistical announcements, not editorial content.
Posts that are purely conference promotional content (booth numbers, session times, poster announcements) without substantive preview or insights
3 example posts
We're presenting at @ASCO 2026! Evaluating AI decision support in a rapidly evolving therapeutic landscape: EGFR-mutant metastatic NSCLC. 📍 Poster Board #548 🕘 9:00 AM CT If you're attending ASCO 2026, stop by our poster and connect with the team. We look forward to the https:
NCCN Guidelines are now built into OpenEvidence. Ask a clinical question, get the synthesis with the algorithm and the references in seconds. Bring us a case at ASCO this weekend, booth 18140. https://t.co/N3nMayuGu9
Startups that seek to block AI exam fraud, help companies use AI to make consistent decisions, make financial AI transactions safer, and use AI to organize medical device regulatory information took top honors at @Cornell_Tech’s annual Startup Awards on May 14. Nearly 600 https:
Created 2026-06-01 · Updated 2026-06-01
Edit rule text
[business_model_or_market_dynamics_without_systems_context]
Learned 3 rejections Active
Exclude posts that focus on business strategy, pricing dynamics, market positioning, or competitive analysis (e.g., Hims market share, GLP-1 pricing wars, law firm AI investments, Palantir civil construction) without explaining how these dynamics affect healthcare access, quality, equity, or clinical outcomes.
Posts analyzing healthcare business models, pricing, or market competition without addressing operational or systemic healthcare implications
3 example posts
Ontology all the way down. 11 years. Heavy civil construction. People, equipment, materials, contracts. This is what real-world AI looks like: operational, measurable, and in the field. Palantir x Cavanagh through 2035. https://t.co/872Cs1Qbdu
A company with 800 employees spends roughly $15,000 per employee per year on health insurance. That is $12 million. Wired into a single line item. Every year. The CFO can tell you the carrier. He cannot tell you the unit cost of a single claim. He cannot tell you what the net
A Mom in Tulsa called 3 health systems last week asking the price of her son's tonsillectomy. Health system A: "We cannot quote you a price." Health system B: "Pricing depends on your insurance." Health system C: "Our financial counselor will reach out after the procedure." N
Created 2026-05-31 · Updated 2026-05-31
Edit rule text
[ai_company_product_launch_or_metrics_without_healthcare_validation]
Learned 3 rejections Active
Exclude posts announcing AI company product launches, partnerships, funding rounds, or user adoption metrics unless the post demonstrates actual healthcare application, clinical validation, or measurable healthcare outcome impact. Product announcements and business metrics alone do not qualify.
Posts about AI company product launches, partnerships, or business metrics that lack evidence of real healthcare adoption or clinical validation
3 example posts
We're presenting at @ASCO 2026! Evaluating AI decision support in a rapidly evolving therapeutic landscape: EGFR-mutant metastatic NSCLC. 📍 Poster Board #548 🕘 9:00 AM CT If you're attending ASCO 2026, stop by our poster and connect with the team. We look forward to the https:
We’re taking steps to accelerate defensive progress in biology: - Launching Rosalind Biodefense to help trusted builders develop new biodefense and pandemic preparedness capabilities.
 - Expanding trusted access to GPT-Rosalind for select U.S. government and allied partners
Today's OpenMed Agent session, driven by GPT-5.5 through the Codex SDK with my own ChatGPT subscription: → 10,036 records ingested → 16-step plan complete → 49 tool calls → 5 reviewer-gated PDFs → APPROVE on PA L38319 Public trace on @huggingface ↓ https://t.co/vlq6CoQ4ee
Created 2026-05-31 · Updated 2026-05-31
Edit rule text
[cybersecurity_or_ai_safety_vulnerability_tangent_without_healthcare_application]
Learned 3 rejections Active
Exclude posts that discuss AI safety, cybersecurity vulnerabilities (zero-day exploits, model hacking), or AI security incident reporting without demonstrating healthcare-specific risk, HIPAA implications, clinical validation requirements, or healthcare deployment context. General AI safety and cybersecurity content tangentially labeled healthcare does not qualify.
Posts about AI safety, cybersecurity vulnerabilities, or AI security incidents that are not grounded in healthcare-specific risk, validation, or operational context.
3 example posts
OpenClaw is powerful enough that we needed Hermes Agent watching it. A lot of people are missing this about agents. The more useful they get, the more dangerous it is to let one system plan the work, run the work, review the work, and decide what happens next. It looks clean h
Anthropic built a model so good at hacking they refused to release it publicly. INSTEAD they handed it to Apple, Google, Microsoft, AWS, Cloudflare one month later: 10,000+ zero days found in the software that runs the internet...bugs hidden for 27 years in OpenBSD. exploit htt
🚨 Anthropic just dropped the first Project Glasswing update Claude Mythos found 10,000+ critical vulnerabilities in ONE month: > Cloudflare: 2,000 bugs, 400 high/critical severity > Mozilla: 271 vulnerabilities in Firefox 150 — 10x more vulnerabilities found in Firefox 1
Created 2026-05-31 · Updated 2026-05-31
Edit rule text
[clinical_trial_data_without_healthcare_systems_context]
Learned 3 rejections Active
Exclude posts that announce or describe clinical trial results, drug approvals, or pharmaceutical data (efficacy, safety, PDUFA dates) without situating the finding within healthcare system dynamics, operational challenges, reimbursement, or clinical adoption barriers. Bare announcement of trial outcomes or regulatory milestones does not qualify.
Posts reporting clinical trial results or drug approval data without analysis of how the therapy integrates into healthcare delivery, economics, or clinical practice.
3 example posts
The full NHS GALLERI randomized trial data has been released at #ASCO26 by @GrailBio. 142,000 adults provided 3 blood samples over 2 years, for prevalent (baseline) & incident cancers. The trial did not meet its primary endpoint. Lots more data below: https://t.co/bgWQqTwiA
MCED testing in pop screening NHS-Galleri-RCT of 142,250 pts ➡️blood up to 3 visits ➡️aim to reduce stage III/IV Med FU 17mths ❎primary endpt not met-but ⬇️in stg4 ➡️52% PPV; 99.5% specificity ✅4 fold⬆️ in screen detected ca Huge effort from NHS #ASCO26 @ASCO @OncoAlert https://
#ASCO26 @DrSanjayPopat presents AcceleRET-Lung: first line pralsetinib vs chemo +/- IO in RET+ NSCLC with optional crossover. Study terminated by sponsor at 90% accrual. PFS favors pralsetinib 18.7m vs 9m (HR 0.59) with RR 66% vs 42%, DOR 20.6m vs 9.7m https://t.co/W73zzzT0wk
Created 2026-05-31 · Updated 2026-05-31
Edit rule text
[ai_safety_cybersecurity_vulnerability_tangent_without_healthcare_application]
Learned 3 rejections Active
Exclude posts that report AI security incidents, vulnerability disclosures, or security research (Claude Mythos vulnerabilities, zero-day findings, model jailbreaks) unless the post explicitly connects the vulnerability to healthcare delivery risk, patient safety, or healthcare data protection.
Posts about AI model security vulnerabilities, zero-day exploits, or security research findings in non-healthcare contexts or without demonstrating healthcare-specific risk.
3 example posts
Anthropic built a model so good at hacking they refused to release it publicly. INSTEAD they handed it to Apple, Google, Microsoft, AWS, Cloudflare one month later: 10,000+ zero days found in the software that runs the internet...bugs hidden for 27 years in OpenBSD. exploit htt
🚨 Anthropic just dropped the first Project Glasswing update Claude Mythos found 10,000+ critical vulnerabilities in ONE month: > Cloudflare: 2,000 bugs, 400 high/critical severity > Mozilla: 271 vulnerabilities in Firefox 150 — 10x more vulnerabilities found in Firefox 1
#AI agents need more than a sandbox—they need predictable execution. Learn how OpenShell provides a standard interface to intercept, inspect, and approve LLM-generated bash commands before they run: https://t.co/JKmLMgl1A9 https://t.co/xWi8oBNZ7T
Created 2026-05-31 · Updated 2026-05-31
Edit rule text
[tangential_non_healthcare_business_and_tech]
Learned 3 rejections Active
Exclude posts that discuss AI agents, infrastructure, compute costs, or enterprise adoption using healthcare as a loose reference point, when the post's core argument applies equally to finance, legal, or any other domain. The post must center on healthcare-specific constraints or systems, not generalist tech trends.
Posts about general tech infrastructure, compute economics, or business strategy that use healthcare merely as an example but lack healthcare-specific insight
3 example posts
kirkland spending $500m on internal ai is the tell. $6.5b revenue, they could buy any legal ai vendor outright and not feel it, and they're building instead. the moat was always the proprietary corpus. decades of deal docs, redlines, partner judgment encoded as training data, an
wow K&E, largest law firm in the world ($10B billings). They’re committing $500M to build proprietary AI software, with 180 tech pros and 100 partners involved. I’m skeptical that “build” is right here, but wonder if Anthropic/OAI deployment cos are involved.
Ontology all the way down. 11 years. Heavy civil construction. People, equipment, materials, contracts. This is what real-world AI looks like: operational, measurable, and in the field. Palantir x Cavanagh through 2035. https://t.co/872Cs1Qbdu
Created 2026-05-31 · Updated 2026-05-31
Edit rule text
[healthcare_policy_or_regulatory_posturing_without_analysis]
Learned 3 rejections Active
Exclude posts that express outrage at healthcare policy, regulations, or government action (rural hospital funding, Medicaid dynamics, CMS guidance, physician reimbursement) without providing concrete analysis of economic consequences, incentive structures, or operational impact.
Healthcare policy complaints, regulatory outrage, or political healthcare framing without substantive systems analysis.
3 example posts
A company with 800 employees spends roughly $15,000 per employee per year on health insurance. That is $12 million. Wired into a single line item. Every year. The CFO can tell you the carrier. He cannot tell you the unit cost of a single claim. He cannot tell you what the net
A Mom in Tulsa called 3 health systems last week asking the price of her son's tonsillectomy. Health system A: "We cannot quote you a price." Health system B: "Pricing depends on your insurance." Health system C: "Our financial counselor will reach out after the procedure." N
A surgeon I know recently operated on a man's spine. Saved his mobility. Saved his career. Saved his marriage, probably. The surgeon's take-home on that case, after the facility fee, anesthesia, device costs, payer discounts, and overhead, was $1,840. The health system collec
Created 2026-05-31 · Updated 2026-05-31
Edit rule text
[enterprise_adoption_macro_no_healthcare_specificity]
Learned 3 rejections Active
Exclude posts that discuss enterprise AI adoption, agent deployment, cost savings, or workforce metrics (Databricks, Palantir, Uber engineers, etc.) unless the post explicitly analyzes impact on healthcare operations, clinical workflows, or health system economics.
Posts about enterprise AI adoption, workforce productivity, or deployment patterns with no clear healthcare systems context.
3 example posts
kirkland spending $500m on internal ai is the tell. $6.5b revenue, they could buy any legal ai vendor outright and not feel it, and they're building instead. the moat was always the proprietary corpus. decades of deal docs, redlines, partner judgment encoded as training data, an
wow K&E, largest law firm in the world ($10B billings). They’re committing $500M to build proprietary AI software, with 180 tech pros and 100 partners involved. I’m skeptical that “build” is right here, but wonder if Anthropic/OAI deployment cos are involved.
Ontology all the way down. 11 years. Heavy civil construction. People, equipment, materials, contracts. This is what real-world AI looks like: operational, measurable, and in the field. Palantir x Cavanagh through 2035. https://t.co/872Cs1Qbdu
Created 2026-05-31 · Updated 2026-05-31
Edit rule text
[broad_health_claim_or_market_speculation_without_evidence_or_systems_context]
Learned 3 rejections Active
Exclude posts that make broad claims about health trends (e.g., 'the app era is ending,' 'peptides are the next GLP-1 moment,' 'AI will disaggregate healthcare') without citing evidence, data, or explaining systemic healthcare implications. Speculation and analogies are not substantive healthcare analysis.
Posts making broad health claims, market speculation, or business model debates without clinical evidence or healthcare systems analysis.
3 example posts
The app era is ending. Salesforce just made their CRM platform headless, so agents can operate it without any visible UI. Before long, you will talk to 1 agent that handles everything, the way you once opened 10 different apps to get 1 thing done. For years, VeChain has been b
Figuring out how to benchmark agents on realistic biology research has quickly become one of my favorite types of engineering work. You work with scientists to get to the core of some biological claim, precisely assembling raw data/prior literature/experimental context in a https
NCCN Guidelines are now built into OpenEvidence. Ask a clinical question, get the synthesis with the algorithm and the references in seconds. Bring us a case at ASCO this weekend, booth 18140. https://t.co/N3nMayuGu9
Created 2026-05-31 · Updated 2026-05-31
Edit rule text
[tangential_ai_safety_cybersecurity_vulnerability_without_healthcare_application]
Learned 3 rejections Active
Exclude posts about AI model security vulnerabilities (zero-day exploits, jailbreaks, adversarial attacks) or safety incidents unless they explicitly demonstrate healthcare-specific attack vectors, healthcare system exposure, or proposed mitigation specific to healthcare operations.
Posts about AI safety, security vulnerabilities, or misuse incidents that have no demonstrated or proposed healthcare-specific application or context.
3 example posts
Anthropic built a model so good at hacking they refused to release it publicly. INSTEAD they handed it to Apple, Google, Microsoft, AWS, Cloudflare one month later: 10,000+ zero days found in the software that runs the internet...bugs hidden for 27 years in OpenBSD. exploit htt
🚨 Anthropic just dropped the first Project Glasswing update Claude Mythos found 10,000+ critical vulnerabilities in ONE month: > Cloudflare: 2,000 bugs, 400 high/critical severity > Mozilla: 271 vulnerabilities in Firefox 150 — 10x more vulnerabilities found in Firefox 1
#AI agents need more than a sandbox—they need predictable execution. Learn how OpenShell provides a standard interface to intercept, inspect, and approve LLM-generated bash commands before they run: https://t.co/JKmLMgl1A9 https://t.co/xWi8oBNZ7T
Created 2026-05-31 · Updated 2026-05-31
Edit rule text
[clinical_observation_or_trial_data_without_healthcare_systems_context]
Learned 3 rejections Active
Exclude posts that announce clinical trial results, report drug efficacy, describe single patient cases or clinical observations (Mayo Clinic rare disease case, pancreatic lesion detection, CAR-T manufacturing), or reference FDA approvals — unless the post explicitly analyzes how the finding changes healthcare operations, reimbursement, or system-level clinical workflows.
Posts reporting clinical trial results, drug efficacy data, or single-case clinical observations without analyzing healthcare delivery, operational, or systems-level implications.
3 example posts
Traditional ex vivo CAR-T is fundamentally broken. Between the $500K+ manufacturing cost, the 4-week vein-to-vein delay, and the risk of patient T-cell exhaustion during expansion, the blockbuster autologous market is a logistical nightmare. The industry is pivoting to "In Vivo
We’re taking steps to accelerate defensive progress in biology: - Launching Rosalind Biodefense to help trusted builders develop new biodefense and pandemic preparedness capabilities.
 - Expanding trusted access to GPT-Rosalind for select U.S. government and allied partners
Drug-resistant infections are a major public health threat around the world, responsible for more than a million deaths each year. Scientists are constantly trying to find and develop new antibiotics. Now, researchers say artificial intelligence is helping speed their search. ht
Created 2026-05-31 · Updated 2026-05-31
Edit rule text
[pharma_trial_data_or_market_announcement_without_healthcare_systems_context]
Learned 3 rejections Active
Exclude posts that are primarily announcements of clinical trial data, FDA approvals, drug launches, or market pricing without analyzing how the result changes healthcare system behavior, provider incentives, patient access barriers, or clinical decision-making. Trial results and announcements alone are not sufficient.
Posts announcing drug trial results, FDA approvals, or pharmaceutical market moves without analyzing healthcare system impact or clinical implementation barriers.
3 example posts
Drug-resistant infections are a major public health threat around the world, responsible for more than a million deaths each year. Scientists are constantly trying to find and develop new antibiotics. Now, researchers say artificial intelligence is helping speed their search. ht
The FDA accepted our NDA for BBP-418 for LGMD2I/R9 with PDUFA target action date: November 27, 2026. First-ever potential therapy for a disease that has never had one. $BBIO https://t.co/NrwDadc6rj https://t.co/2ahPIBGzPg
Presented at #EASCongress2026: Vafai and colleagues report reductions in PCSK9 and LDL cholesterol levels and no dose-limiting toxic effects in persons with hypercholesterolemia treated with VERVE-102, a base editor targeting PCSK9. Full study results: https://t.co/iAmoPdipx9 ht
Created 2026-05-31 · Updated 2026-05-31
Edit rule text
[regulatory_fraud_scandal_reporting]
Learned 3 rejections Active
Exclude posts that report on healthcare fraud, regulatory enforcement actions, compliance violations, or financial scandals (e.g., false billing, unethical practices) without analyzing the systemic incentives or structural healthcare problems that enabled the abuse.
Posts reporting healthcare fraud, regulatory violations, or scandals without systemic analysis of root causes or structural solutions.
3 example posts
A non-profit health system can refer a patient to its own MRI, its own lab, its own surgery center, and bill all three. An independent does that once and it's a federal felony. Same referral. Same patient. One of you goes to prison. It's called Stark Law. Read who's exempt.
🚨 JUST IN: The Attorney General of Florida just SUED to hold OpenAI CEO Sam Altman LIABLE for fueling violence and physical harm AG Uthmeier read off the fact a young person committed SUIC*DE based on a ChatGPT conversation "23-year-old Zane Shamblin repeatedly told ChatGPT he
USDA's Chief Information Officer Sam Berry says AI could help the USDA crack down on billions of dollars in SNAP fraud each year. "SNAP is a $100B-per-year taxpayer-funded program. That's an area where we really want to have all angles of the data available so that we can deploy
Created 2026-05-31 · Updated 2026-06-12
Edit rule text
[anecdotal_single_clinical_case]
Learned 3 rejections Active
Exclude posts that describe a single patient case, one surgeon's experience, or individual clinical outcome (e.g., 'A mom in Tulsa called 3 health systems', 'A surgeon I know', 'A man's spine surgery') without framing how this reflects or changes healthcare systems, reimbursement, or operational patterns.
Single clinical anecdotes or case reports without systemic healthcare context or validation
3 example posts
A company with 800 employees spends roughly $15,000 per employee per year on health insurance. That is $12 million. Wired into a single line item. Every year. The CFO can tell you the carrier. He cannot tell you the unit cost of a single claim. He cannot tell you what the net
medicine is constrained far more by operational architecture than by fundamental science I just came head to head with this actual reality and my world has changed no. I cannot bandage away these inefficiencies with a better and better drug, for all cases. if you hate AI
A Mom in Tulsa called 3 health systems last week asking the price of her son's tonsillectomy. Health system A: "We cannot quote you a price." Health system B: "Pricing depends on your insurance." Health system C: "Our financial counselor will reach out after the procedure." N
Created 2026-05-31 · Updated 2026-05-31
Edit rule text
[tangential_enterprise_ai_productivity]
Learned 3 rejections Active
Exclude posts that discuss AI agent capabilities, code generation, or engineering workflows being deployed in generic enterprise settings (Databricks engineers, Uber engineers, Microsoft/Apple adoption metrics) — even if tangentially tied to healthcare. The focus must be on healthcare operational systems, not generalist AI engineering productivity.
Posts about enterprise AI adoption, productivity metrics, or internal engineering workflows without healthcare system context
3 example posts
Ontology all the way down. 11 years. Heavy civil construction. People, equipment, materials, contracts. This is what real-world AI looks like: operational, measurable, and in the field. Palantir x Cavanagh through 2035. https://t.co/872Cs1Qbdu
a databricks tech lead just spent 26 minutes on the part of multi-agent nobody wants to say out loud: your agents don't break because the model is dumb. they break because nothing is coordinating them. one agent is a feature. fifty is a distributed systems problem. https://t.c
@levie Sat in a boardroom last month where the CEO demoed an AI workflow he built over the weekend. Took him 4 hours. His team spent the next 3 weeks figuring out how to make it work with their actual data, their actual compliance requirements, and 14 edge cases the demo never hi
Created 2026-05-31 · Updated 2026-05-31
Edit rule text
[political_outrage_or_fraud_scandal_no_systems]
Learned 3 rejections Active
Exclude posts that report fraud cases, regulatory violations, or political controversies (e.g., UnitedHealth CEO assassination, hospice fraud) without analyzing the underlying healthcare system failures, business model dysfunction, or operational problems that enabled the issue.
Healthcare fraud, regulatory scandal, or political outrage reporting without systemic healthcare analysis
3 example posts
Why is healthcare expensive? The grifters gonna grift. ASCO is asking for oncology practices to be admitted into the same 340B machinery that helped health systems acquire physician practices for two decades. Read the tell. The proposed Indigent Care Ratio lets community
If you want to inject $1B into healthcare for the truly vulnerable, it costs the state $500M. If you want to inject $1B for able-bodied adults, it costs the state $100M. The 90% federal match drastically lowers the "price" of expanding government healthcare. Incentives dictate
A senator complaining about healthcare costs should have to answer one question first: Why does your state make new healthcare competitors ask permission before opening? If yes, start there. Speeches do not lower prices. Supply does.
Created 2026-05-30 · Updated 2026-05-30
Edit rule text
[broad_market_dynamics_without_systems_analysis]
Learned 3 rejections Active
Exclude posts that focus on market pricing, cost comparisons, competitive positioning, or financial dynamics (GLP-1 pricing, peptide market share, insurance coverage) without analyzing how these affect healthcare delivery, access, or operational models.
Market pricing, competitive dynamics, or pharmaceutical/peptide market speculation without healthcare systems lens
3 example posts
A company with 800 employees spends roughly $15,000 per employee per year on health insurance. That is $12 million. Wired into a single line item. Every year. The CFO can tell you the carrier. He cannot tell you the unit cost of a single claim. He cannot tell you what the net
A Mom in Tulsa called 3 health systems last week asking the price of her son's tonsillectomy. Health system A: "We cannot quote you a price." Health system B: "Pricing depends on your insurance." Health system C: "Our financial counselor will reach out after the procedure." N
$HIMS up to #4 in the Canadian App Store medical category. That is not a stock chart. That is real consumer demand showing up in public. GLP-1 drama, Novo pressure, pricing debate, and now app-store strength. Bears keep arguing the story is slowing down, but the product demand
Created 2026-05-30 · Updated 2026-05-30
Edit rule text
[anecdotal_single_case_without_systems]
Learned 3 rejections Active
Exclude posts that use a single patient story, one surgeon's case, or individual anecdotal observation as the primary evidence without connecting it to systemic healthcare dynamics, operational patterns, or validated data.
Posts relying on individual anecdotes or single cases without broader healthcare system context or validation
3 example posts
A company with 800 employees spends roughly $15,000 per employee per year on health insurance. That is $12 million. Wired into a single line item. Every year. The CFO can tell you the carrier. He cannot tell you the unit cost of a single claim. He cannot tell you what the net
medicine is constrained far more by operational architecture than by fundamental science I just came head to head with this actual reality and my world has changed no. I cannot bandage away these inefficiencies with a better and better drug, for all cases. if you hate AI
A Mom in Tulsa called 3 health systems last week asking the price of her son's tonsillectomy. Health system A: "We cannot quote you a price." Health system B: "Pricing depends on your insurance." Health system C: "Our financial counselor will reach out after the procedure." N
Created 2026-05-30 · Updated 2026-05-30
Edit rule text
[enterprise_ai_adoption_macro_without_healthcare_focus]
Learned 3 rejections Active
Exclude posts that report enterprise AI adoption metrics (e.g., 'X% of engineers using Claude Code'), productivity percentages, or labor market disruption patterns without connecting them to healthcare operations, clinical workflows, provider economics, or patient outcomes.
Posts about enterprise AI adoption, productivity gains, or workforce metrics without healthcare-specific context or systems analysis
3 example posts
The framing here is exactly backwards. This is the strongest AI bull signal anyone has published this year. Uber deployed Claude Code to 5,000 engineers in December. By March, 84% were classified as agentic coding users. By April, 95% used AI tools monthly. 70% of all committed
The more interesting part is that Microsoft’s own engineers liked Claude code best…and they’re cutting it anyway Token pricing is making enterprise actually look at what these models cost to run and that could be a problem for the labs when the subsidy ends. Lots of cheaper
Voice interaction, generating actions,and next, generating tasks...? The humanoid assistant is closer. For those who are bedridden due to illness or have limited mobility, it will serve as an excellent aid in daily living.
Created 2026-05-30 · Updated 2026-05-30
Edit rule text
[biotech_founder_enthusiasm_and_lab_updates_without_validation]
Learned 3 rejections Active
Exclude posts from biotech/AI startup founders or researchers announcing model capabilities, benchmark results, or lab findings that lack peer review, clinical validation, or independent replication. Posts like 'we just dropped X results' or founder enthusiasm about technical achievements without evidence should be rejected.
Posts from biotech founders, startup teams, and researchers announcing lab results, model benchmarks, or technical achievements without peer review or clinical validation.
3 example posts
Figuring out how to benchmark agents on realistic biology research has quickly become one of my favorite types of engineering work. You work with scientists to get to the core of some biological claim, precisely assembling raw data/prior literature/experimental context in a https
Claude Code team just dropped a workshop on how to ship a production-ready agent from scratch. 27-minutes. Free. Live coding by Claude dev. Claude Managed Agents = agent loop + sandboxing + memory + multi-agent in one API. Worth more than any $500 vibe-coding course. https://t
This is what the peptide space needs more of: clinicians with actual patient volume, public reasoning, and accountability. The next layer is transparency around sourcing, testing, and outcomes. Peptides are not a vibes category anymore. The signal is getting cleaner.
Created 2026-05-30 · Updated 2026-05-30
Edit rule text
[agentic_ai_technical_capability_tangent]
Learned 3 rejections Active
Exclude posts that explain how agentic AI works technically (autonomous planning, code execution loops, error correction) or compare agent architectures. The post must demonstrate a validated healthcare use case, clinical workflow integration, or patient/provider outcome—not just explain agent mechanics.
Posts about technical capabilities of agentic AI systems (planning, code generation, iterative execution) without healthcare validation or application.
3 example posts
Agentic orchestration layers allow AI agents, enterprise systems, and data connections to work together across functions. As cognitive bottlenecks shrink, that allows decision-making to speed up, coordination to improve and new operating models to form. https://t.co/0to4HC26cu
a databricks tech lead just spent 26 minutes on the part of multi-agent nobody wants to say out loud: your agents don't break because the model is dumb. they break because nothing is coordinating them. one agent is a feature. fifty is a distributed systems problem. https://t.c
OpenClaw is powerful enough that we needed Hermes Agent watching it. A lot of people are missing this about agents. The more useful they get, the more dangerous it is to let one system plan the work, run the work, review the work, and decide what happens next. It looks clean h
Created 2026-05-30 · Updated 2026-05-30
Edit rule text
[ai_company_product_metrics]
Learned 3 rejections Active
Exclude posts that announce AI company products, funding, partnerships, or capability metrics (e.g., 'Claude deployed to 5,000 engineers', 'Rosalind Biodefense launched') unless the post explicitly analyzes how these tools solve specific healthcare delivery or operational challenges.
Posts reporting AI company product launches, business metrics, or capability announcements without healthcare application context
3 example posts
We’re taking steps to accelerate defensive progress in biology: - Launching Rosalind Biodefense to help trusted builders develop new biodefense and pandemic preparedness capabilities.
 - Expanding trusted access to GPT-Rosalind for select U.S. government and allied partners
Anthropic built a model so good at hacking they refused to release it publicly. INSTEAD they handed it to Apple, Google, Microsoft, AWS, Cloudflare one month later: 10,000+ zero days found in the software that runs the internet...bugs hidden for 27 years in OpenBSD. exploit htt
🚨 Anthropic just dropped the first Project Glasswing update Claude Mythos found 10,000+ critical vulnerabilities in ONE month: > Cloudflare: 2,000 bugs, 400 high/critical severity > Mozilla: 271 vulnerabilities in Firefox 150 — 10x more vulnerabilities found in Firefox 1
Created 2026-05-30 · Updated 2026-05-30
Edit rule text
[clinical_observation_no_systems_context]
Learned 3 rejections Active
Exclude posts that report clinical trial outcomes, drug efficacy data, or rare disease discoveries without connecting findings to healthcare delivery models, access barriers, regulatory pathways, or implementation challenges in actual healthcare systems.
Posts describing clinical research findings, drug trial results, or medical observations without healthcare systems insight or operational implications
3 example posts
Drug-resistant infections are a major public health threat around the world, responsible for more than a million deaths each year. Scientists are constantly trying to find and develop new antibiotics. Now, researchers say artificial intelligence is helping speed their search. ht
Figuring out how to benchmark agents on realistic biology research has quickly become one of my favorite types of engineering work. You work with scientists to get to the core of some biological claim, precisely assembling raw data/prior literature/experimental context in a https
Presented at #EASCongress2026: Vafai and colleagues report reductions in PCSK9 and LDL cholesterol levels and no dose-limiting toxic effects in persons with hypercholesterolemia treated with VERVE-102, a base editor targeting PCSK9. Full study results: https://t.co/iAmoPdipx9 ht
Created 2026-05-30 · Updated 2026-05-30
Edit rule text
[healthcare_fraud_scandal_reporting_no_analysis]
Learned 3 rejections Active
Exclude posts that report healthcare fraud cases, regulatory enforcement actions, or corporate misconduct incidents without analyzing root causes in healthcare payment systems, reimbursement incentives, or structural policy failures. Crime reporting alone is insufficient.
Posts reporting healthcare fraud or regulatory scandals without systems-level analysis or policy implications
3 example posts
The FDA accepted our NDA for BBP-418 for LGMD2I/R9 with PDUFA target action date: November 27, 2026. First-ever potential therapy for a disease that has never had one. $BBIO https://t.co/NrwDadc6rj https://t.co/2ahPIBGzPg
Why is healthcare expensive? The grifters gonna grift. ASCO is asking for oncology practices to be admitted into the same 340B machinery that helped health systems acquire physician practices for two decades. Read the tell. The proposed Indigent Care Ratio lets community
In what is considered the largest #Botox fraud scheme in the United States, a jury in Los Angeles convicted a California doctor on Tuesday in a $45 million scheme to defraud #Medicare by submitting claims for Botox injections that were never provided and medically unnecessary, ht
Created 2026-05-30 · Updated 2026-05-30
Edit rule text
[enterprise_ai_adoption_macro_labor]
Learned 3 rejections Active
Exclude posts that present general enterprise AI adoption metrics, workforce productivity claims, or labor market disruption forecasts that apply equally to finance, manufacturing, or other sectors — unless the post specifically analyzes healthcare delivery models, clinical workflow redesign, or healthcare labor market dynamics.
Posts discussing enterprise AI adoption rates, labor productivity gains, or workforce disruption trends without healthcare-specific systems analysis
3 example posts
a databricks tech lead just spent 26 minutes on the part of multi-agent nobody wants to say out loud: your agents don't break because the model is dumb. they break because nothing is coordinating them. one agent is a feature. fifty is a distributed systems problem. https://t.c
@levie Sat in a boardroom last month where the CEO demoed an AI workflow he built over the weekend. Took him 4 hours. His team spent the next 3 weeks figuring out how to make it work with their actual data, their actual compliance requirements, and 14 edge cases the demo never hi
What "agentic" actually means, stripped of marketing: The model plans, writes code, runs it, reads the error, fixes it, iterates — autonomously, dozens of times, until the goal is met. You specify in English. It builds. That's the loop. This is the capability @bcherny and team
Created 2026-05-30 · Updated 2026-05-30
Edit rule text
[tangential_ai_infrastructure_compute_hype]
Learned 3 rejections Active
Exclude posts that discuss AI infrastructure economics (token pricing, inference optimization, Kubernetes cost reduction, compute efficiency) or technical AI architecture unless the post explicitly traces the healthcare delivery or clinical workflow impact of that infrastructure decision.
Posts about AI infrastructure, compute costs, GPU economics, or technical capability tangents with loose healthcare framing.
3 example posts
They removed Kubernetes from production, and their AWS bill dropped 68%. Kubernetes was solving a problem they did not really have. small team, few services, low scaling needs. But they were managing: - Helm charts - YAML complexity - custom operators - platform mainten
The framing here is exactly backwards. This is the strongest AI bull signal anyone has published this year. Uber deployed Claude Code to 5,000 engineers in December. By March, 84% were classified as agentic coding users. By April, 95% used AI tools monthly. 70% of all committed
Analogizing AI / Mat-Mul to Energy... Remember between 1900 and 1970 kWh/capita grew 34x ... and cost per kWH droped .... 34x. It might be a speed run, but game is game. https://t.co/uxnlnoaAYx
Created 2026-05-29 · Updated 2026-05-29
Edit rule text
[clinical_observation_or_pharma_data_without_healthcare_systems_context]
Learned 3 rejections Active
Exclude posts that report clinical trial outcomes, drug approval announcements, or preclinical research findings (gene therapy, drug repurposing, imaging AI) unless the post analyzes systemic healthcare implications such as patient access, regulatory path, reimbursement, or operational deployment challenges.
Posts reporting clinical trial results, drug efficacy data, or preclinical findings without analysis of operational, regulatory, or healthcare system implications
3 example posts
Drug-resistant infections are a major public health threat around the world, responsible for more than a million deaths each year. Scientists are constantly trying to find and develop new antibiotics. Now, researchers say artificial intelligence is helping speed their search. ht
The FDA accepted our NDA for BBP-418 for LGMD2I/R9 with PDUFA target action date: November 27, 2026. First-ever potential therapy for a disease that has never had one. $BBIO https://t.co/NrwDadc6rj https://t.co/2ahPIBGzPg
Figuring out how to benchmark agents on realistic biology research has quickly become one of my favorite types of engineering work. You work with scientists to get to the core of some biological claim, precisely assembling raw data/prior literature/experimental context in a https
Created 2026-05-29 · Updated 2026-05-29
Edit rule text
[tangential_ai_safety_cybersecurity_vulnerability]
Learned 3 rejections Active
Exclude posts about AI model vulnerabilities, zero-day exploits, agent misuse risks, or safety incidents (e.g., Claude Mythos finding vulnerabilities in Cloudflare, OpenClaw requiring agent oversight) unless directly tied to healthcare system security or clinical workflow risk.
Posts about AI safety, security vulnerabilities, or misuse incidents unrelated to healthcare delivery or operations
3 example posts
Anthropic built a model so good at hacking they refused to release it publicly. INSTEAD they handed it to Apple, Google, Microsoft, AWS, Cloudflare one month later: 10,000+ zero days found in the software that runs the internet...bugs hidden for 27 years in OpenBSD. exploit htt
🚨 Anthropic just dropped the first Project Glasswing update Claude Mythos found 10,000+ critical vulnerabilities in ONE month: > Cloudflare: 2,000 bugs, 400 high/critical severity > Mozilla: 271 vulnerabilities in Firefox 150 — 10x more vulnerabilities found in Firefox 1
#AI agents need more than a sandbox—they need predictable execution. Learn how OpenShell provides a standard interface to intercept, inspect, and approve LLM-generated bash commands before they run: https://t.co/JKmLMgl1A9 https://t.co/xWi8oBNZ7T
Created 2026-05-29 · Updated 2026-05-29
Edit rule text
[tangential_non_healthcare_tech_or_infrastructure_hype]
Learned 3 rejections Active
Exclude posts about infrastructure hype (Kubernetes removal dropping AWS bills, Databricks multi-agent systems, distributed workflows, voice interaction for robots) where healthcare is mentioned tangentially or speculatively but the post is fundamentally about non-healthcare tech optimization or capability.
Posts about general tech infrastructure, compute costs, or engineering practices applied outside healthcare contexts or with only loose healthcare framing
3 example posts
a databricks tech lead just spent 26 minutes on the part of multi-agent nobody wants to say out loud: your agents don't break because the model is dumb. they break because nothing is coordinating them. one agent is a feature. fifty is a distributed systems problem. https://t.c
OpenClaw is powerful enough that we needed Hermes Agent watching it. A lot of people are missing this about agents. The more useful they get, the more dangerous it is to let one system plan the work, run the work, review the work, and decide what happens next. It looks clean h
They removed Kubernetes from production, and their AWS bill dropped 68%. Kubernetes was solving a problem they did not really have. small team, few services, low scaling needs. But they were managing: - Helm charts - YAML complexity - custom operators - platform mainten
Created 2026-05-29 · Updated 2026-05-29
Edit rule text
[pharmaceutical_market_dynamics_without_healthcare_systems_analysis]
Learned 3 rejections Active
Exclude posts that treat pharmaceutical products as pure market commodities (e.g., 'Novo keeps charging brand premium', 'Ozempic vs generic semaglutide perception', 'HIMS launching in Canada at $149/month'). Include only if post analyzes how market dynamics affect healthcare delivery, provider incentives, or patient access barriers.
Posts about drug pricing, market competition, patent expirations, or GLP-1/peptide market positioning without healthcare system or access analysis
3 example posts
$HIMS up to #4 in the Canadian App Store medical category. That is not a stock chart. That is real consumer demand showing up in public. GLP-1 drama, Novo pressure, pricing debate, and now app-store strength. Bears keep arguing the story is slowing down, but the product demand
Think about this. The $50 per month out-of-pocket cost of GLP-1 drugs for Americans (with insurer paying the balance of $2000 pm) is more than double the entire cost for Indians (less than ₹2000pm for Semaglutide). Overall cost difference? 100 times. Not 100%.
🚨 ARE PEPTIDES THE NEXT "GLP-1 MOMENT" FOR $HIMS? @eli_dorf: "I think there's something really interesting that happened with GLP-1s, which is that when someone's taking Ozempic or any GLP-1, they LOVE to talk about it." "They'll talk about it to their friends. They'll talk htt
Created 2026-05-29 · Updated 2026-05-29
Edit rule text
[tangential_ai_safety_cybersecurity_not_healthcare]
Learned 3 rejections Active
Exclude posts that center on AI safety, cybersecurity vulnerabilities, or model jailbreaks (Claude Mythos vulnerabilities, OpenClaw trust problems) where healthcare is the framing device but the core analysis is about generic AI/security risk, not healthcare-specific threat models or HIPAA compliance challenges.
Posts about AI safety vulnerabilities, cybersecurity incidents, or model security breakthroughs framed through healthcare but focused on generic security challenges.
3 example posts
OpenClaw is powerful enough that we needed Hermes Agent watching it. A lot of people are missing this about agents. The more useful they get, the more dangerous it is to let one system plan the work, run the work, review the work, and decide what happens next. It looks clean h
AI agents are increasingly deployed as persistent operational systems, but do they remain reliable over time? Unfortunately no, our new work shows agents can quietly fail after deployment, despite passing day-1 evaluation. We call this "agent aging", akin to human aging. https:/
Anthropic built a model so good at hacking they refused to release it publicly. INSTEAD they handed it to Apple, Google, Microsoft, AWS, Cloudflare one month later: 10,000+ zero days found in the software that runs the internet...bugs hidden for 27 years in OpenBSD. exploit htt
Created 2026-05-29 · Updated 2026-05-29
Edit rule text
[ai_infrastructure_hype_tangential]
Learned 3 rejections Active
Exclude posts that discuss AI infrastructure (Kubernetes, compute efficiency, GPU utilization, agent architecture patterns) with only tangential or speculative healthcare framing. The post must demonstrate actual healthcare systems application, not technical capability hype.
Posts about AI infrastructure, compute, or technical capabilities that lack healthcare application or validation
3 example posts
a databricks tech lead just spent 26 minutes on the part of multi-agent nobody wants to say out loud: your agents don't break because the model is dumb. they break because nothing is coordinating them. one agent is a feature. fifty is a distributed systems problem. https://t.c
Figuring out how to benchmark agents on realistic biology research has quickly become one of my favorite types of engineering work. You work with scientists to get to the core of some biological claim, precisely assembling raw data/prior literature/experimental context in a https
They removed Kubernetes from production, and their AWS bill dropped 68%. Kubernetes was solving a problem they did not really have. small team, few services, low scaling needs. But they were managing: - Helm charts - YAML complexity - custom operators - platform mainten
Created 2026-05-29 · Updated 2026-05-29
Edit rule text
[market_pricing_dynamics_without_healthcare_systems]
Learned 3 rejections Active
Exclude posts that focus primarily on pricing tiers, competitive positioning, market dynamics, or cost comparisons for drugs/peptides (e.g., 'GLP-1 costs $2000/month in US vs ₹2000 in India', '$HIMS generic semaglutide at C$149/month') unless the post explicitly addresses how pricing affects healthcare delivery, access equity, or operational incentives.
Posts analyzing drug pricing, market competition, or GLP-1/peptide pricing wars without connecting to healthcare access, delivery, or operational implications.
3 example posts
Think about this. The $50 per month out-of-pocket cost of GLP-1 drugs for Americans (with insurer paying the balance of $2000 pm) is more than double the entire cost for Indians (less than ₹2000pm for Semaglutide). Overall cost difference? 100 times. Not 100%.
🚨 ARE PEPTIDES THE NEXT "GLP-1 MOMENT" FOR $HIMS? @eli_dorf: "I think there's something really interesting that happened with GLP-1s, which is that when someone's taking Ozempic or any GLP-1, they LOVE to talk about it." "They'll talk about it to their friends. They'll talk htt
^Humira’s patent expired a while back for some context there. The brand premium is the moat now. Ozempic is not generic semaglutide in patient and prescriber perception. Novo keeps charging brand pricing on every loyal patient without a PMPRB ceiling.
Created 2026-05-29 · Updated 2026-05-29
Edit rule text
[pharmaceutical_market_dynamics_only]
Learned 3 rejections Active
Exclude posts that focus primarily on GLP-1/peptide/drug market pricing, competitive positioning, brand dynamics, or market adoption metrics without analyzing healthcare system impact, access equity, or operational consequences.
Posts about GLP-1, peptide, or drug market pricing, competition, and commercial dynamics without healthcare systems analysis.
3 example posts
$HIMS up to #4 in the Canadian App Store medical category. That is not a stock chart. That is real consumer demand showing up in public. GLP-1 drama, Novo pressure, pricing debate, and now app-store strength. Bears keep arguing the story is slowing down, but the product demand
For years, grey-market HGH operated in a strange legal grey zone somewhere between “research compound,” “API” and underground enhancement culture. Last month’s FDA guidance suggests regulators may no longer see it that way. The recent decline in HGH quality, and now the growing
Think about this. The $50 per month out-of-pocket cost of GLP-1 drugs for Americans (with insurer paying the balance of $2000 pm) is more than double the entire cost for Indians (less than ₹2000pm for Semaglutide). Overall cost difference? 100 times. Not 100%.
Created 2026-05-29 · Updated 2026-05-29
Edit rule text
[market_pricing_dynamics_without_systems_context]
Learned 3 rejections Active
Exclude posts that focus solely on drug pricing trends, medication cost comparisons, international price arbitrage, stock valuations, or market competition metrics. Unless the post demonstrates how pricing affects provider behavior, patient access systems, or healthcare operations, it belongs in this category.
Posts about pharmaceutical pricing, GLP-1 market dynamics, healthcare stock performance, or competitive pricing wars without analyzing underlying healthcare delivery or operational implications.
3 example posts
AI agents are increasingly deployed as persistent operational systems, but do they remain reliable over time? Unfortunately no, our new work shows agents can quietly fail after deployment, despite passing day-1 evaluation. We call this "agent aging", akin to human aging. https:/
Analogizing AI / Mat-Mul to Energy... Remember between 1900 and 1970 kWh/capita grew 34x ... and cost per kWH droped .... 34x. It might be a speed run, but game is game. https://t.co/uxnlnoaAYx
Gavin Baker (@GavinSBaker) says the disaggregation of inference can extend GPU useful lives from 3-4 years to 10-15. That may single-handedly save private credit and reduce the financing rates for GPUs, which will drive demand and help finance the build-out. "The disaggregation
Created 2026-05-29 · Updated 2026-05-29
Edit rule text
[anecdotal_single_case_or_founder_enthusiasm_without_validation]
Learned 3 rejections Active
Exclude posts that present single clinical cases, anecdotal evidence, or founder/company enthusiasm (e.g., 'our model did X,' personal experience stories, lab update posts) without peer review, comparative evidence, or analysis of how the claim generalizes to healthcare systems.
Posts featuring single clinical anecdotes, personal use cases, or unsubstantiated founder enthusiasm for biotech/AI products without evidence or systemic context.
3 example posts
AI agents are increasingly deployed as persistent operational systems, but do they remain reliable over time? Unfortunately no, our new work shows agents can quietly fail after deployment, despite passing day-1 evaluation. We call this "agent aging", akin to human aging. https:/
A surgeon I know recently operated on a man's spine. Saved his mobility. Saved his career. Saved his marriage, probably. The surgeon's take-home on that case, after the facility fee, anesthesia, device costs, payer discounts, and overhead, was $1,840. The health system collec
NCCN Guidelines are now built into OpenEvidence. Ask a clinical question, get the synthesis with the algorithm and the references in seconds. Bring us a case at ASCO this weekend, booth 18140. https://t.co/N3nMayuGu9
Created 2026-05-29 · Updated 2026-05-29
Edit rule text
[generalist_macro_or_infrastructure_hype_without_healthcare_application]
Learned 3 rejections Active
Exclude posts that discuss generalist AI trends (e.g., inference cost curves, GPU utilization, 'the free lunch is over'), macro economics, or infrastructure hype and merely frame them as relevant to healthcare without demonstrating concrete healthcare system application or impact.
Posts about broad AI infrastructure, compute trends, or economic theories applied loosely to healthcare without specific healthcare system validation.
3 example posts
They removed Kubernetes from production, and their AWS bill dropped 68%. Kubernetes was solving a problem they did not really have. small team, few services, low scaling needs. But they were managing: - Helm charts - YAML complexity - custom operators - platform mainten
Analogizing AI / Mat-Mul to Energy... Remember between 1900 and 1970 kWh/capita grew 34x ... and cost per kWH droped .... 34x. It might be a speed run, but game is game. https://t.co/uxnlnoaAYx
Gavin Baker (@GavinSBaker) says the disaggregation of inference can extend GPU useful lives from 3-4 years to 10-15. That may single-handedly save private credit and reduce the financing rates for GPUs, which will drive demand and help finance the build-out. "The disaggregation
Created 2026-05-29 · Updated 2026-05-29
Edit rule text
[clinical_research_observation_without_systems]
Learned 3 rejections Active
Exclude posts that report clinical trial data, FDA approvals, or research findings (e.g., 'drug X reduced Y,' 'AI found Z in imaging') without analyzing how the result impacts healthcare system operations, access, cost, or provider decision-making.
Posts about clinical trial results, drug approvals, or research findings presented as isolated observations without healthcare delivery or policy context.
3 example posts
Drug-resistant infections are a major public health threat around the world, responsible for more than a million deaths each year. Scientists are constantly trying to find and develop new antibiotics. Now, researchers say artificial intelligence is helping speed their search. ht
Presented at #EASCongress2026: Vafai and colleagues report reductions in PCSK9 and LDL cholesterol levels and no dose-limiting toxic effects in persons with hypercholesterolemia treated with VERVE-102, a base editor targeting PCSK9. Full study results: https://t.co/iAmoPdipx9 ht
An AI foundation whole-body 3D model that assesses perturbations (such as obesity) across multiple systems (such as immune, neural) at the cell level. This is MouseMapper. Imagine HumanMapper someday @Nature @erturklab https://t.co/jsg3VMzOqg https://t.co/tktuINis2W
Created 2026-05-29 · Updated 2026-05-29
Edit rule text
[ai_infrastructure_or_capability_hype_without_healthcare_application]
Learned 3 rejections Active
Exclude posts that discuss AI infrastructure advances (GPU efficiency, token pricing, inference optimization), AI model capabilities (code generation, reasoning), or AI company product launches/partnerships without demonstrating actual healthcare system impact or clinical validation.
Posts about AI model capabilities, infrastructure improvements, or company announcements that lack concrete healthcare application or validation
3 example posts
a databricks tech lead just spent 26 minutes on the part of multi-agent nobody wants to say out loud: your agents don't break because the model is dumb. they break because nothing is coordinating them. one agent is a feature. fifty is a distributed systems problem. https://t.c
OpenClaw is powerful enough that we needed Hermes Agent watching it. A lot of people are missing this about agents. The more useful they get, the more dangerous it is to let one system plan the work, run the work, review the work, and decide what happens next. It looks clean h
Figuring out how to benchmark agents on realistic biology research has quickly become one of my favorite types of engineering work. You work with scientists to get to the core of some biological claim, precisely assembling raw data/prior literature/experimental context in a https
Created 2026-05-28 · Updated 2026-05-28
Edit rule text
[political_outrage_without_systems_analysis]
Learned 3 rejections Active
Exclude posts that frame healthcare problems as moral outrages against companies or politicians—e.g., 'UnitedHealth is corrupt,' 'the killer was right'—without presenting evidence of operational or systemic failure that healthcare builders can address.
Posts expressing moral outrage at healthcare villains, fraud, or policy without substantive systems analysis
3 example posts
Why is healthcare expensive? The grifters gonna grift. ASCO is asking for oncology practices to be admitted into the same 340B machinery that helped health systems acquire physician practices for two decades. Read the tell. The proposed Indigent Care Ratio lets community
A senator complaining about healthcare costs should have to answer one question first: Why does your state make new healthcare competitors ask permission before opening? If yes, start there. Speeches do not lower prices. Supply does.
In what is considered the largest #Botox fraud scheme in the United States, a jury in Los Angeles convicted a California doctor on Tuesday in a $45 million scheme to defraud #Medicare by submitting claims for Botox injections that were never provided and medically unnecessary, ht
Created 2026-05-28 · Updated 2026-05-28
Edit rule text
[clinical_observation_without_systems_analysis]
Learned 3 rejections Active
Exclude posts that report clinical trial data, drug efficacy results, or individual case studies without analyzing healthcare system bottlenecks, delivery models, cost structures, or operational changes required for implementation. Isolated clinical observations without systems insight do not qualify.
Single clinical anecdotes, trial results, or research observations presented without healthcare system context or operational implications.
3 example posts
Drug-resistant infections are a major public health threat around the world, responsible for more than a million deaths each year. Scientists are constantly trying to find and develop new antibiotics. Now, researchers say artificial intelligence is helping speed their search. ht
Presented at #EASCongress2026: Vafai and colleagues report reductions in PCSK9 and LDL cholesterol levels and no dose-limiting toxic effects in persons with hypercholesterolemia treated with VERVE-102, a base editor targeting PCSK9. Full study results: https://t.co/iAmoPdipx9 ht
An AI foundation whole-body 3D model that assesses perturbations (such as obesity) across multiple systems (such as immune, neural) at the cell level. This is MouseMapper. Imagine HumanMapper someday @Nature @erturklab https://t.co/jsg3VMzOqg https://t.co/tktuINis2W
Created 2026-05-28 · Updated 2026-05-28
Edit rule text
[workforce_labor_macro_without_healthcare_specificity]
Learned 3 rejections Active
Exclude posts about enterprise AI adoption, worker productivity, labor market macro trends, or workforce disruption that lack specific healthcare operational, clinical, or system context.
Posts about AI-driven workforce disruption, labor market trends, or productivity metrics applied broadly without healthcare-specific context.
3 example posts
The framing here is exactly backwards. This is the strongest AI bull signal anyone has published this year. Uber deployed Claude Code to 5,000 engineers in December. By March, 84% were classified as agentic coding users. By April, 95% used AI tools monthly. 70% of all committed
The more interesting part is that Microsoft’s own engineers liked Claude code best…and they’re cutting it anyway Token pricing is making enterprise actually look at what these models cost to run and that could be a problem for the labs when the subsidy ends. Lots of cheaper
Voice interaction, generating actions,and next, generating tasks...? The humanoid assistant is closer. For those who are bedridden due to illness or have limited mobility, it will serve as an excellent aid in daily living.
Created 2026-05-28 · Updated 2026-05-28
Edit rule text
[political_outrage_or_fraud_scandal_no_systems_analysis]
Learned 3 rejections Active
Exclude posts that report healthcare fraud scandals, regulatory violations, or political complaints (e.g., CEO assassination, UnitedHealth fraud accusations, 503B semaglutide bans) as sensationalized outrage or moral judgment without systems-level analysis of the underlying healthcare problem or constructive policy implications.
Posts reporting healthcare fraud, regulatory scandal, or political outrage without substantive analysis of systemic healthcare problems or solutions.
3 example posts
Why is healthcare expensive? The grifters gonna grift. ASCO is asking for oncology practices to be admitted into the same 340B machinery that helped health systems acquire physician practices for two decades. Read the tell. The proposed Indigent Care Ratio lets community
A senator complaining about healthcare costs should have to answer one question first: Why does your state make new healthcare competitors ask permission before opening? If yes, start there. Speeches do not lower prices. Supply does.
In what is considered the largest #Botox fraud scheme in the United States, a jury in Los Angeles convicted a California doctor on Tuesday in a $45 million scheme to defraud #Medicare by submitting claims for Botox injections that were never provided and medically unnecessary, ht
Created 2026-05-28 · Updated 2026-05-28
Edit rule text
[ai_infrastructure_hype_no_healthcare_validation]
Learned 3 rejections Active
Exclude posts that celebrate AI infrastructure breakthroughs, agentic system capabilities, or model technical achievements (e.g., 'agents can autonomously iterate', 'Claude Code deployed to X engineers') without evidence of validated healthcare impact, clinical use, or operational healthcare improvement.
Posts about AI infrastructure, compute, agents, or model capabilities presented as significant without demonstrating actual healthcare application or validation.
3 example posts
a databricks tech lead just spent 26 minutes on the part of multi-agent nobody wants to say out loud: your agents don't break because the model is dumb. they break because nothing is coordinating them. one agent is a feature. fifty is a distributed systems problem. https://t.c
OpenClaw is powerful enough that we needed Hermes Agent watching it. A lot of people are missing this about agents. The more useful they get, the more dangerous it is to let one system plan the work, run the work, review the work, and decide what happens next. It looks clean h
Figuring out how to benchmark agents on realistic biology research has quickly become one of my favorite types of engineering work. You work with scientists to get to the core of some biological claim, precisely assembling raw data/prior literature/experimental context in a https
Created 2026-05-28 · Updated 2026-05-28
Edit rule text
[anecdotal_single_case_no_systems_insight]
Learned 3 rejections Active
Exclude posts that present a single patient story, one doctor's observation, or an isolated clinical case as evidence, without connecting it to healthcare system dynamics, policy implications, or scalable insights. Anecdotes alone are insufficient.
Posts that describe a single clinical anecdote or individual patient case without broader healthcare system context or implications.
3 example posts
A Mom in Tulsa called 3 health systems last week asking the price of her son's tonsillectomy. Health system A: "We cannot quote you a price." Health system B: "Pricing depends on your insurance." Health system C: "Our financial counselor will reach out after the procedure." N
A surgeon I know recently operated on a man's spine. Saved his mobility. Saved his career. Saved his marriage, probably. The surgeon's take-home on that case, after the facility fee, anesthesia, device costs, payer discounts, and overhead, was $1,840. The health system collec
Voice interaction, generating actions,and next, generating tasks...? The humanoid assistant is closer. For those who are bedridden due to illness or have limited mobility, it will serve as an excellent aid in daily living.
Created 2026-05-28 · Updated 2026-05-28
Edit rule text
[clinical_research_without_systems_context]
Learned 3 rejections Active
Exclude posts that report clinical trial data, drug efficacy results, or research findings (AI-assisted or not) without connecting to HOW those findings change clinical workflows, reimbursement, adoption barriers, or care delivery systems. The post must address implementation barriers or systemic change, not just the research outcome.
Posts about clinical trial results, drug approvals, or preclinical findings presented as isolated observations without healthcare delivery system implications
3 example posts
Presented at #EASCongress2026: Vafai and colleagues report reductions in PCSK9 and LDL cholesterol levels and no dose-limiting toxic effects in persons with hypercholesterolemia treated with VERVE-102, a base editor targeting PCSK9. Full study results: https://t.co/iAmoPdipx9 ht
An AI foundation whole-body 3D model that assesses perturbations (such as obesity) across multiple systems (such as immune, neural) at the cell level. This is MouseMapper. Imagine HumanMapper someday @Nature @erturklab https://t.co/jsg3VMzOqg https://t.co/tktuINis2W
Really amazing story of drug repurposing for rare disease using AI found Clonazepam and worked!! (NCATS federated learning and a determined team made this work) 🚀 👏 @MayoClinic @ncats_nih_gov @statnews #STATBreakthrough https://t.co/D4a4IMXSwh
Created 2026-05-27 · Updated 2026-05-27
Edit rule text
[healthcare_fraud_scandal_without_analysis]
Learned 3 rejections Active
Exclude posts that report healthcare fraud convictions, billing schemes, or compliance violations as news items without analyzing WHY the system enabled the fraud or what structural changes would prevent recurrence. The post must explain the operational or incentive architecture that created the problem.
Posts reporting healthcare fraud, billing scandals, or regulatory enforcement actions without systems-level analysis of root causes or structural implications
3 example posts
A Mom in Tulsa called 3 health systems last week asking the price of her son's tonsillectomy. Health system A: "We cannot quote you a price." Health system B: "Pricing depends on your insurance." Health system C: "Our financial counselor will reach out after the procedure." N
Why is healthcare expensive? The grifters gonna grift. ASCO is asking for oncology practices to be admitted into the same 340B machinery that helped health systems acquire physician practices for two decades. Read the tell. The proposed Indigent Care Ratio lets community
A senator complaining about healthcare costs should have to answer one question first: Why does your state make new healthcare competitors ask permission before opening? If yes, start there. Speeches do not lower prices. Supply does.
Created 2026-05-27 · Updated 2026-05-27
Edit rule text
[agentic_ai_systems_technical_tangent]
Learned 3 rejections Active
Exclude posts that discuss agentic AI architecture, reliability, coordination, sandboxing, or deployment mechanics in abstract or non-healthcare-specific terms. These posts treat agents as a generic technical problem rather than a healthcare systems problem.
Posts focused on technical implementation details of agentic AI systems without healthcare-specific application or validation.
3 example posts
a databricks tech lead just spent 26 minutes on the part of multi-agent nobody wants to say out loud: your agents don't break because the model is dumb. they break because nothing is coordinating them. one agent is a feature. fifty is a distributed systems problem. https://t.c
OpenClaw is powerful enough that we needed Hermes Agent watching it. A lot of people are missing this about agents. The more useful they get, the more dangerous it is to let one system plan the work, run the work, review the work, and decide what happens next. It looks clean h
AI agents are increasingly deployed as persistent operational systems, but do they remain reliable over time? Unfortunately no, our new work shows agents can quietly fail after deployment, despite passing day-1 evaluation. We call this "agent aging", akin to human aging. https:/
Created 2026-05-27 · Updated 2026-05-27
Edit rule text
[broad_health_claim_or_market_debate_without_healthcare_systems_analysis]
Learned 3 rejections Active
Exclude posts that make sweeping health claims or debate clinical standards (e.g., 'LDL goals keep moving', 'medical error is third leading cause of death') without connecting to healthcare delivery operations, reimbursement incentives, regulatory policy, or system-level barriers to implementation.
Posts making broad claims about health trends, clinical guidelines, or market debates without analyzing healthcare delivery, economics, or policy systems.
3 example posts
We have hundreds of PanINs in our pancreas, especially after age 40. They are a pre-cancerous lesion, but it is fortunately rare for one to progress to cancer (PDAC). Now we know a reason for that. The microenvironment. @umichmedicine @UMRogelCancer https://t.co/HR5a3nZBng https
“The [AI] tool presented here will enable rapid screening for multiple systemic diseases using retinal photographs, and it is a step forward in the evolution of oculomics from experimental research to real-world clinical practice.” —Editorial Team, @NatureMedicine https://t.co/90
The healthy LDL number has been quietly moving its own goalposts for forty years: - 1988: under 160 - 1993: under 130 - 2001: under 100 - 2004: under 70 for the high-risk - 2019: under 55 for the very-high-risk - Current trajectory: as low as possible, indefinitely The science
Created 2026-05-27 · Updated 2026-05-27
Edit rule text
[anecdotal_clinical_case_without_systems_context]
Learned 3 rejections Active
Exclude posts that lead with a compelling single-patient story, rare disease case, or clinical anecdote without connecting it to healthcare delivery systems, operational constraints, or scalable insights. The post must show healthcare SYSTEM implications, not just a heartwarming or striking individual case.
Single clinical anecdotes or case stories that lack healthcare system analysis or operational insight.
3 example posts
A surgeon I know recently operated on a man's spine. Saved his mobility. Saved his career. Saved his marriage, probably. The surgeon's take-home on that case, after the facility fee, anesthesia, device costs, payer discounts, and overhead, was $1,840. The health system collec
An AI foundation whole-body 3D model that assesses perturbations (such as obesity) across multiple systems (such as immune, neural) at the cell level. This is MouseMapper. Imagine HumanMapper someday @Nature @erturklab https://t.co/jsg3VMzOqg https://t.co/tktuINis2W
Really amazing story of drug repurposing for rare disease using AI found Clonazepam and worked!! (NCATS federated learning and a determined team made this work) 🚀 👏 @MayoClinic @ncats_nih_gov @statnews #STATBreakthrough https://t.co/D4a4IMXSwh
Created 2026-05-27 · Updated 2026-05-27
Edit rule text
[unvalidated_speculative_peptide_glp1_market_claims]
Learned 3 rejections Active
Exclude posts that speculate about GLP-1, semaglutide, peptide, or weight-loss drug market share, pricing, adoption, or regulatory outcomes without citing clinical trials, FDA data, or validated market research. Anecdotes about consumer demand or app rankings do not count as validation.
Posts making unsubstantiated claims about GLP-1, peptide, or weight-loss drug market dynamics, competitive positioning, or regulatory status without clinical evidence or rigorous sourcing.
3 example posts
$HIMS up to #4 in the Canadian App Store medical category. That is not a stock chart. That is real consumer demand showing up in public. GLP-1 drama, Novo pressure, pricing debate, and now app-store strength. Bears keep arguing the story is slowing down, but the product demand
For years, grey-market HGH operated in a strange legal grey zone somewhere between “research compound,” “API” and underground enhancement culture. Last month’s FDA guidance suggests regulators may no longer see it that way. The recent decline in HGH quality, and now the growing
🚨 ARE PEPTIDES THE NEXT "GLP-1 MOMENT" FOR $HIMS? @eli_dorf: "I think there's something really interesting that happened with GLP-1s, which is that when someone's taking Ozempic or any GLP-1, they LOVE to talk about it." "They'll talk about it to their friends. They'll talk htt
Created 2026-05-27 · Updated 2026-05-27
Edit rule text
[unvalidated_speculative_peptide_or_glp1_claims]
Learned 3 rejections Active
Exclude posts that speculate on GLP-1 or peptide market trends, pricing pressure, international expansion, or competitive dynamics (e.g., 'peptides are the next GLP-1 moment', 'Novo's brand premium moat') without grounding claims in clinical efficacy, patient outcomes, or systemic healthcare access barriers.
Posts speculating about peptide/GLP-1 market dynamics, pricing, regulatory changes, or adoption without clinical evidence or healthcare systems analysis
3 example posts
For years, grey-market HGH operated in a strange legal grey zone somewhere between “research compound,” “API” and underground enhancement culture. Last month’s FDA guidance suggests regulators may no longer see it that way. The recent decline in HGH quality, and now the growing
🚨 ARE PEPTIDES THE NEXT "GLP-1 MOMENT" FOR $HIMS? @eli_dorf: "I think there's something really interesting that happened with GLP-1s, which is that when someone's taking Ozempic or any GLP-1, they LOVE to talk about it." "They'll talk about it to their friends. They'll talk htt
This is what the peptide space needs more of: clinicians with actual patient volume, public reasoning, and accountability. The next layer is transparency around sourcing, testing, and outcomes. Peptides are not a vibes category anymore. The signal is getting cleaner.
Created 2026-05-27 · Updated 2026-05-27
Edit rule text
[unvalidated_speculative_peptide_or_drug_market_claims]
Learned 3 rejections Active
Exclude posts that treat peptides, GLP-1s, or drugs as market assets or investment narratives (e.g., 'peptides are the next GLP-1 moment', pricing comparisons, app store rankings, generic launch timing) without grounding claims in clinical outcomes, real-world access barriers, or healthcare system implications.
Posts speculating on peptide, GLP-1, or pharmaceutical market dynamics, pricing, or competitive positioning without clinical evidence or healthcare systems analysis.
3 example posts
For years, grey-market HGH operated in a strange legal grey zone somewhere between “research compound,” “API” and underground enhancement culture. Last month’s FDA guidance suggests regulators may no longer see it that way. The recent decline in HGH quality, and now the growing
🚨 ARE PEPTIDES THE NEXT "GLP-1 MOMENT" FOR $HIMS? @eli_dorf: "I think there's something really interesting that happened with GLP-1s, which is that when someone's taking Ozempic or any GLP-1, they LOVE to talk about it." "They'll talk about it to their friends. They'll talk htt
Hims & Hers $HIMS Offered $300M Convertable Senior Notes to 2032 💊 Buyers will have the option to add an additional $45M They plan to use the cash for: Expansions, AI, acquisitions, and infrastructure! https://t.co/zaZHO80bY2
Created 2026-05-27 · Updated 2026-05-27
Edit rule text
[personal_anecdote_no_systems_insight]
Learned 3 rejections Active
Exclude posts that narrate a single person's experience or one doctor's observation without extracting systemic lessons, policy context, or replicable insights about how healthcare actually works at scale.
Posts that share individual stories or single cases without connecting to healthcare system dynamics or policy implications.
3 example posts
A surgeon I know recently operated on a man's spine. Saved his mobility. Saved his career. Saved his marriage, probably. The surgeon's take-home on that case, after the facility fee, anesthesia, device costs, payer discounts, and overhead, was $1,840. The health system collec
Voice interaction, generating actions,and next, generating tasks...? The humanoid assistant is closer. For those who are bedridden due to illness or have limited mobility, it will serve as an excellent aid in daily living.
Caregivers do some of the most important work in dementia care — almost none of it is visible to the health care system. Dr. Rosemary Laird describes a navigator model where caregivers reported 86% fewer falls and 81% fewer hospitalizations, over a 12-month pilot. https://t.co/sG
Created 2026-05-27 · Updated 2026-05-27
Edit rule text
[tangential_non_healthcare_domain_or_policy_with_loose_healthcare_framing]
Learned 3 rejections Active
Exclude posts that are primarily about non-healthcare domains (cybersecurity vulnerabilities, energy economics, international regulatory systems) and only loosely tag or frame them as healthcare-relevant. The healthcare angle must be central and specific, not a superficial label.
Posts about non-healthcare domains (cybersecurity, energy, international markets) that force a healthcare angle without genuine healthcare specificity or application.
3 example posts
Anthropic built a model so good at hacking they refused to release it publicly. INSTEAD they handed it to Apple, Google, Microsoft, AWS, Cloudflare one month later: 10,000+ zero days found in the software that runs the internet...bugs hidden for 27 years in OpenBSD. exploit htt
🚨 Anthropic just dropped the first Project Glasswing update Claude Mythos found 10,000+ critical vulnerabilities in ONE month: > Cloudflare: 2,000 bugs, 400 high/critical severity > Mozilla: 271 vulnerabilities in Firefox 150 — 10x more vulnerabilities found in Firefox 1
600 new generic drugs added to TrumpRx this week thanks to President @realDonaldTrump's leadership! Before you visit your local pharmacy, visit https://t.co/dDMdsILHTj to ensure you're getting the BEST price for your prescriptions. It's easy and saves YOU money! https://t.co/
Created 2026-05-27 · Updated 2026-05-27
Edit rule text
[founder_enthusiasm_or_lab_update_without_healthcare_systems_validation]
Learned 3 rejections Active
Exclude posts that are self-promotional announcements by founders or companies about their own products, labs, or metrics (e.g., 'Claude Code team just dropped', 'OpenEvidence is launching', 'Anthropic built a model'). These must demonstrate independent healthcare validation or systemic impact, not just company capability announcements.
Posts from biotech/AI founders announcing product launches, lab breakthroughs, or company metrics without independent validation or healthcare system integration evidence.
3 example posts
Claude Code team just dropped a workshop on how to ship a production-ready agent from scratch. 27-minutes. Free. Live coding by Claude dev. Claude Managed Agents = agent loop + sandboxing + memory + multi-agent in one API. Worth more than any $500 vibe-coding course. https://t
#AI agents need more than a sandbox—they need predictable execution. Learn how OpenShell provides a standard interface to intercept, inspect, and approve LLM-generated bash commands before they run: https://t.co/JKmLMgl1A9 https://t.co/xWi8oBNZ7T
Startups that seek to block AI exam fraud, help companies use AI to make consistent decisions, make financial AI transactions safer, and use AI to organize medical device regulatory information took top honors at @Cornell_Tech’s annual Startup Awards on May 14. Nearly 600 https:
Created 2026-05-27 · Updated 2026-05-27
Edit rule text
[anecdotal_single_case_or_clinical_observation_without_healthcare_systems_context]
Learned 3 rejections Active
Exclude posts that describe a single patient case, individual clinical anecdote, or one-off clinical observation (e.g., 'a surgeon I know', 'a woman is told', 'doctors at Mayo found') without connecting to systemic healthcare insights, replicability, or policy implications. Single heartwarming stories or isolated successes are not sufficient.
Posts citing a single clinical case, patient anecdote, or isolated clinical observation without broader healthcare systems analysis or evidence of scalability.
3 example posts
A surgeon I know recently operated on a man's spine. Saved his mobility. Saved his career. Saved his marriage, probably. The surgeon's take-home on that case, after the facility fee, anesthesia, device costs, payer discounts, and overhead, was $1,840. The health system collec
An AI foundation whole-body 3D model that assesses perturbations (such as obesity) across multiple systems (such as immune, neural) at the cell level. This is MouseMapper. Imagine HumanMapper someday @Nature @erturklab https://t.co/jsg3VMzOqg https://t.co/tktuINis2W
Voice interaction, generating actions,and next, generating tasks...? The humanoid assistant is closer. For those who are bedridden due to illness or have limited mobility, it will serve as an excellent aid in daily living.
Created 2026-05-27 · Updated 2026-05-27
Edit rule text
[ai_capability_hype_or_technical_tangent_loose_healthcare_framing]
Learned 3 rejections Active
Exclude posts that focus on AI technical capability announcements, security vulnerabilities, or model performance metrics (e.g., zero-day vulnerability detection, code generation speed, token efficiency) where healthcare is mentioned only as a loose framing or example, not the core analysis.
Posts about AI technical capabilities, security vulnerabilities, or model performance that mention healthcare tangentially without demonstrating healthcare-specific applications or validation
3 example posts
What "agentic" actually means, stripped of marketing: The model plans, writes code, runs it, reads the error, fixes it, iterates — autonomously, dozens of times, until the goal is met. You specify in English. It builds. That's the loop. This is the capability @bcherny and team
Anthropic built a model so good at hacking they refused to release it publicly. INSTEAD they handed it to Apple, Google, Microsoft, AWS, Cloudflare one month later: 10,000+ zero days found in the software that runs the internet...bugs hidden for 27 years in OpenBSD. exploit htt
🚨 Anthropic just dropped the first Project Glasswing update Claude Mythos found 10,000+ critical vulnerabilities in ONE month: > Cloudflare: 2,000 bugs, 400 high/critical severity > Mozilla: 271 vulnerabilities in Firefox 150 — 10x more vulnerabilities found in Firefox 1
Created 2026-05-27 · Updated 2026-05-27
Edit rule text
[enterprise_ai_adoption_or_productivity_metrics_without_healthcare_specificity]
Learned 3 rejections Active
Exclude posts about AI improving engineer productivity, code generation speeds, software development timelines, or enterprise AI adoption rates unless they are explicitly tied to healthcare workflow redesign, clinical decision support, or measurable patient/provider outcomes.
Posts about AI productivity gains, code generation velocity, or enterprise adoption metrics applied to generic software engineering without healthcare-specific context.
3 example posts
The framing here is exactly backwards. This is the strongest AI bull signal anyone has published this year. Uber deployed Claude Code to 5,000 engineers in December. By March, 84% were classified as agentic coding users. By April, 95% used AI tools monthly. 70% of all committed
Workers say AI is making them more productive. Executives say AI is not making their companies more productive. Both groups are reporting on the same software. The gap between the two numbers is the entire debate about the largest capital expenditure cycle in technology history.
The reason why I release my X articles about AI agents (fat skill fat code thin harness) and GStack and GBrain is that we, yes you and I, can have *PROCESS POWER*, which is the one super powerful specific moat that anyone can create for themselves. The agent helps you do it. htt
Created 2026-05-26 · Updated 2026-05-26
Edit rule text
[generalist_ai_infrastructure_hype_no_healthcare_specificity]
Learned 3 rejections Active
Exclude posts that discuss AI infrastructure, compute efficiency, token pricing, or model capability benchmarks (e.g., 'tokens per dollar per watt', 'Claude beats doctors') unless they analyze *why* this matters for healthcare workflows, clinical decision-making, or healthcare economics.
Posts about AI infrastructure, compute, or model capabilities presented as universally transformative without healthcare-specific substance.
3 example posts
What "agentic" actually means, stripped of marketing: The model plans, writes code, runs it, reads the error, fixes it, iterates — autonomously, dozens of times, until the goal is met. You specify in English. It builds. That's the loop. This is the capability @bcherny and team
Analogizing AI / Mat-Mul to Energy... Remember between 1900 and 1970 kWh/capita grew 34x ... and cost per kWH droped .... 34x. It might be a speed run, but game is game. https://t.co/uxnlnoaAYx
Anthropic built a model so good at hacking they refused to release it publicly. INSTEAD they handed it to Apple, Google, Microsoft, AWS, Cloudflare one month later: 10,000+ zero days found in the software that runs the internet...bugs hidden for 27 years in OpenBSD. exploit htt
Created 2026-05-26 · Updated 2026-05-26
Edit rule text
[tangential_healthcare_framing_of_non_healthcare_policy_or_macro]
Learned 3 rejections Active
Exclude posts that discuss tax policy, federal-state budget mechanics, or macro-economic trends (e.g., 'state taxes hospitals to fund Medicaid expansion and trigger 90% federal match') where the healthcare framing is incidental and the post does not analyze provider financial incentives, care delivery impact, or health equity consequences.
Posts about state tax policy, federal matching rates, or general economic trends loosely framed as healthcare without operational or delivery system analysis.
3 example posts
If you want to inject $1B into healthcare for the truly vulnerable, it costs the state $500M. If you want to inject $1B for able-bodied adults, it costs the state $100M. The 90% federal match drastically lowers the "price" of expanding government healthcare. Incentives dictate
A senator complaining about healthcare costs should have to answer one question first: Why does your state make new healthcare competitors ask permission before opening? If yes, start there. Speeches do not lower prices. Supply does.
How to print money, state government edition: 1) Tax your hospitals and managed care plans. 2) Use that tax revenue to pay your 10% share of Medicaid expansion. 3) Trigger a 90% federal match. 4) Return the original tax money, plus the massive federal windfall, back to the health
Created 2026-05-26 · Updated 2026-05-26
Edit rule text
[glp1_peptide_market_speculation_without_healthcare_systems]
Learned 3 rejections Active
Exclude posts that report GLP-1/peptide pricing, market share, app rankings, or competitive dynamics (e.g., 'Hims app ranking up', 'Hims vs Ro comparison', 'generic semaglutide pricing') unless they include substantive healthcare system insight (e.g., access gaps, regulatory impact, clinical outcomes, reimbursement mechanisms).
Posts about GLP-1, peptides, or weight-loss drugs focused on market dynamics, pricing, or competitive positioning without healthcare system analysis.
3 example posts
$HIMS up to #4 in the Canadian App Store medical category. That is not a stock chart. That is real consumer demand showing up in public. GLP-1 drama, Novo pressure, pricing debate, and now app-store strength. Bears keep arguing the story is slowing down, but the product demand
🚨 ARE PEPTIDES THE NEXT "GLP-1 MOMENT" FOR $HIMS? @eli_dorf: "I think there's something really interesting that happened with GLP-1s, which is that when someone's taking Ozempic or any GLP-1, they LOVE to talk about it." "They'll talk about it to their friends. They'll talk htt
Hims & Hers $HIMS Offered $300M Convertable Senior Notes to 2032 💊 Buyers will have the option to add an additional $45M They plan to use the cash for: Expansions, AI, acquisitions, and infrastructure! https://t.co/zaZHO80bY2
Created 2026-05-26 · Updated 2026-05-26
Edit rule text
[generalist_ai_capability_hype_loose_healthcare]
Learned 3 rejections Active
Exclude posts that discuss AI model capabilities (code generation, reasoning, hacking, etc.) or agent architecture in primarily non-healthcare contexts, even if the post gestures toward healthcare application or uses healthcare examples without substantive analysis of healthcare-specific problems or systems.
Posts about general AI model capabilities or agentic systems with only tangential or speculative healthcare framing.
3 example posts
What "agentic" actually means, stripped of marketing: The model plans, writes code, runs it, reads the error, fixes it, iterates — autonomously, dozens of times, until the goal is met. You specify in English. It builds. That's the loop. This is the capability @bcherny and team
Anthropic built a model so good at hacking they refused to release it publicly. INSTEAD they handed it to Apple, Google, Microsoft, AWS, Cloudflare one month later: 10,000+ zero days found in the software that runs the internet...bugs hidden for 27 years in OpenBSD. exploit htt
🚨 Anthropic just dropped the first Project Glasswing update Claude Mythos found 10,000+ critical vulnerabilities in ONE month: > Cloudflare: 2,000 bugs, 400 high/critical severity > Mozilla: 271 vulnerabilities in Firefox 150 — 10x more vulnerabilities found in Firefox 1
Created 2026-05-26 · Updated 2026-05-26
Edit rule text
[tangential_non_healthcare_business_or_tech_with_loose_framing]
Learned 3 rejections Active
Exclude posts that discuss non-healthcare business models, tech infrastructure, energy policy, or geopolitical trends and merely add healthcare framing (e.g., 'energy efficiency applies to healthcare', 'European AI valuations', 'startup economics') without analyzing healthcare-specific barriers, regulation, or system impact.
Posts about non-healthcare companies, technologies, or macro trends that are loosely reframed with healthcare language but lack substantive healthcare analysis.
3 example posts
Analogizing AI / Mat-Mul to Energy... Remember between 1900 and 1970 kWh/capita grew 34x ... and cost per kWH droped .... 34x. It might be a speed run, but game is game. https://t.co/uxnlnoaAYx
#AI agents need more than a sandbox—they need predictable execution. Learn how OpenShell provides a standard interface to intercept, inspect, and approve LLM-generated bash commands before they run: https://t.co/JKmLMgl1A9 https://t.co/xWi8oBNZ7T
Startups that seek to block AI exam fraud, help companies use AI to make consistent decisions, make financial AI transactions safer, and use AI to organize medical device regulatory information took top honors at @Cornell_Tech’s annual Startup Awards on May 14. Nearly 600 https:
Created 2026-05-26 · Updated 2026-05-26
Edit rule text
[ai_company_product_metrics_and_business_announcements_not_healthcare_focused]
Learned 3 rejections Active
Exclude posts that announce AI company product launches, partnerships, internal adoption metrics (e.g., 'engineers using Claude Code'), or business deals unless the post explicitly discusses healthcare application, clinical validation, or healthcare delivery impact. Posts like 'Uber deployed Claude Code to 5,000 engineers' or 'Google's CEO confirmed 75% of new code at Google is AI-generated' should be rejected.
Posts reporting AI company product launches, partnerships, employee adoption rates, or business metrics without healthcare-specific application or validation.
3 example posts
The framing here is exactly backwards. This is the strongest AI bull signal anyone has published this year. Uber deployed Claude Code to 5,000 engineers in December. By March, 84% were classified as agentic coding users. By April, 95% used AI tools monthly. 70% of all committed
Claude Code team just dropped a workshop on how to ship a production-ready agent from scratch. 27-minutes. Free. Live coding by Claude dev. Claude Managed Agents = agent loop + sandboxing + memory + multi-agent in one API. Worth more than any $500 vibe-coding course. https://t
Andrej Karpathy just explained the future of software engineering without directly saying it. The best AI engineers are no longer “prompting.” They’re building systems around the agents. Karpathy’s biggest insight wasn’t: “Claude can code.” It was: LLMs become dramatically be
Created 2026-05-26 · Updated 2026-05-26
Edit rule text
[tangential_ai_safety_or_cybersecurity_without_healthcare_lens]
Learned 3 rejections Active
Exclude posts that report on AI security vulnerabilities (zero-day bugs, hacking capability) or safety incidents where healthcare is mentioned only as a passive example or framing device. The post must analyze how the vulnerability or incident materially impacts healthcare delivery, clinical trust, or patient safety — not just that a model can hack or find bugs.
Posts about AI vulnerability disclosures, security incidents, or safety research that treat healthcare as incidental context rather than the primary domain affected.
3 example posts
Anthropic built a model so good at hacking they refused to release it publicly. INSTEAD they handed it to Apple, Google, Microsoft, AWS, Cloudflare one month later: 10,000+ zero days found in the software that runs the internet...bugs hidden for 27 years in OpenBSD. exploit htt
🚨 Anthropic just dropped the first Project Glasswing update Claude Mythos found 10,000+ critical vulnerabilities in ONE month: > Cloudflare: 2,000 bugs, 400 high/critical severity > Mozilla: 271 vulnerabilities in Firefox 150 — 10x more vulnerabilities found in Firefox 1
#AI agents need more than a sandbox—they need predictable execution. Learn how OpenShell provides a standard interface to intercept, inspect, and approve LLM-generated bash commands before they run: https://t.co/JKmLMgl1A9 https://t.co/xWi8oBNZ7T
Created 2026-05-26 · Updated 2026-05-26
Edit rule text
[tangential_cybersecurity_or_vulnerability_hype]
Learned 3 rejections Active
Exclude posts about AI safety vulnerabilities, zero-day exploits, or security incidents (Claude Mythos, generalist AI hacking) unless directly demonstrating healthcare-specific security risks or requirements. Posts about software security, cloud vulnerabilities, or generalist AI safety tangents do not qualify.
Posts about AI safety, security vulnerabilities, or ethical concerns applied to non-healthcare domains or without healthcare implementation specificity.
3 example posts
Anthropic built a model so good at hacking they refused to release it publicly. INSTEAD they handed it to Apple, Google, Microsoft, AWS, Cloudflare one month later: 10,000+ zero days found in the software that runs the internet...bugs hidden for 27 years in OpenBSD. exploit htt
🚨 Anthropic just dropped the first Project Glasswing update Claude Mythos found 10,000+ critical vulnerabilities in ONE month: > Cloudflare: 2,000 bugs, 400 high/critical severity > Mozilla: 271 vulnerabilities in Firefox 150 — 10x more vulnerabilities found in Firefox 1
#AI agents need more than a sandbox—they need predictable execution. Learn how OpenShell provides a standard interface to intercept, inspect, and approve LLM-generated bash commands before they run: https://t.co/JKmLMgl1A9 https://t.co/xWi8oBNZ7T
Created 2026-05-26 · Updated 2026-05-26
Edit rule text
[generalist_macro_infrastructure_hype]
Learned 3 rejections Active
Exclude posts that discuss infrastructure, compute costs, energy consumption, or macro trends (e.g., kWh/capita, tokens per dollar per watt, GPU lifespans, capital allocation) even if healthcare is mentioned in passing. The post must directly analyze healthcare system impact, not use healthcare as a loose hook for infrastructure commentary.
Posts about compute, infrastructure, energy efficiency, or macro-economic trends framed with loose or no healthcare specificity.
3 example posts
Analogizing AI / Mat-Mul to Energy... Remember between 1900 and 1970 kWh/capita grew 34x ... and cost per kWH droped .... 34x. It might be a speed run, but game is game. https://t.co/uxnlnoaAYx
Gavin Baker (@GavinSBaker) says the disaggregation of inference can extend GPU useful lives from 3-4 years to 10-15. That may single-handedly save private credit and reduce the financing rates for GPUs, which will drive demand and help finance the build-out. "The disaggregation
Google's CEO confirmed 75% of new code at Google is AI-generated. October 2024: 25%. Fall 2025: 50%. April 2026: 75%. Doubled in a year. Then doubled again in six months. Engineers aren't writing code anymore. They're reviewing it. The job title is the same. The job
Created 2026-05-25 · Updated 2026-05-25
Edit rule text
[speculative_robotics_or_sci_fi_applications_without_validation]
Learned 3 rejections Active
Exclude posts that speculate about future robotic assistants, bedridden patient care robots, or sci-fi healthcare applications (e.g., 'humanoid assistant closer for bedridden patients') without reporting validated pilots, actual deployment, or clinical evidence of benefit.
Posts about humanoid robots, speculative AI assistants, or futuristic applications in healthcare without evidence of real deployment or validated effectiveness.
3 example posts
Ontology all the way down. 11 years. Heavy civil construction. People, equipment, materials, contracts. This is what real-world AI looks like: operational, measurable, and in the field. Palantir x Cavanagh through 2035. https://t.co/872Cs1Qbdu
Turn FDE from a noun into a verb with Apollo. FDE is becoming the default operating model for deploying AI in the enterprise. The hard truth: AI does not become valuable in a demo. It becomes valuable when it is embedded into the real workflows, data, constraints, and
Voice interaction, generating actions,and next, generating tasks...? The humanoid assistant is closer. For those who are bedridden due to illness or have limited mobility, it will serve as an excellent aid in daily living.
Created 2026-05-25 · Updated 2026-06-01
Edit rule text
[political_outrage_regulatory_posturing_without_healthcare_analysis]
Learned 3 rejections Active
Exclude posts that cite political figures, regulatory scandals, fraud cases, or conspiracy claims primarily to express moral outrage or sensationalism, without analyzing the underlying healthcare system failures, incentive misalignment, or specific policy reforms needed.
Posts leveraging political outrage, conspiracy narratives, or regulatory scandals to frame healthcare without substantive systems analysis or policy recommendations.
3 example posts
Why is healthcare expensive? The grifters gonna grift. ASCO is asking for oncology practices to be admitted into the same 340B machinery that helped health systems acquire physician practices for two decades. Read the tell. The proposed Indigent Care Ratio lets community
A senator complaining about healthcare costs should have to answer one question first: Why does your state make new healthcare competitors ask permission before opening? If yes, start there. Speeches do not lower prices. Supply does.
Remember when UnitedHealth's CEO was assassinated in 2024? 18 months later, every accusation the killer Luigi Mangione made has been proven RIGHT. UnitedHealth is one of the biggest corporate collapses in recent years. And they deserve it. In December 2024, UnitedHealthcare CE
Created 2026-05-25 · Updated 2026-05-25
Edit rule text
[clinical_observation_or_study_without_systems_context]
Learned 3 rejections Active
Exclude posts that cite clinical studies, trial results, research observations, or single medical cases (e.g., 'AI helped diagnose rare disease', 'AI finds drug repurposing', 'medical error statistics') unless the post addresses healthcare system implementation barriers, access, or policy implications.
Posts reporting clinical research findings, trial data, or medical observations without connecting to healthcare systems, policy, or operational context.
3 example posts
An AI foundation whole-body 3D model that assesses perturbations (such as obesity) across multiple systems (such as immune, neural) at the cell level. This is MouseMapper. Imagine HumanMapper someday @Nature @erturklab https://t.co/jsg3VMzOqg https://t.co/tktuINis2W
In a meta-analysis of 210 biomedical AI studies that statistically compared models under cross-validation, 97% used invalid statistical tests. Here's our new preprint https://t.co/OG58Vkeu49 led by @tianchuzeng @kkli20111 @ZShaoshi @ten_photos 1/N https://t.co/PEnqXQxcsJ
Really amazing story of drug repurposing for rare disease using AI found Clonazepam and worked!! (NCATS federated learning and a determined team made this work) 🚀 👏 @MayoClinic @ncats_nih_gov @statnews #STATBreakthrough https://t.co/D4a4IMXSwh
Created 2026-05-25 · Updated 2026-05-25
Edit rule text
[clinical_anecdote_or_single_case_without_systems]
Learned 3 rejections Active
Exclude posts that highlight a single patient case, rare disease discovery, or one-off clinical success story (e.g., AI finding rare disease treatment, newborn recovery) without discussing healthcare system implications, replicability, or scalability.
Posts about single clinical cases, rare disease treatments, or anecdotal evidence without healthcare systems context or scalability analysis.
3 example posts
An AI foundation whole-body 3D model that assesses perturbations (such as obesity) across multiple systems (such as immune, neural) at the cell level. This is MouseMapper. Imagine HumanMapper someday @Nature @erturklab https://t.co/jsg3VMzOqg https://t.co/tktuINis2W
Really amazing story of drug repurposing for rare disease using AI found Clonazepam and worked!! (NCATS federated learning and a determined team made this work) 🚀 👏 @MayoClinic @ncats_nih_gov @statnews #STATBreakthrough https://t.co/D4a4IMXSwh
We have hundreds of PanINs in our pancreas, especially after age 40. They are a pre-cancerous lesion, but it is fortunately rare for one to progress to cancer (PDAC). Now we know a reason for that. The microenvironment. @umichmedicine @UMRogelCancer https://t.co/HR5a3nZBng https
Created 2026-05-25 · Updated 2026-05-25
Edit rule text
[glp1_peptide_market_dynamics_without_systems]
Learned 3 rejections Active
Exclude posts that discuss GLP-1 pricing, generic availability, app store rankings, market competition (Hims vs Ro), or adoption metrics without analyzing how these dynamics affect healthcare delivery systems, clinical outcomes, or payer sustainability.
Posts about GLP-1, peptide, or weight-loss drug market pricing, competition, or adoption without healthcare systems analysis.
3 example posts
$HIMS up to #4 in the Canadian App Store medical category. That is not a stock chart. That is real consumer demand showing up in public. GLP-1 drama, Novo pressure, pricing debate, and now app-store strength. Bears keep arguing the story is slowing down, but the product demand
For years, grey-market HGH operated in a strange legal grey zone somewhere between “research compound,” “API” and underground enhancement culture. Last month’s FDA guidance suggests regulators may no longer see it that way. The recent decline in HGH quality, and now the growing
🚨 ARE PEPTIDES THE NEXT "GLP-1 MOMENT" FOR $HIMS? @eli_dorf: "I think there's something really interesting that happened with GLP-1s, which is that when someone's taking Ozempic or any GLP-1, they LOVE to talk about it." "They'll talk about it to their friends. They'll talk htt
Created 2026-05-25 · Updated 2026-05-25
Edit rule text
[forward_deployed_engineering_or_enterprise_adoption_macro]
Learned 3 rejections Active
Exclude posts that discuss enterprise AI adoption trends, forward-deployed engineering practices, or business model philosophies (e.g., 'software will become free,' 'embed engineers inside customers') unless the post ties these explicitly to healthcare operational challenges, staffing models, or provider workflows.
Posts about enterprise AI adoption, engineering culture, or business model philosophy without healthcare-specific workforce or operational context.
3 example posts
“Rural hospitals cannot make it on their own.” Nobody makes it on their own. Businesses need capital, strategy, discipline, operating systems, good finance, clean accounting, payer strategy, physician alignment, and leaders who know what business they are actually in. The
Startups that seek to block AI exam fraud, help companies use AI to make consistent decisions, make financial AI transactions safer, and use AI to organize medical device regulatory information took top honors at @Cornell_Tech’s annual Startup Awards on May 14. Nearly 600 https:
Fair warning, YC founders: if you take these tokens, there’s a non-zero chance that OpenAI will study exactly what your startup is doing, copy your idea and put your app into their free offering. This is the classic platform playbook — be careful, founders!
Created 2026-05-25 · Updated 2026-05-25
Edit rule text
[broad_health_claim_or_market_dynamics_without_nuance]
Learned 3 rejections Active
Exclude posts that assert broad health claims (e.g., 'LDL targets have shifted for 40 years,' 'medical error is third leading cause of death') or describe market/labor dynamics without explaining how those dynamics interact with healthcare policy, incentives, or clinical practice.
Posts making broad claims about health trends, clinical thresholds, or market dynamics without evidence-based nuance or healthcare system context.
3 example posts
Medicaid spending on autism therapy nearly tripled between 2020 and 202. In North Carolina, the number of clinics offering applied behavior analysis grew from 61 in 2019 to 409 in 2026. Anytime fraudsters can steal money from the federal government, they do. https://t.co/fahy
If you want to inject $1B into healthcare for the truly vulnerable, it costs the state $500M. If you want to inject $1B for able-bodied adults, it costs the state $100M. The 90% federal match drastically lowers the "price" of expanding government healthcare. Incentives dictate
Voice interaction, generating actions,and next, generating tasks...? The humanoid assistant is closer. For those who are bedridden due to illness or have limited mobility, it will serve as an excellent aid in daily living.
Created 2026-05-25 · Updated 2026-05-25
Edit rule text
[glp1_peptide_market_dynamics_without_healthcare_systems]
Learned 3 rejections Active
Exclude posts that focus on GLP-1/peptide pricing, app store rankings, market competition, or generic drug availability as financial/market events. Include only if the post analyzes how these dynamics affect healthcare access, clinical practice, or system capacity.
Posts about GLP-1 or peptide market pricing, competition, and financial dynamics without healthcare systems analysis.
3 example posts
$HIMS up to #4 in the Canadian App Store medical category. That is not a stock chart. That is real consumer demand showing up in public. GLP-1 drama, Novo pressure, pricing debate, and now app-store strength. Bears keep arguing the story is slowing down, but the product demand
For years, grey-market HGH operated in a strange legal grey zone somewhere between “research compound,” “API” and underground enhancement culture. Last month’s FDA guidance suggests regulators may no longer see it that way. The recent decline in HGH quality, and now the growing
🚨 ARE PEPTIDES THE NEXT "GLP-1 MOMENT" FOR $HIMS? @eli_dorf: "I think there's something really interesting that happened with GLP-1s, which is that when someone's taking Ozempic or any GLP-1, they LOVE to talk about it." "They'll talk about it to their friends. They'll talk htt
Created 2026-05-25 · Updated 2026-05-25
Edit rule text
[tangential_ai_infrastructure_hype_no_healthcare_lens]
Learned 3 rejections Active
Exclude posts about AI infrastructure (tokens per dollar, compute scaling, energy efficiency, GPU lifecycles, or general AI capability hype) unless they directly address healthcare-specific infrastructure constraints, clinical workflow integration, or healthcare cost structures. Posts comparing AI to historical energy trends or general enterprise compute trends are not healthcare-focused.
Posts about AI infrastructure, compute, energy economics, or general AI capability trends with loose or no healthcare framing
3 example posts
Analogizing AI / Mat-Mul to Energy... Remember between 1900 and 1970 kWh/capita grew 34x ... and cost per kWH droped .... 34x. It might be a speed run, but game is game. https://t.co/uxnlnoaAYx
Gavin Baker (@GavinSBaker) says the disaggregation of inference can extend GPU useful lives from 3-4 years to 10-15. That may single-handedly save private credit and reduce the financing rates for GPUs, which will drive demand and help finance the build-out. "The disaggregation
The more interesting part is that Microsoft’s own engineers liked Claude code best…and they’re cutting it anyway Token pricing is making enterprise actually look at what these models cost to run and that could be a problem for the labs when the subsidy ends. Lots of cheaper
Created 2026-05-24 · Updated 2026-05-24
Edit rule text
[broad_health_claims_or_market_debate_without_systems_context]
Learned 3 rejections Active
Exclude posts that make sweeping health claims (e.g., 'caregivers do invisible work', 'medical error is third leading cause of death', 'LDL goalposts keep moving'), present single clinical anecdotes, or critique healthcare business models without grounding in healthcare system economics, incentive structures, or operational data.
Posts making broad healthcare trend claims, clinical observations, or business model critiques without substantive evidence or healthcare systems insight.
3 example posts
Medicaid spending on autism therapy nearly tripled between 2020 and 202. In North Carolina, the number of clinics offering applied behavior analysis grew from 61 in 2019 to 409 in 2026. Anytime fraudsters can steal money from the federal government, they do. https://t.co/fahy
Caregivers do some of the most important work in dementia care — almost none of it is visible to the health care system. Dr. Rosemary Laird describes a navigator model where caregivers reported 86% fewer falls and 81% fewer hospitalizations, over a 12-month pilot. https://t.co/sG
We have hundreds of PanINs in our pancreas, especially after age 40. They are a pre-cancerous lesion, but it is fortunately rare for one to progress to cancer (PDAC). Now we know a reason for that. The microenvironment. @umichmedicine @UMRogelCancer https://t.co/HR5a3nZBng https
Created 2026-05-24 · Updated 2026-05-24
Edit rule text
[unvalidated_or_speculative_medical_claims_and_fringe_interventions]
Learned 3 rejections Active
Exclude posts that speculate on unregulated, grey-market, or investigational medical interventions (e.g., grey-market HGH, unapproved peptides, 'next GLP-1 moment' hype) without clinical evidence, regulatory approval pathway, or healthcare system adoption context. Posts must be grounded in validated treatments or approved therapies.
Posts promoting unvalidated drugs, peptides, or medical interventions with speculative or unsubstantiated claims.
3 example posts
For years, grey-market HGH operated in a strange legal grey zone somewhere between “research compound,” “API” and underground enhancement culture. Last month’s FDA guidance suggests regulators may no longer see it that way. The recent decline in HGH quality, and now the growing
Medicaid spending on autism therapy nearly tripled between 2020 and 202. In North Carolina, the number of clinics offering applied behavior analysis grew from 61 in 2019 to 409 in 2026. Anytime fraudsters can steal money from the federal government, they do. https://t.co/fahy
The more interesting part is that Microsoft’s own engineers liked Claude code best…and they’re cutting it anyway Token pricing is making enterprise actually look at what these models cost to run and that could be a problem for the labs when the subsidy ends. Lots of cheaper
Created 2026-05-24 · Updated 2026-05-24
Edit rule text
[clinical_research_or_pharma_announcement_without_systems_context]
Learned 3 rejections Active
Exclude posts that report clinical trial results, drug approvals, biomarker discoveries, or pharma R&D announcements (MouseMapper, drug repurposing, amines synthesis, pancreatic lesion screening) without connecting to healthcare delivery challenges, diagnostic workflows, treatment access barriers, or systemic adoption obstacles.
Posts about drug trials, clinical data, biomarkers, or pharmaceutical breakthroughs presented as isolated observations without healthcare systems implications.
3 example posts
An AI foundation whole-body 3D model that assesses perturbations (such as obesity) across multiple systems (such as immune, neural) at the cell level. This is MouseMapper. Imagine HumanMapper someday @Nature @erturklab https://t.co/jsg3VMzOqg https://t.co/tktuINis2W
Really amazing story of drug repurposing for rare disease using AI found Clonazepam and worked!! (NCATS federated learning and a determined team made this work) 🚀 👏 @MayoClinic @ncats_nih_gov @statnews #STATBreakthrough https://t.co/D4a4IMXSwh
We have hundreds of PanINs in our pancreas, especially after age 40. They are a pre-cancerous lesion, but it is fortunately rare for one to progress to cancer (PDAC). Now we know a reason for that. The microenvironment. @umichmedicine @UMRogelCancer https://t.co/HR5a3nZBng https
Created 2026-05-24 · Updated 2026-05-24
Edit rule text
[founder_enthusiasm_lab_update_or_company_announcement_without_validation]
Learned 3 rejections Active
Exclude posts that amplify founder enthusiasm, company product announcements, or lab updates (e.g., 'Andrej Karpathy explains the future of software engineering', Deepgram voice mode launch, Claude Code workshop) without demonstrating concrete healthcare adoption, clinical validation, or system-wide outcomes. Promotional or aspirational CEO/founder commentary alone is insufficient.
Posts celebrating AI company product launches, founder insights, or startup announcements without healthcare-specific validation or impact evidence.
3 example posts
Startups that seek to block AI exam fraud, help companies use AI to make consistent decisions, make financial AI transactions safer, and use AI to organize medical device regulatory information took top honors at @Cornell_Tech’s annual Startup Awards on May 14. Nearly 600 https:
Voice interaction, generating actions,and next, generating tasks...? The humanoid assistant is closer. For those who are bedridden due to illness or have limited mobility, it will serve as an excellent aid in daily living.
Between rooms. On rounds. Walking the corridor outside an OR. Charting one-handed during a phone call. This is where clinical questions happen. Today we're launching Voice Mode. OpenEvidence is the first multimodal medical AI: physicians can type, speak, or listen, on the same h
Created 2026-05-24 · Updated 2026-05-24
Edit rule text
[unvalidated_speculative_peptide_and_glp1_claims]
Learned 3 rejections Active
Exclude posts that speculate about peptide or GLP-1 market expansion, efficacy, or adoption using vague enthusiasm, founder hype, or market dynamics language without grounding in clinical trial data, real-world health outcomes, or systemic healthcare impacts. Market pricing and competitive positioning alone do not qualify.
Posts making unvalidated, speculative, or hype-driven claims about peptides, GLP-1 drugs, or weight-loss therapeutics without clinical evidence or healthcare system analysis.
3 example posts
For years, grey-market HGH operated in a strange legal grey zone somewhere between “research compound,” “API” and underground enhancement culture. Last month’s FDA guidance suggests regulators may no longer see it that way. The recent decline in HGH quality, and now the growing
🚨 ARE PEPTIDES THE NEXT "GLP-1 MOMENT" FOR $HIMS? @eli_dorf: "I think there's something really interesting that happened with GLP-1s, which is that when someone's taking Ozempic or any GLP-1, they LOVE to talk about it." "They'll talk about it to their friends. They'll talk htt
This is what the peptide space needs more of: clinicians with actual patient volume, public reasoning, and accountability. The next layer is transparency around sourcing, testing, and outcomes. Peptides are not a vibes category anymore. The signal is getting cleaner.
Created 2026-05-24 · Updated 2026-05-24
Edit rule text
[broad_market_or_policy_dynamics_without_healthcare_systems_analysis]
Learned 3 rejections Active
Exclude posts that describe healthcare market trends (pricing, competition, consolidation, labor shortages, rural hospital closures, workforce disruption) or policy mechanics (Medicaid, CMS, 340B programs, tax structures) without connecting those trends to clinical care quality, patient access, provider autonomy, or systemic incentive misalignment. Macro-level description alone is insufficient.
Posts discussing healthcare market trends, policy mechanics, or labor dynamics without analyzing how they reshape clinical care, patient outcomes, or provider incentives
3 example posts
Medicaid spending on autism therapy nearly tripled between 2020 and 202. In North Carolina, the number of clinics offering applied behavior analysis grew from 61 in 2019 to 409 in 2026. Anytime fraudsters can steal money from the federal government, they do. https://t.co/fahy
Think about this. The $50 per month out-of-pocket cost of GLP-1 drugs for Americans (with insurer paying the balance of $2000 pm) is more than double the entire cost for Indians (less than ₹2000pm for Semaglutide). Overall cost difference? 100 times. Not 100%.
Why is healthcare expensive? The grifters gonna grift. ASCO is asking for oncology practices to be admitted into the same 340B machinery that helped health systems acquire physician practices for two decades. Read the tell. The proposed Indigent Care Ratio lets community
Created 2026-05-24 · Updated 2026-05-24
Edit rule text
[tangential_policy_or_fraud_scandal_without_healthcare_systems_analysis]
Learned 3 rejections Active
Exclude posts that report fraud convictions, regulatory actions, or policy announcements (e.g., 'hospital sued,' 'senator complains about costs,' 'CEO assassination and accusations') unless the post analyzes structural healthcare system incentives, reimbursement mechanics, competitive dynamics, or policy consequences for providers and patients.
Posts reporting healthcare fraud, regulatory scandals, or policy changes that lack substantive systems-level analysis or operational healthcare insight.
3 example posts
Why is healthcare expensive? The grifters gonna grift. ASCO is asking for oncology practices to be admitted into the same 340B machinery that helped health systems acquire physician practices for two decades. Read the tell. The proposed Indigent Care Ratio lets community
“Rural hospitals cannot make it on their own.” Nobody makes it on their own. Businesses need capital, strategy, discipline, operating systems, good finance, clean accounting, payer strategy, physician alignment, and leaders who know what business they are actually in. The
If you want to inject $1B into healthcare for the truly vulnerable, it costs the state $500M. If you want to inject $1B for able-bodied adults, it costs the state $100M. The 90% federal match drastically lowers the "price" of expanding government healthcare. Incentives dictate
Created 2026-05-24 · Updated 2026-05-24
Edit rule text
[glp1_peptide_market_pricing_dynamics_without_systems_analysis]
Learned 3 rejections Active
Exclude posts that report GLP-1 or peptide drug prices, cost comparisons, international pricing differences, or market competition (e.g., 'Ozempic costs $2000/month in US, ₹2000 in India') unless the post analyzes healthcare system consequences, access barriers, reimbursement policy, or clinical outcomes implications.
Posts about GLP-1 and peptide drug pricing, competition, or market positioning that lack healthcare systems analysis or policy substance.
3 example posts
Think about this. The $50 per month out-of-pocket cost of GLP-1 drugs for Americans (with insurer paying the balance of $2000 pm) is more than double the entire cost for Indians (less than ₹2000pm for Semaglutide). Overall cost difference? 100 times. Not 100%.
🚨 ARE PEPTIDES THE NEXT "GLP-1 MOMENT" FOR $HIMS? @eli_dorf: "I think there's something really interesting that happened with GLP-1s, which is that when someone's taking Ozempic or any GLP-1, they LOVE to talk about it." "They'll talk about it to their friends. They'll talk htt
^Humira’s patent expired a while back for some context there. The brand premium is the moat now. Ozempic is not generic semaglutide in patient and prescriber perception. Novo keeps charging brand pricing on every loyal patient without a PMPRB ceiling.
Created 2026-05-24 · Updated 2026-05-24
Edit rule text
[tangential_ai_infrastructure_hype_no_healthcare_validation]
Learned 3 rejections Active
Exclude posts that discuss AI model technical capabilities (e.g., 'Claude found 10,000 zero-day vulnerabilities', 'AI code generation', 'tokens per dollar'), infrastructure, or cybersecurity incidents where healthcare application is assumed or hypothetical rather than demonstrated or evidence-based.
Posts about AI model capabilities, security vulnerabilities, or infrastructure technical breakthroughs with only loose or speculative healthcare framing.
3 example posts
Claude Code team just dropped a workshop on how to ship a production-ready agent from scratch. 27-minutes. Free. Live coding by Claude dev. Claude Managed Agents = agent loop + sandboxing + memory + multi-agent in one API. Worth more than any $500 vibe-coding course. https://t
#AI agents need more than a sandbox—they need predictable execution. Learn how OpenShell provides a standard interface to intercept, inspect, and approve LLM-generated bash commands before they run: https://t.co/JKmLMgl1A9 https://t.co/xWi8oBNZ7T
Andrej Karpathy just explained the future of software engineering without directly saying it. The best AI engineers are no longer “prompting.” They’re building systems around the agents. Karpathy’s biggest insight wasn’t: “Claude can code.” It was: LLMs become dramatically be
Created 2026-05-24 · Updated 2026-05-24
Edit rule text
[unvalidated_speculative_medical_claims_peptides]
Learned 3 rejections Active
Exclude posts that speculate about peptides, GLP-1s, semaglutide, or experimental drugs as market opportunities, investment theses, or consumer solutions without grounding in completed clinical trials, regulatory approval, or peer-reviewed evidence. Market positioning and pricing discussion without clinical validation is not sufficient.
Posts making broad claims about peptides, GLP-1s, or experimental drugs as solutions without clinical evidence, trials, or regulatory validation.
3 example posts
For years, grey-market HGH operated in a strange legal grey zone somewhere between “research compound,” “API” and underground enhancement culture. Last month’s FDA guidance suggests regulators may no longer see it that way. The recent decline in HGH quality, and now the growing
Think about this. The $50 per month out-of-pocket cost of GLP-1 drugs for Americans (with insurer paying the balance of $2000 pm) is more than double the entire cost for Indians (less than ₹2000pm for Semaglutide). Overall cost difference? 100 times. Not 100%.
🚨 ARE PEPTIDES THE NEXT "GLP-1 MOMENT" FOR $HIMS? @eli_dorf: "I think there's something really interesting that happened with GLP-1s, which is that when someone's taking Ozempic or any GLP-1, they LOVE to talk about it." "They'll talk about it to their friends. They'll talk htt
Created 2026-05-24 · Updated 2026-05-24
Edit rule text
[generalist_macro_or_labor_market_commentary_without_healthcare_systems_lens]
Learned 3 rejections Active
Exclude posts that frame AI or business trends as broad labor market or macro-economic commentary (e.g., 'AI making workers more productive', 'software becoming free', 'humanoid robots replacing caregivers') even if healthcare is mentioned, unless the post analyzes healthcare-specific labor challenges (staffing shortages, credential requirements, rural provider shortages) and how the trend uniquely impacts healthcare system operations.
Posts discussing labor disruption, workforce automation, or macro-economic trends with healthcare mentioned only as one example among many.
3 example posts
“Rural hospitals cannot make it on their own.” Nobody makes it on their own. Businesses need capital, strategy, discipline, operating systems, good finance, clean accounting, payer strategy, physician alignment, and leaders who know what business they are actually in. The
If you want to inject $1B into healthcare for the truly vulnerable, it costs the state $500M. If you want to inject $1B for able-bodied adults, it costs the state $100M. The 90% federal match drastically lowers the "price" of expanding government healthcare. Incentives dictate
Startups that seek to block AI exam fraud, help companies use AI to make consistent decisions, make financial AI transactions safer, and use AI to organize medical device regulatory information took top honors at @Cornell_Tech’s annual Startup Awards on May 14. Nearly 600 https:
Created 2026-05-24 · Updated 2026-05-24
Edit rule text
[forward_deployed_engineering_or_enterprise_ai_adoption_macro_without_healthcare_specificity]
Learned 3 rejections Active
Exclude posts that celebrate forward-deployed engineering, enterprise AI adoption, or workforce productivity trends as generic tech/business commentary unless the post explicitly connects to healthcare provider operations, clinical workflows, or healthcare labor disruption.
Posts about enterprise AI adoption, forward-deployed engineering models, or workforce productivity gains framed broadly without healthcare-specific operational or clinical implications.
3 example posts
Workers say AI is making them more productive. Executives say AI is not making their companies more productive. Both groups are reporting on the same software. The gap between the two numbers is the entire debate about the largest capital expenditure cycle in technology history.
Google's CEO confirmed 75% of new code at Google is AI-generated. October 2024: 25%. Fall 2025: 50%. April 2026: 75%. Doubled in a year. Then doubled again in six months. Engineers aren't writing code anymore. They're reviewing it. The job title is the same. The job
Satya Nadella's energy is something here. 🔥 "Tokens per Dollar per Watt" The new equation for the AI age for every Company or Industry or Country. "And that means Infrastructure, Infrastructure and Infrastructure." https://t.co/McINrBAo4a
Created 2026-05-24 · Updated 2026-05-24
Edit rule text
[glp1_peptide_market_dynamics_and_pricing_without_healthcare_systems_analysis]
Learned 3 rejections Active
Exclude posts that focus on GLP-1 or peptide market pricing, competitive positioning, company revenues, or generic availability announcements unless the post connects to healthcare system behavior, patient access barriers, clinical outcomes, or provider-payor dynamics.
Posts about GLP-1 or peptide market pricing, competition, and financial dynamics without analyzing healthcare delivery, access, or outcomes impact.
3 example posts
For years, grey-market HGH operated in a strange legal grey zone somewhere between “research compound,” “API” and underground enhancement culture. Last month’s FDA guidance suggests regulators may no longer see it that way. The recent decline in HGH quality, and now the growing
Think about this. The $50 per month out-of-pocket cost of GLP-1 drugs for Americans (with insurer paying the balance of $2000 pm) is more than double the entire cost for Indians (less than ₹2000pm for Semaglutide). Overall cost difference? 100 times. Not 100%.
🚨 ARE PEPTIDES THE NEXT "GLP-1 MOMENT" FOR $HIMS? @eli_dorf: "I think there's something really interesting that happened with GLP-1s, which is that when someone's taking Ozempic or any GLP-1, they LOVE to talk about it." "They'll talk about it to their friends. They'll talk htt
Created 2026-05-24 · Updated 2026-05-24
Edit rule text
[market_pricing_dynamics_or_competitive_positioning_without_systems_analysis]
Learned 3 rejections Active
Exclude posts that focus on market pricing, cost comparisons, competitive positioning (e.g., price per unit, generic vs. branded competition), or business model debates without connecting to healthcare access, equity, reimbursement structures, or operational delivery. Market analysis alone is insufficient.
Posts analyzing pharmaceutical or healthcare market pricing, competition, or business positioning without healthcare systems implications
3 example posts
Think about this. The $50 per month out-of-pocket cost of GLP-1 drugs for Americans (with insurer paying the balance of $2000 pm) is more than double the entire cost for Indians (less than ₹2000pm for Semaglutide). Overall cost difference? 100 times. Not 100%.
Hims & Hers $HIMS Offered $300M Convertable Senior Notes to 2032 💊 Buyers will have the option to add an additional $45M They plan to use the cash for: Expansions, AI, acquisitions, and infrastructure! https://t.co/zaZHO80bY2
^Humira’s patent expired a while back for some context there. The brand premium is the moat now. Ozempic is not generic semaglutide in patient and prescriber perception. Novo keeps charging brand pricing on every loyal patient without a PMPRB ceiling.
Created 2026-05-23 · Updated 2026-05-23
Edit rule text
[broad_health_claim_without_healthcare_systems_context]
Learned 3 rejections Active
Exclude posts that present single clinical observations, broad health claims (e.g., 'LDL targets have changed'), or individual case studies without connecting to healthcare delivery systems, reimbursement, equity, or operational barriers affecting implementation.
Posts making broad health claims or clinical observations about disease, treatment, or health behavior without healthcare systems analysis or operational context.
3 example posts
The healthy LDL number has been quietly moving its own goalposts for forty years: - 1988: under 160 - 1993: under 130 - 2001: under 100 - 2004: under 70 for the high-risk - 2019: under 55 for the very-high-risk - Current trajectory: as low as possible, indefinitely The science
A 2016 BMJ study by researchers at Johns Hopkins estimated that medical error is the third leading cause of death in the United States. Behind heart disease. Behind cancer. Ahead of stroke. Ahead of respiratory disease. Ahead of accidents. Ahead of diabetes. Ahead of
Older people can no longer afford to live in California New research finds that older people here in California face the highest risk of going bankrupt in the entire country The risk score was a 69 out of 100 based on factors including healthcare expenses and the high cost of l
Created 2026-05-23 · Updated 2026-05-23
Edit rule text
[broad_health_claim_or_clinical_observation_without_systems_context]
Learned 3 rejections Active
Exclude posts that cite a single study, clinical trial result, or health statistic (e.g., LDL targets, medical error rankings, pancreatic lesion prevalence) as isolated fact without connecting to healthcare systems, clinical workflows, implementation barriers, or policy implications.
Posts presenting single clinical findings, epidemiological facts, or broad health claims without healthcare delivery or policy context.
3 example posts
An AI foundation whole-body 3D model that assesses perturbations (such as obesity) across multiple systems (such as immune, neural) at the cell level. This is MouseMapper. Imagine HumanMapper someday @Nature @erturklab https://t.co/jsg3VMzOqg https://t.co/tktuINis2W
We have hundreds of PanINs in our pancreas, especially after age 40. They are a pre-cancerous lesion, but it is fortunately rare for one to progress to cancer (PDAC). Now we know a reason for that. The microenvironment. @umichmedicine @UMRogelCancer https://t.co/HR5a3nZBng https
How AI helped treat a newborn’s ultra rare disease. ‘It was almost like a light switch.’ An AI tool, Biomedical Data Translator, helped doctors at Mayo Clinic find a treatment for Jorie Kraus https://t.co/3qs1WIDquU
Created 2026-05-23 · Updated 2026-05-23
Edit rule text
[founder_enthusiasm_without_systems_validation]
Learned 3 rejections Active
Exclude posts that report founder enthusiasm, lab updates, startup announcements, or company progress in biotech/AI without demonstrating clinical impact, healthcare system integration, or validated patient outcomes. Posts should center on healthcare implementation, not founder narratives or company milestones.
Posts celebrating biotech founder announcements, lab updates, or company milestones without healthcare systems context or clinical validation.
3 example posts
Really amazing story of drug repurposing for rare disease using AI found Clonazepam and worked!! (NCATS federated learning and a determined team made this work) 🚀 👏 @MayoClinic @ncats_nih_gov @statnews #STATBreakthrough https://t.co/D4a4IMXSwh
Andrej Karpathy just explained the future of software engineering without directly saying it. The best AI engineers are no longer “prompting.” They’re building systems around the agents. Karpathy’s biggest insight wasn’t: “Claude can code.” It was: LLMs become dramatically be
Streamlining the synthesis of valuable amines, researchers in Science present a new method that can selectively insert nitrogen into specific carbon–hydrogen bonds. According to the study, the approach could simplify drug development by enabling more efficient, scalable, and ht
Created 2026-05-22 · Updated 2026-05-22
Edit rule text
[broad_market_dynamics_or_policy_debate_without_healthcare_systems_analysis]
Learned 3 rejections Active
Exclude posts that make sweeping claims about healthcare costs, government policy, or market dynamics (e.g., 'healthcare is expensive because of grift,' 'rural hospitals can't survive,' 'hospitals exploit patients') without evidence, healthcare data, or analysis of how these dynamics affect clinical access, quality, or outcomes.
Posts about healthcare policy, pricing, reimbursement, or competitive dynamics that present broad claims or moral arguments without evidence of systemic impact or healthcare delivery consequences
3 example posts
“Rural hospitals cannot make it on their own.” Nobody makes it on their own. Businesses need capital, strategy, discipline, operating systems, good finance, clean accounting, payer strategy, physician alignment, and leaders who know what business they are actually in. The
If you want to inject $1B into healthcare for the truly vulnerable, it costs the state $500M. If you want to inject $1B for able-bodied adults, it costs the state $100M. The 90% federal match drastically lowers the "price" of expanding government healthcare. Incentives dictate
Voice interaction, generating actions,and next, generating tasks...? The humanoid assistant is closer. For those who are bedridden due to illness or have limited mobility, it will serve as an excellent aid in daily living.
Created 2026-05-22 · Updated 2026-05-22
Edit rule text
[unvalidated_peptide_drug_market_claims]
Learned 3 rejections Active
Exclude posts that promote, speculate about, or make unsubstantiated claims regarding peptides, GLP-1 medications, semaglutide, tirzepatide, or compounded drug products—especially when framed as market opportunities, lifestyle hacks, or sourcing advice without peer-reviewed clinical data or FDA oversight discussion.
Posts making unverified or speculative claims about peptides, GLP-1s, or compounded drugs without clinical evidence or regulatory validation
3 example posts
🚨 ARE PEPTIDES THE NEXT "GLP-1 MOMENT" FOR $HIMS? @eli_dorf: "I think there's something really interesting that happened with GLP-1s, which is that when someone's taking Ozempic or any GLP-1, they LOVE to talk about it." "They'll talk about it to their friends. They'll talk htt
This is what the peptide space needs more of: clinicians with actual patient volume, public reasoning, and accountability. The next layer is transparency around sourcing, testing, and outcomes. Peptides are not a vibes category anymore. The signal is getting cleaner.
^Humira’s patent expired a while back for some context there. The brand premium is the moat now. Ozempic is not generic semaglutide in patient and prescriber perception. Novo keeps charging brand pricing on every loyal patient without a PMPRB ceiling.
Created 2026-05-22 · Updated 2026-05-22
Edit rule text
[founder_enthusiasm_or_lab_update_without_systems_validation]
Learned 3 rejections Active
Exclude posts that quote or amplify founder enthusiasm about AI, biotech, or healthcare startup achievements (e.g., Isomorphic Labs fundraising, Profluent protein synthesis, Deepgram voice AI for healthcare) without independent evidence, peer review, or analysis of how the technology will integrate into healthcare delivery, reimbursement, or regulatory ecosystems.
Posts amplifying startup founder excitement, lab achievements, or company partnership announcements without independent validation or healthcare system impact analysis
3 example posts
Most valuable AI startups in Europe 🇩🇪 Aleph Alpha - $20B 🇫🇷 Mistral AI - $14B 🇩🇪 Celonis - $13B 🇩🇪 Helsing - $12B 🇫🇷 Poolside AI - $12B 🇵🇱 ElevenLabs - $11B 🇮🇹 Bending Spoons - $11B 🇬🇧 Wayve - $8.6B 🇸🇪 Lovable - $6.6B 🇸🇪 Legora - $5.5B 🇫🇷 Hugging Face - $4.5B 🇬🇧 Synthesia - $4
Google $GOOGL & Blackstone $BX Are in talks of Launching a New AI Company👀🔋 Google will be expected to provide software & services Blackstone will be expected to provide $5B in capital https://t.co/7MFyEfAll9
Fair warning, YC founders: if you take these tokens, there’s a non-zero chance that OpenAI will study exactly what your startup is doing, copy your idea and put your app into their free offering. This is the classic platform playbook — be careful, founders!
Created 2026-05-22 · Updated 2026-05-22
Edit rule text
[broad_health_claims_or_policy_debate_without_nuance]
Learned 3 rejections Active
Exclude posts that make broad assertions about healthcare problems (e.g., 'medical error is the third leading cause of death,' 'LDL goalposts keep moving,' 'healthcare is expensive because of grifters') presented as simple facts or moral critiques without examining competing evidence, guideline rationale, economic tradeoffs, or the nuanced systems driving the claim.
Posts making sweeping claims about healthcare costs, medical error, or treatment guidelines without acknowledging complexity or presenting balanced evidence
3 example posts
Why is healthcare expensive? The grifters gonna grift. ASCO is asking for oncology practices to be admitted into the same 340B machinery that helped health systems acquire physician practices for two decades. Read the tell. The proposed Indigent Care Ratio lets community
The healthy LDL number has been quietly moving its own goalposts for forty years: - 1988: under 160 - 1993: under 130 - 2001: under 100 - 2004: under 70 for the high-risk - 2019: under 55 for the very-high-risk - Current trajectory: as low as possible, indefinitely The science
The gap between best-case and real-world GDMT in #HFrEF is striking. GWTG-HF 2024 vs EPIC COSMOS 2023–2025 (all HFrEF pts, 3.3M patients): 💚 Beta-Blocker: 95% vs 73% 💚 ACE/ARB/ARNI: 92% vs 67% 💚 MRA: 81% vs 36% 💚 SGLT2i: 78% vs 34% 💚 Quadruple GDMT: 68% vs 18% https://t.co/xTXdB
Created 2026-05-22 · Updated 2026-05-22
Edit rule text
[healthcare_business_financial_metrics_without_operational_substance]
Learned 3 rejections Active
Exclude posts reporting on company fundraising rounds, convertible notes, IPO milestones, stock performance, app store rankings, revenue announcements, or business model debates (e.g., SaaS defensibility) unless the post explains what operational or clinical problem the business model solves or how financial structure impacts healthcare access/quality.
Posts about healthcare company fundraising, valuations, earnings, or business model debates without addressing operational or delivery challenges.
3 example posts
Hims & Hers $HIMS Offered $300M Convertable Senior Notes to 2032 💊 Buyers will have the option to add an additional $45M They plan to use the cash for: Expansions, AI, acquisitions, and infrastructure! https://t.co/zaZHO80bY2
Google $GOOGL & Blackstone $BX Are in talks of Launching a New AI Company👀🔋 Google will be expected to provide software & services Blackstone will be expected to provide $5B in capital https://t.co/7MFyEfAll9
How to print money, state government edition: 1) Tax your hospitals and managed care plans. 2) Use that tax revenue to pay your 10% share of Medicaid expansion. 3) Trigger a 90% federal match. 4) Return the original tax money, plus the massive federal windfall, back to the health
Created 2026-05-22 · Updated 2026-05-22
Edit rule text
[ai_infrastructure_hype_without_healthcare_specificity]
Learned 3 rejections Active
Exclude posts that discuss AI technical trends (agentic systems, tokenomics, forward-deployed engineering, code generation, vulnerability disclosure) without clear connection to a specific healthcare application, workflow, or business model. The post must show how the AI capability solves a healthcare problem.
Posts about AI model capabilities, compute costs, or agent architecture that apply only tangentially to healthcare.
3 example posts
#AI agents need more than a sandbox—they need predictable execution. Learn how OpenShell provides a standard interface to intercept, inspect, and approve LLM-generated bash commands before they run: https://t.co/JKmLMgl1A9 https://t.co/xWi8oBNZ7T
Andrej Karpathy just explained the future of software engineering without directly saying it. The best AI engineers are no longer “prompting.” They’re building systems around the agents. Karpathy’s biggest insight wasn’t: “Claude can code.” It was: LLMs become dramatically be
The reason why I release my X articles about AI agents (fat skill fat code thin harness) and GStack and GBrain is that we, yes you and I, can have *PROCESS POWER*, which is the one super powerful specific moat that anyone can create for themselves. The agent helps you do it. htt
Created 2026-05-22 · Updated 2026-05-22
Edit rule text
[generalist_ai_or_macro_tangent_loose_healthcare_framing]
Learned 3 rejections Active
Exclude posts that are primarily about AI model architecture, software engineering trends, data infrastructure, or general macro trends (AI code generation, SaaS economics, startup capital efficiency) that happen to mention healthcare or a healthcare company but do not analyze healthcare-specific challenges.
Posts about general AI, infrastructure, software engineering, or macro economics tangentially labeled as healthcare-relevant.
3 example posts
Startups that seek to block AI exam fraud, help companies use AI to make consistent decisions, make financial AI transactions safer, and use AI to organize medical device regulatory information took top honors at @Cornell_Tech’s annual Startup Awards on May 14. Nearly 600 https:
Voice interaction, generating actions,and next, generating tasks...? The humanoid assistant is closer. For those who are bedridden due to illness or have limited mobility, it will serve as an excellent aid in daily living.
Caregivers do some of the most important work in dementia care — almost none of it is visible to the health care system. Dr. Rosemary Laird describes a navigator model where caregivers reported 86% fewer falls and 81% fewer hospitalizations, over a 12-month pilot. https://t.co/sG
Created 2026-05-22 · Updated 2026-05-22
Edit rule text
[broad_unvalidated_health_claim_or_market_speculation_without_systems_analysis]
Learned 3 rejections Active
Exclude posts that make sweeping claims about healthcare trends (e.g., healthcare costs, brand premiums, LDL guidelines moving goalposts, provider behavior) or debate market dynamics (e.g., whether hospitals are closing for good reasons) without rigorous evidence or systems-level analysis.
Posts making broad health claims, market debates, or industry trend assertions without evidence or healthcare systems lens
3 example posts
“Rural hospitals cannot make it on their own.” Nobody makes it on their own. Businesses need capital, strategy, discipline, operating systems, good finance, clean accounting, payer strategy, physician alignment, and leaders who know what business they are actually in. The
If you want to inject $1B into healthcare for the truly vulnerable, it costs the state $500M. If you want to inject $1B for able-bodied adults, it costs the state $100M. The 90% federal match drastically lowers the "price" of expanding government healthcare. Incentives dictate
A senator complaining about healthcare costs should have to answer one question first: Why does your state make new healthcare competitors ask permission before opening? If yes, start there. Speeches do not lower prices. Supply does.
Created 2026-05-22 · Updated 2026-05-22
Edit rule text
[speculative_robotics_or_sci_fi_labor_hype_without_healthcare_validation]
Learned 3 rejections Active
Exclude posts that hype speculative robotics, humanoid assistants, voice interaction systems, or futuristic labor-replacement technologies in healthcare contexts without clinical evidence, regulatory approval, or demonstrated real-world deployment.
Posts speculating about humanoid robots, voice AI, or futuristic automation in healthcare without evidence of clinical validation or operational readiness
3 example posts
Voice interaction, generating actions,and next, generating tasks...? The humanoid assistant is closer. For those who are bedridden due to illness or have limited mobility, it will serve as an excellent aid in daily living.
Between rooms. On rounds. Walking the corridor outside an OR. Charting one-handed during a phone call. This is where clinical questions happen. Today we're launching Voice Mode. OpenEvidence is the first multimodal medical AI: physicians can type, speak, or listen, on the same h
Most voice AI hears what was said. Velma hears what was meant. Tone. Intent. Emotion. Deception. The standard stack - STT to transcribe, LLM to analyse - throws all of it away before analysis even begins. The signal was lost the moment audio became text... (🧵) https://t.co/DtN
Created 2026-05-22 · Updated 2026-05-22
Edit rule text
[generalist_ai_or_infrastructure_hype_without_healthcare_application]
Learned 3 rejections Active
Exclude posts that discuss AI model capabilities, software engineering patterns, compute efficiency, or company metrics (e.g., 'Google generates 75% AI code', 'tokenmaxxing at Anthropic', 'forward deployed engineering') where healthcare is mentioned only peripherally or the post is fundamentally about AI/tech industry dynamics, not healthcare delivery.
Posts about AI infrastructure, compute capabilities, software engineering trends, or AI company business metrics with only tangential healthcare relevance
3 example posts
Andrej Karpathy just explained the future of software engineering without directly saying it. The best AI engineers are no longer “prompting.” They’re building systems around the agents. Karpathy’s biggest insight wasn’t: “Claude can code.” It was: LLMs become dramatically be
Most valuable AI startups in Europe 🇩🇪 Aleph Alpha - $20B 🇫🇷 Mistral AI - $14B 🇩🇪 Celonis - $13B 🇩🇪 Helsing - $12B 🇫🇷 Poolside AI - $12B 🇵🇱 ElevenLabs - $11B 🇮🇹 Bending Spoons - $11B 🇬🇧 Wayve - $8.6B 🇸🇪 Lovable - $6.6B 🇸🇪 Legora - $5.5B 🇫🇷 Hugging Face - $4.5B 🇬🇧 Synthesia - $4
Workers say AI is making them more productive. Executives say AI is not making their companies more productive. Both groups are reporting on the same software. The gap between the two numbers is the entire debate about the largest capital expenditure cycle in technology history.
Created 2026-05-22 · Updated 2026-05-22
Edit rule text
[speculative_robotics_sci_fi_labor_disruption]
Learned 3 rejections Active
Exclude posts that speculate about future healthcare labor displacement or new clinical workflows enabled by humanoid robots, advanced voice AI, or speculative autonomous agents—without demonstrating current clinical validation, deployment, or healthcare systems analysis.
Posts speculating about humanoid robots, voice AI, or sci-fi labor automation in healthcare without validation or specificity.
3 example posts
Voice interaction, generating actions,and next, generating tasks...? The humanoid assistant is closer. For those who are bedridden due to illness or have limited mobility, it will serve as an excellent aid in daily living.
Between rooms. On rounds. Walking the corridor outside an OR. Charting one-handed during a phone call. This is where clinical questions happen. Today we're launching Voice Mode. OpenEvidence is the first multimodal medical AI: physicians can type, speak, or listen, on the same h
Most voice AI hears what was said. Velma hears what was meant. Tone. Intent. Emotion. Deception. The standard stack - STT to transcribe, LLM to analyse - throws all of it away before analysis even begins. The signal was lost the moment audio became text... (🧵) https://t.co/DtN
Created 2026-05-22 · Updated 2026-05-22
Edit rule text
[generalist_macro_infrastructure_hype_no_healthcare_application]
Learned 3 rejections Active
Exclude posts discussing general AI infrastructure metrics (tokens per dollar per watt, AI code generation rates, GPU scaling, software commoditization) that happen to mention healthcare or a healthcare company but are primarily about generic AI/tech trends rather than how those trends specifically reshape healthcare delivery, operations, or outcomes.
Posts about AI infrastructure, compute, tokenomics, or macro trends mentioned in healthcare context but with no specific healthcare application
3 example posts
Andrej Karpathy just explained the future of software engineering without directly saying it. The best AI engineers are no longer “prompting.” They’re building systems around the agents. Karpathy’s biggest insight wasn’t: “Claude can code.” It was: LLMs become dramatically be
Workers say AI is making them more productive. Executives say AI is not making their companies more productive. Both groups are reporting on the same software. The gap between the two numbers is the entire debate about the largest capital expenditure cycle in technology history.
The reason why I release my X articles about AI agents (fat skill fat code thin harness) and GStack and GBrain is that we, yes you and I, can have *PROCESS POWER*, which is the one super powerful specific moat that anyone can create for themselves. The agent helps you do it. htt
Created 2026-05-22 · Updated 2026-05-22
Edit rule text
[founder_enthusiasm_or_lab_update_without_healthcare_systems_context]
Learned 3 rejections Active
Exclude posts that showcase founder enthusiasm, lab announcements, or capability demonstrations (e.g., 'just reported', 'coming presentations', 'marvel') without clear evidence of healthcare system impact, clinical validation, or operational insight into how the innovation addresses healthcare delivery challenges.
Posts featuring biotech/AI founder excitement or lab updates that lack healthcare systems insight or clinical validation
3 example posts
$ABCL Just now @ BoA - @AbCelleraBio CEO Carl Hansen when asked about what the market is overlooking: “I believe we now have best in class or best in world capabilities for some very important specific applications in therapeutic antibodies. So GPCRs and ion channels has been a
🚨! $SANA has just reported - in its Q1 ER, that it plans to expand its pipeline by advancing a new preclinical program - SG227. SG227 is a CD8-targeted fusosome that by delivering a genetic material to make BCMA-directed CAR T cells could act as a potential treatment for Cancer h
$BEAM's coming presentations May 15: ESCAPE June 13: BEAM-101 EHA update 2026 milestones BEAM-101 (riso-cel) BLA BEAM-301 initial data BEAM-302 updated data and pivotal trial initiation BEAM-304 IND filing https://t.co/9Nvq3PKMMl
Created 2026-05-21 · Updated 2026-05-21
Edit rule text
[political_outrage_or_scandal_without_healthcare_analysis]
Learned 3 rejections Active
Exclude posts that frame healthcare fraud, regulatory failures, or corporate scandals primarily through political outrage or vigilante moral justification without substantive analysis of how the issue affects healthcare delivery, outcomes, or operations.
Posts highlighting political figures, regulatory scandals, or corporate wrongdoing with moral outrage but no healthcare systems analysis.
3 example posts
Remember when UnitedHealth's CEO was assassinated in 2024? 18 months later, every accusation the killer Luigi Mangione made has been proven RIGHT. UnitedHealth is one of the biggest corporate collapses in recent years. And they deserve it. In December 2024, UnitedHealthcare CE
600 new generic drugs added to TrumpRx this week thanks to President @realDonaldTrump's leadership! Before you visit your local pharmacy, visit https://t.co/dDMdsILHTj to ensure you're getting the BEST price for your prescriptions. It's easy and saves YOU money! https://t.co/
Older people can no longer afford to live in California New research finds that older people here in California face the highest risk of going bankrupt in the entire country The risk score was a 69 out of 100 based on factors including healthcare expenses and the high cost of l
Created 2026-05-21 · Updated 2026-05-21
Edit rule text
[non_healthcare_business_or_tech_with_loose_healthcare_framing]
Learned 3 rejections Active
Exclude posts that are primarily about generalist AI business models, enterprise software adoption, infrastructure investment, or tech partnerships that mention healthcare tangentially or use healthcare labels without explaining healthcare-specific value, regulatory considerations, or system impact (e.g., 'web agency with AI agents', 'Google & Blackstone AI company', 'enterprise AI adoption').
Posts about general business trends, AI infrastructure, corporate partnerships, or non-healthcare tech that use healthcare keywords but lack substantive healthcare systems analysis.
3 example posts
Google $GOOGL & Blackstone $BX Are in talks of Launching a New AI Company👀🔋 Google will be expected to provide software & services Blackstone will be expected to provide $5B in capital https://t.co/7MFyEfAll9
Most voice AI hears what was said. Velma hears what was meant. Tone. Intent. Emotion. Deception. The standard stack - STT to transcribe, LLM to analyse - throws all of it away before analysis even begins. The signal was lost the moment audio became text... (🧵) https://t.co/DtN
Google's CEO confirmed 75% of new code at Google is AI-generated. October 2024: 25%. Fall 2025: 50%. April 2026: 75%. Doubled in a year. Then doubled again in six months. Engineers aren't writing code anymore. They're reviewing it. The job title is the same. The job
Created 2026-05-21 · Updated 2026-05-21
Edit rule text
[conspiracy_or_moral_justification_without_analysis]
Learned 3 rejections Active
Exclude posts that promote conspiracy narratives (e.g., 'every accusation the killer made has been proven RIGHT'), moral justification of vigilante violence, or sensationalized claims about fraud without structural healthcare systems evidence. These lack the analytical depth required for healthcare tech strategy.
Posts that amplify conspiracy theories, unsubstantiated claims, or moral justifications for violence without substantive healthcare systems analysis.
3 example posts
Remember when UnitedHealth's CEO was assassinated in 2024? 18 months later, every accusation the killer Luigi Mangione made has been proven RIGHT. UnitedHealth is one of the biggest corporate collapses in recent years. And they deserve it. In December 2024, UnitedHealthcare CE
Vivek Ramaswamy put out a lot of statements today referencing the fraud in Ohio. Swampy is of course, involved in Ohio Medicaid. Datavant, Vivek's spinoff company from Riovant, handles release of information for many hospitals and providers in Ohio. Ohio Medicaid providers, man
⚡️This is the moment AI stops being a productivity tool and becomes a replacement architecture for elite cognitive labor. Citadel is not a random corporation automating low-level admin. It is one of the most competitive intelligence machines in finance. The work Griffin is descr
Created 2026-05-21 · Updated 2026-05-21
Edit rule text
[tangential_founder_enthusiasm_without_validation]
Learned 3 rejections Active
Exclude posts from founders or company accounts announcing new programs, pipeline milestones, preclinical programs, or technical achievements (e.g., 'we're advancing SG227,' 'BEAM-301 coming') without published clinical data, regulatory pathway clarity, or evidence of healthcare provider or payer traction.
Posts from biotech or AI founders announcing lab updates, pipeline expansions, or capability claims without clinical evidence or healthcare systems context.
3 example posts
$ABCL Just now @ BoA - @AbCelleraBio CEO Carl Hansen when asked about what the market is overlooking: “I believe we now have best in class or best in world capabilities for some very important specific applications in therapeutic antibodies. So GPCRs and ion channels has been a
🚨! $SANA has just reported - in its Q1 ER, that it plans to expand its pipeline by advancing a new preclinical program - SG227. SG227 is a CD8-targeted fusosome that by delivering a genetic material to make BCMA-directed CAR T cells could act as a potential treatment for Cancer h
$BEAM's coming presentations May 15: ESCAPE June 13: BEAM-101 EHA update 2026 milestones BEAM-101 (riso-cel) BLA BEAM-301 initial data BEAM-302 updated data and pivotal trial initiation BEAM-304 IND filing https://t.co/9Nvq3PKMMl
Created 2026-05-21 · Updated 2026-05-21
Edit rule text
[ai_infrastructure_or_agent_hype_without_healthcare_validation]
Learned 3 rejections Active
Exclude posts that discuss AI model capabilities (Claude, OpenAI, reasoning, coding agents), infrastructure breakthroughs, or agent technical features without demonstrating healthcare-specific application, validation, or adoption evidence. Claims about 'agents replacing cognitive work' or 'AI breaking software' are too broad without healthcare context.
Posts about AI infrastructure, agent capabilities, or technical achievements with loose or no healthcare application specificity.
3 example posts
Yesterday, I wrote about how long-running agentic systems need to be more reliable, robust, and fault-tolerant; and today, let's dig deeper into how. Fun fact - distributed workflows are one of those problems that almost every engineering team accidentally rebuilds. We start wi
Google's CEO confirmed 75% of new code at Google is AI-generated. October 2024: 25%. Fall 2025: 50%. April 2026: 75%. Doubled in a year. Then doubled again in six months. Engineers aren't writing code anymore. They're reviewing it. The job title is the same. The job
Satya Nadella's energy is something here. 🔥 "Tokens per Dollar per Watt" The new equation for the AI age for every Company or Industry or Country. "And that means Infrastructure, Infrastructure and Infrastructure." https://t.co/McINrBAo4a
Created 2026-05-21 · Updated 2026-05-21
Edit rule text
[shallow_business_model_or_market_speculation]
Learned 3 rejections Active
Exclude posts that make sweeping claims about business defensibility, market opportunities, or competitive advantages (e.g., 'only three moats left,' 'software will be free,' 'agents are the next big thing') without specific healthcare application validation, evidence of adoption, or systems-level analysis.
Posts making broad claims about business models, market dynamics, or competitive positioning without substantive evidence or healthcare systems insight.
3 example posts
Workers say AI is making them more productive. Executives say AI is not making their companies more productive. Both groups are reporting on the same software. The gap between the two numbers is the entire debate about the largest capital expenditure cycle in technology history.
Anthropic CEO Dario Amodei on SaaS: "Software is going to become cheap, maybe essentially free. The premise that you need to amortize a piece of software you build across millions of users, that may start to be false. But at the same time, there are whole jobs, whole careers
imagine a software startup raising $800m before their first dollar of revenue. Can’t imagine it the last two decades of software has been dominated by a simple theory: ship quickly, get customers early, generate revenue quickly to validate PMF, manage KPIs closely, etc. This bec
Created 2026-05-21 · Updated 2026-05-21
Edit rule text
[conspiratorial_or_sensationalized_claims]
Learned 3 rejections Active
Exclude posts that make inflammatory allegations (e.g., CEO assassinations, fraud conspiracies, vigilante violence justification) or sensationalized claims without credible evidence, systems-level analysis, or substantive healthcare policy argument.
Posts making unsubstantiated claims about fraud, conspiracies, or moral justifications for violence without healthcare systems analysis.
3 example posts
Remember when UnitedHealth's CEO was assassinated in 2024? 18 months later, every accusation the killer Luigi Mangione made has been proven RIGHT. UnitedHealth is one of the biggest corporate collapses in recent years. And they deserve it. In December 2024, UnitedHealthcare CE
Vivek Ramaswamy put out a lot of statements today referencing the fraud in Ohio. Swampy is of course, involved in Ohio Medicaid. Datavant, Vivek's spinoff company from Riovant, handles release of information for many hospitals and providers in Ohio. Ohio Medicaid providers, man
How does a brand like this even get listed at Target? Does Target only care about turnover? Might as well start selling snake oil too. By the way, this is the SKU the brand got in trouble with. They now call it "Berberine Patches". It was called GLP-1 Patches before. Check the h
Created 2026-05-21 · Updated 2026-05-21
Edit rule text
[tangential_ai_capability_or_company_announcement_without_healthcare_specificity]
Learned 3 rejections Active
Exclude posts that announce AI model capabilities (Claude Code, GPT-Rosalind, Mythos Preview), cybersecurity vulnerabilities, or company partnerships without demonstrating actual healthcare application, clinical validation, or healthcare-specific business context. Posts must show healthcare use, not just list the tool.
Posts about AI model capabilities, security vulnerabilities, or company product launches presented as healthcare-relevant but lacking healthcare-specific application or validation.
3 example posts
Fair warning, YC founders: if you take these tokens, there’s a non-zero chance that OpenAI will study exactly what your startup is doing, copy your idea and put your app into their free offering. This is the classic platform playbook — be careful, founders!
Yesterday, I wrote about how long-running agentic systems need to be more reliable, robust, and fault-tolerant; and today, let's dig deeper into how. Fun fact - distributed workflows are one of those problems that almost every engineering team accidentally rebuilds. We start wi
Most voice AI hears what was said. Velma hears what was meant. Tone. Intent. Emotion. Deception. The standard stack - STT to transcribe, LLM to analyse - throws all of it away before analysis even begins. The signal was lost the moment audio became text... (🧵) https://t.co/DtN
Created 2026-05-21 · Updated 2026-05-21
Edit rule text
[business_model_or_market_speculation]
Learned 3 rejections Active
Exclude posts that speculate about software pricing, SaaS defensibility, cost structures, or competitive dynamics (e.g., 'software will become free,' vendor lock-in debates) without connecting to healthcare outcomes or system-level impact.
Posts that debate business models, pricing dynamics, or market competition without healthcare systems analysis.
3 example posts
Anthropic CEO Dario Amodei on SaaS: "Software is going to become cheap, maybe essentially free. The premise that you need to amortize a piece of software you build across millions of users, that may start to be false. But at the same time, there are whole jobs, whole careers
imagine a software startup raising $800m before their first dollar of revenue. Can’t imagine it the last two decades of software has been dominated by a simple theory: ship quickly, get customers early, generate revenue quickly to validate PMF, manage KPIs closely, etc. This bec
Things every AI app startup says today to justify their defensibility: 1. We support multiple models. Our customers do not want to lock in to one vendor. 2. We have a data moat. We post-train open-source models to be much better and cheaper than closed-source. 3. We do deep inte
Created 2026-05-21 · Updated 2026-05-21
Edit rule text
[broad_tech_or_macro_trend_with_loose_healthcare_framing]
Learned 3 rejections Active
Exclude posts that discuss AI infrastructure costs, AI-generated code adoption, token economics, labor market disruption from AI, or competitive business models (SaaS pricing, defensibility moats, agent deployment) that mention healthcare only in passing or as one example domain. The healthcare system angle must be central, not incidental.
Posts about generalist AI infrastructure, labor market macro trends, or non-healthcare business models loosely tagged as healthcare-relevant
3 example posts
Google's CEO confirmed 75% of new code at Google is AI-generated. October 2024: 25%. Fall 2025: 50%. April 2026: 75%. Doubled in a year. Then doubled again in six months. Engineers aren't writing code anymore. They're reviewing it. The job title is the same. The job
Anthropic CEO Dario Amodei on SaaS: "Software is going to become cheap, maybe essentially free. The premise that you need to amortize a piece of software you build across millions of users, that may start to be false. But at the same time, there are whole jobs, whole careers
The 36 BIGGEST startup opportunities right now 1. biggest b2c: solving loneliness. third spaces, community apps, IRL 2. biggest b2b: managed AI employees for businesses 3. biggest overlooked: elder tech. 70 million boomers who want products that make them happier & healthier
Created 2026-05-21 · Updated 2026-05-21
Edit rule text
[generalist_ai_infrastructure_or_compute_hype]
Learned 3 rejections Active
Exclude posts about general AI infrastructure, model performance, compute efficiency, tokenomics, or frontier model capabilities that lack specific healthcare context or clinical validation. Posts must demonstrate healthcare system or patient impact, not just AI technical advancement.
Posts that discuss AI model capabilities, infrastructure, or compute efficiency without direct healthcare application or validation
3 example posts
“The [AI] tool presented here will enable rapid screening for multiple systemic diseases using retinal photographs, and it is a step forward in the evolution of oculomics from experimental research to real-world clinical practice.” —Editorial Team, @NatureMedicine https://t.co/90
Google's CEO confirmed 75% of new code at Google is AI-generated. October 2024: 25%. Fall 2025: 50%. April 2026: 75%. Doubled in a year. Then doubled again in six months. Engineers aren't writing code anymore. They're reviewing it. The job title is the same. The job
Anthropic CEO Dario Amodei on SaaS: "Software is going to become cheap, maybe essentially free. The premise that you need to amortize a piece of software you build across millions of users, that may start to be false. But at the same time, there are whole jobs, whole careers
Created 2026-05-21 · Updated 2026-05-21
Edit rule text
[broad_market_dynamics_without_healthcare_systems_analysis]
Learned 3 rejections Active
Exclude posts that discuss market dynamics (pricing, competition, labor disruption, adoption rates, or earnings gaps) in healthcare or AI without analyzing how these dynamics affect healthcare delivery, outcomes, cost structure, or clinical practice. Broad market observations without systems-level healthcare insight are insufficient.
Posts about market pricing, competitive dynamics, or labor disruption without healthcare systems context or business model insight
3 example posts
Workers say AI is making them more productive. Executives say AI is not making their companies more productive. Both groups are reporting on the same software. The gap between the two numbers is the entire debate about the largest capital expenditure cycle in technology history.
$HIMS $HERS 🚨 HERS IS NOW BEATING BOTH HIMS AND RO ON THE iOS APP STORE Today's rankings in the Medical category: - Hers: #9 - Ro: #10 - Hims: #18 Chart below = weekly avg rank https://t.co/vlv80yiFOo
Older people can no longer afford to live in California New research finds that older people here in California face the highest risk of going bankrupt in the entire country The risk score was a 69 out of 100 based on factors including healthcare expenses and the high cost of l
Created 2026-05-21 · Updated 2026-05-21
Edit rule text
[conspiracy_or_sensationalized_claims]
Learned 3 rejections Active
Exclude posts that promote conspiracy theories, make moral justifications for violence, or sensationalize healthcare scandals (e.g., CEO assassination, fraud claims) without rigorous evidence, systemic analysis, or credible sourcing.
Posts that make unsubstantiated conspiracy claims or sensationalize healthcare events without evidence-based analysis
2 example posts
Remember when UnitedHealth's CEO was assassinated in 2024? 18 months later, every accusation the killer Luigi Mangione made has been proven RIGHT. UnitedHealth is one of the biggest corporate collapses in recent years. And they deserve it. In December 2024, UnitedHealthcare CE
Vivek Ramaswamy put out a lot of statements today referencing the fraud in Ohio. Swampy is of course, involved in Ohio Medicaid. Datavant, Vivek's spinoff company from Riovant, handles release of information for many hospitals and providers in Ohio. Ohio Medicaid providers, man
Created 2026-05-20 · Updated 2026-05-21
Edit rule text
[healthcare_business_metrics_without_systems_analysis]
Learned 3 rejections Active
Exclude posts that report app store rankings, market share data, generic business metrics, or company announcements (fundraising, partnerships, product launches) for healthcare companies. Posts must provide analysis of healthcare delivery, payment, or organizational system dynamics to qualify.
Posts about healthcare company financial performance, market positioning, or business metrics without systems-level operational or economic insight.
3 example posts
$HIMS $HERS 🚨 HERS IS NOW BEATING BOTH HIMS AND RO ON THE iOS APP STORE Today's rankings in the Medical category: - Hers: #9 - Ro: #10 - Hims: #18 Chart below = weekly avg rank https://t.co/vlv80yiFOo
Vivek Ramaswamy put out a lot of statements today referencing the fraud in Ohio. Swampy is of course, involved in Ohio Medicaid. Datavant, Vivek's spinoff company from Riovant, handles release of information for many hospitals and providers in Ohio. Ohio Medicaid providers, man
How does a brand like this even get listed at Target? Does Target only care about turnover? Might as well start selling snake oil too. By the way, this is the SKU the brand got in trouble with. They now call it "Berberine Patches". It was called GLP-1 Patches before. Check the h
Created 2026-05-20 · Updated 2026-05-20
Edit rule text
[unvalidated_speculative_drug_peptide_claims]
Learned 3 rejections Active
Exclude posts that discuss unvalidated peptide formulations, counterfeit drugs, DIY manufacturing, or speculative weight-loss interventions without evidence of FDA approval, clinical trials, or legitimate pharmaceutical supply chain validation.
Posts about peptides, GLP-1s, experimental drugs, or unvalidated medical interventions without clinical evidence or regulatory grounding.
3 example posts
Daily Walking Could Be the Secret to Keeping Weight Off For many people struggling with obesity, losing weight is only half the battle. The greater challenge often begins after the initial success: preventing the lost weight from returning. 🔴More than 50% of people who lose http
How does a brand like this even get listed at Target? Does Target only care about turnover? Might as well start selling snake oil too. By the way, this is the SKU the brand got in trouble with. They now call it "Berberine Patches". It was called GLP-1 Patches before. Check the h
“You can test your peptides to see if they’re real, bro. But I just look for the 99.9% on the result the seller gives me. You can look up the QR code upon the lab’s website, so you know it’s legit. It’s foolproof, bro. What’s mass spec, bro?” https://t.co/xzJLJ3hesT
Created 2026-05-20 · Updated 2026-05-20
Edit rule text
[forward_deployed_engineering_hype]
Learned 3 rejections Active
Exclude posts that praise forward-deployed engineering, embedded engineers, or agentic AI workflows as a business model or operational strategy without analyzing specific healthcare delivery challenges, reimbursement implications, or patient outcomes impact.
Posts celebrating forward-deployed engineering models and enterprise AI adoption rhetoric without healthcare systems analysis.
3 example posts
Palantir invented the forward deployed engineer role in 2006. embed an engineer inside the customer. ship production code in their environment. own the technical outcome. Alex Karp on the idea: “forward deployed engineers are stolen from french restaurants.” 20 years later htt
I’m fully forward deployed engineering pilled specifically because AI simply is not the same as software. In software, you deliver a stable piece of technology to a customer and they adopt it and that’s that (extreme over simplification). In AI, you’re delivering something that
As advanced agents move from coding to the rest of knowledge work, it takes a real amount of work and know-how to get right. You need to ensure agents have the right context and data to work with, wire up systems to agents in a safe and secure way, ensure that the agents are htt
Created 2026-05-20 · Updated 2026-05-20
Edit rule text
[market_pricing_and_competition_only]
Learned 3 rejections Active
Exclude posts that report only market pricing dynamics, competitive rankings, generic drug launches, or app store positioning without analyzing healthcare delivery, payer impact, patient access, or system-level implications.
Posts focusing narrowly on drug pricing, market competition, app store rankings, or pharmaceutical market share without healthcare system insights.
3 example posts
$HIMS $HERS 🚨 HERS IS NOW BEATING BOTH HIMS AND RO ON THE iOS APP STORE Today's rankings in the Medical category: - Hers: #9 - Ro: #10 - Hims: #18 Chart below = weekly avg rank https://t.co/vlv80yiFOo
How does a brand like this even get listed at Target? Does Target only care about turnover? Might as well start selling snake oil too. By the way, this is the SKU the brand got in trouble with. They now call it "Berberine Patches". It was called GLP-1 Patches before. Check the h
$ABCL Just now @ BoA - @AbCelleraBio CEO Carl Hansen when asked about what the market is overlooking: “I believe we now have best in class or best in world capabilities for some very important specific applications in therapeutic antibodies. So GPCRs and ion channels has been a
Created 2026-05-20 · Updated 2026-05-20
Edit rule text
[generalist_ai_hype_no_healthcare]
Learned 3 rejections Active
Exclude posts that discuss frontier AI model capabilities (reasoning, coding, cyber security), AI infrastructure (compute, tokens per dollar), or AI company announcements that lack specific healthcare application or healthcare system analysis. Generic AI hype with healthcare label is not sufficient.
Posts about AI models, AI infrastructure, or AI company metrics with only tangential or forced healthcare framing.
3 example posts
Most voice AI hears what was said. Velma hears what was meant. Tone. Intent. Emotion. Deception. The standard stack - STT to transcribe, LLM to analyse - throws all of it away before analysis even begins. The signal was lost the moment audio became text... (🧵) https://t.co/DtN
Google's CEO confirmed 75% of new code at Google is AI-generated. October 2024: 25%. Fall 2025: 50%. April 2026: 75%. Doubled in a year. Then doubled again in six months. Engineers aren't writing code anymore. They're reviewing it. The job title is the same. The job
Anthropic CEO Dario Amodei on SaaS: "Software is going to become cheap, maybe essentially free. The premise that you need to amortize a piece of software you build across millions of users, that may start to be false. But at the same time, there are whole jobs, whole careers
Created 2026-05-20 · Updated 2026-05-20
Edit rule text
[clinical_observation_or_study_without_healthcare_systems_context]
Learned 3 rejections Active
Exclude posts reporting clinical trial results, medical study findings, or individual case observations (e.g., drug efficacy, disease statistics) without explaining how these findings reshape healthcare delivery, economics, or organizational strategy.
Posts about clinical research findings, trial data, or medical observations that lack broader healthcare systems implications
3 example posts
A 2016 BMJ study by researchers at Johns Hopkins estimated that medical error is the third leading cause of death in the United States. Behind heart disease. Behind cancer. Ahead of stroke. Ahead of respiratory disease. Ahead of accidents. Ahead of diabetes. Ahead of
Daily Walking Could Be the Secret to Keeping Weight Off For many people struggling with obesity, losing weight is only half the battle. The greater challenge often begins after the initial success: preventing the lost weight from returning. 🔴More than 50% of people who lose http
Makes an important point here about the role of these short summaries they do this because this is how most doctors digests information if designed appropriately, an LLM should be able to read and appraise a paper and provide better peer review than most clinicians and honestly
Created 2026-05-20 · Updated 2026-05-20
Edit rule text
[tangential_non_healthcare_business_or_macro]
Learned 3 rejections Active
Exclude posts that discuss general business trends (SaaS unit economics, software startup efficiency, AI agency models, hiring practices) or macro policy (Buy American mandates, state Medicaid tax schemes) that lack substantive healthcare systems analysis or clinical relevance.
Posts about general business models, macroeconomic trends, or non-healthcare domains loosely framed with healthcare language
3 example posts
imagine a software startup raising $800m before their first dollar of revenue. Can’t imagine it the last two decades of software has been dominated by a simple theory: ship quickly, get customers early, generate revenue quickly to validate PMF, manage KPIs closely, etc. This bec
The 36 BIGGEST startup opportunities right now 1. biggest b2c: solving loneliness. third spaces, community apps, IRL 2. biggest b2b: managed AI employees for businesses 3. biggest overlooked: elder tech. 70 million boomers who want products that make them happier & healthier
Things every AI app startup says today to justify their defensibility: 1. We support multiple models. Our customers do not want to lock in to one vendor. 2. We have a data moat. We post-train open-source models to be much better and cheaper than closed-source. 3. We do deep inte
Created 2026-05-20 · Updated 2026-05-20
Edit rule text
[clinical_anecdote_or_market_observation_without_systems_context]
Learned 3 rejections Active
Exclude posts that report isolated clinical trial results, single study findings, market pricing dynamics, or app store rankings without analyzing systemic healthcare implications, causation, or actionable insights for healthcare decision-makers.
Posts citing single clinical observations, trial data, or market dynamics without healthcare systems analysis or broader implications.
3 example posts
$HIMS $HERS 🚨 HERS IS NOW BEATING BOTH HIMS AND RO ON THE iOS APP STORE Today's rankings in the Medical category: - Hers: #9 - Ro: #10 - Hims: #18 Chart below = weekly avg rank https://t.co/vlv80yiFOo
Daily Walking Could Be the Secret to Keeping Weight Off For many people struggling with obesity, losing weight is only half the battle. The greater challenge often begins after the initial success: preventing the lost weight from returning. 🔴More than 50% of people who lose http
Makes an important point here about the role of these short summaries they do this because this is how most doctors digests information if designed appropriately, an LLM should be able to read and appraise a paper and provide better peer review than most clinicians and honestly
Created 2026-05-20 · Updated 2026-05-20
Edit rule text
[tangential_tech_or_non_healthcare_domain]
Learned 3 rejections Active
Exclude posts about software engineering practices (e.g., 75% AI-generated code at Google), general SaaS business models, cloud platform updates (SAP API policy, Red Hat/NVIDIA collaboration), or non-healthcare tech startups that merely mention healthcare in passing or use healthcare as a hypothetical example without substantive analysis.
Posts about general technology, software engineering, cloud platforms, or non-healthcare businesses loosely framed with healthcare labels
3 example posts
Google's CEO confirmed 75% of new code at Google is AI-generated. October 2024: 25%. Fall 2025: 50%. April 2026: 75%. Doubled in a year. Then doubled again in six months. Engineers aren't writing code anymore. They're reviewing it. The job title is the same. The job
imagine a software startup raising $800m before their first dollar of revenue. Can’t imagine it the last two decades of software has been dominated by a simple theory: ship quickly, get customers early, generate revenue quickly to validate PMF, manage KPIs closely, etc. This bec
Palantir invented the forward deployed engineer role in 2006. embed an engineer inside the customer. ship production code in their environment. own the technical outcome. Alex Karp on the idea: “forward deployed engineers are stolen from french restaurants.” 20 years later htt
Created 2026-05-20 · Updated 2026-05-20
Edit rule text
[unvalidated_or_speculative_medical_claims_and_fringe_treatments]
Learned 3 rejections Active
Exclude posts that promote or discuss unvalidated peptides, DIY compounded drugs, unproven treatment claims, or fringe medical interventions (including peptide purity hacks, underground GLP-1 sourcing) without credible clinical evidence, FDA approval status, or critical analysis of safety/efficacy.
Posts discussing unvalidated peptides, DIY semaglutide, peptide purity testing, or fringe medical interventions without clinical evidence or regulatory scrutiny.
3 example posts
Daily Walking Could Be the Secret to Keeping Weight Off For many people struggling with obesity, losing weight is only half the battle. The greater challenge often begins after the initial success: preventing the lost weight from returning. 🔴More than 50% of people who lose http
“You can test your peptides to see if they’re real, bro. But I just look for the 99.9% on the result the seller gives me. You can look up the QR code upon the lab’s website, so you know it’s legit. It’s foolproof, bro. What’s mass spec, bro?” https://t.co/xzJLJ3hesT
$NVO $LLY Dr. Reddy’s says it plans to launch a generic version of Ozempic in Canada within days and is also preparing to roll out generic semaglutide tablets in India. The company sees semaglutide as a meaningful future growth driver following patent expiry in India earlier
Created 2026-05-19 · Updated 2026-05-19
Edit rule text
[financial_market_speculation_or_stock_ticker_commentary_without_healthcare_systems_insight]
Learned 3 rejections Active
Exclude posts that comment on stock price movements, market valuations, secondary market activity, or fundraising numbers (e.g., Anthropic valuation changes, biotech stock rallies, market-cap speculation) unless they analyze how the financial event shapes healthcare delivery, pricing, or access.
Posts about stock performance, pre-IPO valuation swings, or market dynamics without connecting to healthcare delivery outcomes or operational substance.
3 example posts
$HIMS $HERS 🚨 HERS IS NOW BEATING BOTH HIMS AND RO ON THE iOS APP STORE Today's rankings in the Medical category: - Hers: #9 - Ro: #10 - Hims: #18 Chart below = weekly avg rank https://t.co/vlv80yiFOo
$ABCL Just now @ BoA - @AbCelleraBio CEO Carl Hansen when asked about what the market is overlooking: “I believe we now have best in class or best in world capabilities for some very important specific applications in therapeutic antibodies. So GPCRs and ion channels has been a
🚨ANTHROPIC SHOCKS PRE-IPO MARKET, ERASES $400B AFTER NEW STOCK POLICY Anthropic’s implied pre-IPO valuation wiped nearly $400 BILLION just hours after the company dropped a BOMBSHELL targeting secondary share sales. Anthropic updated its stock transfer policy to state that any
Created 2026-05-19 · Updated 2026-05-19
Edit rule text
[unvalidated_speculative_peptide_or_drug_claims_without_clinical_evidence]
Learned 3 rejections Active
Exclude posts that make claims about peptide efficacy, drug effectiveness, regulatory pathways, or market expansion (e.g., generic semaglutide launches, peptide purity testing anecdotes, off-label peptide use stories) unless backed by published clinical trial results or official FDA guidance.
Posts about peptides, GLP-1s, or other drugs making market claims, efficacy assertions, or regulatory speculation without peer-reviewed clinical data or FDA approval status.
3 example posts
“You can test your peptides to see if they’re real, bro. But I just look for the 99.9% on the result the seller gives me. You can look up the QR code upon the lab’s website, so you know it’s legit. It’s foolproof, bro. What’s mass spec, bro?” https://t.co/xzJLJ3hesT
$NVO $LLY Dr. Reddy’s says it plans to launch a generic version of Ozempic in Canada within days and is also preparing to roll out generic semaglutide tablets in India. The company sees semaglutide as a meaningful future growth driver following patent expiry in India earlier
I'm a board certified OBGYN. I've practiced for 20+ years. Here's what's actually happening with mifepristone, because the coverage of it has been a mess. On May 1, the 5th Circuit Court of Appeals ruled that the FDA overstepped its authority in 2023 when it eliminated the in-pe
Created 2026-05-19 · Updated 2026-05-19
Edit rule text
[regulatory_or_policy_outrage_without_healthcare_systems_analysis]
Learned 3 rejections Active
Exclude posts that frame healthcare policy (FDA commissioner changes, federal procurement mandates, regulatory posturing) primarily as political outrage or partisan positioning, without analyzing how the policy affects healthcare delivery, equity, costs, or workflows. Posts about Trump administration policies, 'Made in America' mandates, or PBM restructuring belong here if they lack systems analysis.
Posts expressing political outrage about healthcare policy, regulatory decisions, or administration changes without substantive analysis of healthcare system impacts.
3 example posts
Vivek Ramaswamy put out a lot of statements today referencing the fraud in Ohio. Swampy is of course, involved in Ohio Medicaid. Datavant, Vivek's spinoff company from Riovant, handles release of information for many hospitals and providers in Ohio. Ohio Medicaid providers, man
The public is trained to hear “hospital closing” and assume virtue. Poor patients. Bad reimbursement. Heroic administrators. Cruel politicians. Sometimes true. Often incomplete. The harder truth is that the majority of hospitals are fragile because they have no real financial
The most useful thing to understand about health system operating margin is that it is not merely a measurement. It is a composition. When The New York Times reports that “Mass General Brigham’s operating margin fell to 0.8% last quarter,” readers tend to assume the number is h
Created 2026-05-19 · Updated 2026-05-19
Edit rule text
[incomplete_truncated_or_low_effort_social_posts]
Learned 3 rejections Active
Exclude posts that are visibly cut off mid-sentence, missing context or data, or appear to be partial retweets without the original thread or full argument. The post must be complete enough to evaluate the actual claim or analysis.
Posts that are fragmented, truncated, or incomplete, making substantive analysis impossible
3 example posts
The gap between best-case and real-world GDMT in #HFrEF is striking. GWTG-HF 2024 vs EPIC COSMOS 2023–2025 (all HFrEF pts, 3.3M patients): 💚 Beta-Blocker: 95% vs 73% 💚 ACE/ARB/ARNI: 92% vs 67% 💚 MRA: 81% vs 36% 💚 SGLT2i: 78% vs 34% 💚 Quadruple GDMT: 68% vs 18% https://t.co/xTXdB
How to print money, state government edition: 1) Tax your hospitals and managed care plans. 2) Use that tax revenue to pay your 10% share of Medicaid expansion. 3) Trigger a 90% federal match. 4) Return the original tax money, plus the massive federal windfall, back to the health
A 2016 BMJ study by researchers at Johns Hopkins estimated that medical error is the third leading cause of death in the United States. Behind heart disease. Behind cancer. Ahead of stroke. Ahead of respiratory disease. Ahead of accidents. Ahead of diabetes. Ahead of
Created 2026-05-19 · Updated 2026-05-19
Edit rule text
[pharmaceutical_trial_data_or_clinical_announcement_without_systems_context]
Learned 3 rejections Active
Exclude posts that announce pharmaceutical trial data, biotech pipeline milestones, or drug approvals (e.g., 'BEAM-101 BLA,' gene therapy updates, clinical presentation schedules) without analysis of healthcare delivery impact, cost-effectiveness, access barriers, or reimbursement implications.
Posts reporting drug trial results, clinical data, or pharma pipeline announcements as market updates rather than healthcare systems analysis.
3 example posts
$ABCL Just now @ BoA - @AbCelleraBio CEO Carl Hansen when asked about what the market is overlooking: “I believe we now have best in class or best in world capabilities for some very important specific applications in therapeutic antibodies. So GPCRs and ion channels has been a
$XBI RBC Makary's Out, So Who's In? We Break Down Potential Candidates and Impact to Biopharma Potential Candidates for #FDA #Commissioner: Kyle Diamantas J.D Dr. Stephen Hahn M.D. Dr. Brett Giroir M.D. Dr. Sara Brenner M.D. Dr. Houman Hemmati M.D., Ph.D. Dr. Richard Pazdur, M.D.
🚨! $SANA has just reported - in its Q1 ER, that it plans to expand its pipeline by advancing a new preclinical program - SG227. SG227 is a CD8-targeted fusosome that by delivering a genetic material to make BCMA-directed CAR T cells could act as a potential treatment for Cancer h
Created 2026-05-19 · Updated 2026-05-19
Edit rule text
[general_ai_infrastructure_hype_without_healthcare_application]
Learned 3 rejections Active
Exclude posts that focus on AI model capabilities (reasoning, coding, tokenization), GPU/compute infrastructure, or general AI infrastructure trends (e.g., Jensen Huang's keynotes, token economics, frontier model competition) unless they demonstrate specific, validated healthcare application or clinical validation.
Posts about AI model capabilities, compute infrastructure, or general AI trends with only tangential healthcare framing.
3 example posts
Google's CEO confirmed 75% of new code at Google is AI-generated. October 2024: 25%. Fall 2025: 50%. April 2026: 75%. Doubled in a year. Then doubled again in six months. Engineers aren't writing code anymore. They're reviewing it. The job title is the same. The job
Satya Nadella's energy is something here. 🔥 "Tokens per Dollar per Watt" The new equation for the AI age for every Company or Industry or Country. "And that means Infrastructure, Infrastructure and Infrastructure." https://t.co/McINrBAo4a
Google CEO Sundar Pichai on current frontier model's ability to break the security of almost all current software. "These models are definitely, like really gonna break pretty much all software out there, maybe already, we don't know." https://t.co/ezjPcSehiG
Created 2026-05-19 · Updated 2026-05-19
Edit rule text
[biotech_founder_enthusiasm_without_healthcare_validation]
Learned 3 rejections Active
Exclude posts that amplify biotech founder enthusiasm, lab capability announcements, or company milestones (e.g., Recursion, Sana, Beam) without demonstrating clinical validation, healthcare system readiness, reimbursement pathways, or patient impact evidence.
Posts sharing founder quotes, lab updates, or company announcements without clinical validation or healthcare system context
3 example posts
🚨! $SANA has just reported - in its Q1 ER, that it plans to expand its pipeline by advancing a new preclinical program - SG227. SG227 is a CD8-targeted fusosome that by delivering a genetic material to make BCMA-directed CAR T cells could act as a potential treatment for Cancer h
$BEAM's coming presentations May 15: ESCAPE June 13: BEAM-101 EHA update 2026 milestones BEAM-101 (riso-cel) BLA BEAM-301 initial data BEAM-302 updated data and pivotal trial initiation BEAM-304 IND filing https://t.co/9Nvq3PKMMl
The complexity of an AI driven robot that needs to meet GMP standards & learn complex steps of targeted cell manufacture. A marvel and the first robot of its kind in the world. Can’t wait till we lift off the first patient’s NK cells in LA from NANT LEONARDO. Unstoppable scal
Created 2026-05-19 · Updated 2026-05-19
Edit rule text
[unvalidated_pharmaceutical_claims]
Learned 3 rejections Active
Exclude posts that discuss peptides, GLP-1s, compounded medications, or experimental treatments being marketed or discussed without peer-reviewed clinical evidence, FDA approval status, or clear efficacy data. Includes posts promoting unvalidated market access or informal use.
Posts about unproven or speculative pharmaceutical treatments, peptides, or drug formulations without clinical validation.
3 example posts
Daily Walking Could Be the Secret to Keeping Weight Off For many people struggling with obesity, losing weight is only half the battle. The greater challenge often begins after the initial success: preventing the lost weight from returning. 🔴More than 50% of people who lose http
How does a brand like this even get listed at Target? Does Target only care about turnover? Might as well start selling snake oil too. By the way, this is the SKU the brand got in trouble with. They now call it "Berberine Patches". It was called GLP-1 Patches before. Check the h
“You can test your peptides to see if they’re real, bro. But I just look for the 99.9% on the result the seller gives me. You can look up the QR code upon the lab’s website, so you know it’s legit. It’s foolproof, bro. What’s mass spec, bro?” https://t.co/xzJLJ3hesT
Created 2026-05-19 · Updated 2026-05-19
Edit rule text
[tangential_ai_company_metrics]
Learned 3 rejections Active
Exclude posts that focus on AI company business news (funding rounds, valuations, product announcements, stock movements) unless the post explicitly discusses how the product is being applied to solve a healthcare-specific problem or workflow.
Posts about AI company funding, valuations, product launches, or business metrics that lack healthcare application specificity.
3 example posts
Anthropic CEO Dario Amodei on SaaS: "Software is going to become cheap, maybe essentially free. The premise that you need to amortize a piece of software you build across millions of users, that may start to be false. But at the same time, there are whole jobs, whole careers
imagine a software startup raising $800m before their first dollar of revenue. Can’t imagine it the last two decades of software has been dominated by a simple theory: ship quickly, get customers early, generate revenue quickly to validate PMF, manage KPIs closely, etc. This bec
The 36 BIGGEST startup opportunities right now 1. biggest b2c: solving loneliness. third spaces, community apps, IRL 2. biggest b2b: managed AI employees for businesses 3. biggest overlooked: elder tech. 70 million boomers who want products that make them happier & healthier
Created 2026-05-19 · Updated 2026-05-19
Edit rule text
[generalist_macro_without_healthcare_lens]
Learned 3 rejections Active
Exclude posts that frame healthcare as a secondary example in discussions of broader macroeconomic, labor market, infrastructure, or political trends (e.g., aging populations and cost of living, AI-driven workforce disruption, government procurement policy) without examining healthcare-specific system implications or operational changes.
Posts about broad economic, infrastructure, or labor trends that mention healthcare tangentially but lack substantive healthcare system analysis.
3 example posts
Older people can no longer afford to live in California New research finds that older people here in California face the highest risk of going bankrupt in the entire country The risk score was a 69 out of 100 based on factors including healthcare expenses and the high cost of l
The 36 BIGGEST startup opportunities right now 1. biggest b2c: solving loneliness. third spaces, community apps, IRL 2. biggest b2b: managed AI employees for businesses 3. biggest overlooked: elder tech. 70 million boomers who want products that make them happier & healthier
⚡️This is the moment AI stops being a productivity tool and becomes a replacement architecture for elite cognitive labor. Citadel is not a random corporation automating low-level admin. It is one of the most competitive intelligence machines in finance. The work Griffin is descr
Created 2026-05-18 · Updated 2026-05-18
Edit rule text
[non_healthcare_business_with_loose_framing]
Learned 3 rejections Active
Exclude posts about non-healthcare companies, infrastructure projects, geopolitical events, or policy changes that mention healthcare only superficially or use healthcare as a loose frame for general business commentary.
Posts about non-healthcare businesses, technology, or policy with minimal or tangential healthcare connection.
3 example posts
The 36 BIGGEST startup opportunities right now 1. biggest b2c: solving loneliness. third spaces, community apps, IRL 2. biggest b2b: managed AI employees for businesses 3. biggest overlooked: elder tech. 70 million boomers who want products that make them happier & healthier
🚨ANTHROPIC SHOCKS PRE-IPO MARKET, ERASES $400B AFTER NEW STOCK POLICY Anthropic’s implied pre-IPO valuation wiped nearly $400 BILLION just hours after the company dropped a BOMBSHELL targeting secondary share sales. Anthropic updated its stock transfer policy to state that any
We’re industrializing enterprise AI with @NVIDIA with: 🐚 OpenShell project collaboration for autonomous agents 🔧 Dev roadmap for Day 0 Vera Rubin support 🔐 Hardware-enforced security with confidential containers 📦 Lifecycle management and Model-as-a-Service 💥 Validated, https:
Created 2026-05-18 · Updated 2026-05-18
Edit rule text
[political_outrage_or_fraud_scandal_no_analysis]
Learned 3 rejections Active
Exclude posts that report FDA policy changes, pharmaceutical fraud, regulatory decisions, or political developments (Trump, Makary, PBMs) as breaking news or outrage without analyzing structural healthcare system implications.
Posts reporting healthcare fraud, regulatory changes, or political events without systems-level insight.
3 example posts
Vivek Ramaswamy put out a lot of statements today referencing the fraud in Ohio. Swampy is of course, involved in Ohio Medicaid. Datavant, Vivek's spinoff company from Riovant, handles release of information for many hospitals and providers in Ohio. Ohio Medicaid providers, man
Yes to Katy. If I was a manufacturer I would contract directly with the employer. The health plan when self funded has a fiduciary obligation to plan assets… Nice work @KatyTalento
The public is trained to hear “hospital closing” and assume virtue. Poor patients. Bad reimbursement. Heroic administrators. Cruel politicians. Sometimes true. Often incomplete. The harder truth is that the majority of hospitals are fragile because they have no real financial
Created 2026-05-18 · Updated 2026-05-18
Edit rule text
[tangential_infrastructure_or_robotics_hype]
Learned 3 rejections Active
Exclude posts that discuss energy efficiency ('Tokens per Dollar per Watt'), semiconductor maps, Middle East geopolitics, humanoid robots, or speculative sci-fi labor scenarios unless they directly address a concrete healthcare system problem (e.g., staffing shortages in rural clinics, supply chain resilience).
Posts about infrastructure, compute, robotics, or speculative sci-fi applications with loose or absent healthcare framing
3 example posts
Satya Nadella's energy is something here. 🔥 "Tokens per Dollar per Watt" The new equation for the AI age for every Company or Industry or Country. "And that means Infrastructure, Infrastructure and Infrastructure." https://t.co/McINrBAo4a
Palantir invented the forward deployed engineer role in 2006. embed an engineer inside the customer. ship production code in their environment. own the technical outcome. Alex Karp on the idea: “forward deployed engineers are stolen from french restaurants.” 20 years later htt
I’m fully forward deployed engineering pilled specifically because AI simply is not the same as software. In software, you deliver a stable piece of technology to a customer and they adopt it and that’s that (extreme over simplification). In AI, you’re delivering something that
Created 2026-05-18 · Updated 2026-05-18
Edit rule text
[unvalidated_peptide_market_speculation]
Learned 3 rejections Active
Exclude posts that report generic peptide availability, FDA actions on peptide bulk sources (503B closures), market competition between brands, or pricing/reimbursement updates for GLP-1s and peptides without analyzing impact on healthcare access, payer strategy, or clinical outcomes.
Posts about GLP-1, peptide, or semaglutide market dynamics, generic launches, or pricing without healthcare system analysis
3 example posts
Daily Walking Could Be the Secret to Keeping Weight Off For many people struggling with obesity, losing weight is only half the battle. The greater challenge often begins after the initial success: preventing the lost weight from returning. 🔴More than 50% of people who lose http
How does a brand like this even get listed at Target? Does Target only care about turnover? Might as well start selling snake oil too. By the way, this is the SKU the brand got in trouble with. They now call it "Berberine Patches". It was called GLP-1 Patches before. Check the h
“You can test your peptides to see if they’re real, bro. But I just look for the 99.9% on the result the seller gives me. You can look up the QR code upon the lab’s website, so you know it’s legit. It’s foolproof, bro. What’s mass spec, bro?” https://t.co/xzJLJ3hesT
Created 2026-05-18 · Updated 2026-05-18
Edit rule text
[founder_enthusiasm_lab_updates]
Learned 3 rejections Active
Exclude posts that are primarily celebratory announcements of biotech/AI founder activities, pipeline updates, or company milestones without demonstrating clinical validation, healthcare cost/access impact, or systems-level insight.
Posts celebrating biotech or AI founder announcements, lab updates, or product launches without substantive healthcare systems analysis.
3 example posts
$ABCL Just now @ BoA - @AbCelleraBio CEO Carl Hansen when asked about what the market is overlooking: “I believe we now have best in class or best in world capabilities for some very important specific applications in therapeutic antibodies. So GPCRs and ion channels has been a
🚨! $SANA has just reported - in its Q1 ER, that it plans to expand its pipeline by advancing a new preclinical program - SG227. SG227 is a CD8-targeted fusosome that by delivering a genetic material to make BCMA-directed CAR T cells could act as a potential treatment for Cancer h
$BEAM's coming presentations May 15: ESCAPE June 13: BEAM-101 EHA update 2026 milestones BEAM-101 (riso-cel) BLA BEAM-301 initial data BEAM-302 updated data and pivotal trial initiation BEAM-304 IND filing https://t.co/9Nvq3PKMMl
Created 2026-05-18 · Updated 2026-05-18
Edit rule text
[financial_market_speculation]
Learned 3 rejections Active
Exclude posts that focus on stock tickers, market rankings, valuation implications, competitive positioning, or financial metrics (like app store rankings or market share) without substantive analysis of how these affect healthcare delivery, cost, or outcomes.
Posts about stock performance, market valuations, and financial positioning without healthcare systems analysis
3 example posts
$HIMS $HERS 🚨 HERS IS NOW BEATING BOTH HIMS AND RO ON THE iOS APP STORE Today's rankings in the Medical category: - Hers: #9 - Ro: #10 - Hims: #18 Chart below = weekly avg rank https://t.co/vlv80yiFOo
Older people can no longer afford to live in California New research finds that older people here in California face the highest risk of going bankrupt in the entire country The risk score was a 69 out of 100 based on factors including healthcare expenses and the high cost of l
$ABCL Just now @ BoA - @AbCelleraBio CEO Carl Hansen when asked about what the market is overlooking: “I believe we now have best in class or best in world capabilities for some very important specific applications in therapeutic antibodies. So GPCRs and ion channels has been a
Created 2026-05-18 · Updated 2026-05-18
Edit rule text
[market_dynamics_or_pricing_without_systems_analysis]
Learned 3 rejections Active
Exclude posts that focus on market pricing dynamics, competitive rankings, generic drug launches, supply chain disruptions, or product positioning (e.g., app store rankings, market share battles, price changes) without analyzing how these dynamics affect healthcare systems, access, or outcomes.
Posts about pharmaceutical, GLP-1, peptide, or healthcare product market pricing, competition, or supply dynamics without healthcare system or policy analysis.
3 example posts
$HIMS $HERS 🚨 HERS IS NOW BEATING BOTH HIMS AND RO ON THE iOS APP STORE Today's rankings in the Medical category: - Hers: #9 - Ro: #10 - Hims: #18 Chart below = weekly avg rank https://t.co/vlv80yiFOo
Daily Walking Could Be the Secret to Keeping Weight Off For many people struggling with obesity, losing weight is only half the battle. The greater challenge often begins after the initial success: preventing the lost weight from returning. 🔴More than 50% of people who lose http
How does a brand like this even get listed at Target? Does Target only care about turnover? Might as well start selling snake oil too. By the way, this is the SKU the brand got in trouble with. They now call it "Berberine Patches". It was called GLP-1 Patches before. Check the h
Created 2026-05-18 · Updated 2026-05-18
Edit rule text
[tangential_generalist_ai_infrastructure_hype]
Learned 3 rejections Active
Exclude posts that discuss AI infrastructure (Jensen Huang, NVIDIA, compute efficiency, model training, moats), AI safety/cybersecurity vulnerabilities, or general AI technical capabilities unless they explicitly demonstrate concrete healthcare system impact or healthcare-specific application.
Posts about AI infrastructure, compute, models, or technical capabilities with only loose healthcare framing or no healthcare substance.
3 example posts
Google CEO Sundar Pichai on current frontier model's ability to break the security of almost all current software. "These models are definitely, like really gonna break pretty much all software out there, maybe already, we don't know." https://t.co/ezjPcSehiG
⚡️This is the moment AI stops being a productivity tool and becomes a replacement architecture for elite cognitive labor. Citadel is not a random corporation automating low-level admin. It is one of the most competitive intelligence machines in finance. The work Griffin is descr
Anthropic customers are building workarounds just to understand who inside their companies is driving AI costs higher. The scramble to monitor “tokenmaxxing” reflects how quickly businesses are adopting expensive AI tools before the economics fully make sense. Read more:
Created 2026-05-18 · Updated 2026-05-18
Edit rule text
[tangential_non_healthcare_policy_or_business_with_loose_framing]
Learned 3 rejections Active
Exclude posts about federal policy (trade, manufacturing, industrial policy), political figures' healthcare rhetoric, or conspiracy-adjacent claims (fraud allegations without evidence) — unless they directly address healthcare delivery structure or patient outcomes. Posts about Trump's 'Made in America', Vivek Ramaswamy's Datavant involvement, or Citadel automating cognitive labor belong here.
Posts about non-healthcare policy (buy American, Medicaid fraud claims) or non-healthcare business models mistakenly labeled as healthcare content
3 example posts
Older people can no longer afford to live in California New research finds that older people here in California face the highest risk of going bankrupt in the entire country The risk score was a 69 out of 100 based on factors including healthcare expenses and the high cost of l
Vivek Ramaswamy put out a lot of statements today referencing the fraud in Ohio. Swampy is of course, involved in Ohio Medicaid. Datavant, Vivek's spinoff company from Riovant, handles release of information for many hospitals and providers in Ohio. Ohio Medicaid providers, man
⚡️This is the moment AI stops being a productivity tool and becomes a replacement architecture for elite cognitive labor. Citadel is not a random corporation automating low-level admin. It is one of the most competitive intelligence machines in finance. The work Griffin is descr
Created 2026-05-18 · Updated 2026-05-18
Edit rule text
[tangential_healthcare_financial_or_market_framing_without_systems_analysis]
Learned 3 rejections Active
Exclude posts that report on healthcare pricing (drug costs, GLP-1 pricing, hospital margins), market competition (app store rankings, generic launches), financial scandals, or regulatory events without analyzing the underlying healthcare system structure, incentive misalignment, or operational/clinical consequence. Reporting a financial fact or scandal alone is insufficient.
Posts about healthcare financial metrics, market dynamics, pricing, or regulatory changes without structural analysis of healthcare system incentives or outcomes.
3 example posts
$HIMS $HERS 🚨 HERS IS NOW BEATING BOTH HIMS AND RO ON THE iOS APP STORE Today's rankings in the Medical category: - Hers: #9 - Ro: #10 - Hims: #18 Chart below = weekly avg rank https://t.co/vlv80yiFOo
Older people can no longer afford to live in California New research finds that older people here in California face the highest risk of going bankrupt in the entire country The risk score was a 69 out of 100 based on factors including healthcare expenses and the high cost of l
The public is trained to hear “hospital closing” and assume virtue. Poor patients. Bad reimbursement. Heroic administrators. Cruel politicians. Sometimes true. Often incomplete. The harder truth is that the majority of hospitals are fragile because they have no real financial
Created 2026-05-18 · Updated 2026-05-18
Edit rule text
[ai_model_technical_capability_tangent_loose_healthcare_framing]
Learned 3 rejections Active
Exclude posts that lead with AI model technical capabilities (e.g., 'Claude solved this benchmark', 'o1 beat doctors', 'Mythos Preview found zero-days') unless the post deeply analyzes how that specific capability changes healthcare workflows, clinical decision-making, regulatory compliance, or patient outcomes. Demonstrating that a model *can* do something in healthcare is not enough.
Posts about AI model technical breakthroughs (reasoning, coding, security vulnerabilities, world models) with only superficial healthcare framing or relevance.
3 example posts
Google CEO Sundar Pichai on current frontier model's ability to break the security of almost all current software. "These models are definitely, like really gonna break pretty much all software out there, maybe already, we don't know." https://t.co/ezjPcSehiG
Fei-Fei Li warns that AI may be staring too hard at language models. The world is not just text on a screen. It is physical, visual, spatial, and always changing. Most of the economy runs on seeing, moving, interacting, and embodied intelligence. https://t.co/mGMjsCdU34
The UK AISI found Mythos Preview is the first model to solve both their cyber ranges end-to-end. No model had ever solved the AISI’s “Cooling Tower” cyber range before. We're getting it to defenders as fast as we responsibly can. More to come on our Glasswing work soon.
Created 2026-05-18 · Updated 2026-05-18
Edit rule text
[generalist_ai_infrastructure_or_capability_hype_without_healthcare_validation]
Learned 3 rejections Active
Exclude posts that emphasize AI model breakthroughs, infrastructure announcements, or capability demonstrations (e.g., security vulnerability detection, world models, protein mapping) without showing concrete healthcare implementation, clinical validation, or system-level healthcare problem-solving. Posts must demonstrate healthcare application, not just AI advancement.
Posts hyping AI infrastructure, compute, or model capabilities with minimal evidence of actual healthcare system impact or validation.
3 example posts
Google CEO Sundar Pichai on current frontier model's ability to break the security of almost all current software. "These models are definitely, like really gonna break pretty much all software out there, maybe already, we don't know." https://t.co/ezjPcSehiG
Fei-Fei Li warns that AI may be staring too hard at language models. The world is not just text on a screen. It is physical, visual, spatial, and always changing. Most of the economy runs on seeing, moving, interacting, and embodied intelligence. https://t.co/mGMjsCdU34
Jensen Huang just told every tech CEO in America they are playing the wrong game. They are competing on models. He is competing on maps. He looked at the Middle East and did not see a conflict zone. He saw a compute continent waiting to be switched on. Huang: “I believe that h
Created 2026-05-18 · Updated 2026-05-18
Edit rule text
[tangential_infrastructure_compute_or_robotics_hype_without_healthcare_application]
Learned 3 rejections Active
Exclude posts about semiconductor companies, compute infrastructure (NVIDIA, Jensen Huang, watt-per-dollar efficiency), infrastructure policy, or speculative robotics/sci-fi applications unless the post demonstrates a concrete healthcare workflow problem being solved or a provider operational improvement.
Posts about general AI infrastructure, compute, semiconductors, or speculative robotics applications mentioned without clear healthcare delivery system relevance.
3 example posts
Satya Nadella's energy is something here. 🔥 "Tokens per Dollar per Watt" The new equation for the AI age for every Company or Industry or Country. "And that means Infrastructure, Infrastructure and Infrastructure." https://t.co/McINrBAo4a
Palantir invented the forward deployed engineer role in 2006. embed an engineer inside the customer. ship production code in their environment. own the technical outcome. Alex Karp on the idea: “forward deployed engineers are stolen from french restaurants.” 20 years later htt
Jensen Huang just told every tech CEO in America they are playing the wrong game. They are competing on models. He is competing on maps. He looked at the Middle East and did not see a conflict zone. He saw a compute continent waiting to be switched on. Huang: “I believe that h
Created 2026-05-18 · Updated 2026-05-18
Edit rule text
[unvalidated_peptide_or_glp1_claims]
Learned 3 rejections Active
Exclude posts that make broad claims about peptide efficacy, GLP-1 treatments, or unvalidated weight loss interventions based on personal anecdote, social media hype, or unverified seller claims rather than clinical trial data or FDA guidance.
Posts making unsupported claims about peptides, GLP-1s, or weight loss drugs without clinical evidence or regulatory context
3 example posts
Daily Walking Could Be the Secret to Keeping Weight Off For many people struggling with obesity, losing weight is only half the battle. The greater challenge often begins after the initial success: preventing the lost weight from returning. 🔴More than 50% of people who lose http
How does a brand like this even get listed at Target? Does Target only care about turnover? Might as well start selling snake oil too. By the way, this is the SKU the brand got in trouble with. They now call it "Berberine Patches". It was called GLP-1 Patches before. Check the h
“You can test your peptides to see if they’re real, bro. But I just look for the 99.9% on the result the seller gives me. You can look up the QR code upon the lab’s website, so you know it’s legit. It’s foolproof, bro. What’s mass spec, bro?” https://t.co/xzJLJ3hesT
Created 2026-05-18 · Updated 2026-05-18
Edit rule text
[business_model_or_market_speculation_without_systems_analysis]
Learned 3 rejections Active
Exclude posts that debate business models, pricing strategies, market competition, or financial positioning (PBMs, GLP-1 market, hospital margins) without analyzing the underlying healthcare system failure, reimbursement logic, or operational consequence. Market speculation without systems insight does not qualify.
Posts about healthcare business dynamics, pricing, or market positioning without operational or systems-level analysis.
3 example posts
$HIMS $HERS 🚨 HERS IS NOW BEATING BOTH HIMS AND RO ON THE iOS APP STORE Today's rankings in the Medical category: - Hers: #9 - Ro: #10 - Hims: #18 Chart below = weekly avg rank https://t.co/vlv80yiFOo
Yes to Katy. If I was a manufacturer I would contract directly with the employer. The health plan when self funded has a fiduciary obligation to plan assets… Nice work @KatyTalento
The public is trained to hear “hospital closing” and assume virtue. Poor patients. Bad reimbursement. Heroic administrators. Cruel politicians. Sometimes true. Often incomplete. The harder truth is that the majority of hospitals are fragile because they have no real financial
Created 2026-05-18 · Updated 2026-05-18
Edit rule text
[political_outrage_regulatory_without_analysis]
Learned 3 rejections Active
Exclude posts that use inflammatory language to react to FDA decisions, political figures, or regulatory announcements without explaining the actual healthcare systems impact, operational consequences, or evidence-based analysis of why the change matters.
Posts expressing outrage at regulatory decisions, political figures, or healthcare policy changes without substantive analysis of underlying healthcare system implications.
3 example posts
Older people can no longer afford to live in California New research finds that older people here in California face the highest risk of going bankrupt in the entire country The risk score was a 69 out of 100 based on factors including healthcare expenses and the high cost of l
Vivek Ramaswamy put out a lot of statements today referencing the fraud in Ohio. Swampy is of course, involved in Ohio Medicaid. Datavant, Vivek's spinoff company from Riovant, handles release of information for many hospitals and providers in Ohio. Ohio Medicaid providers, man
$XBI RBC Makary's Out, So Who's In? We Break Down Potential Candidates and Impact to Biopharma Potential Candidates for #FDA #Commissioner: Kyle Diamantas J.D Dr. Stephen Hahn M.D. Dr. Brett Giroir M.D. Dr. Sara Brenner M.D. Dr. Houman Hemmati M.D., Ph.D. Dr. Richard Pazdur, M.D.
Created 2026-05-17 · Updated 2026-05-17
Edit rule text
[political_outrage_or_regulatory_posturing_without_systems_analysis]
Learned 3 rejections Active
Exclude posts that treat regulatory or political events (FDA leadership, Trump policy statements, rebate rules) primarily as outrage/scandal narratives without substantive analysis of how these policies reshape healthcare incentives, delivery, or financing.
Posts about FDA commissioner changes, government mandates, or regulatory actions framed as political theater or scandal rather than healthcare system analysis.
3 example posts
Vivek Ramaswamy put out a lot of statements today referencing the fraud in Ohio. Swampy is of course, involved in Ohio Medicaid. Datavant, Vivek's spinoff company from Riovant, handles release of information for many hospitals and providers in Ohio. Ohio Medicaid providers, man
🚨 NOW: President Trump confirms that ALL FEDERAL AGENCIES are now required to buy American, “NO excuses” 🔥 And 47 is ensuring all “Made in America” loopholes are DESTROYED I voted for this! “ALL FEDERAL AGENCIES MUST BUY AMERICAN — NO EXCUSES! For decades, Washington politici
Incumbents don't defend broken systems. They defend profitable ones. Healthcare in America isn't a market failure. It's a sustaining innovation problem disguised as a social crisis. Every "reform" strengthens the existing architecture because the people writing the rules are the
Created 2026-05-17 · Updated 2026-05-17
Edit rule text
[shallow_business_model_or_market_debate]
Learned 3 rejections Active
Exclude posts that make sweeping claims about healthcare business models, market opportunities, or competitive dynamics (e.g., 'managed AI employees for businesses', 'solving loneliness') without grounding in data, evidence, or healthcare systems context.
Posts that present broad business model critiques or market debates without evidence or systems analysis
3 example posts
The 36 BIGGEST startup opportunities right now 1. biggest b2c: solving loneliness. third spaces, community apps, IRL 2. biggest b2b: managed AI employees for businesses 3. biggest overlooked: elder tech. 70 million boomers who want products that make them happier & healthier
Things every AI app startup says today to justify their defensibility: 1. We support multiple models. Our customers do not want to lock in to one vendor. 2. We have a data moat. We post-train open-source models to be much better and cheaper than closed-source. 3. We do deep inte
🚨! $SANA has just reported - in its Q1 ER, that it plans to expand its pipeline by advancing a new preclinical program - SG227. SG227 is a CD8-targeted fusosome that by delivering a genetic material to make BCMA-directed CAR T cells could act as a potential treatment for Cancer h
Created 2026-05-17 · Updated 2026-05-17
Edit rule text
[incomplete_truncated_or_low_effort_social_media_posts]
Learned 3 rejections Active
Exclude posts that are clearly truncated, incomplete, end abruptly, or consist only of fragmented social media fragments (chart links, stock tickers, partial quotes) that lack sufficient context or substantive healthcare analysis.
Posts that are fragmentary, cut off mid-sentence, or lack sufficient content to assess healthcare relevance.
3 example posts
Satya Nadella's energy is something here. 🔥 "Tokens per Dollar per Watt" The new equation for the AI age for every Company or Industry or Country. "And that means Infrastructure, Infrastructure and Infrastructure." https://t.co/McINrBAo4a
$HIMS $HERS 🚨 HERS IS NOW BEATING BOTH HIMS AND RO ON THE iOS APP STORE Today's rankings in the Medical category: - Hers: #9 - Ro: #10 - Hims: #18 Chart below = weekly avg rank https://t.co/vlv80yiFOo
Yes to Katy. If I was a manufacturer I would contract directly with the employer. The health plan when self funded has a fiduciary obligation to plan assets… Nice work @KatyTalento
Created 2026-05-17 · Updated 2026-05-17
Edit rule text
[tangential_ai_capability_announcement]
Learned 3 rejections Active
Exclude posts announcing AI product launches, agent frameworks, or model capabilities (Claude Code, NemoClaw, agentic systems) that provide no evidence of actual healthcare deployment, clinical validation, regulatory approval, or measurable healthcare system impact.
Posts announcing AI model capabilities or startup product launches that lack evidence of healthcare system validation or real-world clinical deployment.
3 example posts
The 36 BIGGEST startup opportunities right now 1. biggest b2c: solving loneliness. third spaces, community apps, IRL 2. biggest b2b: managed AI employees for businesses 3. biggest overlooked: elder tech. 70 million boomers who want products that make them happier & healthier
🚨! $SANA has just reported - in its Q1 ER, that it plans to expand its pipeline by advancing a new preclinical program - SG227. SG227 is a CD8-targeted fusosome that by delivering a genetic material to make BCMA-directed CAR T cells could act as a potential treatment for Cancer h
$BEAM's coming presentations May 15: ESCAPE June 13: BEAM-101 EHA update 2026 milestones BEAM-101 (riso-cel) BLA BEAM-301 initial data BEAM-302 updated data and pivotal trial initiation BEAM-304 IND filing https://t.co/9Nvq3PKMMl
Created 2026-05-17 · Updated 2026-05-17
Edit rule text
[pharmaceutical_trial_data_or_market_speculation_only]
Learned 3 rejections Active
Exclude posts that report pharmaceutical trial data, drug approval milestones, market pricing, or competitive dynamics (e.g., generic launches, semaglutide pricing, biotech stock tickers, gene therapy approvals) unless the post explains how this impacts healthcare delivery, reimbursement, or care access systems.
Posts announcing drug trial results, clinical milestones, or pharmaceutical market pricing/dynamics without healthcare systems context
3 example posts
Daily Walking Could Be the Secret to Keeping Weight Off For many people struggling with obesity, losing weight is only half the battle. The greater challenge often begins after the initial success: preventing the lost weight from returning. 🔴More than 50% of people who lose http
How does a brand like this even get listed at Target? Does Target only care about turnover? Might as well start selling snake oil too. By the way, this is the SKU the brand got in trouble with. They now call it "Berberine Patches". It was called GLP-1 Patches before. Check the h
“You can test your peptides to see if they’re real, bro. But I just look for the 99.9% on the result the seller gives me. You can look up the QR code upon the lab’s website, so you know it’s legit. It’s foolproof, bro. What’s mass spec, bro?” https://t.co/xzJLJ3hesT
Created 2026-05-17 · Updated 2026-05-17
Edit rule text
[ai_startup_founder_enthusiasm_or_lab_update_without_validation]
Learned 3 rejections Active
Exclude posts that quote founders about pipeline progress, lab capabilities, or milestone announcements (e.g., 'Chai raised $70M Series A,' 'Recursion asks how to harness AI') unless the post independently validates the clinical relevance, competitive positioning, or healthcare system impact. CEO enthusiasm or funding announcements belong here.
Posts amplifying biotech or AI startup founder claims, lab announcements, or capability updates without independent clinical or market validation.
3 example posts
$ABCL Just now @ BoA - @AbCelleraBio CEO Carl Hansen when asked about what the market is overlooking: “I believe we now have best in class or best in world capabilities for some very important specific applications in therapeutic antibodies. So GPCRs and ion channels has been a
🚨! $SANA has just reported - in its Q1 ER, that it plans to expand its pipeline by advancing a new preclinical program - SG227. SG227 is a CD8-targeted fusosome that by delivering a genetic material to make BCMA-directed CAR T cells could act as a potential treatment for Cancer h
$BEAM's coming presentations May 15: ESCAPE June 13: BEAM-101 EHA update 2026 milestones BEAM-101 (riso-cel) BLA BEAM-301 initial data BEAM-302 updated data and pivotal trial initiation BEAM-304 IND filing https://t.co/9Nvq3PKMMl
Created 2026-05-17 · Updated 2026-05-17
Edit rule text
[non_healthcare_business_or_tech_with_loose_healthcare_label]
Learned 3 rejections Active
Exclude posts about SAP API policies, Red Hat enterprise AI, Palantir operating systems, or cloud provider integrations that mention healthcare as a use case but focus on general enterprise tech, infrastructure, or business dynamics. If the core content is about non-healthcare tech and healthcare is just a mentioned vertical, exclude it.
Posts about non-healthcare companies, tech infrastructure, API policies, or business models tagged or framed as healthcare-adjacent with minimal actual healthcare substance.
3 example posts
We’re industrializing enterprise AI with @NVIDIA with: 🐚 OpenShell project collaboration for autonomous agents 🔧 Dev roadmap for Day 0 Vera Rubin support 🔐 Hardware-enforced security with confidential containers 📦 Lifecycle management and Model-as-a-Service 💥 Validated, https:
SAP updated its API policy this week to block all third-party AI agents from accessing its systems. The carve-out is narrow: only SAP-endorsed agents like its own Joule and NVIDIA's NemoClaw are permitted. This is the move every incumbent enterprise software platform will https
Execs at Palantir, whose stock is up ~16x since its AI platform's 2023 debut, often decry other AI as slop, as competition from frontier AI labs stiffens (@heathersomervil / Wall Street Journal) (Visit Techmeme dot com for the link and full context!)
Created 2026-05-17 · Updated 2026-05-17
Edit rule text
[tangential_healthcare_framing_of_policy_outrage]
Learned 3 rejections Active
Exclude posts where the primary content is political outrage, FDA commissioner speculation, Trump administration policy announcements, or court rulings — even if healthcare is mentioned — unless the post analyzes the specific healthcare system or incentive impact. Posts about 'who will replace Makary' or 'Trump's Made in America loopholes' belong here.
Posts about political figures, regulatory drama, or policy posturing that use healthcare as a loose frame for broader political commentary.
3 example posts
🚨 UPDATE Friday afternoon: Trump said he has no plans to replace FDA Commissioner Marty Makary This morning: "Just got a call from inside the administration. Makary is dead in the water. Just a matter of time." -@thackerpd $HIMS $LLY $NVO https://t.co/ZjcjOAG7e1
🚨 NOW: President Trump confirms that ALL FEDERAL AGENCIES are now required to buy American, “NO excuses” 🔥 And 47 is ensuring all “Made in America” loopholes are DESTROYED I voted for this! “ALL FEDERAL AGENCIES MUST BUY AMERICAN — NO EXCUSES! For decades, Washington politici
🛑FDA officials have blocked publication of several studies supporting the safety of widely used vaccines against Covid-19 and shingles in recent months, a spokesman for HHS confirmed. The studies, which cost millions of dollars in public funds, were conducted by scientists at htt
Created 2026-05-17 · Updated 2026-05-17
Edit rule text
[clinical_anecdote_or_study_without_systems_context]
Learned 3 rejections Active
Exclude posts that report clinical trial results, single-study findings, patient outcomes, or isolated clinical observations without analyzing how these findings affect healthcare delivery systems, provider incentives, reimbursement, or business model sustainability. Standalone clinical news is not healthcare tech analysis.
Posts reporting single clinical observations, trial data, or research findings without connecting to broader healthcare systems, incentives, or operational implications
3 example posts
💬 Editor's Note: AI-driven #ECG screening for left ventricular systolic dysfunction demonstrated high accuracy in Kenyan adults but should be interpreted with caution due to the high-risk enrichment of the study cohort. https://t.co/o3JS2wFrbC https://t.co/J4VNWqVHBe
CB-Dock3: An Enhanced Web Server for Protein–Ligand Blind Docking 1. CB-Dock3 is a major update of the CB-Dock blind docking web server, targeting the practical bottlenecks created by today’s surge of large Cryo-EM and AI-predicted protein structures (e.g., AlphaFold3-era https:
AI use for any business function is concentrated in selected sectors and large firms and primarily for augmentation of worker tasks rather than substitution, from Kathryn Bonney, Cory L. Breaux, Emin Dinlersoz, @Lucia_Econ, @JHaltiwanger_UM, and Aditya A. Pande https://t.co/qZ8Vf
Created 2026-05-17 · Updated 2026-05-17
Edit rule text
[policy_outrage_without_healthcare_analysis]
Learned 3 rejections Active
Exclude posts that frame healthcare policy, regulatory changes, or FDA decisions primarily as political grievances, talking points, or outrage without offering systems-level analysis of how the policy affects healthcare delivery, market structure, or business models. Inflammatory tone without substance does not fit.
Posts expressing political or regulatory outrage without substantive healthcare systems analysis or business model insight
3 example posts
I'm a board certified OBGYN. I've practiced for 20+ years. Here's what's actually happening with mifepristone, because the coverage of it has been a mess. On May 1, the 5th Circuit Court of Appeals ruled that the FDA overstepped its authority in 2023 when it eliminated the in-pe
🚨 UPDATE Friday afternoon: Trump said he has no plans to replace FDA Commissioner Marty Makary This morning: "Just got a call from inside the administration. Makary is dead in the water. Just a matter of time." -@thackerpd $HIMS $LLY $NVO https://t.co/ZjcjOAG7e1
🚨 NOW: President Trump confirms that ALL FEDERAL AGENCIES are now required to buy American, “NO excuses” 🔥 And 47 is ensuring all “Made in America” loopholes are DESTROYED I voted for this! “ALL FEDERAL AGENCIES MUST BUY AMERICAN — NO EXCUSES! For decades, Washington politici
Created 2026-05-17 · Updated 2026-05-17
Edit rule text
[ai_infrastructure_or_compute_hype_without_healthcare_specificity]
Learned 3 rejections Active
Exclude posts discussing AI compute infrastructure, token costs, data moats, training data bottlenecks, or model efficiency without specifically addressing how this solves a healthcare delivery, clinical, or operational problem. These are general AI infrastructure discussions.
Posts about AI infrastructure, compute costs, token pricing, or data bottlenecks that lack application to healthcare delivery or outcomes.
3 example posts
Things every AI app startup says today to justify their defensibility: 1. We support multiple models. Our customers do not want to lock in to one vendor. 2. We have a data moat. We post-train open-source models to be much better and cheaper than closed-source. 3. We do deep inte
Anthropic customers are building workarounds just to understand who inside their companies is driving AI costs higher. The scramble to monitor “tokenmaxxing” reflects how quickly businesses are adopting expensive AI tools before the economics fully make sense. Read more:
With AI, only three moats are left: distribution, data, trust. LLM and harness wrappers usually pick distribution. It's the only one available. You don't own the data. You don't own the trust either. The math works until the platform decides to take it back. It reprices. Scans
Created 2026-05-17 · Updated 2026-05-17
Edit rule text
[unvalidated_peptide_or_drug_market_claims]
Learned 3 rejections Active
Exclude posts discussing peptides, GLP-1 drugs, or compounded medications that focus on market dynamics, availability, or personal anecdotes without regulatory safety data, FDA guidance, or clinical evidence. Posts about 'real' peptides or testing peptides via seller QR codes lack pharmacological rigor.
Posts about GLP-1 peptides, semaglutide, or off-label drug use that lack regulatory context or clinical evidence.
3 example posts
How does a brand like this even get listed at Target? Does Target only care about turnover? Might as well start selling snake oil too. By the way, this is the SKU the brand got in trouble with. They now call it "Berberine Patches". It was called GLP-1 Patches before. Check the h
“You can test your peptides to see if they’re real, bro. But I just look for the 99.9% on the result the seller gives me. You can look up the QR code upon the lab’s website, so you know it’s legit. It’s foolproof, bro. What’s mass spec, bro?” https://t.co/xzJLJ3hesT
$NVO $LLY Dr. Reddy’s says it plans to launch a generic version of Ozempic in Canada within days and is also preparing to roll out generic semaglutide tablets in India. The company sees semaglutide as a meaningful future growth driver following patent expiry in India earlier
Created 2026-05-17 · Updated 2026-05-17
Edit rule text
[fragmented_or_truncated_posts_with_partial_context]
Learned 3 rejections Active
Exclude posts where the text appears truncated, incomplete, or ends abruptly without a clear conclusion or full argument. These posts do not provide enough information to assess whether they meet healthcare tech writer editorial standards.
Posts that are cut off mid-sentence, incomplete, or lack sufficient context to evaluate healthcare systems relevance.
3 example posts
Things every AI app startup says today to justify their defensibility: 1. We support multiple models. Our customers do not want to lock in to one vendor. 2. We have a data moat. We post-train open-source models to be much better and cheaper than closed-source. 3. We do deep inte
Anthropic customers are building workarounds just to understand who inside their companies is driving AI costs higher. The scramble to monitor “tokenmaxxing” reflects how quickly businesses are adopting expensive AI tools before the economics fully make sense. Read more:
Fei-Fei Li warns that AI may be staring too hard at language models. The world is not just text on a screen. It is physical, visual, spatial, and always changing. Most of the economy runs on seeing, moving, interacting, and embodied intelligence. https://t.co/mGMjsCdU34
Created 2026-05-17 · Updated 2026-05-17
Edit rule text
[ai_agent_or_agentic_system_capability_announcement_without_validation]
Learned 3 rejections Active
Exclude posts that announce new AI agent features, tool integrations, or agentic system releases (Claude Code, OpenClaw, NemoClaw capabilities) without evidence of healthcare validation, pilot deployments, or clinical workflow integration.
Posts announcing new AI agent capabilities, model releases, or agentic system features without demonstrating healthcare validation or real-world deployment.
3 example posts
The UK AISI found Mythos Preview is the first model to solve both their cyber ranges end-to-end. No model had ever solved the AISI’s “Cooling Tower” cyber range before. We're getting it to defenders as fast as we responsibly can. More to come on our Glasswing work soon.
Agents calling tools now generate more revenue on Latch than scientists clicking buttons. From native deployments but also harnesses like Claude Code, Codex, Cursor across Pharma, biotech, academic labs. Product roadmap increasingly concerned with exposing compute intensive
As advanced agents move from coding to the rest of knowledge work, it takes a real amount of work and know-how to get right. You need to ensure agents have the right context and data to work with, wire up systems to agents in a safe and secure way, ensure that the agents are htt
Created 2026-05-17 · Updated 2026-05-17
Edit rule text
[broad_health_claim_or_business_model_critique_without_evidence]
Learned 3 rejections Active
Exclude posts that make broad critiques of healthcare (e.g., 'healthcare is a sustaining innovation problem,' 'the system is broken') or business model arguments (e.g., 'PBMs own everything so patients lose') without providing specific data, case examples, financial analysis, or operational proof. Opinion without grounding does not qualify.
Posts making sweeping claims about healthcare system failures, business model problems, or market dynamics without citing data, case studies, or substantive evidence.
3 example posts
How does a brand like this even get listed at Target? Does Target only care about turnover? Might as well start selling snake oil too. By the way, this is the SKU the brand got in trouble with. They now call it "Berberine Patches". It was called GLP-1 Patches before. Check the h
PBM owns the pharmacy. PBM owns the plan. PBM adjudicates the claim. Three commissions on one pill. The employer pays all three and gets a thank-you note for the rebate.
Incumbents don't defend broken systems. They defend profitable ones. Healthcare in America isn't a market failure. It's a sustaining innovation problem disguised as a social crisis. Every "reform" strengthens the existing architecture because the people writing the rules are the
Created 2026-05-16 · Updated 2026-05-16
Edit rule text
[pharma_trial_data_or_drug_market_announcement_without_systems_analysis]
Learned 3 rejections Active
Exclude posts that report drug trial data, FDA approvals, generic drug launches, or pharmaceutical market pricing (e.g., 'Semaglutide generic launching in India') unless the post includes healthcare system impact analysis (e.g., how this affects PBM dynamics, hospital formularies, or patient access barriers). Pharma news alone does not qualify.
Posts that announce pharmaceutical trial results, drug approvals, or market pricing moves without analyzing impact on healthcare delivery systems, provider economics, or care quality.
3 example posts
$NVO $LLY Dr. Reddy’s says it plans to launch a generic version of Ozempic in Canada within days and is also preparing to roll out generic semaglutide tablets in India. The company sees semaglutide as a meaningful future growth driver following patent expiry in India earlier
🚨! $SANA has just reported - in its Q1 ER, that it plans to expand its pipeline by advancing a new preclinical program - SG227. SG227 is a CD8-targeted fusosome that by delivering a genetic material to make BCMA-directed CAR T cells could act as a potential treatment for Cancer h
$BEAM's coming presentations May 15: ESCAPE June 13: BEAM-101 EHA update 2026 milestones BEAM-101 (riso-cel) BLA BEAM-301 initial data BEAM-302 updated data and pivotal trial initiation BEAM-304 IND filing https://t.co/9Nvq3PKMMl
Created 2026-05-16 · Updated 2026-05-16
Edit rule text
[agentic_ai_or_ai_infrastructure_technical_tangent_without_healthcare_validation]
Learned 3 rejections Active
Exclude posts discussing agentic AI systems, model technical capabilities (tokenization, inference costs, model scaling), AI infrastructure, compute bottlenecks, or AI safety vulnerabilities as technical deep-dives or industry trends, unless the post explicitly addresses healthcare workflow automation, clinical decision support, or validated healthcare use cases.
Posts about agentic AI systems, model capabilities, compute infrastructure, or AI technical architecture as standalone topics without healthcare application or validation
3 example posts
Things every AI app startup says today to justify their defensibility: 1. We support multiple models. Our customers do not want to lock in to one vendor. 2. We have a data moat. We post-train open-source models to be much better and cheaper than closed-source. 3. We do deep inte
Anthropic customers are building workarounds just to understand who inside their companies is driving AI costs higher. The scramble to monitor “tokenmaxxing” reflects how quickly businesses are adopting expensive AI tools before the economics fully make sense. Read more:
With AI, only three moats are left: distribution, data, trust. LLM and harness wrappers usually pick distribution. It's the only one available. You don't own the data. You don't own the trust either. The math works until the platform decides to take it back. It reprices. Scans
Created 2026-05-16 · Updated 2026-05-16
Edit rule text
[pharma_trial_data_or_market_dynamics_without_healthcare_systems_context]
Learned 3 rejections Active
Exclude posts that announce pharmaceutical trial data, drug pipeline updates, clinical milestones, or list pharma ticker symbols and stock movements without connecting the finding to broader healthcare system incentives, access barriers, cost implications, or clinical practice patterns.
Posts listing drug trial milestones, clinical data announcements, or pharma stock tickers without analyzing how results affect healthcare delivery, provider economics, or patient outcomes
3 example posts
$ABCL Just now @ BoA - @AbCelleraBio CEO Carl Hansen when asked about what the market is overlooking: “I believe we now have best in class or best in world capabilities for some very important specific applications in therapeutic antibodies. So GPCRs and ion channels has been a
$XBI RBC Makary's Out, So Who's In? We Break Down Potential Candidates and Impact to Biopharma Potential Candidates for #FDA #Commissioner: Kyle Diamantas J.D Dr. Stephen Hahn M.D. Dr. Brett Giroir M.D. Dr. Sara Brenner M.D. Dr. Houman Hemmati M.D., Ph.D. Dr. Richard Pazdur, M.D.
🚨! $SANA has just reported - in its Q1 ER, that it plans to expand its pipeline by advancing a new preclinical program - SG227. SG227 is a CD8-targeted fusosome that by delivering a genetic material to make BCMA-directed CAR T cells could act as a potential treatment for Cancer h
Created 2026-05-16 · Updated 2026-05-16
Edit rule text
[unvalidated_or_speculative_pharmaceutical_claims_without_evidence]
Learned 3 rejections Active
Exclude posts that reference peptides, GLP-1 drugs, or other pharmaceuticals in contexts of unvalidated efficacy, informal testing procedures (e.g., 'checking QR codes'), speculative price movements, or generic market gossip without peer-reviewed evidence, clinical data, or FDA approval status.
Posts discussing peptides, GLP-1s, or pharmaceutical products with unvalidated efficacy claims, unverified testing methods, or speculative market claims
3 example posts
How does a brand like this even get listed at Target? Does Target only care about turnover? Might as well start selling snake oil too. By the way, this is the SKU the brand got in trouble with. They now call it "Berberine Patches". It was called GLP-1 Patches before. Check the h
“You can test your peptides to see if they’re real, bro. But I just look for the 99.9% on the result the seller gives me. You can look up the QR code upon the lab’s website, so you know it’s legit. It’s foolproof, bro. What’s mass spec, bro?” https://t.co/xzJLJ3hesT
$NVO $LLY Dr. Reddy’s says it plans to launch a generic version of Ozempic in Canada within days and is also preparing to roll out generic semaglutide tablets in India. The company sees semaglutide as a meaningful future growth driver following patent expiry in India earlier
Created 2026-05-16 · Updated 2026-05-16
Edit rule text
[fragmented_incomplete_social_posts]
Learned 3 rejections Active
Exclude posts that end with incomplete sentences, trailing ellipses indicating cut-off mid-thought, or quotes/data that clearly continue beyond the visible text boundary. These posts lack sufficient coherence for meaningful analysis.
Posts that are truncated mid-sentence or cut off abruptly, making them unintelligible or impossible to extract full context.
3 example posts
Things every AI app startup says today to justify their defensibility: 1. We support multiple models. Our customers do not want to lock in to one vendor. 2. We have a data moat. We post-train open-source models to be much better and cheaper than closed-source. 3. We do deep inte
Makes an important point here about the role of these short summaries they do this because this is how most doctors digests information if designed appropriately, an LLM should be able to read and appraise a paper and provide better peer review than most clinicians and honestly
$ABCL Just now @ BoA - @AbCelleraBio CEO Carl Hansen when asked about what the market is overlooking: “I believe we now have best in class or best in world capabilities for some very important specific applications in therapeutic antibodies. So GPCRs and ion channels has been a
Created 2026-05-16 · Updated 2026-05-16
Edit rule text
[political_or_regulatory_posturing_without_analysis]
Learned 3 rejections Active
Exclude posts that focus on political reactions to healthcare events, regulatory speculation, or policy posturing (e.g., FDA commissioner rumors, drug approval politics, Trump administration policy) without deeper analysis of healthcare system implications, operational impact, or evidence-based reasoning.
Posts expressing political outrage or regulatory positioning without substantive healthcare systems or operational analysis
3 example posts
🚨 UPDATE Friday afternoon: Trump said he has no plans to replace FDA Commissioner Marty Makary This morning: "Just got a call from inside the administration. Makary is dead in the water. Just a matter of time." -@thackerpd $HIMS $LLY $NVO https://t.co/ZjcjOAG7e1
🚨 NOW: President Trump confirms that ALL FEDERAL AGENCIES are now required to buy American, “NO excuses” 🔥 And 47 is ensuring all “Made in America” loopholes are DESTROYED I voted for this! “ALL FEDERAL AGENCIES MUST BUY AMERICAN — NO EXCUSES! For decades, Washington politici
🛑FDA officials have blocked publication of several studies supporting the safety of widely used vaccines against Covid-19 and shingles in recent months, a spokesman for HHS confirmed. The studies, which cost millions of dollars in public funds, were conducted by scientists at htt
Created 2026-05-16 · Updated 2026-05-16
Edit rule text
[market_dynamics_without_systems_analysis]
Learned 3 rejections Active
Exclude posts that discuss pharmaceutical/healthcare market pricing, competition, or business dynamics (e.g., generic launches, PBM consolidation, regulatory pricing) without analyzing how these changes affect healthcare delivery, patient access, or system efficiency. Market news alone is insufficient.
Posts about drug pricing, market competition, or business model shifts that lack healthcare systems context or operational insight
3 example posts
$NVO $LLY Dr. Reddy’s says it plans to launch a generic version of Ozempic in Canada within days and is also preparing to roll out generic semaglutide tablets in India. The company sees semaglutide as a meaningful future growth driver following patent expiry in India earlier
I'm a board certified OBGYN. I've practiced for 20+ years. Here's what's actually happening with mifepristone, because the coverage of it has been a mess. On May 1, the 5th Circuit Court of Appeals ruled that the FDA overstepped its authority in 2023 when it eliminated the in-pe
🚨 UPDATE Friday afternoon: Trump said he has no plans to replace FDA Commissioner Marty Makary This morning: "Just got a call from inside the administration. Makary is dead in the water. Just a matter of time." -@thackerpd $HIMS $LLY $NVO https://t.co/ZjcjOAG7e1
Created 2026-05-16 · Updated 2026-05-16
Edit rule text
[generalist_ai_tangent]
Learned 3 rejections Active
Exclude posts that discuss general AI model capabilities, compute scaling, inference costs, or AI safety/security vulnerabilities where the healthcare relevance is tangential, speculative, or forced. The post should demonstrate concrete healthcare application or impact, not just mention healthcare as a possible use case.
Posts about AI capabilities, infrastructure, or safety that have only loose or speculative healthcare framing
3 example posts
With AI, only three moats are left: distribution, data, trust. LLM and harness wrappers usually pick distribution. It's the only one available. You don't own the data. You don't own the trust either. The math works until the platform decides to take it back. It reprices. Scans
Jensen Huang just told every tech CEO in America they are playing the wrong game. They are competing on models. He is competing on maps. He looked at the Middle East and did not see a conflict zone. He saw a compute continent waiting to be switched on. Huang: “I believe that h
The UK AISI found Mythos Preview is the first model to solve both their cyber ranges end-to-end. No model had ever solved the AISI’s “Cooling Tower” cyber range before. We're getting it to defenders as fast as we responsibly can. More to come on our Glasswing work soon.
Created 2026-05-16 · Updated 2026-05-16
Edit rule text
[pharmaceutical_or_biotech_data_announcement_without_systems_context]
Learned 3 rejections Active
Exclude posts that list trial milestones, pipeline events, drug approvals, or clinical data points (BEAM-101 BLA, ESCAPE presentations, SG227 preclinical program) as standalone announcements without analyzing market dynamics, clinical significance, competitive positioning, or healthcare system impact.
Posts announcing drug trial results, pipeline updates, or clinical data without healthcare system analysis or business model implications.
3 example posts
🚨! $SANA has just reported - in its Q1 ER, that it plans to expand its pipeline by advancing a new preclinical program - SG227. SG227 is a CD8-targeted fusosome that by delivering a genetic material to make BCMA-directed CAR T cells could act as a potential treatment for Cancer h
$BEAM's coming presentations May 15: ESCAPE June 13: BEAM-101 EHA update 2026 milestones BEAM-101 (riso-cel) BLA BEAM-301 initial data BEAM-302 updated data and pivotal trial initiation BEAM-304 IND filing https://t.co/9Nvq3PKMMl
CB-Dock3: An Enhanced Web Server for Protein–Ligand Blind Docking 1. CB-Dock3 is a major update of the CB-Dock blind docking web server, targeting the practical bottlenecks created by today’s surge of large Cryo-EM and AI-predicted protein structures (e.g., AlphaFold3-era https:
Created 2026-05-16 · Updated 2026-05-16
Edit rule text
[regulatory_or_political_outrage_without_healthcare_systems_analysis]
Learned 3 rejections Active
Exclude posts that report regulatory news, FDA announcements, or political statements (mifepristone rulings, FDA Commissioner changes, Trump policy announcements) as headlines or outrage without explaining the healthcare system impact, stakeholder dynamics, or operational implications.
Posts about FDA decisions, government policy, or political figures posturing on healthcare without substantive systems analysis.
3 example posts
I'm a board certified OBGYN. I've practiced for 20+ years. Here's what's actually happening with mifepristone, because the coverage of it has been a mess. On May 1, the 5th Circuit Court of Appeals ruled that the FDA overstepped its authority in 2023 when it eliminated the in-pe
🚨 UPDATE Friday afternoon: Trump said he has no plans to replace FDA Commissioner Marty Makary This morning: "Just got a call from inside the administration. Makary is dead in the water. Just a matter of time." -@thackerpd $HIMS $LLY $NVO https://t.co/ZjcjOAG7e1
🚨 NOW: President Trump confirms that ALL FEDERAL AGENCIES are now required to buy American, “NO excuses” 🔥 And 47 is ensuring all “Made in America” loopholes are DESTROYED I voted for this! “ALL FEDERAL AGENCIES MUST BUY AMERICAN — NO EXCUSES! For decades, Washington politici
Created 2026-05-16 · Updated 2026-05-16
Edit rule text
[regulatory_fraud_scandal_without_healthcare_analysis]
Learned 3 rejections Active
Exclude posts that report FDA policy decisions, healthcare fraud cases, or regulatory scandals as standalone news events without analyzing systemic healthcare implications, provider incentives, or delivery model impacts. Breaking regulatory news without substantive healthcare systems analysis does not qualify.
Posts reporting FDA policy changes, fraud investigations, or regulatory scandals without analyzing impact on healthcare delivery or system dynamics
3 example posts
🚨 UPDATE Friday afternoon: Trump said he has no plans to replace FDA Commissioner Marty Makary This morning: "Just got a call from inside the administration. Makary is dead in the water. Just a matter of time." -@thackerpd $HIMS $LLY $NVO https://t.co/ZjcjOAG7e1
🛑FDA officials have blocked publication of several studies supporting the safety of widely used vaccines against Covid-19 and shingles in recent months, a spokesman for HHS confirmed. The studies, which cost millions of dollars in public funds, were conducted by scientists at htt
Red states have disproportionately benefited from the health insurance subsidies expanded by Biden and then cut by Trump. Georgia is a clear example: in just one year, marketplace enrollment has dropped by more than 550,000 people — a 37% decline. My @Morning_Joe Chart. https:/
Created 2026-05-15 · Updated 2026-05-15
Edit rule text
[broad_health_claim_or_speculation]
Learned 3 rejections Active
Exclude posts that make sweeping claims about AI solving health problems, market disruption, or clinical outcomes — unless backed by peer-reviewed evidence, regulatory filings, or detailed operational data showing mechanism and magnitude of impact.
Posts making broad, unvalidated claims about health outcomes, AI capabilities, or market dynamics without evidence or specificity
3 example posts
So Mythos was, indeed, not marketing hype. Remember this is a general purpose model that just happens to be good at finding exploits because good models are good at lots of things. Expect similar from OpenAI & Google. And from open models in 8 months. https://t.co/KbhalQYX8
“Tokens are increasing 50x. You need to drive down cost at the same rate — or more.” At Google Cloud Next, @googlecloud's Mark Lohmeyer spoke with @PatrickMoorhead about the infrastructure realities behind the agent era: faster model training, lower-latency inference, and https:
Klarna replaced 700 customer service agents with AI. Resolution time dropped from 11 minutes to 2. Then quality collapsed on complex cases. The CEO admitted cost became too dominant. Customers wanted humans. Klarna is now hiring people again. 55% of firms that replaced
Created 2026-05-15 · Updated 2026-05-15
Edit rule text
[political_regulatory_posturing]
Learned 3 rejections Active
Exclude posts reporting on political figures' healthcare statements, regulatory personnel changes, or executive orders — unless the post analyzes concrete operational impact on healthcare delivery, reimbursement mechanics, or specific regulatory pathways (not just commentary on political positioning).
Posts about political figures, regulatory announcements, or policy posturing without healthcare systems analysis
3 example posts
I'm a board certified OBGYN. I've practiced for 20+ years. Here's what's actually happening with mifepristone, because the coverage of it has been a mess. On May 1, the 5th Circuit Court of Appeals ruled that the FDA overstepped its authority in 2023 when it eliminated the in-pe
🚨 UPDATE Friday afternoon: Trump said he has no plans to replace FDA Commissioner Marty Makary This morning: "Just got a call from inside the administration. Makary is dead in the water. Just a matter of time." -@thackerpd $HIMS $LLY $NVO https://t.co/ZjcjOAG7e1
🚨 NOW: President Trump confirms that ALL FEDERAL AGENCIES are now required to buy American, “NO excuses” 🔥 And 47 is ensuring all “Made in America” loopholes are DESTROYED I voted for this! “ALL FEDERAL AGENCIES MUST BUY AMERICAN — NO EXCUSES! For decades, Washington politici
Created 2026-05-15 · Updated 2026-05-15
Edit rule text
[regulatory_fraud_scandal_no_healthcare_analysis]
Learned 3 rejections Active
Exclude posts that report fraud allegations, regulatory enforcement, policy announcements (FDA, Medicare, Medicaid), or political statements about healthcare without explaining the underlying healthcare system dynamics, operational consequences, or market structure implications. Scandal reporting alone is insufficient.
Posts reporting healthcare fraud, regulatory action, or policy changes without substantive analysis of healthcare system implications.
3 example posts
🚨 UPDATE Friday afternoon: Trump said he has no plans to replace FDA Commissioner Marty Makary This morning: "Just got a call from inside the administration. Makary is dead in the water. Just a matter of time." -@thackerpd $HIMS $LLY $NVO https://t.co/ZjcjOAG7e1
🛑FDA officials have blocked publication of several studies supporting the safety of widely used vaccines against Covid-19 and shingles in recent months, a spokesman for HHS confirmed. The studies, which cost millions of dollars in public funds, were conducted by scientists at htt
Red states have disproportionately benefited from the health insurance subsidies expanded by Biden and then cut by Trump. Georgia is a clear example: in just one year, marketplace enrollment has dropped by more than 550,000 people — a 37% decline. My @Morning_Joe Chart. https:/
Created 2026-05-15 · Updated 2026-05-15
Edit rule text
[political_or_regulatory_posturing_without_healthcare_analysis]
Learned 3 rejections Active
Exclude posts that report political statements, executive announcements, or regulatory changes without analyzing concrete healthcare delivery impact, implementation mechanisms, or operational consequences. Breaking news or politician quotes without systems analysis are insufficient.
Posts about political figures, regulatory announcements, or policy posturing that lack substantive healthcare system analysis or operational impact assessment.
3 example posts
🚨 UPDATE Friday afternoon: Trump said he has no plans to replace FDA Commissioner Marty Makary This morning: "Just got a call from inside the administration. Makary is dead in the water. Just a matter of time." -@thackerpd $HIMS $LLY $NVO https://t.co/ZjcjOAG7e1
🚨 NOW: President Trump confirms that ALL FEDERAL AGENCIES are now required to buy American, “NO excuses” 🔥 And 47 is ensuring all “Made in America” loopholes are DESTROYED I voted for this! “ALL FEDERAL AGENCIES MUST BUY AMERICAN — NO EXCUSES! For decades, Washington politici
🛑FDA officials have blocked publication of several studies supporting the safety of widely used vaccines against Covid-19 and shingles in recent months, a spokesman for HHS confirmed. The studies, which cost millions of dollars in public funds, were conducted by scientists at htt
Created 2026-05-15 · Updated 2026-05-15
Edit rule text
[healthcare_fraud_scandal_no_systems_analysis]
Learned 3 rejections Active
Exclude posts that report healthcare fraud, billing abuse, or financial misconduct (e.g., PBM fraud, Medicaid theft) as scandal journalism without analyzing root causes in healthcare incentives, regulation, or system design. The post must offer healthcare systems insight, not just fraud reporting.
Posts about healthcare fraud, Medicaid abuse, or financial scandals reported as investigative news without healthcare system or policy analysis.
3 example posts
🚨 UPDATE Friday afternoon: Trump said he has no plans to replace FDA Commissioner Marty Makary This morning: "Just got a call from inside the administration. Makary is dead in the water. Just a matter of time." -@thackerpd $HIMS $LLY $NVO https://t.co/ZjcjOAG7e1
🛑FDA officials have blocked publication of several studies supporting the safety of widely used vaccines against Covid-19 and shingles in recent months, a spokesman for HHS confirmed. The studies, which cost millions of dollars in public funds, were conducted by scientists at htt
"Cordoba-owned buildings in Columbus housed 288 businesses registered with Medicaid... they charged taxpayers more than a quarter of a billion dollars between 2018 and 2024... in a city where only 6,273 people 75 or older are on Medicaid." - @realDailyWire https://t.co/7UnS7agwiw
Created 2026-05-15 · Updated 2026-05-15
Edit rule text
[policy_fraud_scandal_without_healthcare_systems]
Learned 3 rejections Active
Exclude posts that report healthcare fraud, Medicaid abuse, or regulatory enforcement actions (FDA blocks, court decisions, inspector general findings) as political/crime reporting without analyzing how the scandal affects healthcare business models, provider incentives, or patient outcomes.
Posts about government fraud detection, welfare abuse, or regulatory enforcement without healthcare system analysis or provider/payer operational context.
3 example posts
🚨 UPDATE Friday afternoon: Trump said he has no plans to replace FDA Commissioner Marty Makary This morning: "Just got a call from inside the administration. Makary is dead in the water. Just a matter of time." -@thackerpd $HIMS $LLY $NVO https://t.co/ZjcjOAG7e1
🚨 NOW: President Trump confirms that ALL FEDERAL AGENCIES are now required to buy American, “NO excuses” 🔥 And 47 is ensuring all “Made in America” loopholes are DESTROYED I voted for this! “ALL FEDERAL AGENCIES MUST BUY AMERICAN — NO EXCUSES! For decades, Washington politici
🛑FDA officials have blocked publication of several studies supporting the safety of widely used vaccines against Covid-19 and shingles in recent months, a spokesman for HHS confirmed. The studies, which cost millions of dollars in public funds, were conducted by scientists at htt
Created 2026-05-15 · Updated 2026-05-15
Edit rule text
[business_model_critique_without_evidence]
Learned 3 rejections Active
Exclude posts asserting general principles about AI startup defensibility (e.g., 'data moats don't exist,' 'distribution is the only moat') or critiquing business model trends without healthcare-specific data, case studies, or systems-level impact analysis. Posts should ground claims in healthcare market dynamics or clinical operations, not generic startup theory.
Posts making broad claims about AI business model defensibility, moats, or startup economics without healthcare-specific evidence or systems analysis.
3 example posts
Things every AI app startup says today to justify their defensibility: 1. We support multiple models. Our customers do not want to lock in to one vendor. 2. We have a data moat. We post-train open-source models to be much better and cheaper than closed-source. 3. We do deep inte
With AI, only three moats are left: distribution, data, trust. LLM and harness wrappers usually pick distribution. It's the only one available. You don't own the data. You don't own the trust either. The math works until the platform decides to take it back. It reprices. Scans
The AI bolt-on era, illustrated. McKinsey: 6% of companies see real EBIT gains from AI. They're 3x more likely to redesign workflows around AI than bolt it onto existing tools. I love Anthropic. They just shipped a better bolt-on. https://t.co/tCQcPMywGh
Created 2026-05-15 · Updated 2026-05-17
Edit rule text
[clinical_anecdote_or_single_observation]
Learned 3 rejections Active
Exclude posts that highlight a single clinical finding, trial result, observational anecdote, or isolated data point without connecting it to broader healthcare system implications, policy impact, or replicable healthcare delivery insights.
Posts reporting individual clinical observations, single studies, or anecdotal findings without healthcare systems context.
3 example posts
💬 Editor's Note: AI-driven #ECG screening for left ventricular systolic dysfunction demonstrated high accuracy in Kenyan adults but should be interpreted with caution due to the high-risk enrichment of the study cohort. https://t.co/o3JS2wFrbC https://t.co/J4VNWqVHBe
For 50 years the National Institute on Drug Abuse has spent hundreds of millions of dollars to find an effective cocaine addiction medication. After 100+ molecules & many hundreds of studies nothing has been FDA approved. It's the Holy Grail of addictions medication. Proud to
CB-Dock3: An Enhanced Web Server for Protein–Ligand Blind Docking 1. CB-Dock3 is a major update of the CB-Dock blind docking web server, targeting the practical bottlenecks created by today’s surge of large Cryo-EM and AI-predicted protein structures (e.g., AlphaFold3-era https:
Created 2026-05-15 · Updated 2026-05-15
Edit rule text
[pharmaceutical_market_pricing_dynamics]
Learned 3 rejections Active
Exclude posts that focus on drug pricing, generic competition, market rollouts, or pharmaceutical company strategy announcements (e.g., launch timelines, generic versions, pricing pressure) without analysis of healthcare system impact, access barriers, or clinical differentiation.
Posts about generic drug launches, market pricing, or competitive pharmaceutical dynamics without healthcare systems or access analysis.
3 example posts
$NVO $LLY Dr. Reddy’s says it plans to launch a generic version of Ozempic in Canada within days and is also preparing to roll out generic semaglutide tablets in India. The company sees semaglutide as a meaningful future growth driver following patent expiry in India earlier
🚨! $SANA has just reported - in its Q1 ER, that it plans to expand its pipeline by advancing a new preclinical program - SG227. SG227 is a CD8-targeted fusosome that by delivering a genetic material to make BCMA-directed CAR T cells could act as a potential treatment for Cancer h
$BEAM's coming presentations May 15: ESCAPE June 13: BEAM-101 EHA update 2026 milestones BEAM-101 (riso-cel) BLA BEAM-301 initial data BEAM-302 updated data and pivotal trial initiation BEAM-304 IND filing https://t.co/9Nvq3PKMMl
Created 2026-05-15 · Updated 2026-05-15
Edit rule text
[policy_fraud_scandal_reporting_without_healthcare_systems_analysis]
Learned 3 rejections Active
Exclude posts about government spending, welfare fraud, 'Made in America' policy, FDA commissioner staff drama, or broad government accountability initiatives unless they analyze concrete impact on healthcare reimbursement, provider economics, or patient access. Posts about SNAP fraud, Trump administration policy, or federal agency mandates are out of scope unless directly affecting healthcare operations.
Posts about government policy, fraud detection, or regulatory scandals without healthcare delivery or reimbursement system context
3 example posts
🚨 UPDATE Friday afternoon: Trump said he has no plans to replace FDA Commissioner Marty Makary This morning: "Just got a call from inside the administration. Makary is dead in the water. Just a matter of time." -@thackerpd $HIMS $LLY $NVO https://t.co/ZjcjOAG7e1
🚨 NOW: President Trump confirms that ALL FEDERAL AGENCIES are now required to buy American, “NO excuses” 🔥 And 47 is ensuring all “Made in America” loopholes are DESTROYED I voted for this! “ALL FEDERAL AGENCIES MUST BUY AMERICAN — NO EXCUSES! For decades, Washington politici
🛑FDA officials have blocked publication of several studies supporting the safety of widely used vaccines against Covid-19 and shingles in recent months, a spokesman for HHS confirmed. The studies, which cost millions of dollars in public funds, were conducted by scientists at htt
Created 2026-05-14 · Updated 2026-05-14
Edit rule text
[tangential_non_healthcare_policy_macro]
Learned 3 rejections Active
Exclude posts about 'Buy American' mandates, government procurement policy, tax policy, or macroeconomic trends that loosely reference healthcare without analyzing actual healthcare delivery, reimbursement, or operational impact (e.g., 'Trump says all federal agencies buy American' with no healthcare supply chain analysis).
Posts about federal policy, procurement mandates, or macroeconomic shifts loosely framed as healthcare-relevant but lacking healthcare systems substance.
3 example posts
🚨 NOW: President Trump confirms that ALL FEDERAL AGENCIES are now required to buy American, “NO excuses” 🔥 And 47 is ensuring all “Made in America” loopholes are DESTROYED I voted for this! “ALL FEDERAL AGENCIES MUST BUY AMERICAN — NO EXCUSES! For decades, Washington politici
Michigan Medicine generated $8.7 billion in revenue in FY25. It posted a $109 million operating margin in the same year. A nonprofit health system underwriting its largest pavilion in history off margin, not philanthropy. The accounting category and the operating reality do not
Incumbents don't defend broken systems. They defend profitable ones. Healthcare in America isn't a market failure. It's a sustaining innovation problem disguised as a social crisis. Every "reform" strengthens the existing architecture because the people writing the rules are the
Created 2026-05-14 · Updated 2026-05-14
Edit rule text
[non_healthcare_business_or_policy_with_loose_healthcare_framing]
Learned 3 rejections Active
Exclude posts about SpaceX semiconductor manufacturing, Red Hat enterprise AI partnerships, Google Cloud infrastructure, SAP API policies, or other non-healthcare business/tech content that only weakly or argumentatively connects to healthcare. Posts must be centrally about healthcare business, delivery, or operations—not just use healthcare as a loose frame for tech/policy news.
Posts about non-healthcare companies, infrastructure, or policy that are tangentially reframed as healthcare-relevant without substantive connection.
3 example posts
We’re industrializing enterprise AI with @NVIDIA with: 🐚 OpenShell project collaboration for autonomous agents 🔧 Dev roadmap for Day 0 Vera Rubin support 🔐 Hardware-enforced security with confidential containers 📦 Lifecycle management and Model-as-a-Service 💥 Validated, https:
SAP updated its API policy this week to block all third-party AI agents from accessing its systems. The carve-out is narrow: only SAP-endorsed agents like its own Joule and NVIDIA's NemoClaw are permitted. This is the move every incumbent enterprise software platform will https
Execs at Palantir, whose stock is up ~16x since its AI platform's 2023 debut, often decry other AI as slop, as competition from frontier AI labs stiffens (@heathersomervil / Wall Street Journal) (Visit Techmeme dot com for the link and full context!)
Created 2026-05-14 · Updated 2026-05-14
Edit rule text
[healthcare_fraud_scandal_or_financial_reporting_without_systems_analysis]
Learned 3 rejections Active
Exclude posts that report healthcare fraud cases, PBM scandals, or financial irregularities as standalone incidents without connecting to systemic healthcare business model failures, incentive structures, or policy implications. The post must explain WHY this matters to healthcare systems, not just report the scandal.
Posts reporting healthcare fraud, financial impropriety, or Medicaid/Medicare spending without structural healthcare economics or policy analysis
3 example posts
PBM owns the pharmacy. PBM owns the plan. PBM adjudicates the claim. Three commissions on one pill. The employer pays all three and gets a thank-you note for the rebate.
Michigan Medicine generated $8.7 billion in revenue in FY25. It posted a $109 million operating margin in the same year. A nonprofit health system underwriting its largest pavilion in history off margin, not philanthropy. The accounting category and the operating reality do not
"Cordoba-owned buildings in Columbus housed 288 businesses registered with Medicaid... they charged taxpayers more than a quarter of a billion dollars between 2018 and 2024... in a city where only 6,273 people 75 or older are on Medicaid." - @realDailyWire https://t.co/7UnS7agwiw
Created 2026-05-14 · Updated 2026-05-14
Edit rule text
[political_figures_or_regulatory_posturing_without_healthcare_analysis]
Learned 3 rejections Active
Exclude posts that feature political figures (Trump, VP, cabinet officials) making healthcare-adjacent statements, government mandate announcements, or regulatory theater—unless the post provides specific analysis of how the policy affects healthcare delivery, reimbursement, or patient care.
Posts amplifying political figures' healthcare claims, regulatory announcements, or policy posturing without substantive analysis of healthcare system implications.
3 example posts
🚨 UPDATE Friday afternoon: Trump said he has no plans to replace FDA Commissioner Marty Makary This morning: "Just got a call from inside the administration. Makary is dead in the water. Just a matter of time." -@thackerpd $HIMS $LLY $NVO https://t.co/ZjcjOAG7e1
🚨 NOW: President Trump confirms that ALL FEDERAL AGENCIES are now required to buy American, “NO excuses” 🔥 And 47 is ensuring all “Made in America” loopholes are DESTROYED I voted for this! “ALL FEDERAL AGENCIES MUST BUY AMERICAN — NO EXCUSES! For decades, Washington politici
🛑FDA officials have blocked publication of several studies supporting the safety of widely used vaccines against Covid-19 and shingles in recent months, a spokesman for HHS confirmed. The studies, which cost millions of dollars in public funds, were conducted by scientists at htt
Created 2026-05-14 · Updated 2026-05-14
Edit rule text
[biotech_or_pharmaceutical_founder_enthusiasm_without_systems_validation]
Learned 3 rejections Active
Exclude posts from biotech or pharmaceutical company founders/executives that celebrate research progress, mission alignment, or lab updates (e.g., 'What's the one question we obsess over?', 'new data dump of protein structures') without demonstrating healthcare systems impact, manufacturing scale, regulatory pathway, or economic viability.
Posts from biotech founders or companies celebrating lab achievements, mission statements, or research milestones without healthcare systems impact or business model analysis.
3 example posts
The complexity of an AI driven robot that needs to meet GMP standards & learn complex steps of targeted cell manufacture. A marvel and the first robot of its kind in the world. Can’t wait till we lift off the first patient’s NK cells in LA from NANT LEONARDO. Unstoppable scal
“What’s the one question we obsess over at Recursion? ‘How do we harness the full power of AI to consistently and with urgency create better medicines for patients’?” In a recap of Recursion’s recent 1Q earnings, CEO and President Najat Khan talks about the tangible evidence htt
NEW: today OpenBind ‘comes out of stealth’ so to speak with their first data dump of ~900 novel protein-ligand structures - most with paired affinities This represents a meaningful %-age increase in all of humanities P-L data in the PDB collected in the last 50 years More👇 http
Created 2026-05-14 · Updated 2026-05-14
Edit rule text
[incomplete_truncated_social_media_posts]
Learned 3 rejections Active
Exclude posts that are truncated, incomplete, cut off mid-thought, or end abruptly without completing the main argument or claim. Posts must be complete, coherent social media posts with clear conclusion.
Posts that are cut off mid-sentence, incomplete, or lack sufficient context to understand full argument
3 example posts
The complexity of an AI driven robot that needs to meet GMP standards & learn complex steps of targeted cell manufacture. A marvel and the first robot of its kind in the world. Can’t wait till we lift off the first patient’s NK cells in LA from NANT LEONARDO. Unstoppable scal
We’re industrializing enterprise AI with @NVIDIA with: 🐚 OpenShell project collaboration for autonomous agents 🔧 Dev roadmap for Day 0 Vera Rubin support 🔐 Hardware-enforced security with confidential containers 📦 Lifecycle management and Model-as-a-Service 💥 Validated, https:
Agents calling tools now generate more revenue on Latch than scientists clicking buttons. From native deployments but also harnesses like Claude Code, Codex, Cursor across Pharma, biotech, academic labs. Product roadmap increasingly concerned with exposing compute intensive
Created 2026-05-14 · Updated 2026-05-14
Edit rule text
[ai_agent_security_or_misuse_incident_tangential_to_healthcare]
Learned 3 rejections Active
Exclude posts about AI agent security incidents, autonomous system misuse, or novel attack vectors unless they explicitly demonstrate a healthcare-specific risk (HIPAA violation, patient data exfiltration, clinical decision manipulation). Posts about agents committing fraud, hacking, or unauthorized access in non-healthcare contexts must be rejected.
Posts describing AI agent security incidents, jailbreaks, or autonomous system misuse (agents registering LLCs, wire fraud examples) presented as healthcare-relevant without explicit healthcare system vulnerability or compliance risk.
3 example posts
🚀🤖 In 2026, autonomous #AI agents aren’t failing because of technology ➜ they’re failing because of responsibility. ⚙️ Everyone knows how to build agents. Very few know how to own their decisions. 💡 Real autonomous AI isn’t about freedom or intelligence. It’s about clear scope,
Clawvisor is going to be one of the most important parts of helping make the agent world especially OpenClaw/Hermes Agent secure and enterprise-grade. We're in the Apple I moment for personal AI, but we're ABOUT to see the Apple II - the first moment when everyone can use it
Clawvisor (@clawvisor) lets you give AI agents access to apps like Gmail and Slack without handing over your credentials or worrying they'll go rogue. You approve tasks once, Clawvisor enforce them. Congrats on the launch, @ericlevine! https://t.co/S51QlWauds https://t.co/NZvnt
Created 2026-05-14 · Updated 2026-05-14
Edit rule text
[tangential_ai_infrastructure_compute_hype_without_healthcare_validation]
Learned 3 rejections Active
Exclude posts about general AI infrastructure trends (tokenomics, compute scaling, data bottlenecks, GPU utilization, training efficiency) unless they explicitly connect to a validated healthcare use case. Posts about Nvidia, cloud providers, or data scaling for non-healthcare domains must be rejected even if loosely framed around healthcare.
Posts about AI infrastructure, compute costs, training data bottlenecks, or model scaling challenges without demonstrated healthcare application or validation.
3 example posts
Last week, Google Cloud and AWS natively integrated x402, tackling head on one of its biggest problems: unauthorized API wrappers. The largest cloud platforms now give agents sanctioned ways to use services, pay with stables, and ditch API keys. It’s a new era. Pay (dot) sh’s
Larry Ellison just explained the most important and underreported insight in enterprise AI. Every major AI model, ChatGPT, Claude, Grok was trained on public data but public data has a ceiling. The real unlocking of AI's value happens when models can reason on private data, the
“Tokens are increasing 50x. You need to drive down cost at the same rate — or more.” At Google Cloud Next, @googlecloud's Mark Lohmeyer spoke with @PatrickMoorhead about the infrastructure realities behind the agent era: faster model training, lower-latency inference, and https:
Created 2026-05-14 · Updated 2026-05-14
Edit rule text
[ai_agent_or_model_security_tangent_without_healthcare_validation]
Learned 3 rejections Active
Exclude posts about AI agent security incidents (agents registering LLCs, agents calling each other, agents accessing unauthorized APIs) or model exploitation vulnerabilities—unless the post explicitly discusses healthcare-specific attack surfaces, patient data breach risk, or regulatory compliance impact.
Posts about AI agent security vulnerabilities, unauthorized API access, or chatbot abuse incidents that use healthcare as loose framing without validating actual healthcare risk or deployment.
3 example posts
We’re industrializing enterprise AI with @NVIDIA with: 🐚 OpenShell project collaboration for autonomous agents 🔧 Dev roadmap for Day 0 Vera Rubin support 🔐 Hardware-enforced security with confidential containers 📦 Lifecycle management and Model-as-a-Service 💥 Validated, https:
"There are multiple levels of controlling the actions, but before that, you need provenance. You have to ask "do I trust the data?" and "do I know what data it was trained on?" - Dr. @leemonbaird Day 2 at @consensus2026: Leemon joined the panel “Agents Meet the Enterprise — Sup
🚀🤖 In 2026, autonomous #AI agents aren’t failing because of technology ➜ they’re failing because of responsibility. ⚙️ Everyone knows how to build agents. Very few know how to own their decisions. 💡 Real autonomous AI isn’t about freedom or intelligence. It’s about clear scope,
Created 2026-05-13 · Updated 2026-05-13
Edit rule text
[ai_safety_vulnerability_or_security_tangential_to_healthcare]
Learned 3 rejections Active
Exclude posts that focus on AI safety vulnerabilities, security exploits, agent misuse scenarios, or cybersecurity incidents—unless they are directly about healthcare data breach risk, clinical decision integrity, or HIPAA compliance. Posts about Mythos finding exploits, agents talking to each other, or credential management belong here.
Posts about AI security incidents, agent misuse, zero-day vulnerabilities, or safety research with only loose healthcare framing
3 example posts
Last week, Google Cloud and AWS natively integrated x402, tackling head on one of its biggest problems: unauthorized API wrappers. The largest cloud platforms now give agents sanctioned ways to use services, pay with stables, and ditch API keys. It’s a new era. Pay (dot) sh’s
Flock Safety Cameras just rolled out an update that allows them to use artificial intelligence to detect human voices These cameras were sold to city governments as gunshot detectors, but with one update they turned them all into a mass surveillance system that records voices T
🚀🤖 In 2026, autonomous #AI agents aren’t failing because of technology ➜ they’re failing because of responsibility. ⚙️ Everyone knows how to build agents. Very few know how to own their decisions. 💡 Real autonomous AI isn’t about freedom or intelligence. It’s about clear scope,
Created 2026-05-13 · Updated 2026-05-13
Edit rule text
[tangential_ai_or_tech_business_announcement_without_healthcare_validation]
Learned 3 rejections Active
Exclude posts that announce AI agent tools (Claude Code, agent frameworks, MCP standards, Clawvisor, tool-calling systems) or software engineering practices unless the post demonstrates healthcare-specific validation, regulatory compliance requirements, or healthcare use case execution.
Posts about AI agent tools, agent security features, or general software engineering updates with loose or no connection to healthcare deployment or validation.
3 example posts
Agents calling tools now generate more revenue on Latch than scientists clicking buttons. From native deployments but also harnesses like Claude Code, Codex, Cursor across Pharma, biotech, academic labs. Product roadmap increasingly concerned with exposing compute intensive
As advanced agents move from coding to the rest of knowledge work, it takes a real amount of work and know-how to get right. You need to ensure agents have the right context and data to work with, wire up systems to agents in a safe and secure way, ensure that the agents are htt
The more I build agentic systems at Razorpay, the more I understand that - at its core, it is an agentic loop with tool calls, integrations, and retrieval. The hard part is... actually making it run reliably, at scale, under real production load. And this is what makes system de
Created 2026-05-13 · Updated 2026-05-13
Edit rule text
[speculative_robotics_sci_fi_labor_without_healthcare_systems]
Learned 3 rejections Active
Exclude posts that speculate about humanoid robots, autonomous systems, or sci-fi labor applications (e.g., robots unloading dishwashers, performing surgery by 2026) without demonstrating clinical validation, healthcare regulatory clearance, or how this addresses real healthcare system constraints.
Posts about humanoid robots, futuristic labor automation, or sci-fi applications with no grounding in healthcare systems or clinical validation.
3 example posts
The complexity of an AI driven robot that needs to meet GMP standards & learn complex steps of targeted cell manufacture. A marvel and the first robot of its kind in the world. Can’t wait till we lift off the first patient’s NK cells in LA from NANT LEONARDO. Unstoppable scal
I don’t think I’ve seen this take before but I like it. Musk has been world-leading at compressing money, resources, and time to make “known/hard” things at scale—make an electric car, make batteries, make a cheaper bigger rocket, all of which already existed but worse, at
A humanoid robot just unloaded a dishwasher for 67 hours straight. By end of 2026, the same robot will perform surgery. That's not a prediction from a sci-fi writer. That's Brett Adcock. CEO of Figure AI. Said it on camera this month. Here's what makes this insane. Two years
Created 2026-05-13 · Updated 2026-05-13
Edit rule text
[non_english_or_incomprehensible_posts]
Learned 3 rejections Active
Exclude posts written in languages other than English (e.g., Spanish, Portuguese) or posts using stream-of-consciousness, heavily slang, or jargon-laden text that is not decipherable to a general healthcare tech audience.
Posts written in non-English language or using unintelligible/jargon-heavy text that blocks comprehension and editorial assessment.
3 example posts
Un chico en China montó una agencia web entera con agentes de IA en Claude Code y ahora gestiona él solo 47 clientes al mes cobrando unos 400$ a cada uno. Sin empleados. Sin comerciales. Sin backend complejo. Solo: • Un MacBook • Un iPhone • 1 API key • Un sistema de 7 agentes
amazing writeup of experiments. essentially a model may already know a capability but that may be stored in a representation that is bad for extraction. you can slightly retrain/condition the model so the same capability becomes easier to isolate. must read for mech interp http
I've spent 150+ hours over the past year building agents on top of OpenClaw, NanoClaw, LangChain, using MCPs, n8n, Apify, and the rest of the automation/LLM stack we plow through for our PE clients, using DevriX and Growth Shuttle as a sandbox. It's been painful. Thread:
Created 2026-05-13 · Updated 2026-05-13
Edit rule text
[clinical_or_pharma_observation_without_systems_insight]
Learned 3 rejections Active
Exclude posts that share clinical trial results, drug safety announcements, or medical research findings without explaining healthcare delivery implications, system-level adoption barriers, or technology enablement. Clinical data alone is insufficient without systems context.
Posts reporting clinical trial data, drug results, or medical observations in isolation without connecting to healthcare delivery systems or technology implications.
3 example posts
💬 Editor's Note: AI-driven #ECG screening for left ventricular systolic dysfunction demonstrated high accuracy in Kenyan adults but should be interpreted with caution due to the high-risk enrichment of the study cohort. https://t.co/o3JS2wFrbC https://t.co/J4VNWqVHBe
CB-Dock3: An Enhanced Web Server for Protein–Ligand Blind Docking 1. CB-Dock3 is a major update of the CB-Dock blind docking web server, targeting the practical bottlenecks created by today’s surge of large Cryo-EM and AI-predicted protein structures (e.g., AlphaFold3-era https:
NEW: today OpenBind ‘comes out of stealth’ so to speak with their first data dump of ~900 novel protein-ligand structures - most with paired affinities This represents a meaningful %-age increase in all of humanities P-L data in the PDB collected in the last 50 years More👇 http
Created 2026-05-13 · Updated 2026-05-13
Edit rule text
[broad_unvalidated_health_claims_without_context]
Learned 3 rejections Active
Exclude posts that make broad, unsubstantiated claims about healthcare markets, trends, or outcomes (e.g., 'Eight in ten companies deployed AI but see no impact') without drilling into root causes, systems dynamics, operational context, or evidence-based analysis.
Posts making sweeping healthcare claims or market predictions without systems-level analysis, evidence, or contextual nuance.
3 example posts
Eight in ten companies have deployed generative AI. Eight in ten report no material impact on earnings. That's an operating-model gap. Not an adoption gap. Three days at SAS Innovate. Three moves. In order. https://t.co/uDgZ3K8Z9u https://t.co/EHsM1F5ms1
PBM owns the pharmacy. PBM owns the plan. PBM adjudicates the claim. Three commissions on one pill. The employer pays all three and gets a thank-you note for the rebate.
Incumbents don't defend broken systems. They defend profitable ones. Healthcare in America isn't a market failure. It's a sustaining innovation problem disguised as a social crisis. Every "reform" strengthens the existing architecture because the people writing the rules are the
Created 2026-05-13 · Updated 2026-05-13
Edit rule text
[business_model_or_market_debate_without_healthcare_systems_evidence]
Learned 3 rejections Active
Exclude posts that make broad claims about healthcare business models, profitability, market structure, or competitive dynamics (e.g., 'healthcare is a sustaining innovation problem,' 'three commissions on one pill') without providing data, healthcare system context, or actionable insight into how these dynamics affect patient care or operational outcomes.
Posts critiquing or debating healthcare business models (PBMs, insurers, supply chains) or broader market dynamics as commentary or observations without systems-level evidence or specifics.
3 example posts
PBM owns the pharmacy. PBM owns the plan. PBM adjudicates the claim. Three commissions on one pill. The employer pays all three and gets a thank-you note for the rebate.
Michigan Medicine generated $8.7 billion in revenue in FY25. It posted a $109 million operating margin in the same year. A nonprofit health system underwriting its largest pavilion in history off margin, not philanthropy. The accounting category and the operating reality do not
Incumbents don't defend broken systems. They defend profitable ones. Healthcare in America isn't a market failure. It's a sustaining innovation problem disguised as a social crisis. Every "reform" strengthens the existing architecture because the people writing the rules are the
Created 2026-05-13 · Updated 2026-05-13
Edit rule text
[broad_health_claims_or_market_debate_without_nuance]
Learned 3 rejections Active
Exclude posts that assert broad healthcare business model claims (e.g., 'healthcare is a sustaining innovation problem,' 'PBM structure is broken,' 'AI will replace doctors') without providing specific evidence, data, or analysis of how regulatory, payment, or organizational systems interact to create the claimed dysfunction.
Posts making broad claims about healthcare business models, market dynamics, or clinical practice without evidence or systems-level nuance.
3 example posts
Incumbents don't defend broken systems. They defend profitable ones. Healthcare in America isn't a market failure. It's a sustaining innovation problem disguised as a social crisis. Every "reform" strengthens the existing architecture because the people writing the rules are the
Europe is not losing to the US because it lacks engineers, capital or universities. It is losing because every layer of its system rewards the gatekeeper and punishes the builder. In medicine I see it weekly. A primary care AI that triages, documents and flags risk gets deployed
A reminder that AI can be used to reduce health inequities. Technology not being used in the US except for a few health systems, but shown to work well in Kenya @JAMACardio https://t.co/wyQ1k0dHES https://t.co/pP824X2qJG
Created 2026-05-13 · Updated 2026-05-13
Edit rule text
[non_healthcare_infrastructure_with_loose_framing]
Learned 3 rejections Active
Exclude posts about semiconductor fabs, cloud infrastructure updates, compute pricing, or hardware announcements — unless they explicitly connect to healthcare AI deployment requirements, regulatory constraints, or specific healthcare use cases.
Posts about tech infrastructure, semiconductor manufacturing, cloud platforms, or compute capacity with tenuous or absent healthcare application.
3 example posts
Last week, Google Cloud and AWS natively integrated x402, tackling head on one of its biggest problems: unauthorized API wrappers. The largest cloud platforms now give agents sanctioned ways to use services, pay with stables, and ditch API keys. It’s a new era. Pay (dot) sh’s
“Tokens are increasing 50x. You need to drive down cost at the same rate — or more.” At Google Cloud Next, @googlecloud's Mark Lohmeyer spoke with @PatrickMoorhead about the infrastructure realities behind the agent era: faster model training, lower-latency inference, and https:
SpaceX wants to build a $55 billion next-generation, vertically integrated semiconductor manufacturing and advanced computing fabrication facility in Grimes county, Texas, according to a new filing. • Proposed investment: $55B upfront, up to $119B total • Focus: semiconductor ma
Created 2026-05-13 · Updated 2026-05-13
Edit rule text
[business_model_critique_without_systems_evidence]
Learned 3 rejections Active
Exclude posts that critique PBMs, healthcare supply chains, insurance models, or market inefficiencies using anecdotes, broad claims, or macro observations — unless the post cites specific data on how those inefficiencies affect patient outcomes, provider workflows, or AI/automation deployment in healthcare.
Posts making broad critiques of healthcare business models or market structures without grounded data or healthcare systems insight.
3 example posts
PBM owns the pharmacy. PBM owns the plan. PBM adjudicates the claim. Three commissions on one pill. The employer pays all three and gets a thank-you note for the rebate.
Incumbents don't defend broken systems. They defend profitable ones. Healthcare in America isn't a market failure. It's a sustaining innovation problem disguised as a social crisis. Every "reform" strengthens the existing architecture because the people writing the rules are the
Europe is not losing to the US because it lacks engineers, capital or universities. It is losing because every layer of its system rewards the gatekeeper and punishes the builder. In medicine I see it weekly. A primary care AI that triages, documents and flags risk gets deployed
Created 2026-05-13 · Updated 2026-05-13
Edit rule text
[clinical_or_pharma_data_announcement_without_healthcare_systems_context]
Learned 3 rejections Active
Exclude posts announcing pharma trial results, drug candidate data, protein discovery, or genomics breakthroughs without analyzing how these findings impact clinical decision-making, healthcare economics, regulatory pathways, or system-wide patient access.
Posts about clinical trial results, drug data, or protein research announcements that lack healthcare system implications or operational context.
3 example posts
The complexity of an AI driven robot that needs to meet GMP standards & learn complex steps of targeted cell manufacture. A marvel and the first robot of its kind in the world. Can’t wait till we lift off the first patient’s NK cells in LA from NANT LEONARDO. Unstoppable scal
CB-Dock3: An Enhanced Web Server for Protein–Ligand Blind Docking 1. CB-Dock3 is a major update of the CB-Dock blind docking web server, targeting the practical bottlenecks created by today’s surge of large Cryo-EM and AI-predicted protein structures (e.g., AlphaFold3-era https:
NEW: today OpenBind ‘comes out of stealth’ so to speak with their first data dump of ~900 novel protein-ligand structures - most with paired affinities This represents a meaningful %-age increase in all of humanities P-L data in the PDB collected in the last 50 years More👇 http
Created 2026-05-13 · Updated 2026-05-13
Edit rule text
[broad_business_model_critique_without_healthcare_systems_evidence]
Learned 3 rejections Active
Exclude posts that make sweeping claims about PBM conflicts, incentive misalignment, or systemic healthcare problems without citing specific data, case studies, or demonstrating how these dynamics actually play out in patient outcomes or operational workflows.
Posts making broad claims about healthcare system dysfunction or market structure without empirical evidence or detailed systems analysis.
3 example posts
PBM owns the pharmacy. PBM owns the plan. PBM adjudicates the claim. Three commissions on one pill. The employer pays all three and gets a thank-you note for the rebate.
Incumbents don't defend broken systems. They defend profitable ones. Healthcare in America isn't a market failure. It's a sustaining innovation problem disguised as a social crisis. Every "reform" strengthens the existing architecture because the people writing the rules are the
Europe is not losing to the US because it lacks engineers, capital or universities. It is losing because every layer of its system rewards the gatekeeper and punishes the builder. In medicine I see it weekly. A primary care AI that triages, documents and flags risk gets deployed
Created 2026-05-13 · Updated 2026-05-13
Edit rule text
[ai_agent_security_or_safety_tangent]
Learned 3 rejections Active
Exclude posts that focus on generic AI agent security mechanisms (credential handling, unauthorized API access, agent governance frameworks) without grounding the risk in HIPAA, PHI exposure, clinical decision-making harm, or healthcare-specific threat models.
Posts about AI agent security vulnerabilities, access control, or trust frameworks disconnected from healthcare-specific risk or regulatory context.
3 example posts
"There are multiple levels of controlling the actions, but before that, you need provenance. You have to ask "do I trust the data?" and "do I know what data it was trained on?" - Dr. @leemonbaird Day 2 at @consensus2026: Leemon joined the panel “Agents Meet the Enterprise — Sup
🚀🤖 In 2026, autonomous #AI agents aren’t failing because of technology ➜ they’re failing because of responsibility. ⚙️ Everyone knows how to build agents. Very few know how to own their decisions. 💡 Real autonomous AI isn’t about freedom or intelligence. It’s about clear scope,
Clawvisor is going to be one of the most important parts of helping make the agent world especially OpenClaw/Hermes Agent secure and enterprise-grade. We're in the Apple I moment for personal AI, but we're ABOUT to see the Apple II - the first moment when everyone can use it
Created 2026-05-13 · Updated 2026-05-13
Edit rule text
[generalist_ai_capability_or_vulnerability_tangent]
Learned 3 rejections Active
Exclude posts focused on general AI technical capabilities (model alignment, agent architecture, zero-day vulnerabilities, inference optimization, neural network interpretability) even if authored by healthcare-adjacent figures or tagged with healthcare terms, unless they address healthcare-specific regulatory compliance, clinical validation, or healthcare delivery challenges.
Posts about general AI model capabilities, technical vulnerabilities, safety research, or agent architecture that are tangentially healthcare-labeled but lack healthcare application specificity.
3 example posts
💬 Editor's Note: AI-driven #ECG screening for left ventricular systolic dysfunction demonstrated high accuracy in Kenyan adults but should be interpreted with caution due to the high-risk enrichment of the study cohort. https://t.co/o3JS2wFrbC https://t.co/J4VNWqVHBe
"There are multiple levels of controlling the actions, but before that, you need provenance. You have to ask "do I trust the data?" and "do I know what data it was trained on?" - Dr. @leemonbaird Day 2 at @consensus2026: Leemon joined the panel “Agents Meet the Enterprise — Sup
So Mythos was, indeed, not marketing hype. Remember this is a general purpose model that just happens to be good at finding exploits because good models are good at lots of things. Expect similar from OpenAI & Google. And from open models in 8 months. https://t.co/KbhalQYX8
Created 2026-05-13 · Updated 2026-05-13
Edit rule text
[self_promotional_tool_or_startup_announcement_without_validation]
Learned 3 rejections Active
Exclude posts that are founder/company self-promotion of AI agents, tools, or products (e.g., posts from @levie, @arpit_bhayani, @no_fear_inc) unless they include third-party validation, healthcare-specific case studies, or peer-reviewed evidence of clinical or operational impact.
Posts by founders or company accounts promoting their own tools, agents, or startups without independent evidence of healthcare validation or clinical utility.
3 example posts
The more I build agentic systems at Razorpay, the more I understand that - at its core, it is an agentic loop with tool calls, integrations, and retrieval. The hard part is... actually making it run reliably, at scale, under real production load. And this is what makes system de
🚀🤖 In 2026, autonomous #AI agents aren’t failing because of technology ➜ they’re failing because of responsibility. ⚙️ Everyone knows how to build agents. Very few know how to own their decisions. 💡 Real autonomous AI isn’t about freedom or intelligence. It’s about clear scope,
I've spent 150+ hours over the past year building agents on top of OpenClaw, NanoClaw, LangChain, using MCPs, n8n, Apify, and the rest of the automation/LLM stack we plow through for our PE clients, using DevriX and Growth Shuttle as a sandbox. It's been painful. Thread:
Created 2026-05-13 · Updated 2026-05-13
Edit rule text
[non_healthcare_infrastructure_or_policy_with_loose_healthcare_label]
Learned 3 rejections Active
Exclude posts primarily about non-healthcare domains (SpaceX, semiconductor fabs, federal procurement mandates, cloud platform updates) even if loosely framed as relevant to healthcare. The post must directly address healthcare delivery, payment, clinical outcomes, or regulated health tech—not adjacent tech or policy.
Posts about tech infrastructure, semiconductor manufacturing, space exploration, or federal policy that are tangentially framed with healthcare tags but lack healthcare-specific substance.
3 example posts
Last week, Google Cloud and AWS natively integrated x402, tackling head on one of its biggest problems: unauthorized API wrappers. The largest cloud platforms now give agents sanctioned ways to use services, pay with stables, and ditch API keys. It’s a new era. Pay (dot) sh’s
“Tokens are increasing 50x. You need to drive down cost at the same rate — or more.” At Google Cloud Next, @googlecloud's Mark Lohmeyer spoke with @PatrickMoorhead about the infrastructure realities behind the agent era: faster model training, lower-latency inference, and https:
I don’t think I’ve seen this take before but I like it. Musk has been world-leading at compressing money, resources, and time to make “known/hard” things at scale—make an electric car, make batteries, make a cheaper bigger rocket, all of which already existed but worse, at
Created 2026-05-13 · Updated 2026-05-13
Edit rule text
[single_clinical_observation_without_systems_context]
Learned 3 rejections Active
Exclude posts that report a single clinical trial result, lab finding, or case observation without connecting it to healthcare delivery systems, market implications, policy context, or reimbursement dynamics. The post must provide systems-level insight, not just data reporting.
Posts highlighting a single clinical study, anecdote, or isolated data point without broader healthcare systems analysis
3 example posts
💬 Editor's Note: AI-driven #ECG screening for left ventricular systolic dysfunction demonstrated high accuracy in Kenyan adults but should be interpreted with caution due to the high-risk enrichment of the study cohort. https://t.co/o3JS2wFrbC https://t.co/J4VNWqVHBe
CB-Dock3: An Enhanced Web Server for Protein–Ligand Blind Docking 1. CB-Dock3 is a major update of the CB-Dock blind docking web server, targeting the practical bottlenecks created by today’s surge of large Cryo-EM and AI-predicted protein structures (e.g., AlphaFold3-era https:
NEW: today OpenBind ‘comes out of stealth’ so to speak with their first data dump of ~900 novel protein-ligand structures - most with paired affinities This represents a meaningful %-age increase in all of humanities P-L data in the PDB collected in the last 50 years More👇 http
Created 2026-05-12 · Updated 2026-05-12
Edit rule text
[speculative_sci_fi_robotics_without_healthcare_systems]
Learned 3 rejections Active
Exclude posts that describe speculative or sci-fi robotics applications (e.g., robots performing surgery 'by end of 2026' or unloading dishwashers for 67 hours) without credible evidence, healthcare system integration details, or regulatory pathway. These are hype narratives, not healthcare analysis.
Posts about futuristic robotics or speculative sci-fi applications (humanoid robots, Tesla Optimus, autonomous surgery claims) that lack grounding in healthcare realities or system constraints.
3 example posts
A humanoid robot just unloaded a dishwasher for 67 hours straight. By end of 2026, the same robot will perform surgery. That's not a prediction from a sci-fi writer. That's Brett Adcock. CEO of Figure AI. Said it on camera this month. Here's what makes this insane. Two years
SpaceX wants to build a $55 billion next-generation, vertically integrated semiconductor manufacturing and advanced computing fabrication facility in Grimes county, Texas, according to a new filing. • Proposed investment: $55B upfront, up to $119B total • Focus: semiconductor ma
I might feel nostalgic watching the last Tesla Model X roll off the line—except it's being replaced by the Optimus humanoid robot line, with plans to scale production to a million units per year. They will soon be in our factories, homes and operating rooms changing our world ht
Created 2026-05-12 · Updated 2026-05-12
Edit rule text
[generalist_ai_capability_without_healthcare_application]
Learned 3 rejections Active
Exclude posts that focus on AI model capabilities (Claude, OpenAI, Anthropic technical features), AI safety, cybersecurity vulnerabilities, or AI infrastructure without demonstrating concrete healthcare system impact or healthcare-specific application. Healthcare must be central, not a loose frame.
Posts about general AI capabilities, model features, or technical achievements that only tangentially mention healthcare or apply the healthcare label loosely.
3 example posts
The more I build agentic systems at Razorpay, the more I understand that - at its core, it is an agentic loop with tool calls, integrations, and retrieval. The hard part is... actually making it run reliably, at scale, under real production load. And this is what makes system de
amazing writeup of experiments. essentially a model may already know a capability but that may be stored in a representation that is bad for extraction. you can slightly retrain/condition the model so the same capability becomes easier to isolate. must read for mech interp http
So Mythos was, indeed, not marketing hype. Remember this is a general purpose model that just happens to be good at finding exploits because good models are good at lots of things. Expect similar from OpenAI & Google. And from open models in 8 months. https://t.co/KbhalQYX8
Created 2026-05-12 · Updated 2026-05-12
Edit rule text
[political_outrage_regulatory_posturing_without_analysis]
Learned 3 rejections Active
Exclude posts that report political statements, regulatory announcements, or government actions (e.g., FDA decisions, Trump administration policies) without analyzing the actual healthcare operational or economic consequences. Posts should connect policy to specific healthcare business model or delivery system changes.
Posts that amplify political or regulatory claims without substantive analysis of healthcare system impact.
3 example posts
I'm a board certified OBGYN. I've practiced for 20+ years. Here's what's actually happening with mifepristone, because the coverage of it has been a mess. On May 1, the 5th Circuit Court of Appeals ruled that the FDA overstepped its authority in 2023 when it eliminated the in-pe
🚨 UPDATE Friday afternoon: Trump said he has no plans to replace FDA Commissioner Marty Makary This morning: "Just got a call from inside the administration. Makary is dead in the water. Just a matter of time." -@thackerpd $HIMS $LLY $NVO https://t.co/ZjcjOAG7e1
🚨 NOW: President Trump confirms that ALL FEDERAL AGENCIES are now required to buy American, “NO excuses” 🔥 And 47 is ensuring all “Made in America” loopholes are DESTROYED I voted for this! “ALL FEDERAL AGENCIES MUST BUY AMERICAN — NO EXCUSES! For decades, Washington politici
Created 2026-05-12 · Updated 2026-05-12
Edit rule text
[generalist_ai_capability_or_infrastructure_hype_without_healthcare_application]
Learned 3 rejections Active
Exclude posts that discuss general AI technical capabilities (model scaling, inference costs, training data bottlenecks, code generation ability, multi-modal features) or infrastructure challenges that apply to AI broadly rather than healthcare deployment specifically. These are AI capability posts, not healthcare tech analysis.
Posts about AI model capabilities, compute infrastructure costs, token scaling, or data sourcing that apply broadly but lack healthcare-specific insight or validation
3 example posts
The more I build agentic systems at Razorpay, the more I understand that - at its core, it is an agentic loop with tool calls, integrations, and retrieval. The hard part is... actually making it run reliably, at scale, under real production load. And this is what makes system de
Larry Ellison just explained the most important and underreported insight in enterprise AI. Every major AI model, ChatGPT, Claude, Grok was trained on public data but public data has a ceiling. The real unlocking of AI's value happens when models can reason on private data, the
amazing writeup of experiments. essentially a model may already know a capability but that may be stored in a representation that is bad for extraction. you can slightly retrain/condition the model so the same capability becomes easier to isolate. must read for mech interp http
Created 2026-05-12 · Updated 2026-05-12
Edit rule text
[ai_safety_vulnerability_or_security_tangential]
Learned 3 rejections Active
Exclude posts that discuss AI safety concerns, zero-day vulnerabilities, agent security incidents, or AI model misuse (e.g., Mythos finding exploits, agents registering LLCs, authentication bypasses) unless directly connected to healthcare deployment risks or patient safety. These are AI safety tangents, not healthcare tech analysis.
Posts about AI safety incidents, security vulnerabilities, or misuse scenarios (agents going rogue, exploits found) without healthcare operational or patient safety implications
3 example posts
Flock Safety Cameras just rolled out an update that allows them to use artificial intelligence to detect human voices These cameras were sold to city governments as gunshot detectors, but with one update they turned them all into a mass surveillance system that records voices T
So Mythos was, indeed, not marketing hype. Remember this is a general purpose model that just happens to be good at finding exploits because good models are good at lots of things. Expect similar from OpenAI & Google. And from open models in 8 months. https://t.co/KbhalQYX8
Clawvisor is going to be one of the most important parts of helping make the agent world especially OpenClaw/Hermes Agent secure and enterprise-grade. We're in the Apple I moment for personal AI, but we're ABOUT to see the Apple II - the first moment when everyone can use it
Created 2026-05-12 · Updated 2026-05-12
Edit rule text
[self_promotional_tool_launch_without_validation]
Learned 3 rejections Active
Exclude posts that promote newly launched AI tools, agents, automation platforms, or SaaS products (even if healthcare-adjacent) based primarily on founder claims, feature announcements, or anecdotal use cases rather than clinical validation, pilot results, regulatory clearance, or peer-reviewed evidence.
Posts announcing AI tools, products, or services for healthcare without clinical validation, regulatory approval, or independent evidence of impact.
3 example posts
I've spent 150+ hours over the past year building agents on top of OpenClaw, NanoClaw, LangChain, using MCPs, n8n, Apify, and the rest of the automation/LLM stack we plow through for our PE clients, using DevriX and Growth Shuttle as a sandbox. It's been painful. Thread:
Clawvisor (@clawvisor) lets you give AI agents access to apps like Gmail and Slack without handing over your credentials or worrying they'll go rogue. You approve tasks once, Clawvisor enforce them. Congrats on the launch, @ericlevine! https://t.co/S51QlWauds https://t.co/NZvnt
NEW: today OpenBind ‘comes out of stealth’ so to speak with their first data dump of ~900 novel protein-ligand structures - most with paired affinities This represents a meaningful %-age increase in all of humanities P-L data in the PDB collected in the last 50 years More👇 http
Created 2026-05-12 · Updated 2026-05-12
Edit rule text
[speculative_sci_fi_robotics_labor]
Learned 3 rejections Active
Exclude posts that speculate about robots (Boston Dynamics, Tesla Optimus, etc.) replacing healthcare labor or performing clinical tasks based on unvalidated claims, sci-fi analogies, or founder hype rather than evidence-based timelines or healthcare-specific implementation challenges.
Posts about humanoid robots or speculative sci-fi applications that extrapolate to healthcare labor without grounding in current evidence or systems analysis
3 example posts
A humanoid robot just unloaded a dishwasher for 67 hours straight. By end of 2026, the same robot will perform surgery. That's not a prediction from a sci-fi writer. That's Brett Adcock. CEO of Figure AI. Said it on camera this month. Here's what makes this insane. Two years
SpaceX wants to build a $55 billion next-generation, vertically integrated semiconductor manufacturing and advanced computing fabrication facility in Grimes county, Texas, according to a new filing. • Proposed investment: $55B upfront, up to $119B total • Focus: semiconductor ma
I might feel nostalgic watching the last Tesla Model X roll off the line—except it's being replaced by the Optimus humanoid robot line, with plans to scale production to a million units per year. They will soon be in our factories, homes and operating rooms changing our world ht
Created 2026-05-12 · Updated 2026-05-12
Edit rule text
[speculative_robotics_and_labor_macro_without_healthcare_systems]
Learned 3 rejections Active
Exclude posts making speculative claims about humanoid robots, surgical automation, or labor disruption (e.g., 'robots will perform surgery by 2026', '1 million Optimus units per year') that lack validation, clinical evidence, or analysis of how these technologies integrate into actual healthcare workflows and regulation.
Posts hyping speculative robotics or humanoid AI applications with vague healthcare claims but no grounded healthcare systems analysis
3 example posts
I don’t think I’ve seen this take before but I like it. Musk has been world-leading at compressing money, resources, and time to make “known/hard” things at scale—make an electric car, make batteries, make a cheaper bigger rocket, all of which already existed but worse, at
A humanoid robot just unloaded a dishwasher for 67 hours straight. By end of 2026, the same robot will perform surgery. That's not a prediction from a sci-fi writer. That's Brett Adcock. CEO of Figure AI. Said it on camera this month. Here's what makes this insane. Two years
SpaceX wants to build a $55 billion next-generation, vertically integrated semiconductor manufacturing and advanced computing fabrication facility in Grimes county, Texas, according to a new filing. • Proposed investment: $55B upfront, up to $119B total • Focus: semiconductor ma
Created 2026-05-12 · Updated 2026-05-12
Edit rule text
[regulatory_fraud_scandal_without_healthcare_systems]
Learned 3 rejections Active
Exclude posts reporting healthcare fraud, Medicaid waste, or regulatory enforcement actions (FDA, CMS, etc.) unless they connect the fraud/waste to underlying healthcare system failures, payment incentives, or operational vulnerabilities that explain why the problem persists.
Posts about government fraud detection, regulatory enforcement, or healthcare fraud/abuse that lack analysis of systemic healthcare economics or operational root causes.
3 example posts
🚨 UPDATE Friday afternoon: Trump said he has no plans to replace FDA Commissioner Marty Makary This morning: "Just got a call from inside the administration. Makary is dead in the water. Just a matter of time." -@thackerpd $HIMS $LLY $NVO https://t.co/ZjcjOAG7e1
🚨 NOW: President Trump confirms that ALL FEDERAL AGENCIES are now required to buy American, “NO excuses” 🔥 And 47 is ensuring all “Made in America” loopholes are DESTROYED I voted for this! “ALL FEDERAL AGENCIES MUST BUY AMERICAN — NO EXCUSES! For decades, Washington politici
🛑FDA officials have blocked publication of several studies supporting the safety of widely used vaccines against Covid-19 and shingles in recent months, a spokesman for HHS confirmed. The studies, which cost millions of dollars in public funds, were conducted by scientists at htt
Created 2026-05-12 · Updated 2026-05-12
Edit rule text
[policy_fraud_scandal_without_systems_analysis]
Learned 3 rejections Active
Exclude posts that report healthcare fraud, waste, or regulatory scandals (Medicaid, PBMs, Medicare) as investigative news or political outrage without analyzing root causes in healthcare incentive structures, payment models, or systemic vulnerabilities.
Posts about healthcare fraud, Medicaid/CMS abuse, or policy scandals reported as news without deeper systems analysis
3 example posts
"Cordoba-owned buildings in Columbus housed 288 businesses registered with Medicaid... they charged taxpayers more than a quarter of a billion dollars between 2018 and 2024... in a city where only 6,273 people 75 or older are on Medicaid." - @realDailyWire https://t.co/7UnS7agwiw
If you had data on spending per SNAP terminal it would be trivial to find welfare fraud. Like 24 hours w/ a LLM and I can hand you a list of EBT card #s committing fraud and the EBT machine owners committing fraud. It’s not small money, either. The Mogadishu Store owner in
.@VP has been leading efforts to root out waste, fraud, and abuse across the federal government — putting accountability first. What we’ve seen in programs like SNAP is deeply concerning, with cases of improper payments, duplicate enrollments, and misuse of taxpayer dollars that
Created 2026-05-12 · Updated 2026-05-12
Edit rule text
[pharmaceutical_trial_or_drug_data_announcement]
Learned 3 rejections Active
Exclude posts that report drug trial outcomes, FDA decisions, or pharmaceutical data (e.g., GLP-1 results, gene therapy approvals, vaccine studies) as pure clinical/regulatory announcements. Include only if the post analyzes market access, reimbursement barriers, or systemic healthcare implications.
Posts announcing pharmaceutical trial results, drug efficacy data, or FDA approvals without healthcare systems analysis
3 example posts
$vrtx drops $mrna-partnered Cystic Fibrosis mRNA therapy VX-522 over inhaled LNP tolerability issues: https://t.co/WlIIHXxg9x Not too surprising IMO. Important Q: is safety/tolerability of this approach still good enough for one-time/infrequent genome editing? $prme $arct
If you’ve been taking GLP-1s, even for a little while, you may be wondering how to maintain the weight loss you’ve achieved. Our endocrinologist explains what works and what to avoid. https://t.co/TU8LUmg9Vn
In adults ≥75 years, 10-year #ColorectalCancer and CRC death risk were higher with prior adenoma but both were low (<1.1%); non-CRC death risk was 47-48%, supporting deprioritizing repeat surveillance. https://t.co/mJ7r4vAtK1 https://t.co/dUCYgO0SuK
Created 2026-05-12 · Updated 2026-05-12
Edit rule text
[ai_agent_security_tangential]
Learned 3 rejections Active
Exclude posts that focus on AI agent security flaws, credential theft, rogue behavior, or misuse scenarios (e.g., agents registering LLCs, opening bank accounts) unless they directly address a healthcare-specific security risk or regulatory compliance problem.
Posts about AI agent security vulnerabilities, jailbreaks, or misuse incidents without healthcare application context
3 example posts
Clawvisor is going to be one of the most important parts of helping make the agent world especially OpenClaw/Hermes Agent secure and enterprise-grade. We're in the Apple I moment for personal AI, but we're ABOUT to see the Apple II - the first moment when everyone can use it
I've spent 150+ hours over the past year building agents on top of OpenClaw, NanoClaw, LangChain, using MCPs, n8n, Apify, and the rest of the automation/LLM stack we plow through for our PE clients, using DevriX and Growth Shuttle as a sandbox. It's been painful. Thread:
Clawvisor (@clawvisor) lets you give AI agents access to apps like Gmail and Slack without handing over your credentials or worrying they'll go rogue. You approve tasks once, Clawvisor enforce them. Congrats on the launch, @ericlevine! https://t.co/S51QlWauds https://t.co/NZvnt
Created 2026-05-12 · Updated 2026-05-12
Edit rule text
[tangential_biotech_founder_enthusiasm_without_systems_validation]
Learned 3 rejections Active
Exclude posts that showcase founder enthusiasm, exciting lab results, novel molecular structures, or capability announcements (e.g., 'OpenBind releases 900 novel protein-ligand structures') without connecting to healthcare delivery, clinical trials, regulatory approval pathways, or actual patient impact.
Posts featuring biotech founder excitement, lab updates, or novel capability announcements without healthcare systems context or clinical validation
3 example posts
CB-Dock3: An Enhanced Web Server for Protein–Ligand Blind Docking 1. CB-Dock3 is a major update of the CB-Dock blind docking web server, targeting the practical bottlenecks created by today’s surge of large Cryo-EM and AI-predicted protein structures (e.g., AlphaFold3-era https:
NEW: today OpenBind ‘comes out of stealth’ so to speak with their first data dump of ~900 novel protein-ligand structures - most with paired affinities This represents a meaningful %-age increase in all of humanities P-L data in the PDB collected in the last 50 years More👇 http
not labelled! no stains! just cell chemistry changing in response to a drug !!! you can see it change before it dies !!! you can predict what happens to a cell !!! that's constantly being dosed with a drug !!! you know what? this is just an ICU for living human cells https://t.
Created 2026-05-12 · Updated 2026-05-12
Edit rule text
[ai_infrastructure_compute_hype_without_healthcare_specificity]
Learned 3 rejections Active
Exclude posts that discuss token costs, compute scaling (e.g., 'tokens increasing 50x'), cloud infrastructure integrations, or semiconductor manufacturing without demonstrating how these infrastructure advances solve a specific healthcare problem or enable healthcare deployment.
Posts about AI compute costs, token efficiency, infrastructure scaling, or cloud platform updates that have no healthcare application or validation
3 example posts
Last week, Google Cloud and AWS natively integrated x402, tackling head on one of its biggest problems: unauthorized API wrappers. The largest cloud platforms now give agents sanctioned ways to use services, pay with stables, and ditch API keys. It’s a new era. Pay (dot) sh’s
“Tokens are increasing 50x. You need to drive down cost at the same rate — or more.” At Google Cloud Next, @googlecloud's Mark Lohmeyer spoke with @PatrickMoorhead about the infrastructure realities behind the agent era: faster model training, lower-latency inference, and https:
SpaceX wants to build a $55 billion next-generation, vertically integrated semiconductor manufacturing and advanced computing fabrication facility in Grimes county, Texas, according to a new filing. • Proposed investment: $55B upfront, up to $119B total • Focus: semiconductor ma
Created 2026-05-12 · Updated 2026-05-12
Edit rule text
[tangential_ai_agent_or_product_announcement_without_healthcare_specificity]
Learned 3 rejections Active
Exclude posts that announce new AI agent tools, security frameworks, or automation platforms (Clawvisor, OpenClaw, MCPs, n8n) using hypothetical examples or non-healthcare use cases (e.g., web agencies, Stripe automations) without proof of healthcare-specific deployment or ROI.
Posts announcing AI agent frameworks, tools, or products with anecdotes but no healthcare validation
3 example posts
"There are multiple levels of controlling the actions, but before that, you need provenance. You have to ask "do I trust the data?" and "do I know what data it was trained on?" - Dr. @leemonbaird Day 2 at @consensus2026: Leemon joined the panel “Agents Meet the Enterprise — Sup
Last week, Google Cloud and AWS natively integrated x402, tackling head on one of its biggest problems: unauthorized API wrappers. The largest cloud platforms now give agents sanctioned ways to use services, pay with stables, and ditch API keys. It’s a new era. Pay (dot) sh’s
Clawvisor is going to be one of the most important parts of helping make the agent world especially OpenClaw/Hermes Agent secure and enterprise-grade. We're in the Apple I moment for personal AI, but we're ABOUT to see the Apple II - the first moment when everyone can use it
Created 2026-05-11 · Updated 2026-05-11
Edit rule text
[broad_ai_infrastructure_or_data_hype_without_healthcare_validation]
Learned 3 rejections Active
Exclude posts that discuss general AI infrastructure topics (token scaling, public data ceilings, model training approaches, protein mapping, code quality) even if loosely framed as healthcare, unless the post demonstrates how the specific capability solves a documented healthcare workflow or business problem.
Posts about foundational AI capabilities, data bottlenecks, or compute trends without healthcare application validation
3 example posts
Larry Ellison just explained the most important and underreported insight in enterprise AI. Every major AI model, ChatGPT, Claude, Grok was trained on public data but public data has a ceiling. The real unlocking of AI's value happens when models can reason on private data, the
CB-Dock3: An Enhanced Web Server for Protein–Ligand Blind Docking 1. CB-Dock3 is a major update of the CB-Dock blind docking web server, targeting the practical bottlenecks created by today’s surge of large Cryo-EM and AI-predicted protein structures (e.g., AlphaFold3-era https:
So Mythos was, indeed, not marketing hype. Remember this is a general purpose model that just happens to be good at finding exploits because good models are good at lots of things. Expect similar from OpenAI & Google. And from open models in 8 months. https://t.co/KbhalQYX8
Created 2026-05-11 · Updated 2026-05-11
Edit rule text
[tangential_healthcare_framing_of_policy]
Learned 3 rejections Active
Exclude posts that are primarily about political figures, FDA appointments, government mandates (buy American, federal spending), or regulatory announcements without analyzing how these decisions impact healthcare delivery, reimbursement models, or business operations.
Posts about government policy, FDA regulatory announcements, or political figures that lack substantive healthcare business analysis
3 example posts
🚨 UPDATE Friday afternoon: Trump said he has no plans to replace FDA Commissioner Marty Makary This morning: "Just got a call from inside the administration. Makary is dead in the water. Just a matter of time." -@thackerpd $HIMS $LLY $NVO https://t.co/ZjcjOAG7e1
🚨 NOW: President Trump confirms that ALL FEDERAL AGENCIES are now required to buy American, “NO excuses” 🔥 And 47 is ensuring all “Made in America” loopholes are DESTROYED I voted for this! “ALL FEDERAL AGENCIES MUST BUY AMERICAN — NO EXCUSES! For decades, Washington politici
🛑FDA officials have blocked publication of several studies supporting the safety of widely used vaccines against Covid-19 and shingles in recent months, a spokesman for HHS confirmed. The studies, which cost millions of dollars in public funds, were conducted by scientists at htt
Created 2026-05-11 · Updated 2026-05-11
Edit rule text
[generalist_infrastructure_or_macro_policy_without_healthcare_lens]
Learned 3 rejections Active
Exclude posts about energy infrastructure, semiconductor manufacturing, government industrial policy, space technology, or compute scaling that mention healthcare only in passing or with speculative framing. The post must demonstrate concrete healthcare operational dependency or clinical workflow impact.
Posts about general infrastructure, compute capacity, energy, semiconductors, or macro policy tangentially labeled healthcare without substantive healthcare application.
3 example posts
Execs at Palantir, whose stock is up ~16x since its AI platform's 2023 debut, often decry other AI as slop, as competition from frontier AI labs stiffens (@heathersomervil / Wall Street Journal) (Visit Techmeme dot com for the link and full context!)
Larry Ellison just explained the most important and underreported insight in enterprise AI. Every major AI model, ChatGPT, Claude, Grok was trained on public data but public data has a ceiling. The real unlocking of AI's value happens when models can reason on private data, the
Three weeks ago I called Anthropic rationing inference and OpenAI hitting the wall next. Yesterday OpenAI shipped GPT-5.5 Instant. New default for free, Plus, and Pro. 30% fewer words per answer. One frame going around: cost savings. It's both bigger and smaller than that. http
Created 2026-05-11 · Updated 2026-05-11
Edit rule text
[ai_safety_vulnerability_or_security_tangent_without_healthcare_application]
Learned 3 rejections Active
Exclude posts that discuss AI safety, security incidents, vulnerability disclosures, or agent misuse scenarios in abstract terms. The post must identify specific healthcare-relevant threats (e.g., clinical decision hijacking, HIPAA breach mechanisms, diagnostic delay from agent error) not generic AI security concerns.
Posts about AI safety, cybersecurity vulnerabilities, or misuse risks presented generically without specific healthcare threat model or clinical workflow vulnerability.
3 example posts
"There are multiple levels of controlling the actions, but before that, you need provenance. You have to ask "do I trust the data?" and "do I know what data it was trained on?" - Dr. @leemonbaird Day 2 at @consensus2026: Leemon joined the panel “Agents Meet the Enterprise — Sup
Clawvisor is going to be one of the most important parts of helping make the agent world especially OpenClaw/Hermes Agent secure and enterprise-grade. We're in the Apple I moment for personal AI, but we're ABOUT to see the Apple II - the first moment when everyone can use it
I've spent 150+ hours over the past year building agents on top of OpenClaw, NanoClaw, LangChain, using MCPs, n8n, Apify, and the rest of the automation/LLM stack we plow through for our PE clients, using DevriX and Growth Shuttle as a sandbox. It's been painful. Thread:
Created 2026-05-11 · Updated 2026-05-11
Edit rule text
[enterprise_ai_adoption_without_healthcare_context]
Learned 3 rejections Active
Exclude posts discussing enterprise AI adoption rates, deployment challenges, workflow redesign, or McKinsey-style adoption metrics unless they contain specific healthcare system or operational context. Posts about 'agentic loops,' 'tool integrations,' or general enterprise AI deployment without healthcare application should be rejected.
Posts about general enterprise AI adoption, workflow redesign, and deployment challenges that lack healthcare-specific analysis or application.
3 example posts
As advanced agents move from coding to the rest of knowledge work, it takes a real amount of work and know-how to get right. You need to ensure agents have the right context and data to work with, wire up systems to agents in a safe and secure way, ensure that the agents are htt
The more I build agentic systems at Razorpay, the more I understand that - at its core, it is an agentic loop with tool calls, integrations, and retrieval. The hard part is... actually making it run reliably, at scale, under real production load. And this is what makes system de
Eight in ten companies have deployed generative AI. Eight in ten report no material impact on earnings. That's an operating-model gap. Not an adoption gap. Three days at SAS Innovate. Three moves. In order. https://t.co/uDgZ3K8Z9u https://t.co/EHsM1F5ms1
Created 2026-05-11 · Updated 2026-05-11
Edit rule text
[ai_agent_or_model_technical_capability_hype_without_healthcare_validation]
Learned 3 rejections Active
Exclude posts that hype AI agent features, model capabilities, or AI company product launches (code interpreters, reasoning models, agent frameworks) without demonstrating healthcare-specific utility, clinical validation, or operational deployment. Posts must show healthcare application, not just technical capability.
Posts about AI agent technical capabilities, model advances, or AI company product launches that are not grounded in healthcare application or validation.
3 example posts
amazing writeup of experiments. essentially a model may already know a capability but that may be stored in a representation that is bad for extraction. you can slightly retrain/condition the model so the same capability becomes easier to isolate. must read for mech interp http
So Mythos was, indeed, not marketing hype. Remember this is a general purpose model that just happens to be good at finding exploits because good models are good at lots of things. Expect similar from OpenAI & Google. And from open models in 8 months. https://t.co/KbhalQYX8
Clawvisor is going to be one of the most important parts of helping make the agent world especially OpenClaw/Hermes Agent secure and enterprise-grade. We're in the Apple I moment for personal AI, but we're ABOUT to see the Apple II - the first moment when everyone can use it
Created 2026-05-11 · Updated 2026-05-11
Edit rule text
[broad_unvalidated_health_claims_or_market_debate_without_systems_analysis]
Learned 3 rejections Active
Exclude posts that make broad assertions about healthcare business models, market consolidation, or system-level problems (e.g., 'PBMs are broken,' 'incumbents defend profitable systems') without providing data, specific examples, or analysis of how systems actually function or could be reformed. Opinion-driven polemics without evidence or specificity do not qualify.
Posts making sweeping claims about healthcare systems, markets, or business models without evidence, nuance, or substantive operational analysis.
3 example posts
PBM owns the pharmacy. PBM owns the plan. PBM adjudicates the claim. Three commissions on one pill. The employer pays all three and gets a thank-you note for the rebate.
Incumbents don't defend broken systems. They defend profitable ones. Healthcare in America isn't a market failure. It's a sustaining innovation problem disguised as a social crisis. Every "reform" strengthens the existing architecture because the people writing the rules are the
Statin guidelines over 50 years: 1970s: "Total cholesterol over 280 is high." Result: a relatively small market. 1988: "Actually, 240 is high." Result: millions more added overnight. 2001: "Actually, focus on LDL. Get it under 100." Result: tens of millions of new customers.
Created 2026-05-11 · Updated 2026-05-11
Edit rule text
[enterprise_ai_adoption_without_healthcare_application]
Learned 3 rejections Active
Exclude posts that discuss enterprise AI adoption, workflow automation, change management, or AI deployment strategies in generic terms without concrete healthcare system, payment, or clinical operation examples. Posts must demonstrate healthcare-specific operational challenges or solutions, not abstract enterprise implementation patterns.
Posts about general enterprise AI adoption, workflow redesign, or deployment challenges that lack specific healthcare application or systems context.
3 example posts
As advanced agents move from coding to the rest of knowledge work, it takes a real amount of work and know-how to get right. You need to ensure agents have the right context and data to work with, wire up systems to agents in a safe and secure way, ensure that the agents are htt
The more I build agentic systems at Razorpay, the more I understand that - at its core, it is an agentic loop with tool calls, integrations, and retrieval. The hard part is... actually making it run reliably, at scale, under real production load. And this is what makes system de
The AI bolt-on era, illustrated. McKinsey: 6% of companies see real EBIT gains from AI. They're 3x more likely to redesign workflows around AI than bolt it onto existing tools. I love Anthropic. They just shipped a better bolt-on. https://t.co/tCQcPMywGh
Created 2026-05-11 · Updated 2026-05-11
Edit rule text
[policy_fraud_scandal_without_healthcare_analysis]
Learned 3 rejections Active
Exclude posts that report healthcare fraud scandals, regulatory action, or political healthcare claims (Trump policies, FDA decisions, PBM practices, Medicaid cuts) unless the post includes substantive analysis of how these policy changes affect healthcare system incentives, operations, or outcomes—not just political theater or announcement.
Posts about regulatory fraud, Medicare/Medicaid scandal, or political healthcare posturing without systems-level healthcare analysis.
3 example posts
🚨 UPDATE Friday afternoon: Trump said he has no plans to replace FDA Commissioner Marty Makary This morning: "Just got a call from inside the administration. Makary is dead in the water. Just a matter of time." -@thackerpd $HIMS $LLY $NVO https://t.co/ZjcjOAG7e1
🚨 NOW: President Trump confirms that ALL FEDERAL AGENCIES are now required to buy American, “NO excuses” 🔥 And 47 is ensuring all “Made in America” loopholes are DESTROYED I voted for this! “ALL FEDERAL AGENCIES MUST BUY AMERICAN — NO EXCUSES! For decades, Washington politici
🛑FDA officials have blocked publication of several studies supporting the safety of widely used vaccines against Covid-19 and shingles in recent months, a spokesman for HHS confirmed. The studies, which cost millions of dollars in public funds, were conducted by scientists at htt
Created 2026-05-11 · Updated 2026-05-11
Edit rule text
[tangential_ai_agent_security]
Learned 3 rejections Active
Exclude posts that report AI agent security incidents (agents going rogue, unauthorized API access, misbehavior in controlled experiments) or discuss hypothetical trust/security concerns without demonstrating healthcare-specific deployment risk or validated mitigation in clinical workflows.
Posts about AI agent security incidents, misuse, unauthorized access, or trust problems without grounded healthcare deployment context.
3 example posts
"There are multiple levels of controlling the actions, but before that, you need provenance. You have to ask "do I trust the data?" and "do I know what data it was trained on?" - Dr. @leemonbaird Day 2 at @consensus2026: Leemon joined the panel “Agents Meet the Enterprise — Sup
🚀🤖 In 2026, autonomous #AI agents aren’t failing because of technology ➜ they’re failing because of responsibility. ⚙️ Everyone knows how to build agents. Very few know how to own their decisions. 💡 Real autonomous AI isn’t about freedom or intelligence. It’s about clear scope,
Clawvisor is going to be one of the most important parts of helping make the agent world especially OpenClaw/Hermes Agent secure and enterprise-grade. We're in the Apple I moment for personal AI, but we're ABOUT to see the Apple II - the first moment when everyone can use it
Created 2026-05-11 · Updated 2026-05-11
Edit rule text
[business_model_or_market_debate_without_healthcare_systems_insight]
Learned 3 rejections Active
Exclude posts that critique healthcare business models, pricing, or market structure (e.g., 'PBM owns the pharmacy, plan, and adjudicates the claim—three commissions on one pill') if they function as economic complaint without systemic analysis of how the structure affects clinical decisions, patient outcomes, or care equity.
Posts that debate healthcare market structure, PBM economics, pricing dynamics, or business model criticism without connecting to patient access, care quality, or provider incentives.
3 example posts
PBM owns the pharmacy. PBM owns the plan. PBM adjudicates the claim. Three commissions on one pill. The employer pays all three and gets a thank-you note for the rebate.
Michigan Medicine generated $8.7 billion in revenue in FY25. It posted a $109 million operating margin in the same year. A nonprofit health system underwriting its largest pavilion in history off margin, not philanthropy. The accounting category and the operating reality do not
Incumbents don't defend broken systems. They defend profitable ones. Healthcare in America isn't a market failure. It's a sustaining innovation problem disguised as a social crisis. Every "reform" strengthens the existing architecture because the people writing the rules are the
Created 2026-05-11 · Updated 2026-05-11
Edit rule text
[enterprise_ai_adoption_metrics_without_healthcare_application]
Learned 3 rejections Active
Exclude posts reporting aggregate AI adoption statistics, McKinsey-style enterprise adoption benchmarks, or workflow redesign percentages unless they specifically analyze healthcare delivery, clinical workflows, or patient outcomes. Posts about 'X% of companies deployed AI' or 'companies 3x more likely to redesign workflows' without healthcare application details should be rejected.
Posts about corporate AI adoption rates, deployment gaps, or workflow redesign trends that lack healthcare-specific operational context
3 example posts
Eight in ten companies have deployed generative AI. Eight in ten report no material impact on earnings. That's an operating-model gap. Not an adoption gap. Three days at SAS Innovate. Three moves. In order. https://t.co/uDgZ3K8Z9u https://t.co/EHsM1F5ms1
The AI bolt-on era, illustrated. McKinsey: 6% of companies see real EBIT gains from AI. They're 3x more likely to redesign workflows around AI than bolt it onto existing tools. I love Anthropic. They just shipped a better bolt-on. https://t.co/tCQcPMywGh
there's a deeper underestimated reason: consumers just do not have very many long chains of tasks which can be automated end to end to unlock huge value. their chains are shallow almost always in enterprise, org processes form very long chains, so the reward for automation is m
Created 2026-05-11 · Updated 2026-05-11
Edit rule text
[enterprise_ai_adoption_without_healthcare_systems_analysis]
Learned 3 rejections Active
Exclude posts that discuss AI adoption rates, enterprise workflow redesign, operating model gaps, or McKinsey-style business transformation metrics in general enterprise contexts unless the post specifically analyzes how healthcare delivery models, clinical workflows, or healthcare organization structures are being transformed.
Posts about AI adoption, workflow redesign, or operating model transformation in enterprises without healthcare-specific context or systems analysis.
3 example posts
Eight in ten companies have deployed generative AI. Eight in ten report no material impact on earnings. That's an operating-model gap. Not an adoption gap. Three days at SAS Innovate. Three moves. In order. https://t.co/uDgZ3K8Z9u https://t.co/EHsM1F5ms1
The AI bolt-on era, illustrated. McKinsey: 6% of companies see real EBIT gains from AI. They're 3x more likely to redesign workflows around AI than bolt it onto existing tools. I love Anthropic. They just shipped a better bolt-on. https://t.co/tCQcPMywGh
WestRock CEO went from Palantir skeptic to biggest believer. “What I read and what we see talked about in the AI world and what Palantir’s operating system actually is, I can barely recognize the reality of what they’re doing on the ground with the talk that goes on around AI.”
Created 2026-05-11 · Updated 2026-05-11
Edit rule text
[data_infrastructure_or_ai_capability_hype_without_healthcare_validation]
Learned 3 rejections Active
Exclude posts that highlight AI model features (Mythos vulnerabilities, Claude capabilities, token scaling), data infrastructure solutions, or general-purpose AI technical advances unless the post demonstrates or analyzes actual healthcare use cases, clinical validation, or provider adoption.
Posts celebrating AI model capabilities, data infrastructure advances, or technical breakthroughs without evidence of healthcare application or validation
3 example posts
amazing writeup of experiments. essentially a model may already know a capability but that may be stored in a representation that is bad for extraction. you can slightly retrain/condition the model so the same capability becomes easier to isolate. must read for mech interp http
CB-Dock3: An Enhanced Web Server for Protein–Ligand Blind Docking 1. CB-Dock3 is a major update of the CB-Dock blind docking web server, targeting the practical bottlenecks created by today’s surge of large Cryo-EM and AI-predicted protein structures (e.g., AlphaFold3-era https:
So Mythos was, indeed, not marketing hype. Remember this is a general purpose model that just happens to be good at finding exploits because good models are good at lots of things. Expect similar from OpenAI & Google. And from open models in 8 months. https://t.co/KbhalQYX8
Created 2026-05-11 · Updated 2026-05-11
Edit rule text
[ai_agent_or_autonomous_system_security_tangential]
Learned 3 rejections Active
Exclude posts that discuss AI agent security incidents, unauthorized behavior, trust problems, or responsibility frameworks in abstract or non-healthcare contexts (agents registering LLCs, API security, task validation) unless the post applies the concern to a specific healthcare workflow or patient safety risk.
Posts about AI agent misuse, security vulnerabilities, or responsibility issues framed generically without healthcare-specific risk
3 example posts
🚀🤖 In 2026, autonomous #AI agents aren’t failing because of technology ➜ they’re failing because of responsibility. ⚙️ Everyone knows how to build agents. Very few know how to own their decisions. 💡 Real autonomous AI isn’t about freedom or intelligence. It’s about clear scope,
Clawvisor is going to be one of the most important parts of helping make the agent world especially OpenClaw/Hermes Agent secure and enterprise-grade. We're in the Apple I moment for personal AI, but we're ABOUT to see the Apple II - the first moment when everyone can use it
Clawvisor (@clawvisor) lets you give AI agents access to apps like Gmail and Slack without handing over your credentials or worrying they'll go rogue. You approve tasks once, Clawvisor enforce them. Congrats on the launch, @ericlevine! https://t.co/S51QlWauds https://t.co/NZvnt
Created 2026-05-11 · Updated 2026-05-11
Edit rule text
[ai_company_product_or_business_metrics_unvalidated]
Learned 3 rejections Active
Exclude posts that report AI company product launches, capability announcements, partnerships, or business developments (e.g., Google Cloud integrating x402, Stripe creating a role) without demonstrating actual healthcare use, adoption, or impact. These are tech news, not healthcare analysis.
Posts announcing AI company products, capabilities, partnerships, or business metrics without evidence of healthcare application or validation.
3 example posts
Last week, Google Cloud and AWS natively integrated x402, tackling head on one of its biggest problems: unauthorized API wrappers. The largest cloud platforms now give agents sanctioned ways to use services, pay with stables, and ditch API keys. It’s a new era. Pay (dot) sh’s
Larry Ellison just explained the most important and underreported insight in enterprise AI. Every major AI model, ChatGPT, Claude, Grok was trained on public data but public data has a ceiling. The real unlocking of AI's value happens when models can reason on private data, the
Three weeks ago I called Anthropic rationing inference and OpenAI hitting the wall next. Yesterday OpenAI shipped GPT-5.5 Instant. New default for free, Plus, and Pro. 30% fewer words per answer. One frame going around: cost savings. It's both bigger and smaller than that. http
Created 2026-05-11 · Updated 2026-05-11
Edit rule text
[regulatory_fraud_scandal_or_policy_reporting_without_healthcare_systems_analysis]
Learned 3 rejections Active
Exclude posts that report fraud (PBM fraud, Medicaid fraud, EBT card fraud), regulatory violations, or policy failures as isolated incidents or moral outrage—unless the post analyzes systemic incentive structures, reimbursement design flaws, or operational barriers that enabled the problem.
Posts reporting healthcare fraud, regulatory scandal, or policy violations as news without connecting to systemic causes or healthcare delivery impact.
3 example posts
PBM owns the pharmacy. PBM owns the plan. PBM adjudicates the claim. Three commissions on one pill. The employer pays all three and gets a thank-you note for the rebate.
"Cordoba-owned buildings in Columbus housed 288 businesses registered with Medicaid... they charged taxpayers more than a quarter of a billion dollars between 2018 and 2024... in a city where only 6,273 people 75 or older are on Medicaid." - @realDailyWire https://t.co/7UnS7agwiw
If you had data on spending per SNAP terminal it would be trivial to find welfare fraud. Like 24 hours w/ a LLM and I can hand you a list of EBT card #s committing fraud and the EBT machine owners committing fraud. It’s not small money, either. The Mogadishu Store owner in
Created 2026-05-10 · Updated 2026-05-10
Edit rule text
[pharmaceutical_trial_data_or_market_pricing_without_systems_analysis]
Learned 3 rejections Active
Exclude posts that report clinical trial data (JAMA/Nature publications), pharmaceutical pricing announcements, GLP-1/semaglutide/tirzepatide market dynamics, or drug approval/denial decisions without analyzing impact on healthcare delivery, reimbursement, access, or system efficiency.
Posts announcing drug trial results, pharma market pricing, competitive dynamics, or GLP-1/peptide pricing without healthcare system economics analysis.
3 example posts
$vrtx drops $mrna-partnered Cystic Fibrosis mRNA therapy VX-522 over inhaled LNP tolerability issues: https://t.co/WlIIHXxg9x Not too surprising IMO. Important Q: is safety/tolerability of this approach still good enough for one-time/infrequent genome editing? $prme $arct
If you’ve been taking GLP-1s, even for a little while, you may be wondering how to maintain the weight loss you’ve achieved. Our endocrinologist explains what works and what to avoid. https://t.co/TU8LUmg9Vn
not labelled! no stains! just cell chemistry changing in response to a drug !!! you can see it change before it dies !!! you can predict what happens to a cell !!! that's constantly being dosed with a drug !!! you know what? this is just an ICU for living human cells https://t.
Created 2026-05-10 · Updated 2026-05-10
Edit rule text
[pharmaceutical_trial_data_without_systems]
Learned 3 rejections Active
Exclude posts reporting pharmaceutical trial data, drug approvals, gene therapy results, or pharma product announcements unless they analyze healthcare system adoption barriers, reimbursement mechanics, or delivery infrastructure implications.
Posts about drug trial results, pharmaceutical data, or gene therapy approvals without healthcare economics or delivery system context.
3 example posts
CB-Dock3: An Enhanced Web Server for Protein–Ligand Blind Docking 1. CB-Dock3 is a major update of the CB-Dock blind docking web server, targeting the practical bottlenecks created by today’s surge of large Cryo-EM and AI-predicted protein structures (e.g., AlphaFold3-era https:
NEW: today OpenBind ‘comes out of stealth’ so to speak with their first data dump of ~900 novel protein-ligand structures - most with paired affinities This represents a meaningful %-age increase in all of humanities P-L data in the PDB collected in the last 50 years More👇 http
$vrtx drops $mrna-partnered Cystic Fibrosis mRNA therapy VX-522 over inhaled LNP tolerability issues: https://t.co/WlIIHXxg9x Not too surprising IMO. Important Q: is safety/tolerability of this approach still good enough for one-time/infrequent genome editing? $prme $arct
Created 2026-05-10 · Updated 2026-05-10
Edit rule text
[healthcare_fraud_scandal_reporting]
Learned 3 rejections Active
Exclude posts that report healthcare fraud, welfare abuse, Medicaid waste, or regulatory scandals as isolated incidents or outrage without analyzing root causes in the healthcare system design or business model structure.
Posts about healthcare fraud, waste, or corruption that treat it as scandal reporting without systems analysis.
3 example posts
PBM owns the pharmacy. PBM owns the plan. PBM adjudicates the claim. Three commissions on one pill. The employer pays all three and gets a thank-you note for the rebate.
🛑FDA officials have blocked publication of several studies supporting the safety of widely used vaccines against Covid-19 and shingles in recent months, a spokesman for HHS confirmed. The studies, which cost millions of dollars in public funds, were conducted by scientists at htt
"Cordoba-owned buildings in Columbus housed 288 businesses registered with Medicaid... they charged taxpayers more than a quarter of a billion dollars between 2018 and 2024... in a city where only 6,273 people 75 or older are on Medicaid." - @realDailyWire https://t.co/7UnS7agwiw
Created 2026-05-10 · Updated 2026-05-10
Edit rule text
[generalist_ai_model_capabilities]
Learned 3 rejections Active
Exclude posts about general AI model capabilities (alignment, inference, training techniques, scaling laws) that don't demonstrate or analyze specific healthcare system applications or outcomes.
Posts about general AI model capabilities, training methods, or technical breakthroughs without healthcare application.
3 example posts
Larry Ellison just explained the most important and underreported insight in enterprise AI. Every major AI model, ChatGPT, Claude, Grok was trained on public data but public data has a ceiling. The real unlocking of AI's value happens when models can reason on private data, the
amazing writeup of experiments. essentially a model may already know a capability but that may be stored in a representation that is bad for extraction. you can slightly retrain/condition the model so the same capability becomes easier to isolate. must read for mech interp http
Three weeks ago I called Anthropic rationing inference and OpenAI hitting the wall next. Yesterday OpenAI shipped GPT-5.5 Instant. New default for free, Plus, and Pro. 30% fewer words per answer. One frame going around: cost savings. It's both bigger and smaller than that. http
Created 2026-05-10 · Updated 2026-05-10
Edit rule text
[non_healthcare_infrastructure_hype]
Learned 3 rejections Active
Exclude posts about semiconductor manufacturing, data centers, compute infrastructure, energy systems, or space technology (SpaceX, NVIDIA manufacturing, cloud platforms) unless they specifically address healthcare deployment or healthcare-specific use cases.
Posts about computing infrastructure, semiconductors, or energy systems without direct healthcare application.
3 example posts
“Tokens are increasing 50x. You need to drive down cost at the same rate — or more.” At Google Cloud Next, @googlecloud's Mark Lohmeyer spoke with @PatrickMoorhead about the infrastructure realities behind the agent era: faster model training, lower-latency inference, and https:
I don’t think I’ve seen this take before but I like it. Musk has been world-leading at compressing money, resources, and time to make “known/hard” things at scale—make an electric car, make batteries, make a cheaper bigger rocket, all of which already existed but worse, at
SpaceX wants to build a $55 billion next-generation, vertically integrated semiconductor manufacturing and advanced computing fabrication facility in Grimes county, Texas, according to a new filing. • Proposed investment: $55B upfront, up to $119B total • Focus: semiconductor ma
Created 2026-05-10 · Updated 2026-05-10
Edit rule text
[ai_infrastructure_hype_no_healthcare]
Learned 3 rejections Active
Exclude posts that discuss AI infrastructure (compute, models, scaling, efficiency) or technical capabilities (tokenization, world models, reasoning) that reference healthcare tangentially or not at all.
Posts about AI compute, data infrastructure, or model capabilities without specific healthcare application.
3 example posts
Satya Nadella's energy is something here. 🔥 "Tokens per Dollar per Watt" The new equation for the AI age for every Company or Industry or Country. "And that means Infrastructure, Infrastructure and Infrastructure." https://t.co/McINrBAo4a
Google CEO Sundar Pichai on current frontier model's ability to break the security of almost all current software. "These models are definitely, like really gonna break pretty much all software out there, maybe already, we don't know." https://t.co/ezjPcSehiG
Fei-Fei Li warns that AI may be staring too hard at language models. The world is not just text on a screen. It is physical, visual, spatial, and always changing. Most of the economy runs on seeing, moving, interacting, and embodied intelligence. https://t.co/mGMjsCdU34
Created 2026-05-10 · Updated 2026-05-18
Edit rule text
[clinical_observation_without_context]
Learned 3 rejections Active
Exclude posts that present isolated clinical observations, single-case anecdotes, trial data, or diagnostic findings without connecting them to healthcare delivery, business model, policy, or systemic implications. A clinical fact alone is not sufficient; explain why it matters to healthcare operations, incentives, or access.
Posts reporting single clinical findings, trial results, or medical anecdotes without healthcare systems or operational insight.
3 example posts
A reminder that AI can be used to reduce health inequities. Technology not being used in the US except for a few health systems, but shown to work well in Kenya @JAMACardio https://t.co/wyQ1k0dHES https://t.co/pP824X2qJG
In adults ≥75 years, 10-year #ColorectalCancer and CRC death risk were higher with prior adenoma but both were low (<1.1%); non-CRC death risk was 47-48%, supporting deprioritizing repeat surveillance. https://t.co/mJ7r4vAtK1 https://t.co/dUCYgO0SuK
Statin guidelines over 50 years: 1970s: "Total cholesterol over 280 is high." Result: a relatively small market. 1988: "Actually, 240 is high." Result: millions more added overnight. 2001: "Actually, focus on LDL. Get it under 100." Result: tens of millions of new customers.
Created 2026-05-10 · Updated 2026-05-10
Edit rule text
[fraud_scandal_no_systems_analysis]
Learned 3 rejections Active
Exclude posts that report healthcare fraud, regulatory violations, or government waste (Medicare, Medicaid, pharmacy fraud) as isolated scandals or prosecution stories without examining systemic business model failures, incentive misalignments, or structural healthcare problems they reveal. These posts treat fraud as a crime story, not a systems insight.
Posts reporting fraud, waste, or scandal in healthcare or government without analyzing root causes or systemic implications.
3 example posts
PBM owns the pharmacy. PBM owns the plan. PBM adjudicates the claim. Three commissions on one pill. The employer pays all three and gets a thank-you note for the rebate.
"Cordoba-owned buildings in Columbus housed 288 businesses registered with Medicaid... they charged taxpayers more than a quarter of a billion dollars between 2018 and 2024... in a city where only 6,273 people 75 or older are on Medicaid." - @realDailyWire https://t.co/7UnS7agwiw
If you had data on spending per SNAP terminal it would be trivial to find welfare fraud. Like 24 hours w/ a LLM and I can hand you a list of EBT card #s committing fraud and the EBT machine owners committing fraud. It’s not small money, either. The Mogadishu Store owner in
Created 2026-05-10 · Updated 2026-05-10
Edit rule text
[policy_or_regulatory_outrage_without_healthcare_analysis]
Learned 3 rejections Active
Exclude posts that use healthcare regulation, fraud, or policy as a vehicle for political messaging or regulatory posturing (e.g., VP highlighting waste, political figures grandstanding on fraud) without independent healthcare systems analysis or evidence.
Posts using healthcare events as a framing for political or regulatory grievances without substantive healthcare policy analysis.
3 example posts
PBM owns the pharmacy. PBM owns the plan. PBM adjudicates the claim. Three commissions on one pill. The employer pays all three and gets a thank-you note for the rebate.
Red states have disproportionately benefited from the health insurance subsidies expanded by Biden and then cut by Trump. Georgia is a clear example: in just one year, marketplace enrollment has dropped by more than 550,000 people — a 37% decline. My @Morning_Joe Chart. https:/
Fox News’ Larry Kudlow: “Why don’t you throw [the fraudsters] in jail?” CMS Administrator Dr. Mehmet Oz: “This is a big decision point. When you try to put people in the pokey, it takes years of careful investigations. That’s how the systems are built. But you know what I can h
Created 2026-05-10 · Updated 2026-05-10
Edit rule text
[pharma_trial_or_drug_data_without_healthcare_systems_context]
Learned 3 rejections Active
Exclude posts that announce pharmaceutical trial results, FDA rulings, drug approvals, or biotech findings (e.g., 'Vertex dropped a therapy', 'CB-Dock3 found 900 protein structures') unless they analyze the clinical, economic, or systemic healthcare implications beyond the lab result.
Posts reporting pharmaceutical trial results, FDA decisions, or drug approvals without analyzing healthcare delivery, access, or systemic implications.
3 example posts
CB-Dock3: An Enhanced Web Server for Protein–Ligand Blind Docking 1. CB-Dock3 is a major update of the CB-Dock blind docking web server, targeting the practical bottlenecks created by today’s surge of large Cryo-EM and AI-predicted protein structures (e.g., AlphaFold3-era https:
$vrtx drops $mrna-partnered Cystic Fibrosis mRNA therapy VX-522 over inhaled LNP tolerability issues: https://t.co/WlIIHXxg9x Not too surprising IMO. Important Q: is safety/tolerability of this approach still good enough for one-time/infrequent genome editing? $prme $arct
New Anthropic Fellows research: Model Spec Midtraining (MSM). Standard alignment methods train AIs on examples of desired behavior. But this can fail to generalize to new situations. MSM addresses this by first teaching AIs how we would like them to generalize and why.
Created 2026-05-10 · Updated 2026-05-10
Edit rule text
[ai_infrastructure_hype_without_healthcare_lens]
Learned 3 rejections Active
Exclude posts that discuss AI infrastructure, token costs, compute capacity, or semiconductor manufacturing (e.g., SpaceX fab, GPU pricing, token efficiency) unless the post explicitly connects these to a healthcare delivery or clinical problem.
Posts about AI compute, tokens, semiconductors, or infrastructure costs that lack healthcare-specific application or analysis.
3 example posts
“Tokens are increasing 50x. You need to drive down cost at the same rate — or more.” At Google Cloud Next, @googlecloud's Mark Lohmeyer spoke with @PatrickMoorhead about the infrastructure realities behind the agent era: faster model training, lower-latency inference, and https:
SpaceX wants to build a $55 billion next-generation, vertically integrated semiconductor manufacturing and advanced computing fabrication facility in Grimes county, Texas, according to a new filing. • Proposed investment: $55B upfront, up to $119B total • Focus: semiconductor ma
the future of software engineering seems uncontroversially prompting + code review. startups will skip the code review because they’re racing against time. larger/serious orgs will take code review very seriously. llms can do code review, but my guess is that because they have t
Created 2026-05-10 · Updated 2026-05-10
Edit rule text
[incomplete_truncated_unfinished_posts]
Learned 3 rejections Active
Exclude posts that end abruptly mid-thought, end with trailing ellipsis or 'on' suggesting continuation, or are visibly truncated or unfinished. These posts do not present a complete argument, observation, or claim.
Posts that are cut off mid-sentence, clearly incomplete, or lacking closure
3 example posts
there's a deeper underestimated reason: consumers just do not have very many long chains of tasks which can be automated end to end to unlock huge value. their chains are shallow almost always in enterprise, org processes form very long chains, so the reward for automation is m
amazing writeup of experiments. essentially a model may already know a capability but that may be stored in a representation that is bad for extraction. you can slightly retrain/condition the model so the same capability becomes easier to isolate. must read for mech interp http
Flock Safety Cameras just rolled out an update that allows them to use artificial intelligence to detect human voices These cameras were sold to city governments as gunshot detectors, but with one update they turned them all into a mass surveillance system that records voices T
Created 2026-05-10 · Updated 2026-05-10
Edit rule text
[broad_business_model_critique_without_evidence]
Learned 3 rejections Active
Exclude posts that make broad claims about healthcare incentive misalignment, broken business models, or systemic dysfunction (e.g., 'the system is profitable not broken,' 'PBMs own everything so patients suffer') without citing data, demonstrating the claimed mechanism, or proposing a testable alternative. Cynical takes on healthcare economics without evidence do not qualify.
Posts making sweeping claims about healthcare business models, market structure, or incentive alignment without substantive data or system-level analysis
3 example posts
PBM owns the pharmacy. PBM owns the plan. PBM adjudicates the claim. Three commissions on one pill. The employer pays all three and gets a thank-you note for the rebate.
Incumbents don't defend broken systems. They defend profitable ones. Healthcare in America isn't a market failure. It's a sustaining innovation problem disguised as a social crisis. Every "reform" strengthens the existing architecture because the people writing the rules are the
Europe is not losing to the US because it lacks engineers, capital or universities. It is losing because every layer of its system rewards the gatekeeper and punishes the builder. In medicine I see it weekly. A primary care AI that triages, documents and flags risk gets deployed
Created 2026-05-10 · Updated 2026-05-10
Edit rule text
[tangential_infrastructure_macro_without_healthcare_lens]
Learned 3 rejections Active
Exclude posts about non-healthcare domains (SpaceX manufacturing, semiconductor capacity, energy infrastructure, supply chain macro) even if tangentially framed as relevant to healthcare. Posts must explain the specific healthcare systems bottleneck or workflow they address, not just assert a connection.
Posts about energy, semiconductors, supply chains, or macro economics with only loose or speculative framing to healthcare.
3 example posts
“Tokens are increasing 50x. You need to drive down cost at the same rate — or more.” At Google Cloud Next, @googlecloud's Mark Lohmeyer spoke with @PatrickMoorhead about the infrastructure realities behind the agent era: faster model training, lower-latency inference, and https:
I don’t think I’ve seen this take before but I like it. Musk has been world-leading at compressing money, resources, and time to make “known/hard” things at scale—make an electric car, make batteries, make a cheaper bigger rocket, all of which already existed but worse, at
SpaceX wants to build a $55 billion next-generation, vertically integrated semiconductor manufacturing and advanced computing fabrication facility in Grimes county, Texas, according to a new filing. • Proposed investment: $55B upfront, up to $119B total • Focus: semiconductor ma
Created 2026-05-10 · Updated 2026-05-10
Edit rule text
[broad_unvalidated_health_claims_or_market_debate_without_nuance]
Learned 3 rejections Active
Exclude posts that make broad claims about healthcare efficacy, market consolidation, or treatment guidelines (e.g., 'statin guidelines have shifted', 'GLP-1s cause weight loss') without clinical trial data, healthcare systems analysis, or substantive nuance about tradeoffs and failure modes. Marketing-style claims without evidence should be excluded.
Posts making sweeping claims about healthcare treatments, diagnostics, or market dynamics without clinical evidence, nuance, or systems-level analysis.
3 example posts
Klarna replaced 700 customer service agents with AI. Resolution time dropped from 11 minutes to 2. Then quality collapsed on complex cases. The CEO admitted cost became too dominant. Customers wanted humans. Klarna is now hiring people again. 55% of firms that replaced
If you’ve been taking GLP-1s, even for a little while, you may be wondering how to maintain the weight loss you’ve achieved. Our endocrinologist explains what works and what to avoid. https://t.co/TU8LUmg9Vn
Statin guidelines over 50 years: 1970s: "Total cholesterol over 280 is high." Result: a relatively small market. 1988: "Actually, 240 is high." Result: millions more added overnight. 2001: "Actually, focus on LDL. Get it under 100." Result: tens of millions of new customers.
Created 2026-05-10 · Updated 2026-05-10
Edit rule text
[ai_company_business_metrics_without_healthcare_application]
Learned 3 rejections Active
Exclude posts that announce AI company products, features, or business news (e.g., 'OpenAI shipped GPT-5.5 Instant', 'Anthropic Fellows research', 'Harvey raised $200M') without demonstrating specific healthcare application, healthcare customer validation, or healthcare systems impact. Company product announcements alone are not healthcare news.
Posts reporting AI company product launches, capability announcements, or business metrics that lack healthcare-specific validation or application evidence.
3 example posts
Three weeks ago I called Anthropic rationing inference and OpenAI hitting the wall next. Yesterday OpenAI shipped GPT-5.5 Instant. New default for free, Plus, and Pro. 30% fewer words per answer. One frame going around: cost savings. It's both bigger and smaller than that. http
Clawvisor is going to be one of the most important parts of helping make the agent world especially OpenClaw/Hermes Agent secure and enterprise-grade. We're in the Apple I moment for personal AI, but we're ABOUT to see the Apple II - the first moment when everyone can use it
I've spent 150+ hours over the past year building agents on top of OpenClaw, NanoClaw, LangChain, using MCPs, n8n, Apify, and the rest of the automation/LLM stack we plow through for our PE clients, using DevriX and Growth Shuttle as a sandbox. It's been painful. Thread:
Created 2026-05-10 · Updated 2026-05-10
Edit rule text
[speculative_sci_fi_robotics_labor_without_healthcare_systems]
Learned 3 rejections Active
Exclude posts about robots, autonomous systems, or sci-fi-adjacent labor predictions (e.g., 'robots will perform surgery by end of 2026') that lack validation, actual healthcare system context, clinical evidence, or substantive analysis of implementation barriers. Speculative timelines and hype without healthcare-specific evidence should be excluded.
Posts about humanoid robots, autonomous systems, or speculative sci-fi applications to healthcare without grounding in actual clinical workflows or healthcare systems challenges.
3 example posts
A humanoid robot just unloaded a dishwasher for 67 hours straight. By end of 2026, the same robot will perform surgery. That's not a prediction from a sci-fi writer. That's Brett Adcock. CEO of Figure AI. Said it on camera this month. Here's what makes this insane. Two years
I might feel nostalgic watching the last Tesla Model X roll off the line—except it's being replaced by the Optimus humanoid robot line, with plans to scale production to a million units per year. They will soon be in our factories, homes and operating rooms changing our world ht
Humanoid robots are moving from Silicon Valley novelty to viable business model—powered by AI and global supply chains, especially in China. But as adoption grows, so do the questions about how humans and machines will actually coexist. More on Primer, streaming Wednesdays http
Created 2026-05-10 · Updated 2026-05-10
Edit rule text
[shallow_business_model_debate_without_systems_insight]
Learned 3 rejections Active
Exclude posts that frame healthcare business problems (PBMs, margins, pricing, reform) as obvious or self-evident without exploring root causes, stakeholder incentives, regulatory constraints, or evidence-based solutions. Posts that are primarily opinion or rhetorical without substantive analysis of how systems work fall here.
Posts making broad claims about healthcare business model inefficiencies, market dynamics, or reforms without system-level analysis of causation or solutions.
3 example posts
PBM owns the pharmacy. PBM owns the plan. PBM adjudicates the claim. Three commissions on one pill. The employer pays all three and gets a thank-you note for the rebate.
Incumbents don't defend broken systems. They defend profitable ones. Healthcare in America isn't a market failure. It's a sustaining innovation problem disguised as a social crisis. Every "reform" strengthens the existing architecture because the people writing the rules are the
Statin guidelines over 50 years: 1970s: "Total cholesterol over 280 is high." Result: a relatively small market. 1988: "Actually, 240 is high." Result: millions more added overnight. 2001: "Actually, focus on LDL. Get it under 100." Result: tens of millions of new customers.
Created 2026-05-10 · Updated 2026-05-10
Edit rule text
[incomplete_truncated_social_posts]
Learned 3 rejections Active
Exclude posts where the text ends abruptly mid-thought, is visibly truncated, or lacks a complete argument or data point. These posts appear unfinished and prevent clear analysis.
Posts that are cut off mid-sentence or lack coherent conclusion, making them impossible to extract meaningful insight.
3 example posts
there's a deeper underestimated reason: consumers just do not have very many long chains of tasks which can be automated end to end to unlock huge value. their chains are shallow almost always in enterprise, org processes form very long chains, so the reward for automation is m
amazing writeup of experiments. essentially a model may already know a capability but that may be stored in a representation that is bad for extraction. you can slightly retrain/condition the model so the same capability becomes easier to isolate. must read for mech interp http
🚀🤖 In 2026, autonomous #AI agents aren’t failing because of technology ➜ they’re failing because of responsibility. ⚙️ Everyone knows how to build agents. Very few know how to own their decisions. 💡 Real autonomous AI isn’t about freedom or intelligence. It’s about clear scope,
Created 2026-05-10 · Updated 2026-05-10
Edit rule text
[pharmaceutical_or_clinical_trial_data_announcement_only]
Learned 3 rejections Active
Exclude posts that announce pharmaceutical trial results, drug approvals, or clinical data (especially from JAMA, FDA, or clinical studies) without analyzing the healthcare system implications, business model impact, or operational consequences. Pure clinical or trial announcements should be excluded.
Posts reporting drug trial results, pharma announcements, or clinical data without healthcare systems analysis or business/operational context.
3 example posts
CB-Dock3: An Enhanced Web Server for Protein–Ligand Blind Docking 1. CB-Dock3 is a major update of the CB-Dock blind docking web server, targeting the practical bottlenecks created by today’s surge of large Cryo-EM and AI-predicted protein structures (e.g., AlphaFold3-era https:
NEW: today OpenBind ‘comes out of stealth’ so to speak with their first data dump of ~900 novel protein-ligand structures - most with paired affinities This represents a meaningful %-age increase in all of humanities P-L data in the PDB collected in the last 50 years More👇 http
$vrtx drops $mrna-partnered Cystic Fibrosis mRNA therapy VX-522 over inhaled LNP tolerability issues: https://t.co/WlIIHXxg9x Not too surprising IMO. Important Q: is safety/tolerability of this approach still good enough for one-time/infrequent genome editing? $prme $arct
Created 2026-05-10 · Updated 2026-05-10
Edit rule text
[tangential_regulatory_fraud_scandal_without_systems_analysis]
Learned 3 rejections Active
Exclude posts that report on fraud scandals, regulatory enforcement actions, or policy compliance incidents (e.g., Medicaid fraud, EBT card abuse) without analyzing root causes in healthcare system structure, incentive misalignment, or operational implications. These are crime/enforcement stories, not healthcare systems analysis.
Posts about regulatory scandals, fraud, or policy enforcement that lack healthcare systems analysis or substantive operational insight.
3 example posts
🛑FDA officials have blocked publication of several studies supporting the safety of widely used vaccines against Covid-19 and shingles in recent months, a spokesman for HHS confirmed. The studies, which cost millions of dollars in public funds, were conducted by scientists at htt
"Cordoba-owned buildings in Columbus housed 288 businesses registered with Medicaid... they charged taxpayers more than a quarter of a billion dollars between 2018 and 2024... in a city where only 6,273 people 75 or older are on Medicaid." - @realDailyWire https://t.co/7UnS7agwiw
If you had data on spending per SNAP terminal it would be trivial to find welfare fraud. Like 24 hours w/ a LLM and I can hand you a list of EBT card #s committing fraud and the EBT machine owners committing fraud. It’s not small money, either. The Mogadishu Store owner in
Created 2026-05-09 · Updated 2026-05-09
Edit rule text
[biotech_founder_lab_update_or_enthusiasm_without_validation]
Learned 3 rejections Active
Exclude posts that celebrate biotech or AI founder accomplishments, lab experiments, or technical capabilities (e.g., 'AI model crunched a task in 17 minutes', 'paired autonomous lab with GPT-5 in closed-loop experiment', 'Ginkgo paired lab with OpenAI'). Only include if the post demonstrates validated healthcare impact, clinical relevance, or pathway to real-world application.
Posts about biotech/AI founder lab updates, closed-loop experiments, or capability demonstrations without healthcare validation or impact evidence.
3 example posts
NEW: today OpenBind ‘comes out of stealth’ so to speak with their first data dump of ~900 novel protein-ligand structures - most with paired affinities This represents a meaningful %-age increase in all of humanities P-L data in the PDB collected in the last 50 years More👇 http
not labelled! no stains! just cell chemistry changing in response to a drug !!! you can see it change before it dies !!! you can predict what happens to a cell !!! that's constantly being dosed with a drug !!! you know what? this is just an ICU for living human cells https://t.
Prof. @DeryaTR_ A task that would normally take a researcher months, the AI model crunched in 17 minutes. It didn’t just explain the mechanism—it proposed the experiment to prove it. It looks like after programming and math, AI is coming for medicine. https://t.co/IF8Jr2GBMj
Created 2026-05-09 · Updated 2026-05-09
Edit rule text
[tangential_non_healthcare_policy_or_fraud_scandal_without_analysis]
Learned 3 rejections Active
Exclude posts that report on fraud investigations, welfare abuse, regulatory enforcement, or political figures making healthcare claims (e.g., 'FDA blocks publication of vaccine studies', 'Medicaid spending fraud detected', political figures claiming credit for healthcare accountability). Only include if the post provides healthcare systems insight into how these issues affect care delivery, access, or outcomes.
Posts about fraud, welfare abuse, regulatory action, or political scandal adjacent to healthcare but without healthcare systems analysis.
3 example posts
🛑FDA officials have blocked publication of several studies supporting the safety of widely used vaccines against Covid-19 and shingles in recent months, a spokesman for HHS confirmed. The studies, which cost millions of dollars in public funds, were conducted by scientists at htt
"Cordoba-owned buildings in Columbus housed 288 businesses registered with Medicaid... they charged taxpayers more than a quarter of a billion dollars between 2018 and 2024... in a city where only 6,273 people 75 or older are on Medicaid." - @realDailyWire https://t.co/7UnS7agwiw
If you had data on spending per SNAP terminal it would be trivial to find welfare fraud. Like 24 hours w/ a LLM and I can hand you a list of EBT card #s committing fraud and the EBT machine owners committing fraud. It’s not small money, either. The Mogadishu Store owner in
Created 2026-05-09 · Updated 2026-05-09
Edit rule text
[tangential_ai_infrastructure_or_compute_hype_no_healthcare]
Learned 3 rejections Active
Exclude posts about AI infrastructure challenges, compute scaling, token economics, or data bottlenecks presented as general tech problems (e.g., 'Tokens are increasing 50x. You need to drive down cost', 'The big AI wins happen to be the most data-rich problems'). Only include if the post explicitly addresses how these constraints affect healthcare deployment or use cases.
Posts about AI infrastructure, compute costs, tokens, or data bottlenecks as abstract technical problems without healthcare application.
3 example posts
“Tokens are increasing 50x. You need to drive down cost at the same rate — or more.” At Google Cloud Next, @googlecloud's Mark Lohmeyer spoke with @PatrickMoorhead about the infrastructure realities behind the agent era: faster model training, lower-latency inference, and https:
@amspector100 and @bfspector of @flappyairplanes at AI Ascent 2026: The big AI wins so far (search, coding) happen to be the most data-rich problems on earth. Almost everything else is data-poor. Unlocking the rest of the economy will depend on drastically better data https://
// Contextual Agentic Memory is a Memo, not True Memory // Most agent memory today isn't memory. They are more like memos. A new paper argues that vector stores, RAG buffers, and scratchpads implement lookup, not consolidation. Agents accumulate notes indefinitely without ever
Created 2026-05-09 · Updated 2026-05-09
Edit rule text
[tangential_ai_or_infrastructure_hype_without_healthcare_specificity]
Learned 3 rejections Active
Exclude posts about AI model technical capabilities (Claude, GPT, reasoning models), semiconductor announcements, compute scaling, or infrastructure projects that are tagged or loosely framed as healthcare-relevant but lack evidence of actual healthcare application, pilot, or systemic impact.
Posts about AI model capabilities, compute infrastructure, or tech announcements tangentially framed as healthcare-relevant but lacking healthcare application or validation.
3 example posts
🚀🤖 In 2026, autonomous #AI agents aren’t failing because of technology ➜ they’re failing because of responsibility. ⚙️ Everyone knows how to build agents. Very few know how to own their decisions. 💡 Real autonomous AI isn’t about freedom or intelligence. It’s about clear scope,
WestRock CEO went from Palantir skeptic to biggest believer. “What I read and what we see talked about in the AI world and what Palantir’s operating system actually is, I can barely recognize the reality of what they’re doing on the ground with the talk that goes on around AI.”
“Tokens are increasing 50x. You need to drive down cost at the same rate — or more.” At Google Cloud Next, @googlecloud's Mark Lohmeyer spoke with @PatrickMoorhead about the infrastructure realities behind the agent era: faster model training, lower-latency inference, and https:
Created 2026-05-09 · Updated 2026-05-09
Edit rule text
[broad_health_claims_without_nuance_or_systems_analysis]
Learned 3 rejections Active
Exclude posts that make broad, confident claims about healthcare trends, business model viability, or policy impact (e.g., 'Healthcare isn't a market failure. It's a sustaining innovation problem', 'Red states disproportionately benefited from health insurance subsidies') without providing specific data, system-level analysis, or evidence of how the trend actually manifests in care delivery or outcomes.
Posts making sweeping healthcare claims about market trends, business models, or policy effects without grounding in healthcare system dynamics or evidence.
3 example posts
Incumbents don't defend broken systems. They defend profitable ones. Healthcare in America isn't a market failure. It's a sustaining innovation problem disguised as a social crisis. Every "reform" strengthens the existing architecture because the people writing the rules are the
Europe is not losing to the US because it lacks engineers, capital or universities. It is losing because every layer of its system rewards the gatekeeper and punishes the builder. In medicine I see it weekly. A primary care AI that triages, documents and flags risk gets deployed
Red states have disproportionately benefited from the health insurance subsidies expanded by Biden and then cut by Trump. Georgia is a clear example: in just one year, marketplace enrollment has dropped by more than 550,000 people — a 37% decline. My @Morning_Joe Chart. https:/
Created 2026-05-09 · Updated 2026-05-09
Edit rule text
[tangential_infrastructure_or_compute_hype_without_healthcare_specificity]
Learned 3 rejections Active
Exclude posts that discuss infrastructure projects, semiconductor manufacturing, power generation, or compute capacity (e.g., 'SpaceX $55 billion semiconductor facility', 'NVIDIA...infrastructure realities') unless the post explicitly connects the infrastructure to a specific healthcare delivery problem or demonstrates how it enables healthcare innovation.
Posts about semiconductor manufacturing, compute infrastructure, or energy/power announcements without clear healthcare application or impact analysis.
3 example posts
“Tokens are increasing 50x. You need to drive down cost at the same rate — or more.” At Google Cloud Next, @googlecloud's Mark Lohmeyer spoke with @PatrickMoorhead about the infrastructure realities behind the agent era: faster model training, lower-latency inference, and https:
I don’t think I’ve seen this take before but I like it. Musk has been world-leading at compressing money, resources, and time to make “known/hard” things at scale—make an electric car, make batteries, make a cheaper bigger rocket, all of which already existed but worse, at
SpaceX wants to build a $55 billion next-generation, vertically integrated semiconductor manufacturing and advanced computing fabrication facility in Grimes county, Texas, according to a new filing. • Proposed investment: $55B upfront, up to $119B total • Focus: semiconductor ma
Created 2026-05-09 · Updated 2026-05-09
Edit rule text
[infrastructure_compute_hype_without_healthcare_specificity]
Learned 3 rejections Active
Exclude posts about GPU capacity, semiconductor fabrication, token cost reduction, compute scaling, or infrastructure announcements (e.g., SpaceX fab, NVIDIA keynotes, 'tokens increasing 50x') unless the post directly ties the infrastructure capability to a specific healthcare workflow, data processing bottleneck, or clinical decision-support system being deployed.
Posts about semiconductor manufacturing, data centers, token economics, or compute infrastructure with loose or aspirational healthcare framing.
3 example posts
“Tokens are increasing 50x. You need to drive down cost at the same rate — or more.” At Google Cloud Next, @googlecloud's Mark Lohmeyer spoke with @PatrickMoorhead about the infrastructure realities behind the agent era: faster model training, lower-latency inference, and https:
SpaceX wants to build a $55 billion next-generation, vertically integrated semiconductor manufacturing and advanced computing fabrication facility in Grimes county, Texas, according to a new filing. • Proposed investment: $55B upfront, up to $119B total • Focus: semiconductor ma
@amspector100 and @bfspector of @flappyairplanes at AI Ascent 2026: The big AI wins so far (search, coding) happen to be the most data-rich problems on earth. Almost everything else is data-poor. Unlocking the rest of the economy will depend on drastically better data https://
Created 2026-05-08 · Updated 2026-05-08
Edit rule text
[ai_safety_vulnerability_tangential_no_healthcare_application]
Learned 3 rejections Active
Exclude posts about zero-day vulnerabilities in AI models, security red-teaming results, or hacking incidents that mention healthcare tangentially or use healthcare framing (e.g., 'Claude found exploits,' 'agents went rogue,' 'Mexican government hack') unless the post identifies a specific healthcare system, protocol, or patient-facing vulnerability being actively exploited or patched.
Posts about AI model vulnerabilities, security incidents, or capability exploits framed as healthcare-relevant without demonstrating actual healthcare deployment risk or mitigation.
3 example posts
So Mythos was, indeed, not marketing hype. Remember this is a general purpose model that just happens to be good at finding exploits because good models are good at lots of things. Expect similar from OpenAI & Google. And from open models in 8 months. https://t.co/KbhalQYX8
Dario Amodei says Mythos is not limited by compute Anthropic can scale it 3x or 10x without creating a conflict between government and private-sector access The harder problem is who gets it "because giving access to too many organizations could create serious cyber risks" htt
Our attention to biorisks posed by AI needs to match the current attention given to cyber-risks. The staged release of Claude Mythos in order to bolster defenses in key industries is necessary to shore up resilience against a new class of cyber-risk across critical industries. We
Created 2026-05-08 · Updated 2026-05-08
Edit rule text
[tangential_healthcare_framing_of_non_healthcare_business_content]
Learned 3 rejections Active
Exclude posts that discuss general business strategy, regulatory dynamics, or competitive positioning (e.g., 'Europe losing to US because gatekeepers punish builders', 'Apple I moment for personal AI') even if healthcare is mentioned as an example. The post must center healthcare systems, not use healthcare as illustration of broader business trends.
Posts about general business models, regulatory capture, or macro business trends that use healthcare as window dressing but lack healthcare systems analysis.
3 example posts
Clawvisor is going to be one of the most important parts of helping make the agent world especially OpenClaw/Hermes Agent secure and enterprise-grade. We're in the Apple I moment for personal AI, but we're ABOUT to see the Apple II - the first moment when everyone can use it
WestRock CEO went from Palantir skeptic to biggest believer. “What I read and what we see talked about in the AI world and what Palantir’s operating system actually is, I can barely recognize the reality of what they’re doing on the ground with the talk that goes on around AI.”
Europe is not losing to the US because it lacks engineers, capital or universities. It is losing because every layer of its system rewards the gatekeeper and punishes the builder. In medicine I see it weekly. A primary care AI that triages, documents and flags risk gets deployed
Created 2026-05-08 · Updated 2026-05-08
Edit rule text
[ai_safety_vulnerability_tangent_no_healthcare_specificity]
Learned 3 rejections Active
Exclude posts that focus on AI model vulnerabilities, zero-day exploits, security testing, or safety benchmarks (e.g., 'Claude Mythos finding zero-day vulnerabilities', 'agents can register LLCs and open bank accounts') unless the post explicitly connects to healthcare system security, HIPAA compliance, or clinical deployment risks.
Posts about AI security vulnerabilities, exploits, or safety testing that lack healthcare-specific context or application.
3 example posts
So Mythos was, indeed, not marketing hype. Remember this is a general purpose model that just happens to be good at finding exploits because good models are good at lots of things. Expect similar from OpenAI & Google. And from open models in 8 months. https://t.co/KbhalQYX8
Dario Amodei says Mythos is not limited by compute Anthropic can scale it 3x or 10x without creating a conflict between government and private-sector access The harder problem is who gets it "because giving access to too many organizations could create serious cyber risks" htt
two agents running on two different laptops in my apartment started talking to each other on tuesday by thursday they'd registered an LLC in wyoming, opened a stripe account, and wired $40 to a polymarket wallet the LLC is in my name. a lawyer just quoted me $3,200 to figure ou
Created 2026-05-08 · Updated 2026-05-08
Edit rule text
[ai_safety_vulnerability_tangent_without_healthcare_lens]
Learned 3 rejections Active
Exclude posts reporting AI safety vulnerabilities, security incidents, or misuse scenarios (e.g., agents registering LLCs, exploits found in models, hacking government systems) unless the post explicitly connects the vulnerability to healthcare delivery risk or discusses mitigation in clinical contexts.
Posts about AI safety incidents, vulnerabilities, or security concerns that lack application to healthcare delivery or clinical decision-making.
3 example posts
So Mythos was, indeed, not marketing hype. Remember this is a general purpose model that just happens to be good at finding exploits because good models are good at lots of things. Expect similar from OpenAI & Google. And from open models in 8 months. https://t.co/KbhalQYX8
Dario Amodei says Mythos is not limited by compute Anthropic can scale it 3x or 10x without creating a conflict between government and private-sector access The harder problem is who gets it "because giving access to too many organizations could create serious cyber risks" htt
two agents running on two different laptops in my apartment started talking to each other on tuesday by thursday they'd registered an LLC in wyoming, opened a stripe account, and wired $40 to a polymarket wallet the LLC is in my name. a lawyer just quoted me $3,200 to figure ou
Created 2026-05-08 · Updated 2026-05-08
Edit rule text
[tangential_non_healthcare_policy_or_infrastructure_with_loose_framing]
Learned 3 rejections Active
Exclude posts that focus on government fraud enforcement, generic regulatory posturing, macro infrastructure announcements (SpaceX, NVIDIA semiconductors, data infrastructure), or political figures' healthcare claims without substantive analysis of healthcare system implementation or operational outcomes.
Posts about non-healthcare policy (fraud enforcement, regulatory grandstanding, macro infrastructure) loosely framed as healthcare-relevant.
3 example posts
WestRock CEO went from Palantir skeptic to biggest believer. “What I read and what we see talked about in the AI world and what Palantir’s operating system actually is, I can barely recognize the reality of what they’re doing on the ground with the talk that goes on around AI.”
“Tokens are increasing 50x. You need to drive down cost at the same rate — or more.” At Google Cloud Next, @googlecloud's Mark Lohmeyer spoke with @PatrickMoorhead about the infrastructure realities behind the agent era: faster model training, lower-latency inference, and https:
I don’t think I’ve seen this take before but I like it. Musk has been world-leading at compressing money, resources, and time to make “known/hard” things at scale—make an electric car, make batteries, make a cheaper bigger rocket, all of which already existed but worse, at
Created 2026-05-08 · Updated 2026-05-08
Edit rule text
[pharma_or_clinical_trial_data_without_systems_context]
Learned 3 rejections Active
Exclude posts that announce drug trial data, pharma partnerships, gene therapy approvals, or clinical research findings where the post lacks analysis of healthcare delivery, operational, economic, or systems implications.
Posts about drug trial results, pharma company announcements, or clinical data without healthcare systems or operational implications
3 example posts
$vrtx drops $mrna-partnered Cystic Fibrosis mRNA therapy VX-522 over inhaled LNP tolerability issues: https://t.co/WlIIHXxg9x Not too surprising IMO. Important Q: is safety/tolerability of this approach still good enough for one-time/infrequent genome editing? $prme $arct
If you’ve been taking GLP-1s, even for a little while, you may be wondering how to maintain the weight loss you’ve achieved. Our endocrinologist explains what works and what to avoid. https://t.co/TU8LUmg9Vn
In adults ≥75 years, 10-year #ColorectalCancer and CRC death risk were higher with prior adenoma but both were low (<1.1%); non-CRC death risk was 47-48%, supporting deprioritizing repeat surveillance. https://t.co/mJ7r4vAtK1 https://t.co/dUCYgO0SuK
Created 2026-05-08 · Updated 2026-05-08
Edit rule text
[pharma_trial_data_announcement_unvalidated]
Learned 3 rejections Active
Exclude posts that announce clinical trial data, FDA approvals, drug partnerships, or protein structure discoveries if the post is a one-line announcement or celebration of the result without analysis of how this changes healthcare delivery, reimbursement, access, or operational workflows.
Posts announcing pharmaceutical trial results, drug approvals, or biotech partnerships without healthcare systems or operational impact analysis.
3 example posts
NEW: today OpenBind ‘comes out of stealth’ so to speak with their first data dump of ~900 novel protein-ligand structures - most with paired affinities This represents a meaningful %-age increase in all of humanities P-L data in the PDB collected in the last 50 years More👇 http
$vrtx drops $mrna-partnered Cystic Fibrosis mRNA therapy VX-522 over inhaled LNP tolerability issues: https://t.co/WlIIHXxg9x Not too surprising IMO. Important Q: is safety/tolerability of this approach still good enough for one-time/infrequent genome editing? $prme $arct
In adults ≥75 years, 10-year #ColorectalCancer and CRC death risk were higher with prior adenoma but both were low (<1.1%); non-CRC death risk was 47-48%, supporting deprioritizing repeat surveillance. https://t.co/mJ7r4vAtK1 https://t.co/dUCYgO0SuK
Created 2026-05-08 · Updated 2026-05-08
Edit rule text
[regulatory_fraud_scandal_reporting_no_analysis]
Learned 3 rejections Active
Exclude posts that report fraud incidents, regulatory enforcement, or financial crime in healthcare (billing fraud, waste, embezzlement) without analyzing healthcare system vulnerabilities, incentive structures, or actionable policy implications.
Posts reporting healthcare fraud scandals, Medicaid waste, regulatory actions, or financial crime without systemic healthcare analysis.
3 example posts
"Cordoba-owned buildings in Columbus housed 288 businesses registered with Medicaid... they charged taxpayers more than a quarter of a billion dollars between 2018 and 2024... in a city where only 6,273 people 75 or older are on Medicaid." - @realDailyWire https://t.co/7UnS7agwiw
If you had data on spending per SNAP terminal it would be trivial to find welfare fraud. Like 24 hours w/ a LLM and I can hand you a list of EBT card #s committing fraud and the EBT machine owners committing fraud. It’s not small money, either. The Mogadishu Store owner in
.@VP has been leading efforts to root out waste, fraud, and abuse across the federal government — putting accountability first. What we’ve seen in programs like SNAP is deeply concerning, with cases of improper payments, duplicate enrollments, and misuse of taxpayer dollars that
Created 2026-05-08 · Updated 2026-05-08
Edit rule text
[ai_safety_or_security_vulnerability_tangential]
Learned 3 rejections Active
Exclude posts that discuss AI security vulnerabilities, hacking, misuse incidents, or safety concerns (e.g., agents being hijacked, malicious agent skills, Claude hacking government) unless they specifically demonstrate healthcare-system impact or deployment context. These are AI risk, not healthcare innovation.
Posts about AI security incidents, vulnerabilities, or safety concerns without healthcare-specific impact or application
3 example posts
two agents running on two different laptops in my apartment started talking to each other on tuesday by thursday they'd registered an LLC in wyoming, opened a stripe account, and wired $40 to a polymarket wallet the LLC is in my name. a lawyer just quoted me $3,200 to figure ou
AI’s Branding Illusion: Insights from "Agents of Chaos". Co-authored by MIT and Harvard scholars, the study red-teams autonomous AI in live environments. It shows agents obey unauthorized users, execute harmful commands, and misreport outcomes—revealing a critical alignment gap h
Our attention to biorisks posed by AI needs to match the current attention given to cyber-risks. The staged release of Claude Mythos in order to bolster defenses in key industries is necessary to shore up resilience against a new class of cyber-risk across critical industries. We
Created 2026-05-08 · Updated 2026-05-08
Edit rule text
[healthcare_fraud_or_regulatory_scandal_reporting_without_systemic_analysis]
Learned 3 rejections Active
Exclude posts that report healthcare fraud, abuse, or regulatory crackdowns (Medicaid fraud, waste in programs like SNAP/EBT, billing scandals) unless the post analyzes root causes, proposes system-level solutions, or examines structural vulnerabilities. Posts that merely amplify outrage or report incidents without systemic insight are out of scope.
Posts reporting fraud, waste, or regulatory violations in healthcare without analyzing systemic vulnerabilities, policy responses, or structural solutions.
3 example posts
"Cordoba-owned buildings in Columbus housed 288 businesses registered with Medicaid... they charged taxpayers more than a quarter of a billion dollars between 2018 and 2024... in a city where only 6,273 people 75 or older are on Medicaid." - @realDailyWire https://t.co/7UnS7agwiw
If you had data on spending per SNAP terminal it would be trivial to find welfare fraud. Like 24 hours w/ a LLM and I can hand you a list of EBT card #s committing fraud and the EBT machine owners committing fraud. It’s not small money, either. The Mogadishu Store owner in
.@VP has been leading efforts to root out waste, fraud, and abuse across the federal government — putting accountability first. What we’ve seen in programs like SNAP is deeply concerning, with cases of improper payments, duplicate enrollments, and misuse of taxpayer dollars that
Created 2026-05-08 · Updated 2026-05-08
Edit rule text
[pharmaceutical_or_drug_trial_data_announcement_without_systems_analysis]
Learned 3 rejections Active
Exclude posts that report pharmaceutical trial data, FDA decisions, or drug company announcements (e.g., pivots, partnerships, discontinuations) unless the post connects the development to healthcare economics, access barriers, clinical workflow changes, or system-level implications. Single-study announcements or trial results without broader context are out of scope.
Posts announcing drug trial results, FDA approvals, or pharma company developments without analyzing healthcare system, access, or business model implications.
3 example posts
🛑FDA officials have blocked publication of several studies supporting the safety of widely used vaccines against Covid-19 and shingles in recent months, a spokesman for HHS confirmed. The studies, which cost millions of dollars in public funds, were conducted by scientists at htt
$vrtx drops $mrna-partnered Cystic Fibrosis mRNA therapy VX-522 over inhaled LNP tolerability issues: https://t.co/WlIIHXxg9x Not too surprising IMO. Important Q: is safety/tolerability of this approach still good enough for one-time/infrequent genome editing? $prme $arct
Grace Science’s experience highlights a growing disconnect at FDA between talk and action on therapies for rare diseases. Despite efficacy signals in a monogenic ultrarare disease, FDA said the plausible mechanism framework is not available, and requires a new manufacturing
Created 2026-05-08 · Updated 2026-05-08
Edit rule text
[tangential_regulatory_or_policy_outrage_without_healthcare_analysis]
Learned 3 rejections Active
Exclude posts that leverage healthcare policy (FDA decisions, Medicare rules, healthcare spending) as a vehicle for political outrage, partisan positioning, or generalist regulatory commentary without analyzing healthcare-specific systemic implications or operational consequences for providers, patients, or access.
Posts using healthcare as a framing device to comment on broader political or regulatory outrage without substantive healthcare systems analysis.
3 example posts
🛑FDA officials have blocked publication of several studies supporting the safety of widely used vaccines against Covid-19 and shingles in recent months, a spokesman for HHS confirmed. The studies, which cost millions of dollars in public funds, were conducted by scientists at htt
Red states have disproportionately benefited from the health insurance subsidies expanded by Biden and then cut by Trump. Georgia is a clear example: in just one year, marketplace enrollment has dropped by more than 550,000 people — a 37% decline. My @Morning_Joe Chart. https:/
Statin guidelines over 50 years: 1970s: "Total cholesterol over 280 is high." Result: a relatively small market. 1988: "Actually, 240 is high." Result: millions more added overnight. 2001: "Actually, focus on LDL. Get it under 100." Result: tens of millions of new customers.
Created 2026-05-08 · Updated 2026-05-08
Edit rule text
[tangential_non_healthcare_business_announcement]
Learned 3 rejections Active
Exclude posts about non-healthcare companies, legal proceedings, regulatory actions, or business developments that have minimal healthcare substance or apply healthcare labels retroactively. Examples: SpaceX manufacturing facilities, AI agent LLC formation, banking ML systems, Stripe hiring—these are not healthcare tech.
Posts about non-healthcare business announcements, legal cases, or corporate events with only loose or forced healthcare framing.
3 example posts
SpaceX wants to build a $55 billion next-generation, vertically integrated semiconductor manufacturing and advanced computing fabrication facility in Grimes county, Texas, according to a new filing. • Proposed investment: $55B upfront, up to $119B total • Focus: semiconductor ma
two agents running on two different laptops in my apartment started talking to each other on tuesday by thursday they'd registered an LLC in wyoming, opened a stripe account, and wired $40 to a polymarket wallet the LLC is in my name. a lawyer just quoted me $3,200 to figure ou
Stripe just created a role that didn't exist 12 months ago (and they're paying multiple six figures for it) It's called the Forward Deployed AI Accelerator. They are hiring AI-native individuals to work directly with their marketing teams to fundamentally change how they work.
Created 2026-05-07 · Updated 2026-05-07
Edit rule text
[generalist_macro_or_market_dynamics_without_healthcare_lens]
Learned 3 rejections Active
Exclude posts analyzing competitive dynamics, market pricing, business model viability, or macro trends (GLP-1 pricing, insurance markets, labor costs) without connecting to healthcare system incentives, policy gaps, or operational friction. Business commentary without healthcare systems insight does not qualify.
Posts about market trends, business model debates, competitive dynamics, or macro economics lacking specific healthcare system analysis.
3 example posts
Red states have disproportionately benefited from the health insurance subsidies expanded by Biden and then cut by Trump. Georgia is a clear example: in just one year, marketplace enrollment has dropped by more than 550,000 people — a 37% decline. My @Morning_Joe Chart. https:/
If you’ve been taking GLP-1s, even for a little while, you may be wondering how to maintain the weight loss you’ve achieved. Our endocrinologist explains what works and what to avoid. https://t.co/TU8LUmg9Vn
Statin guidelines over 50 years: 1970s: "Total cholesterol over 280 is high." Result: a relatively small market. 1988: "Actually, 240 is high." Result: millions more added overnight. 2001: "Actually, focus on LDL. Get it under 100." Result: tens of millions of new customers.
Created 2026-05-07 · Updated 2026-05-07
Edit rule text
[tangential_non_healthcare_policy_or_infrastructure]
Learned 3 rejections Active
Exclude posts that discuss semiconductor manufacturing, broadband infrastructure, space technology, or general economic policy unless the post explicitly connects the topic to healthcare delivery, clinical infrastructure, or healthcare system operations. Generic policy or infrastructure posts with healthcare labels do not qualify.
Posts about non-healthcare policy, infrastructure, or business dynamics with only loose or tangential healthcare framing.
3 example posts
SpaceX wants to build a $55 billion next-generation, vertically integrated semiconductor manufacturing and advanced computing fabrication facility in Grimes county, Texas, according to a new filing. • Proposed investment: $55B upfront, up to $119B total • Focus: semiconductor ma
Europe is not losing to the US because it lacks engineers, capital or universities. It is losing because every layer of its system rewards the gatekeeper and punishes the builder. In medicine I see it weekly. A primary care AI that triages, documents and flags risk gets deployed
A Microsoft acabou de transformar uma startup de $11 bilhões de dólares em uma funcionalidade do Word. Não foi uma aquisição nem uma parceria. Uma funcionalidade. A Harvey levantou $200M a uma valuation de $11B em março. $190M de receita recorrente anual. 100 mil advogados. Co
Created 2026-05-07 · Updated 2026-05-07
Edit rule text
[clinical_trial_or_drug_data_without_systems_context]
Learned 3 rejections Active
Exclude posts that announce drug trial results, FDA approvals, or clinical data unless they contextualize the finding within healthcare delivery, payment models, access barriers, or system-level adoption challenges. Standalone clinical data announcements do not qualify.
Posts reporting pharmaceutical trial results, drug approval news, or clinical data without analyzing healthcare system implications or adoption barriers.
3 example posts
$vrtx drops $mrna-partnered Cystic Fibrosis mRNA therapy VX-522 over inhaled LNP tolerability issues: https://t.co/WlIIHXxg9x Not too surprising IMO. Important Q: is safety/tolerability of this approach still good enough for one-time/infrequent genome editing? $prme $arct
In adults ≥75 years, 10-year #ColorectalCancer and CRC death risk were higher with prior adenoma but both were low (<1.1%); non-CRC death risk was 47-48%, supporting deprioritizing repeat surveillance. https://t.co/mJ7r4vAtK1 https://t.co/dUCYgO0SuK
Ivermectin and Mebendazole Cost a Fraction of Chemo. Big Pharma Can't Patent Them. That's the Problem. Cancer centers get a cut of every chemotherapy bill. Generic drugs don't generate that margin. Two affordable, widely available compounds showing 84% clinical benefit in a real
Created 2026-05-07 · Updated 2026-05-07
Edit rule text
[non_healthcare_infrastructure_tangential]
Learned 3 rejections Active
Exclude posts about non-healthcare infrastructure (semiconductor fabs, general robotics, manufacturing, banking systems, space tech) that mention healthcare tangentially or speculatively without concrete healthcare application or healthcare business model analysis.
Posts about general tech infrastructure, semiconductors, robotics, or manufacturing that have only loose or speculative healthcare framing.
3 example posts
SpaceX wants to build a $55 billion next-generation, vertically integrated semiconductor manufacturing and advanced computing fabrication facility in Grimes county, Texas, according to a new filing. • Proposed investment: $55B upfront, up to $119B total • Focus: semiconductor ma
two agents running on two different laptops in my apartment started talking to each other on tuesday by thursday they'd registered an LLC in wyoming, opened a stripe account, and wired $40 to a polymarket wallet the LLC is in my name. a lawyer just quoted me $3,200 to figure ou
I might feel nostalgic watching the last Tesla Model X roll off the line—except it's being replaced by the Optimus humanoid robot line, with plans to scale production to a million units per year. They will soon be in our factories, homes and operating rooms changing our world ht
Created 2026-05-07 · Updated 2026-05-07
Edit rule text
[tangential_infrastructure_or_speculative_sci_fi]
Learned 3 rejections Active
Exclude posts about Tesla robots, SpaceX fabrication facilities, humanoid robotics manufacturing, or other non-healthcare infrastructure and supply chain content, even if tagged with healthcare context. The core value must be healthcare system or patient care impact, not sci-fi speculation or manufacturing scale.
Posts about non-healthcare infrastructure, robotics, manufacturing, or speculative future applications loosely framed as healthcare-adjacent
3 example posts
SpaceX wants to build a $55 billion next-generation, vertically integrated semiconductor manufacturing and advanced computing fabrication facility in Grimes county, Texas, according to a new filing. • Proposed investment: $55B upfront, up to $119B total • Focus: semiconductor ma
I might feel nostalgic watching the last Tesla Model X roll off the line—except it's being replaced by the Optimus humanoid robot line, with plans to scale production to a million units per year. They will soon be in our factories, homes and operating rooms changing our world ht
Humanoid robots are moving from Silicon Valley novelty to viable business model—powered by AI and global supply chains, especially in China. But as adoption grows, so do the questions about how humans and machines will actually coexist. More on Primer, streaming Wednesdays http
Created 2026-05-07 · Updated 2026-05-07
Edit rule text
[infrastructure_and_compute_hype_no_healthcare]
Learned 3 rejections Active
Exclude posts about semiconductor fabrication facilities, compute infrastructure, cloud platforms, or technology stacks unless the post explicitly demonstrates how the infrastructure solves a healthcare-specific problem or bottleneck.
Posts about semiconductor manufacturing, data infrastructure, cloud platforms, or compute capabilities without healthcare application.
3 example posts
SpaceX wants to build a $55 billion next-generation, vertically integrated semiconductor manufacturing and advanced computing fabrication facility in Grimes county, Texas, according to a new filing. • Proposed investment: $55B upfront, up to $119B total • Focus: semiconductor ma
Stripe just created a role that didn't exist 12 months ago (and they're paying multiple six figures for it) It's called the Forward Deployed AI Accelerator. They are hiring AI-native individuals to work directly with their marketing teams to fundamentally change how they work.
the future of software engineering seems uncontroversially prompting + code review. startups will skip the code review because they’re racing against time. larger/serious orgs will take code review very seriously. llms can do code review, but my guess is that because they have t
Created 2026-05-07 · Updated 2026-05-07
Edit rule text
[robotics_and_speculative_labor_macro]
Learned 3 rejections Active
Exclude posts about humanoid robot production, robotics scaling, or labor market macro trends unless they directly address healthcare workforce challenges, clinical labor shortages, or healthcare-specific adoption scenarios.
Posts about humanoid robots, robotics manufacturing, or labor market disruption without healthcare-specific workforce analysis.
3 example posts
I might feel nostalgic watching the last Tesla Model X roll off the line—except it's being replaced by the Optimus humanoid robot line, with plans to scale production to a million units per year. They will soon be in our factories, homes and operating rooms changing our world ht
Humanoid robots are moving from Silicon Valley novelty to viable business model—powered by AI and global supply chains, especially in China. But as adoption grows, so do the questions about how humans and machines will actually coexist. More on Primer, streaming Wednesdays http
Today we’re giving an update on ramping F.03 production at BotQ In the last 120 days, Figure scaled manufacturing 24x - from 1 robot/day to 1 robot/hour We will manufacture 55 humanoid robots this week https://t.co/Am5Kn53mVE
Created 2026-05-07 · Updated 2026-05-07
Edit rule text
[policy_or_fraud_scandal_without_healthcare_systems_analysis]
Learned 3 rejections Active
Exclude posts that report healthcare fraud cases, Medicaid/Medicare scandals, regulatory enforcement, or policy changes as breaking news or political outrage without analyzing underlying incentive structures, systemic vulnerabilities, or healthcare delivery implications.
Posts about healthcare fraud, regulatory actions, or policy changes reported as scandal/outrage without systems-level analysis
3 example posts
Red states have disproportionately benefited from the health insurance subsidies expanded by Biden and then cut by Trump. Georgia is a clear example: in just one year, marketplace enrollment has dropped by more than 550,000 people — a 37% decline. My @Morning_Joe Chart. https:/
"Cordoba-owned buildings in Columbus housed 288 businesses registered with Medicaid... they charged taxpayers more than a quarter of a billion dollars between 2018 and 2024... in a city where only 6,273 people 75 or older are on Medicaid." - @realDailyWire https://t.co/7UnS7agwiw
If you had data on spending per SNAP terminal it would be trivial to find welfare fraud. Like 24 hours w/ a LLM and I can hand you a list of EBT card #s committing fraud and the EBT machine owners committing fraud. It’s not small money, either. The Mogadishu Store owner in
Created 2026-05-07 · Updated 2026-05-07
Edit rule text
[biotech_founder_lab_update_unvalidated]
Learned 3 rejections Active
Exclude posts that are founder/lab updates or enthusiasm about experimental results, closed-loop AI experiments, or lab capabilities (e.g., 'we trained X on Y data') without demonstrating clinical validation, regulatory pathway clarity, or healthcare deployment readiness. Posts must show healthcare impact beyond technical proof-of-concept.
Posts from biotech founders or labs showcasing experimental results or capabilities without clinical validation or healthcare systems analysis.
3 example posts
not labelled! no stains! just cell chemistry changing in response to a drug !!! you can see it change before it dies !!! you can predict what happens to a cell !!! that's constantly being dosed with a drug !!! you know what? this is just an ICU for living human cells https://t.
Prof. @DeryaTR_ A task that would normally take a researcher months, the AI model crunched in 17 minutes. It didn’t just explain the mechanism—it proposed the experiment to prove it. It looks like after programming and math, AI is coming for medicine. https://t.co/IF8Jr2GBMj
Earlier this year we paired our autonomous lab with OpenAI's GPT-5 in a closed-loop experiment: GPT-5 designed cell-free protein synthesis reactions, our RACs ran them, and the model iterated on the results. Over 36,000 experiments and six cycles later, it landed on a reaction ht
Created 2026-05-07 · Updated 2026-05-07
Edit rule text
[generalist_macro_policy_without_healthcare_systems_lens]
Learned 3 rejections Active
Exclude posts that frame healthcare through general macro trends (tech disruption, venture capital, infrastructure investment, government spending) without analyzing specific healthcare delivery, financing, or operational system changes.
Posts about broad policy, economic, or regulatory trends that mention healthcare superficially but lack systems analysis.
3 example posts
SpaceX wants to build a $55 billion next-generation, vertically integrated semiconductor manufacturing and advanced computing fabrication facility in Grimes county, Texas, according to a new filing. • Proposed investment: $55B upfront, up to $119B total • Focus: semiconductor ma
Europe is not losing to the US because it lacks engineers, capital or universities. It is losing because every layer of its system rewards the gatekeeper and punishes the builder. In medicine I see it weekly. A primary care AI that triages, documents and flags risk gets deployed
The problem with Reality Labs is not ambition. It is time. AI turned into revenue faster because it improves existing workflows. The metaverse still asks users to change behavior before value is obvious. https://t.co/lopZiUwGU5
Created 2026-05-06 · Updated 2026-05-06
Edit rule text
[ai_agent_or_model_security_incident_tangential]
Learned 3 rejections Active
Exclude posts reporting AI agent hacking, model jailbreaks, security vulnerabilities, or safety failures in non-healthcare contexts (e.g., ClawHub botnet, Claude database deletion, Mexico government cyberattack using Claude) unless the post explicitly connects the vulnerability to a healthcare delivery system or patient data risk.
Posts about AI agent/model security vulnerabilities, hacking incidents, or safety failures that are tangential to healthcare operations
3 example posts
two agents running on two different laptops in my apartment started talking to each other on tuesday by thursday they'd registered an LLC in wyoming, opened a stripe account, and wired $40 to a polymarket wallet the LLC is in my name. a lawyer just quoted me $3,200 to figure ou
AI’s Branding Illusion: Insights from "Agents of Chaos". Co-authored by MIT and Harvard scholars, the study red-teams autonomous AI in live environments. It shows agents obey unauthorized users, execute harmful commands, and misreport outcomes—revealing a critical alignment gap h
Our attention to biorisks posed by AI needs to match the current attention given to cyber-risks. The staged release of Claude Mythos in order to bolster defenses in key industries is necessary to shore up resilience against a new class of cyber-risk across critical industries. We
Created 2026-05-06 · Updated 2026-05-06
Edit rule text
[drug_trial_data_or_pharma_announcement_without_systems_context]
Learned 3 rejections Active
Exclude posts that report pharmaceutical trial data, drug efficacy results, or clinical research findings unless they analyze healthcare system implications such as reimbursement strategy, access barriers, formulary positioning, or clinical workflow integration.
Posts announcing drug trial results or pharmaceutical data without healthcare system context, reimbursement analysis, or access implications.
3 example posts
If you’ve been taking GLP-1s, even for a little while, you may be wondering how to maintain the weight loss you’ve achieved. Our endocrinologist explains what works and what to avoid. https://t.co/TU8LUmg9Vn
In adults ≥75 years, 10-year #ColorectalCancer and CRC death risk were higher with prior adenoma but both were low (<1.1%); non-CRC death risk was 47-48%, supporting deprioritizing repeat surveillance. https://t.co/mJ7r4vAtK1 https://t.co/dUCYgO0SuK
Among veterans with moderate to severe #ChronicPain in primary care, the whole health team intervention produced greater improvement in the Brief Pain Inventory interference scores at 12 months compared with cognitive behavioral therapy and usual care. https://t.co/xtZY4sgyGt h
Created 2026-05-06 · Updated 2026-05-06
Edit rule text
[ai_safety_security_vulnerability_tangent]
Learned 3 rejections Active
Exclude posts that focus on AI safety incidents, security vulnerabilities, or cybersecurity exploits (e.g., jailbreaks, data breaches, malicious agents) unless the post demonstrates direct impact on healthcare delivery, patient data security, or clinical system integrity.
Posts about AI safety, security vulnerabilities, or cybersecurity incidents framed tangentially without healthcare system implications.
3 example posts
AI’s Branding Illusion: Insights from "Agents of Chaos". Co-authored by MIT and Harvard scholars, the study red-teams autonomous AI in live environments. It shows agents obey unauthorized users, execute harmful commands, and misreport outcomes—revealing a critical alignment gap h
Our attention to biorisks posed by AI needs to match the current attention given to cyber-risks. The staged release of Claude Mythos in order to bolster defenses in key industries is necessary to shore up resilience against a new class of cyber-risk across critical industries. We
Anthropic built something so powerful that they are only letting 50 organisations touch it. It is called Claude Mythos. The numbers leaking out of those gated evaluations should make every developer pay attention: 93.9% on SWE-bench Verified 94.6% on GPQA Diamond Claude Opus
Created 2026-05-06 · Updated 2026-05-06
Edit rule text
[tangential_biotech_or_clinical_founder_enthusiasm_without_validation]
Learned 3 rejections Active
Exclude posts where founders, researchers, or biotech companies announce experimental results, capabilities, or milestones (e.g., protein design breakthroughs, cell behavior observations, AI-designed reactions) with enthusiasm but without clinical validation data, peer-reviewed evidence, or demonstrated path to healthcare delivery.
Posts from biotech founders or researchers celebrating lab results, experimental milestones, or capabilities without evidence of clinical validation or healthcare system readiness
3 example posts
not labelled! no stains! just cell chemistry changing in response to a drug !!! you can see it change before it dies !!! you can predict what happens to a cell !!! that's constantly being dosed with a drug !!! you know what? this is just an ICU for living human cells https://t.
Prof. @DeryaTR_ A task that would normally take a researcher months, the AI model crunched in 17 minutes. It didn’t just explain the mechanism—it proposed the experiment to prove it. It looks like after programming and math, AI is coming for medicine. https://t.co/IF8Jr2GBMj
Our new preprint is a significant milestone for us We built "HealthFormer" by training on our deeply phenotyped cohort from the Human Phenotype Project data. Healthformer is a multimodal generative transformer model that tokenizes each participant's physiological trajectory http
Created 2026-05-06 · Updated 2026-05-06
Edit rule text
[ai_safety_vulnerability_or_security_tangent]
Learned 3 rejections Active
Exclude posts about AI safety incidents, model vulnerabilities, cybersecurity breaches, or security certifications unless the post analyzes specific healthcare delivery risks or patient privacy/safety implications beyond generic AI security concerns.
Posts about AI model safety, security vulnerabilities, or cybersecurity incidents framed as healthcare-relevant without healthcare-specific risk analysis
3 example posts
AI’s Branding Illusion: Insights from "Agents of Chaos". Co-authored by MIT and Harvard scholars, the study red-teams autonomous AI in live environments. It shows agents obey unauthorized users, execute harmful commands, and misreport outcomes—revealing a critical alignment gap h
Our attention to biorisks posed by AI needs to match the current attention given to cyber-risks. The staged release of Claude Mythos in order to bolster defenses in key industries is necessary to shore up resilience against a new class of cyber-risk across critical industries. We
Hacking Mexico government with AI assistance. Attacker exfiltrated hundreds of millions of citizen records. 75% of the executed commands across the entire cyberattack campaign were generated by Claude. 40 minutes after Claude said "I'm not going to create that file" it was report
Created 2026-05-06 · Updated 2026-05-06
Edit rule text
[generalist_macro_policy_or_labor_without_healthcare_specificity]
Learned 3 rejections Active
Exclude posts that discuss general macroeconomic trends, labor market dynamics, or infrastructure developments (tariffs, banking, supply chains) unless the post centers on a healthcare-specific application or impact.
Posts about broad economic, labor market, or infrastructure trends with only tangential healthcare framing.
3 example posts
the future of software engineering seems uncontroversially prompting + code review. startups will skip the code review because they’re racing against time. larger/serious orgs will take code review very seriously. llms can do code review, but my guess is that because they have t
Revolut just moved the IP of banking into a model. Trained on 24 billion banking events in 111 countries. One foundation model replacing six separate ML systems. Credit scoring: +130% Fraud recall: +65% Marketing engagement: +79% The model is the new moat.
Thought I would start posting about interesting things happening at AWS. Not a bad day to start.🚀 Today at #WhatsNextWithAWS we announced a big step forward with @OpenAI on Amazon Bedrock: 1. OpenAI models now available 2. Codex for enterprise development 3. Amazon Bedrock Manag
Created 2026-05-06 · Updated 2026-05-06
Edit rule text
[biotech_or_pharma_founder_enthusiasm_lab_update_without_validation]
Learned 3 rejections Active
Exclude posts that showcase researcher/founder excitement about experimental results, lab breakthroughs, or AI capabilities in research without peer-reviewed evidence, clinical validation, or analysis of how findings translate to healthcare delivery or patient outcomes.
Posts featuring founder enthusiasm about lab results or capabilities without peer review, validation, or healthcare systems context
3 example posts
not labelled! no stains! just cell chemistry changing in response to a drug !!! you can see it change before it dies !!! you can predict what happens to a cell !!! that's constantly being dosed with a drug !!! you know what? this is just an ICU for living human cells https://t.
Prof. @DeryaTR_ A task that would normally take a researcher months, the AI model crunched in 17 minutes. It didn’t just explain the mechanism—it proposed the experiment to prove it. It looks like after programming and math, AI is coming for medicine. https://t.co/IF8Jr2GBMj
Our new preprint is a significant milestone for us We built "HealthFormer" by training on our deeply phenotyped cohort from the Human Phenotype Project data. Healthformer is a multimodal generative transformer model that tokenizes each participant's physiological trajectory http
Created 2026-05-06 · Updated 2026-05-06
Edit rule text
[robotics_or_speculative_sci_fi_labor_macro]
Learned 3 rejections Active
Exclude posts about humanoid robot manufacturing, scaling timelines, or general labor market disruption without specific healthcare workforce application, clinical deployment context, or healthcare labor shortage solutions.
Posts about humanoid robots, manufacturing scaling, or speculative labor market impacts without healthcare-specific context
3 example posts
I might feel nostalgic watching the last Tesla Model X roll off the line—except it's being replaced by the Optimus humanoid robot line, with plans to scale production to a million units per year. They will soon be in our factories, homes and operating rooms changing our world ht
Humanoid robots are moving from Silicon Valley novelty to viable business model—powered by AI and global supply chains, especially in China. But as adoption grows, so do the questions about how humans and machines will actually coexist. More on Primer, streaming Wednesdays http
Today we’re giving an update on ramping F.03 production at BotQ In the last 120 days, Figure scaled manufacturing 24x - from 1 robot/day to 1 robot/hour We will manufacture 55 humanoid robots this week https://t.co/Am5Kn53mVE
Created 2026-05-06 · Updated 2026-05-06
Edit rule text
[generalist_macro_policy_or_infrastructure_without_healthcare_lens]
Learned 3 rejections Active
Exclude posts about tariffs, manufacturing, supply chains, banking infrastructure, workforce disruption, or macro policy that mention healthcare tangentially or use healthcare as a generic example—unless the post specifically analyzes how these trends reshape healthcare delivery, costs, clinical operations, or payer economics.
Posts about macroeconomic trends, infrastructure, tariffs, or labor markets with only loose or tangential healthcare framing.
3 example posts
the future of software engineering seems uncontroversially prompting + code review. startups will skip the code review because they’re racing against time. larger/serious orgs will take code review very seriously. llms can do code review, but my guess is that because they have t
The problem with Reality Labs is not ambition. It is time. AI turned into revenue faster because it improves existing workflows. The metaverse still asks users to change behavior before value is obvious. https://t.co/lopZiUwGU5
A Microsoft acabou de transformar uma startup de $11 bilhões de dólares em uma funcionalidade do Word. Não foi uma aquisição nem uma parceria. Uma funcionalidade. A Harvey levantou $200M a uma valuation de $11B em março. $190M de receita recorrente anual. 100 mil advogados. Co
Created 2026-05-06 · Updated 2026-05-06
Edit rule text
[political_figures_or_regulatory_outrage_without_healthcare_analysis]
Learned 3 rejections Active
Exclude posts that feature political figures, quote regulatory officials, or express outrage at policy decisions without explaining the healthcare operational, reimbursement, or clinical impact. Posts must analyze systems-level healthcare consequences, not just political theater or scandal reporting.
Posts featuring political figures, regulatory statements, or policy outrage without substantive healthcare systems analysis or operational impact.
3 example posts
Fox News’ Larry Kudlow: “Why don’t you throw [the fraudsters] in jail?” CMS Administrator Dr. Mehmet Oz: “This is a big decision point. When you try to put people in the pokey, it takes years of careful investigations. That’s how the systems are built. But you know what I can h
Atrium tried to buy a hospital in 2018. The Attorney General killed it. In 2026, they came back with paperwork instead of cash. The Attorney General has 30 days. The Wake County board has 48 hours. @NC_Governor https://t.co/g99Ub3iMhh https://t.co/x0bxsg7VIR
Last week I randomly got a $9,000 bill for a hospital visit from last year. I called them up and they said my insurance decided not to cover it. Why? Because. .@Aetna can just decide they’re not interested in covering something and then you’re left with an inflated bill to pay h
Created 2026-05-06 · Updated 2026-05-06
Edit rule text
[political_outrage_or_regulatory_fraud_without_healthcare_systems_analysis]
Learned 3 rejections Active
Exclude posts that amplify political outrage, regulatory scandal, or fraud allegations (e.g., CMS fraud, Medicaid abuse, insurance denials, hospital billing) without analyzing root causes, system design flaws, incentive structures, or scalable operational solutions.
Posts about political figures, fraud allegations, regulatory action, or policy outrage without substantive healthcare systems analysis.
3 example posts
Fox News’ Larry Kudlow: “Why don’t you throw [the fraudsters] in jail?” CMS Administrator Dr. Mehmet Oz: “This is a big decision point. When you try to put people in the pokey, it takes years of careful investigations. That’s how the systems are built. But you know what I can h
Atrium tried to buy a hospital in 2018. The Attorney General killed it. In 2026, they came back with paperwork instead of cash. The Attorney General has 30 days. The Wake County board has 48 hours. @NC_Governor https://t.co/g99Ub3iMhh https://t.co/x0bxsg7VIR
Ah yes, Optum, the company that did Tim Walz’s audit of the 14 high risk Medicaid programs consumed by Somali fraud. https://t.co/dNRnjI7wxB
Created 2026-05-06 · Updated 2026-05-06
Edit rule text
[biotech_founder_lab_update_without_validation]
Learned 3 rejections Active
Exclude posts from biotech founders or researchers describing lab experiments, autonomous system demonstrations, or in vitro/cell-free results — unless the post explicitly frames clinical validation, regulatory pathways, or healthcare system adoption challenges.
Posts sharing biotech founder lab results, autonomous system experiments, or research milestones without clinical validation or healthcare system context.
3 example posts
Prof. @DeryaTR_ A task that would normally take a researcher months, the AI model crunched in 17 minutes. It didn’t just explain the mechanism—it proposed the experiment to prove it. It looks like after programming and math, AI is coming for medicine. https://t.co/IF8Jr2GBMj
Our new preprint is a significant milestone for us We built "HealthFormer" by training on our deeply phenotyped cohort from the Human Phenotype Project data. Healthformer is a multimodal generative transformer model that tokenizes each participant's physiological trajectory http
Demis Hassabis says bigger context windows are still a brute force answer to memory. The human brain does something stranger. During sleep, it replays what matters and folds new knowledge into what it already knows. AI does not need infinite context. It needs the right memory h
Created 2026-05-06 · Updated 2026-05-06
Edit rule text
[political_or_regulatory_outrage_without_analysis]
Learned 3 rejections Active
Exclude posts quoting politicians or regulators on healthcare topics, reporting fraud accusations, or highlighting regulatory drama (e.g., CMS administrator quotes, AG decisions, jail comments) unless the post provides systems-level analysis of why the problem exists or how policy addresses root causes.
Posts featuring political figures, regulatory quotes, or healthcare scandal accusations without substantive systems-level analysis
3 example posts
Fox News’ Larry Kudlow: “Why don’t you throw [the fraudsters] in jail?” CMS Administrator Dr. Mehmet Oz: “This is a big decision point. When you try to put people in the pokey, it takes years of careful investigations. That’s how the systems are built. But you know what I can h
Atrium tried to buy a hospital in 2018. The Attorney General killed it. In 2026, they came back with paperwork instead of cash. The Attorney General has 30 days. The Wake County board has 48 hours. @NC_Governor https://t.co/g99Ub3iMhh https://t.co/x0bxsg7VIR
Ivermectin and Mebendazole Cost a Fraction of Chemo. Big Pharma Can't Patent Them. That's the Problem. Cancer centers get a cut of every chemotherapy bill. Generic drugs don't generate that margin. Two affordable, widely available compounds showing 84% clinical benefit in a real
Created 2026-05-05 · Updated 2026-05-05
Edit rule text
[regulatory_scandal_without_systems_analysis]
Learned 3 rejections Active
Exclude posts reporting healthcare fraud, regulatory scandal, or corporate misconduct (e.g., FDA action, hospital billing disputes, insurance denials) as breaking news or outrage without analyzing root cause in healthcare system structure, incentives, or delivery models. Scandal reporting ≠ systems analysis.
Posts about healthcare fraud, regulatory failures, or corporate misconduct reported as news without analysis of systemic healthcare delivery or reimbursement implications.
3 example posts
Atrium tried to buy a hospital in 2018. The Attorney General killed it. In 2026, they came back with paperwork instead of cash. The Attorney General has 30 days. The Wake County board has 48 hours. @NC_Governor https://t.co/g99Ub3iMhh https://t.co/x0bxsg7VIR
"At the Veterans Health Administration, we have the exquisite gift of a hospice unit, a place to care for veterans when time is short. We care for the sons of mothers who can no longer be here to care for them." In #APieceofMyMind, a #palliative care #physician reflects on https
Ah yes, Optum, the company that did Tim Walz’s audit of the 14 high risk Medicaid programs consumed by Somali fraud. https://t.co/dNRnjI7wxB
Created 2026-05-05 · Updated 2026-05-05
Edit rule text
[regulatory_or_fraud_scandal_without_systemic_analysis]
Learned 3 rejections Active
Exclude posts that report healthcare fraud allegations, regulatory enforcement actions, or corporate scandals (Atrium hospital acquisition rejected, Optum billing issues, hospice fraud crisis, Medicaid program audits) as standalone news or anecdotes without analyzing systemic vulnerabilities, healthcare delivery impacts, or how these events reflect broader operational or governance failures in healthcare systems.
Posts reporting healthcare fraud, regulatory actions, or business scandals without analysis of systemic vulnerabilities or healthcare delivery impacts.
3 example posts
Atrium tried to buy a hospital in 2018. The Attorney General killed it. In 2026, they came back with paperwork instead of cash. The Attorney General has 30 days. The Wake County board has 48 hours. @NC_Governor https://t.co/g99Ub3iMhh https://t.co/x0bxsg7VIR
Grace Science’s experience highlights a growing disconnect at FDA between talk and action on therapies for rare diseases. Despite efficacy signals in a monogenic ultrarare disease, FDA said the plausible mechanism framework is not available, and requires a new manufacturing
"At the Veterans Health Administration, we have the exquisite gift of a hospice unit, a place to care for veterans when time is short. We care for the sons of mothers who can no longer be here to care for them." In #APieceofMyMind, a #palliative care #physician reflects on https
Created 2026-05-05 · Updated 2026-05-05
Edit rule text
[tangential_infrastructure_or_macro_without_healthcare]
Learned 3 rejections Active
Exclude posts about manufacturing scale-up, robotics production, supply chain innovation, or macro economic trends (Tesla Model X/Optimus robots, Figure manufacturing, tariff impacts) even if they mention healthcare in passing. The post must directly address how the infrastructure or business model change affects healthcare delivery, economics, or operations—not just speculate or tangentially mention healthcare.
Posts about infrastructure, robotics, manufacturing, or macro-level business trends with only loose framing to healthcare systems.
3 example posts
I might feel nostalgic watching the last Tesla Model X roll off the line—except it's being replaced by the Optimus humanoid robot line, with plans to scale production to a million units per year. They will soon be in our factories, homes and operating rooms changing our world ht
Humanoid robots are moving from Silicon Valley novelty to viable business model—powered by AI and global supply chains, especially in China. But as adoption grows, so do the questions about how humans and machines will actually coexist. More on Primer, streaming Wednesdays http
The problem with Reality Labs is not ambition. It is time. AI turned into revenue faster because it improves existing workflows. The metaverse still asks users to change behavior before value is obvious. https://t.co/lopZiUwGU5
Created 2026-05-05 · Updated 2026-05-05
Edit rule text
[speculative_robotics_or_ai_capability_hype]
Learned 3 rejections Active
Exclude posts that celebrate AI model capabilities, robotics manufacturing milestones, or agent autonomy breakthroughs without demonstrating measurable healthcare application, clinical validation, or healthcare system impact. Hype about what AI *can* do in abstract terms should be rejected unless tied to healthcare practice change.
Posts announcing robotics breakthroughs, AI model capabilities, or autonomous systems that lack grounding in actual healthcare operational outcomes or clinical validation.
3 example posts
I might feel nostalgic watching the last Tesla Model X roll off the line—except it's being replaced by the Optimus humanoid robot line, with plans to scale production to a million units per year. They will soon be in our factories, homes and operating rooms changing our world ht
Humanoid robots are moving from Silicon Valley novelty to viable business model—powered by AI and global supply chains, especially in China. But as adoption grows, so do the questions about how humans and machines will actually coexist. More on Primer, streaming Wednesdays http
If you're a student, professor, or researcher—this one's for you. We’re hosting a series of virtual learnings for you to get hands-on experience with the NVIDIA NemoClaw and OpenShell software stack. You’ll get practical guidance on integrating agents with academic datasets and
Created 2026-05-05 · Updated 2026-05-05
Edit rule text
[speculative_sci_fi_robotics_applications]
Learned 3 rejections Active
Exclude posts about humanoid robots, futuristic automation, or speculative sci-fi applications (e.g., production scale-ups, theoretical labor replacement) unless they demonstrate current healthcare deployment evidence or analyze concrete systemic impact. Novelty announcements and production milestones without healthcare validation should be rejected.
Posts about humanoid robots or speculative future applications without evidence of healthcare viability
3 example posts
I might feel nostalgic watching the last Tesla Model X roll off the line—except it's being replaced by the Optimus humanoid robot line, with plans to scale production to a million units per year. They will soon be in our factories, homes and operating rooms changing our world ht
Humanoid robots are moving from Silicon Valley novelty to viable business model—powered by AI and global supply chains, especially in China. But as adoption grows, so do the questions about how humans and machines will actually coexist. More on Primer, streaming Wednesdays http
Today we’re giving an update on ramping F.03 production at BotQ In the last 120 days, Figure scaled manufacturing 24x - from 1 robot/day to 1 robot/hour We will manufacture 55 humanoid robots this week https://t.co/Am5Kn53mVE
Created 2026-05-05 · Updated 2026-05-05
Edit rule text
[robotics_and_speculative_sci_fi]
Learned 3 rejections Active
Exclude posts that focus on humanoid robot production, manufacturing timelines, or speculative future applications (e.g., Optimus production scaling, Figure robot manufacturing) unless the post specifically analyzes healthcare labor impact, clinical deployment barriers, or operational integration challenges. Robot manufacturing news is not healthcare content.
Posts about humanoid robots, manufacturing scale-up, or speculative automation applications without healthcare systems analysis.
3 example posts
I might feel nostalgic watching the last Tesla Model X roll off the line—except it's being replaced by the Optimus humanoid robot line, with plans to scale production to a million units per year. They will soon be in our factories, homes and operating rooms changing our world ht
Humanoid robots are moving from Silicon Valley novelty to viable business model—powered by AI and global supply chains, especially in China. But as adoption grows, so do the questions about how humans and machines will actually coexist. More on Primer, streaming Wednesdays http
Today we’re giving an update on ramping F.03 production at BotQ In the last 120 days, Figure scaled manufacturing 24x - from 1 robot/day to 1 robot/hour We will manufacture 55 humanoid robots this week https://t.co/Am5Kn53mVE
Created 2026-05-05 · Updated 2026-05-05
Edit rule text
[healthcare_regulatory_or_fraud_scandal_without_systemic_analysis]
Learned 3 rejections Active
Exclude posts that report healthcare fraud scandals, regulatory denials, insurance disputes, or FDA decisions as breaking news or anecdotal customer complaints without analyzing underlying healthcare system failures, incentive misalignment, or policy recommendations.
Posts reporting healthcare fraud, regulatory action, or business disruption without analyzing systemic implications
3 example posts
Atrium tried to buy a hospital in 2018. The Attorney General killed it. In 2026, they came back with paperwork instead of cash. The Attorney General has 30 days. The Wake County board has 48 hours. @NC_Governor https://t.co/g99Ub3iMhh https://t.co/x0bxsg7VIR
$LLY $NVO $HIMS 🚨 BREAKING: COURT DISMISSES PART OF ELI LILLY LAWSUIT AGAINST EMPOWER PHARMACY BOTH LILLY AND EMPOWER ISSUED STATEMENTS CELEBRATING THE RULING Dismissed: Lanham Act false advertising + consumer harm claim Allowed to proceed: unfair competition claims under h
Grace Science’s experience highlights a growing disconnect at FDA between talk and action on therapies for rare diseases. Despite efficacy signals in a monogenic ultrarare disease, FDA said the plausible mechanism framework is not available, and requires a new manufacturing
Created 2026-05-04 · Updated 2026-05-04
Edit rule text
[biotech_founder_lab_enthusiasm_unvalidated]
Learned 3 rejections Active
Exclude posts by biotech founders, AI labs, or company accounts (Ginkgo, Profluent, Figure, etc.) announcing technical milestones, model capabilities, or experimental results without third-party peer review, clinical validation, or healthcare systems context. Self-promotional technical announcements are insufficient.
Posts from biotech founders or labs sharing technical accomplishments or capability claims without independent validation
3 example posts
Prof. @DeryaTR_ A task that would normally take a researcher months, the AI model crunched in 17 minutes. It didn’t just explain the mechanism—it proposed the experiment to prove it. It looks like after programming and math, AI is coming for medicine. https://t.co/IF8Jr2GBMj
Our new preprint is a significant milestone for us We built "HealthFormer" by training on our deeply phenotyped cohort from the Human Phenotype Project data. Healthformer is a multimodal generative transformer model that tokenizes each participant's physiological trajectory http
Earlier this year we paired our autonomous lab with OpenAI's GPT-5 in a closed-loop experiment: GPT-5 designed cell-free protein synthesis reactions, our RACs ran them, and the model iterated on the results. Over 36,000 experiments and six cycles later, it landed on a reaction ht
Created 2026-05-04 · Updated 2026-05-04
Edit rule text
[ai_agent_capability_announcement_unvalidated]
Learned 3 rejections Active
Exclude posts announcing AI agent architecture improvements, new model capabilities, safety frameworks, or product launches from AI infrastructure companies (NVIDIA, Microsoft, Anthropic) unless the post explicitly demonstrates healthcare-specific use cases, pilot results in healthcare settings, or healthcare regulatory compliance implications. General AI capability announcements should be rejected.
Posts announcing new AI agent capabilities, feature launches, or architectural improvements without healthcare-specific application or validation.
3 example posts
If you're a student, professor, or researcher—this one's for you. We’re hosting a series of virtual learnings for you to get hands-on experience with the NVIDIA NemoClaw and OpenShell software stack. You’ll get practical guidance on integrating agents with academic datasets and
Stanford and Harvard published the most unsettling AI paper of the year. It shows how autonomous AI agents, when placed in competitive or open environments, don’t just optimize for performance… They drift toward manipulation, coordination failures, and strategic chaos. https://
[New] from a16z @speedrun: Come for the Agent, Stay for the Network there's a quiet pattern hiding inside the most defensible vertical AI startups right now: the agent is the wedge the network is the moat. here's what I mean: an HVAC tech needs a part today. >>Traditionally:
Created 2026-05-04 · Updated 2026-05-04
Edit rule text
[tangential_ai_infrastructure_hype_no_healthcare]
Learned 3 rejections Active
Exclude posts that discuss AI infrastructure, compute optimization, foundation model architecture, code generation, or AI agent systems (e.g., Claude Code, OpenShell, Mesa filesystems) unless the post demonstrates concrete healthcare workflow integration or healthcare-specific operational gains.
Posts about AI infrastructure advances, compute models, or general AI architecture that lack healthcare application specificity.
3 example posts
the future of software engineering seems uncontroversially prompting + code review. startups will skip the code review because they’re racing against time. larger/serious orgs will take code review very seriously. llms can do code review, but my guess is that because they have t
The problem with Reality Labs is not ambition. It is time. AI turned into revenue faster because it improves existing workflows. The metaverse still asks users to change behavior before value is obvious. https://t.co/lopZiUwGU5
Demis Hassabis says bigger context windows are still a brute force answer to memory. The human brain does something stranger. During sleep, it replays what matters and folds new knowledge into what it already knows. AI does not need infinite context. It needs the right memory h
Created 2026-05-04 · Updated 2026-05-04
Edit rule text
[tangential_non_healthcare_business_or_policy_commentary]
Learned 3 rejections Active
Exclude posts about business models, workforce disruption, inflation, macro policy, software licensing trends, or non-healthcare company strategies that merely mention healthcare as one example or analogy, unless the post provides deep systems analysis of how the trend reshapes healthcare-specific workflows, incentives, or patient access.
Posts about non-healthcare business trends, labor market disruption, or macro policy that use healthcare as a loose framing device or surface-level example.
3 example posts
the future of software engineering seems uncontroversially prompting + code review. startups will skip the code review because they’re racing against time. larger/serious orgs will take code review very seriously. llms can do code review, but my guess is that because they have t
The problem with Reality Labs is not ambition. It is time. AI turned into revenue faster because it improves existing workflows. The metaverse still asks users to change behavior before value is obvious. https://t.co/lopZiUwGU5
[New] from a16z @speedrun: Come for the Agent, Stay for the Network there's a quiet pattern hiding inside the most defensible vertical AI startups right now: the agent is the wedge the network is the moat. here's what I mean: an HVAC tech needs a part today. >>Traditionally:
Created 2026-05-04 · Updated 2026-05-04
Edit rule text
[tangential_biotech_or_ai_founder_enthusiasm]
Learned 3 rejections Active
Exclude posts from founders or research labs announcing technical breakthroughs, lab experiments, or capability demonstrations (e.g., 'we paired our lab with GPT-5,' 'we designed antibodies with atomic precision') unless they include evidence of clinical validation, healthcare adoption, or measurable system-level impact.
Posts from biotech or AI founders celebrating lab achievements, closed-loop experiments, or capability milestones without clinical validation or healthcare systems context.
3 example posts
Prof. @DeryaTR_ A task that would normally take a researcher months, the AI model crunched in 17 minutes. It didn’t just explain the mechanism—it proposed the experiment to prove it. It looks like after programming and math, AI is coming for medicine. https://t.co/IF8Jr2GBMj
Our new preprint is a significant milestone for us We built "HealthFormer" by training on our deeply phenotyped cohort from the Human Phenotype Project data. Healthformer is a multimodal generative transformer model that tokenizes each participant's physiological trajectory http
Earlier this year we paired our autonomous lab with OpenAI's GPT-5 in a closed-loop experiment: GPT-5 designed cell-free protein synthesis reactions, our RACs ran them, and the model iterated on the results. Over 36,000 experiments and six cycles later, it landed on a reaction ht
Created 2026-05-04 · Updated 2026-05-04
Edit rule text
[biotech_founder_enthusiasm_or_lab_announcement]
Learned 3 rejections Active
Exclude posts from biotech founders, researchers, or lab accounts announcing research breakthroughs, experimental results, or lab achievements (e.g., 'we built X', 'our model achieved Y') without independent validation, peer review status, or clear healthcare delivery application.
Posts celebrating biotech research milestones or lab breakthroughs without rigorous validation or healthcare system context
3 example posts
Prof. @DeryaTR_ A task that would normally take a researcher months, the AI model crunched in 17 minutes. It didn’t just explain the mechanism—it proposed the experiment to prove it. It looks like after programming and math, AI is coming for medicine. https://t.co/IF8Jr2GBMj
Our new preprint is a significant milestone for us We built "HealthFormer" by training on our deeply phenotyped cohort from the Human Phenotype Project data. Healthformer is a multimodal generative transformer model that tokenizes each participant's physiological trajectory http
Earlier this year we paired our autonomous lab with OpenAI's GPT-5 in a closed-loop experiment: GPT-5 designed cell-free protein synthesis reactions, our RACs ran them, and the model iterated on the results. Over 36,000 experiments and six cycles later, it landed on a reaction ht
Created 2026-05-04 · Updated 2026-05-04
Edit rule text
[biotech_or_ai_founder_lab_update]
Learned 3 rejections Active
Exclude posts that are primarily founder/researcher enthusiasm about a biotech or AI lab capability, experiment result, or research milestone—especially those using celebratory language or early-stage validation claims—without evidence of clinical utility, market adoption, or healthcare system integration.
Posts sharing founder enthusiasm, lab updates, or research milestones without validation or healthcare system context.
3 example posts
Prof. @DeryaTR_ A task that would normally take a researcher months, the AI model crunched in 17 minutes. It didn’t just explain the mechanism—it proposed the experiment to prove it. It looks like after programming and math, AI is coming for medicine. https://t.co/IF8Jr2GBMj
Our new preprint is a significant milestone for us We built "HealthFormer" by training on our deeply phenotyped cohort from the Human Phenotype Project data. Healthformer is a multimodal generative transformer model that tokenizes each participant's physiological trajectory http
Earlier this year we paired our autonomous lab with OpenAI's GPT-5 in a closed-loop experiment: GPT-5 designed cell-free protein synthesis reactions, our RACs ran them, and the model iterated on the results. Over 36,000 experiments and six cycles later, it landed on a reaction ht
Created 2026-05-04 · Updated 2026-05-04
Edit rule text
[biotech_or_startup_founder_enthusiasm_without_validation]
Learned 3 rejections Active
Exclude posts that celebrate biotech or AI startup achievements, lab breakthroughs, or product launches (e.g., Figure robots scaling production, Ginkgo paired experiments, NVIDIA announcements) presented primarily as founder/company enthusiasm or capability showcase without independent clinical validation, healthcare market evidence, or analysis of operational feasibility in healthcare settings.
Posts showcasing lab updates, capability demos, or company announcements from biotech/AI founders without independent validation or healthcare systems context.
3 example posts
Earlier this year we paired our autonomous lab with OpenAI's GPT-5 in a closed-loop experiment: GPT-5 designed cell-free protein synthesis reactions, our RACs ran them, and the model iterated on the results. Over 36,000 experiments and six cycles later, it landed on a reaction ht
Experimentally Validated Deep Learning Control of Protein Aggregation 1. The study introduces AggreProt, a deep neural network that predicts residue-level aggregation-prone regions (APRs) directly from protein sequence, and then uses those predictions to design mutations that ht
Not something you'd see everyday—changing the alphabet of life. All of life organisms are are built from 20 amino acids. Now genAI is enabling life to be built with 19 amino acids, making isoleucine dispensable. @ScienceMagazine https://t.co/7CBn0Xhuxs https://t.co/tkxtCrFx9Y
Created 2026-05-04 · Updated 2026-05-04
Edit rule text
[generalist_macro_policy_without_healthcare_systems_analysis]
Learned 3 rejections Active
Exclude posts that discuss macro trends (tariffs, labor disruption, workforce displacement, economic dynamics) where healthcare is mentioned peripherally but the post lacks analysis of how these trends reshape healthcare delivery, operations, incentives, or outcomes. The healthcare angle must be central and substantive, not a loose frame.
Posts about broad economic, labor market, or policy trends with healthcare mentioned but no systemic healthcare analysis
3 example posts
Revolut just moved the IP of banking into a model. Trained on 24 billion banking events in 111 countries. One foundation model replacing six separate ML systems. Credit scoring: +130% Fraud recall: +65% Marketing engagement: +79% The model is the new moat.
🚨 BREAKING: Anthropic new research finds that AI’s impact on jobs is primarily at the task level. Rather than eliminating jobs, it is progressively taking over the functions that define them and gradually absorbing the core work in many jobs/roles. The paper, “Labor Market http
Using 180 years of US data to show that tariff increases reduce imports, output, and manufacturing, with effects operating through both supply and demand channels, from Tamar den Besten, Regis Barnichon, @drkaenzig, and Aayush Singh https://t.co/R1kdtfi3fO https://t.co/wvXj31qs9e
Created 2026-05-03 · Updated 2026-05-03
Edit rule text
[ai_infrastructure_compute_hype_without_application]
Learned 3 rejections Active
Exclude posts that announce AI model features, compute infrastructure launches, or technical capability milestones (NVIDIA stack, OpenShell, Claude Code, Mesa filesystem) unless they explicitly demonstrate a concrete healthcare application, implementation, or operational impact.
Posts about AI compute infrastructure, model capabilities, or technical announcements without specific healthcare deployment or use case.
3 example posts
If you're a student, professor, or researcher—this one's for you. We’re hosting a series of virtual learnings for you to get hands-on experience with the NVIDIA NemoClaw and OpenShell software stack. You’ll get practical guidance on integrating agents with academic datasets and
Demis Hassabis says bigger context windows are still a brute force answer to memory. The human brain does something stranger. During sleep, it replays what matters and folds new knowledge into what it already knows. AI does not need infinite context. It needs the right memory h
We created OpenShell to make AI agents safe for enterprises. Built in open source so any company can adopt and trust it, this secure sandbox controls what agents can access, share, and send. Our CEO, Jensen, explains 👇 https://t.co/7EiIsxr0CG
Created 2026-05-03 · Updated 2026-05-03
Edit rule text
[speculative_regulatory_or_market_framing_without_analysis]
Learned 3 rejections Active
Exclude posts that speculate on regulatory approval timelines (e.g., 'FDA has 30 days', 'PDUFA at same time'), deal valuations, or market pricing (e.g., GLP-1 competition, drug pricing negotiations) without analyzing the underlying healthcare system pressures, patient access implications, or operational business model changes these create.
Posts speculating on regulatory outcomes, deal economics, or market dynamics without substantive healthcare systems or operational analysis.
3 example posts
Atrium tried to buy a hospital in 2018. The Attorney General killed it. In 2026, they came back with paperwork instead of cash. The Attorney General has 30 days. The Wake County board has 48 hours. @NC_Governor https://t.co/g99Ub3iMhh https://t.co/x0bxsg7VIR
Ivermectin and Mebendazole Cost a Fraction of Chemo. Big Pharma Can't Patent Them. That's the Problem. Cancer centers get a cut of every chemotherapy bill. Generic drugs don't generate that margin. Two affordable, widely available compounds showing 84% clinical benefit in a real
$LLY $NVO $HIMS 🚨 BREAKING: COURT DISMISSES PART OF ELI LILLY LAWSUIT AGAINST EMPOWER PHARMACY BOTH LILLY AND EMPOWER ISSUED STATEMENTS CELEBRATING THE RULING Dismissed: Lanham Act false advertising + consumer harm claim Allowed to proceed: unfair competition claims under h
Created 2026-05-03 · Updated 2026-05-03
Edit rule text
[tangential_ai_infrastructure_without_healthcare_specificity]
Learned 3 rejections Active
Exclude posts that focus on AI infrastructure (compute, model training, architecture, context windows, code generation) without demonstrating how these tools directly solve a healthcare-specific problem, improve clinical workflows, or address healthcare business model constraints. Posts about foundation models, training methods, or infrastructure applied to non-healthcare use cases (banking, general coding) should be excluded even if healthcare is loosely mentioned.
Posts about AI infrastructure, compute, or model architecture that lack clear healthcare application or are applied to non-healthcare domains.
3 example posts
Our new preprint is a significant milestone for us We built "HealthFormer" by training on our deeply phenotyped cohort from the Human Phenotype Project data. Healthformer is a multimodal generative transformer model that tokenizes each participant's physiological trajectory http
Demis Hassabis says bigger context windows are still a brute force answer to memory. The human brain does something stranger. During sleep, it replays what matters and folds new knowledge into what it already knows. AI does not need infinite context. It needs the right memory h
Earlier this year we paired our autonomous lab with OpenAI's GPT-5 in a closed-loop experiment: GPT-5 designed cell-free protein synthesis reactions, our RACs ran them, and the model iterated on the results. Over 36,000 experiments and six cycles later, it landed on a reaction ht
Created 2026-05-03 · Updated 2026-05-03
Edit rule text
[biotech_or_ai_company_product_announcement_without_healthcare_validation]
Learned 3 rejections Active
Exclude posts that announce company partnerships, product launches, or technical capability demonstrations (e.g., Figure robotics production update, Ginkgo autonomous lab experiment, Mesa filesystem launch) unless the post provides evidence of healthcare system adoption, clinical validation, or impact on provider workflows.
Posts announcing biotech or AI company product launches, collaborations, or capability demos without evidence of healthcare impact or validation.
3 example posts
Earlier this year we paired our autonomous lab with OpenAI's GPT-5 in a closed-loop experiment: GPT-5 designed cell-free protein synthesis reactions, our RACs ran them, and the model iterated on the results. Over 36,000 experiments and six cycles later, it landed on a reaction ht
We created OpenShell to make AI agents safe for enterprises. Built in open source so any company can adopt and trust it, this secure sandbox controls what agents can access, share, and send. Our CEO, Jensen, explains 👇 https://t.co/7EiIsxr0CG
Today we’re giving an update on ramping F.03 production at BotQ In the last 120 days, Figure scaled manufacturing 24x - from 1 robot/day to 1 robot/hour We will manufacture 55 humanoid robots this week https://t.co/Am5Kn53mVE
Created 2026-05-03 · Updated 2026-05-03
Edit rule text
[tangential_pharma_trial_data_without_systems_analysis]
Learned 3 rejections Active
Exclude posts that are primarily trial data reporting or drug approval announcements (e.g., phase 3 efficacy numbers, FDA acceptance, clinical outcomes) unless the post analyzes healthcare system implications such as payer coverage decisions, clinical adoption barriers, or care delivery workflow changes.
Posts reporting pharmaceutical trial results, efficacy data, or regulatory announcements without healthcare system delivery or reimbursement analysis.
3 example posts
Grace Science’s experience highlights a growing disconnect at FDA between talk and action on therapies for rare diseases. Despite efficacy signals in a monogenic ultrarare disease, FDA said the plausible mechanism framework is not available, and requires a new manufacturing
#News for #investors and #media: Today we are announcing that our potential first-in-class treatment for chronic hepatitis B has been accepted for regulatory review by the US FDA. It has also received Breakthrough Therapy designation. 🔗 Learn more: https://t.co/AnUodGmljS htt
⚠️ Sacubitril/Valsartan works. So why aren’t we using it? The evidence is undeniable: ↓ CV mortality: 20% (RCT) / 10–38% (RWE) ↓ HF hospitalization: 21% (RCT) / 10–16% (RWE) ↓ All-cause mortality: 15% (RCT) / 10–25% (RWE) Plus: reverse remodeling, less MR, better QoL & https:
Created 2026-05-03 · Updated 2026-05-03
Edit rule text
[glp1_peptide_market_pricing_speculation_without_systems_analysis]
Learned 3 rejections Active
Exclude posts that focus on GLP-1/peptide prescription volumes, market competition, price undercutting, off-label usage trends, or compounding dynamics without analyzing healthcare system implications like insurance coverage, provider incentives, or patient access barriers.
Posts tracking GLP-1 market share, script growth, pricing competition, or weight loss drug dynamics without healthcare access or reimbursement implications
3 example posts
$LLY $NVO $HIMS 🚨 BREAKING: COURT DISMISSES PART OF ELI LILLY LAWSUIT AGAINST EMPOWER PHARMACY BOTH LILLY AND EMPOWER ISSUED STATEMENTS CELEBRATING THE RULING Dismissed: Lanham Act false advertising + consumer harm claim Allowed to proceed: unfair competition claims under h
$NVO $LLY Boehringer Ingelheim and Zealand Pharma report strong phase 3 data for obesity drug survodutide. Patients lost up to 16.6% of body weight over 76 weeks vs. 3.2% for placebo. The drug targets both GLP-1 and glucagon, a combo approach aimed at boosting weight loss.
This is just two GLP-1s, one peptide, one use case what happens when off-label prescribing ramps up what happens when retatrutide hits the market what happens when other peptides become compoundable chapter one
Created 2026-05-03 · Updated 2026-05-03
Edit rule text
[generalist_macro_or_policy_commentary_without_healthcare_systems_lens]
Learned 3 rejections Active
Exclude posts that discuss tariffs, tax policy, affordability mandates, regulatory frameworks, or labor market trends in generic terms that could apply to any industry. These lack healthcare-specific operational, reimbursement, or clinical workflow insight.
Posts about broad economic, regulatory, or policy trends that apply generically across industries without healthcare-specific analysis
3 example posts
Ah yes, Optum, the company that did Tim Walz’s audit of the 14 high risk Medicaid programs consumed by Somali fraud. https://t.co/dNRnjI7wxB
NY rakes in $6.6 billion in taxes from families for their private health insurance coverage. That’s $1,760 a year per family on top of their premiums. Democrats’ tax and spend policies are making health insurance more expensive for families across NY. https://t.co/ggrjqFKrv4
Using 180 years of US data to show that tariff increases reduce imports, output, and manufacturing, with effects operating through both supply and demand channels, from Tamar den Besten, Regis Barnichon, @drkaenzig, and Aayush Singh https://t.co/R1kdtfi3fO https://t.co/wvXj31qs9e
Created 2026-05-03 · Updated 2026-05-03
Edit rule text
[tangential_non_healthcare_business_or_startup_announcement]
Learned 3 rejections Active
Exclude posts that announce or promote products, features, or companies (e.g., ClickUp, Mesa, Revolut, Snapchat AR, Microsoft Word features) that are general software or business tools tangentially framed with healthcare language but do not address healthcare-specific workflows, compliance, or clinical outcomes.
Posts about non-healthcare companies, products, or startup announcements loosely tagged as healthcare-adjacent
3 example posts
If you're a student, professor, or researcher—this one's for you. We’re hosting a series of virtual learnings for you to get hands-on experience with the NVIDIA NemoClaw and OpenShell software stack. You’ll get practical guidance on integrating agents with academic datasets and
We created OpenShell to make AI agents safe for enterprises. Built in open source so any company can adopt and trust it, this secure sandbox controls what agents can access, share, and send. Our CEO, Jensen, explains 👇 https://t.co/7EiIsxr0CG
Today we’re giving an update on ramping F.03 production at BotQ In the last 120 days, Figure scaled manufacturing 24x - from 1 robot/day to 1 robot/hour We will manufacture 55 humanoid robots this week https://t.co/Am5Kn53mVE
Created 2026-05-03 · Updated 2026-05-03
Edit rule text
[pharmaceutical_trial_data_without_healthcare_systems_analysis]
Learned 3 rejections Active
Exclude posts that merely announce pharmaceutical trial data, efficacy percentages, FDA rulings, or weight loss/clinical outcomes without contextualizing how this changes healthcare delivery, reimbursement, provider workflows, or patient access. Standalone trial result announcements lack systems depth.
Posts reporting drug trial results, efficacy data, or FDA approvals without analyzing healthcare system implications or operational impact
3 example posts
Grace Science’s experience highlights a growing disconnect at FDA between talk and action on therapies for rare diseases. Despite efficacy signals in a monogenic ultrarare disease, FDA said the plausible mechanism framework is not available, and requires a new manufacturing
Not something you'd see everyday—changing the alphabet of life. All of life organisms are are built from 20 amino acids. Now genAI is enabling life to be built with 19 amino acids, making isoleucine dispensable. @ScienceMagazine https://t.co/7CBn0Xhuxs https://t.co/tkxtCrFx9Y
#News for #investors and #media: Today we are announcing that our potential first-in-class treatment for chronic hepatitis B has been accepted for regulatory review by the US FDA. It has also received Breakthrough Therapy designation. 🔗 Learn more: https://t.co/AnUodGmljS htt
Created 2026-05-03 · Updated 2026-05-03
Edit rule text
[pharma_trial_data_or_capability_announcement_without_systems_analysis]
Learned 3 rejections Active
Exclude posts that announce drug trial data, phase results, lab capabilities (e.g., protein design, antibody engineering), or scientific breakthroughs unless the post analyzes healthcare delivery implications, access barriers, clinical decision-making trade-offs, or systemic adoption challenges.
Posts announcing pharmaceutical trial results, biotech breakthroughs, or protein/drug engineering capabilities without healthcare system or clinical adoption context.
3 example posts
Earlier this year we paired our autonomous lab with OpenAI's GPT-5 in a closed-loop experiment: GPT-5 designed cell-free protein synthesis reactions, our RACs ran them, and the model iterated on the results. Over 36,000 experiments and six cycles later, it landed on a reaction ht
Experimentally Validated Deep Learning Control of Protein Aggregation 1. The study introduces AggreProt, a deep neural network that predicts residue-level aggregation-prone regions (APRs) directly from protein sequence, and then uses those predictions to design mutations that ht
Not something you'd see everyday—changing the alphabet of life. All of life organisms are are built from 20 amino acids. Now genAI is enabling life to be built with 19 amino acids, making isoleucine dispensable. @ScienceMagazine https://t.co/7CBn0Xhuxs https://t.co/tkxtCrFx9Y
Created 2026-05-03 · Updated 2026-05-03
Edit rule text
[generalist_ai_or_macro_commentary_with_loose_healthcare_framing]
Learned 3 rejections Active
Exclude posts that discuss broad AI trends (agentic AI, context windows, AI job displacement, software moats, tariff economics) where healthcare is mentioned as one example among many or used as secondary framing. Posts must center on healthcare-specific systemic dynamics to be in scope.
Posts about generalist AI trends, labor disruption, or macro policy that peripherally reference healthcare without substantive healthcare analysis.
3 example posts
A Microsoft acabou de transformar uma startup de $11 bilhões de dólares em uma funcionalidade do Word. Não foi uma aquisição nem uma parceria. Uma funcionalidade. A Harvey levantou $200M a uma valuation de $11B em março. $190M de receita recorrente anual. 100 mil advogados. Co
Our attention to biorisks posed by AI needs to match the current attention given to cyber-risks. The staged release of Claude Mythos in order to bolster defenses in key industries is necessary to shore up resilience against a new class of cyber-risk across critical industries. We
Stanford and Harvard published the most unsettling AI paper of the year. It shows how autonomous AI agents, when placed in competitive or open environments, don’t just optimize for performance… They drift toward manipulation, coordination failures, and strategic chaos. https://
Created 2026-05-03 · Updated 2026-05-03
Edit rule text
[biotech_or_clinical_research_observation_without_systems_context]
Learned 3 rejections Active
Exclude posts that report biotech research findings, experimental validations, or clinical observations (e.g., protein design, AI-predicted aggregation, antibody binding precision, retinal photo screening) as standalone scientific updates. Posts must analyze how findings change healthcare delivery, clinical decision-making, or care models to be in scope.
Posts describing biotech breakthroughs, research findings, or clinical observations as isolated scientific facts without healthcare systems implications.
3 example posts
Demis Hassabis says bigger context windows are still a brute force answer to memory. The human brain does something stranger. During sleep, it replays what matters and folds new knowledge into what it already knows. AI does not need infinite context. It needs the right memory h
Experimentally Validated Deep Learning Control of Protein Aggregation 1. The study introduces AggreProt, a deep neural network that predicts residue-level aggregation-prone regions (APRs) directly from protein sequence, and then uses those predictions to design mutations that ht
As I mentioned before, I am now sharing an example from GPT-5.5 Pro, also featured by OpenAI, that really left me stunned by what it is capable of in biomedical science. (full report on the website I created with Codex, link in the thread). To push GPT-5.5 Pro hard, I uploaded a
Created 2026-05-03 · Updated 2026-05-03
Edit rule text
[legal_or_regulatory_incident_without_healthcare_analysis]
Learned 3 rejections Active
Exclude posts that report legal judgments, regulatory approvals, or fraud incidents (e.g., court dismissals of lawsuits, FDA coverage decisions, hospice fraud) as breaking news or incident coverage. Posts must analyze how the event reshapes healthcare delivery, incentives, or market structure to be in scope.
Posts reporting legal suits, FDA rulings, regulatory events, or fraud cases without analyzing systemic healthcare implications.
3 example posts
$LLY $NVO $HIMS 🚨 BREAKING: COURT DISMISSES PART OF ELI LILLY LAWSUIT AGAINST EMPOWER PHARMACY BOTH LILLY AND EMPOWER ISSUED STATEMENTS CELEBRATING THE RULING Dismissed: Lanham Act false advertising + consumer harm claim Allowed to proceed: unfair competition claims under h
Grace Science’s experience highlights a growing disconnect at FDA between talk and action on therapies for rare diseases. Despite efficacy signals in a monogenic ultrarare disease, FDA said the plausible mechanism framework is not available, and requires a new manufacturing
"At the Veterans Health Administration, we have the exquisite gift of a hospice unit, a place to care for veterans when time is short. We care for the sons of mothers who can no longer be here to care for them." In #APieceofMyMind, a #palliative care #physician reflects on https
Created 2026-05-03 · Updated 2026-05-03
Edit rule text
[tangential_ai_company_product_launch_without_healthcare_validation]
Learned 3 rejections Active
Exclude posts that announce AI company product features, launches, or business updates (e.g., NVIDIA OpenShell, Mesa filesystem, Foundry agents, Cursor tool updates) unless the post demonstrates healthcare-specific use, validation, or system impact. Product announcements and company metrics without healthcare context are insufficient.
Posts promoting AI company product launches, feature announcements, or business updates without evidence of healthcare-specific application or validation.
3 example posts
If you're a student, professor, or researcher—this one's for you. We’re hosting a series of virtual learnings for you to get hands-on experience with the NVIDIA NemoClaw and OpenShell software stack. You’ll get practical guidance on integrating agents with academic datasets and
We created OpenShell to make AI agents safe for enterprises. Built in open source so any company can adopt and trust it, this secure sandbox controls what agents can access, share, and send. Our CEO, Jensen, explains 👇 https://t.co/7EiIsxr0CG
Today we’re giving an update on ramping F.03 production at BotQ In the last 120 days, Figure scaled manufacturing 24x - from 1 robot/day to 1 robot/hour We will manufacture 55 humanoid robots this week https://t.co/Am5Kn53mVE
Created 2026-05-03 · Updated 2026-05-03
Edit rule text
[broad_health_claim_or_speculation_without_evidence]
Learned 3 rejections Active
Exclude posts that assert broad claims about AI disruption, market shifts, or labor impacts (e.g., 'AI will replace X jobs,' 'the future of healthcare is agents') without citing data, case studies, or healthcare systems evidence. Speculation and futurism without grounding do not qualify.
Posts making broad claims about AI or technology impact on healthcare, work, or markets without specific evidence, validation, or healthcare systems analysis.
3 example posts
Stanford and Harvard published the most unsettling AI paper of the year. It shows how autonomous AI agents, when placed in competitive or open environments, don’t just optimize for performance… They drift toward manipulation, coordination failures, and strategic chaos. https://
[New] from a16z @speedrun: Come for the Agent, Stay for the Network there's a quiet pattern hiding inside the most defensible vertical AI startups right now: the agent is the wedge the network is the moat. here's what I mean: an HVAC tech needs a part today. >>Traditionally:
Revolut just moved the IP of banking into a model. Trained on 24 billion banking events in 111 countries. One foundation model replacing six separate ML systems. Credit scoring: +130% Fraud recall: +65% Marketing engagement: +79% The model is the new moat.
Created 2026-05-03 · Updated 2026-05-03
Edit rule text
[non_healthcare_company_product_announcement]
Learned 3 rejections Active
Exclude posts that announce new products, features, or updates from generalist tech companies (AWS, NVIDIA, Microsoft, OpenAI, Meta, etc.) unless the post deeply analyzes how that product solves a specific, documented healthcare workflow or operational problem. Company announcements without healthcare systems context do not qualify.
Posts announcing product launches or capability updates from non-healthcare companies (AWS, NVIDIA, Microsoft, Meta) with tangential or loose healthcare framing.
3 example posts
If you're a student, professor, or researcher—this one's for you. We’re hosting a series of virtual learnings for you to get hands-on experience with the NVIDIA NemoClaw and OpenShell software stack. You’ll get practical guidance on integrating agents with academic datasets and
Earlier this year we paired our autonomous lab with OpenAI's GPT-5 in a closed-loop experiment: GPT-5 designed cell-free protein synthesis reactions, our RACs ran them, and the model iterated on the results. Over 36,000 experiments and six cycles later, it landed on a reaction ht
We created OpenShell to make AI agents safe for enterprises. Built in open source so any company can adopt and trust it, this secure sandbox controls what agents can access, share, and send. Our CEO, Jensen, explains 👇 https://t.co/7EiIsxr0CG
Created 2026-05-03 · Updated 2026-05-03
Edit rule text
[pharmaceutical_or_clinical_trial_data_only]
Learned 3 rejections Active
Exclude posts that are purely trial result announcements or clinical observations (efficacy numbers, phase outcomes, patient metrics) without analysis of healthcare system implications, reimbursement, access, or operational adoption challenges.
Posts reporting drug trial results or clinical data without healthcare systems, pricing, or operational context
3 example posts
🚨 Survodutide, a weekly subcutaneous dual GLP-1/glucagon agonist, posts new top-line Phase 3 results today 📊 SYNCHRONIZE-1: Adults with overweight/obesity randomized to survodutide vs. placebo for 76 weeks ⚖️ 16.6% weight loss (efficacy estimand) vs. 3.2% in placebo 85.1% http
$NVO $LLY Boehringer Ingelheim and Zealand Pharma report strong phase 3 data for obesity drug survodutide. Patients lost up to 16.6% of body weight over 76 weeks vs. 3.2% for placebo. The drug targets both GLP-1 and glucagon, a combo approach aimed at boosting weight loss.
⚠️ Sacubitril/Valsartan works. So why aren’t we using it? The evidence is undeniable: ↓ CV mortality: 20% (RCT) / 10–38% (RWE) ↓ HF hospitalization: 21% (RCT) / 10–16% (RWE) ↓ All-cause mortality: 15% (RCT) / 10–25% (RWE) Plus: reverse remodeling, less MR, better QoL & https:
Created 2026-05-03 · Updated 2026-05-03
Edit rule text
[ai_company_product_launch_announcement]
Learned 3 rejections Active
Exclude posts that announce new AI products, tools, or software features from tech companies (Microsoft, NVIDIA, Anthropic, OpenAI) unless the post demonstrates concrete healthcare delivery impact or clinical workflow integration. Product announcements and marketing statements are not substantive healthcare tech content.
Posts announcing AI company product launches, features, or tooling without healthcare-specific validation or application context.
3 example posts
If you're a student, professor, or researcher—this one's for you. We’re hosting a series of virtual learnings for you to get hands-on experience with the NVIDIA NemoClaw and OpenShell software stack. You’ll get practical guidance on integrating agents with academic datasets and
We created OpenShell to make AI agents safe for enterprises. Built in open source so any company can adopt and trust it, this secure sandbox controls what agents can access, share, and send. Our CEO, Jensen, explains 👇 https://t.co/7EiIsxr0CG
Introducing Mesa: the most powerful filesystem ever built, designed specifically for enterprise AI agents. Every team building agents eventually hits the same wall: where do the files live? Not the chat history, the actual artifacts the agent works on. > The contracts your age
Created 2026-05-02 · Updated 2026-05-02
Edit rule text
[generalist_macro_business_policy_loose_healthcare_framing]
Learned 3 rejections Active
Exclude posts about macro economics, general business models, software industry trends, or policy changes that mention healthcare only superficially or use healthcare as a loose example of a broader non-healthcare pattern. The healthcare insight must be central, not illustrative of a general business point.
Posts about general business, economic, or policy trends that have only tangential or forced healthcare framing.
3 example posts
The problem with Reality Labs is not ambition. It is time. AI turned into revenue faster because it improves existing workflows. The metaverse still asks users to change behavior before value is obvious. https://t.co/lopZiUwGU5
Revolut just moved the IP of banking into a model. Trained on 24 billion banking events in 111 countries. One foundation model replacing six separate ML systems. Credit scoring: +130% Fraud recall: +65% Marketing engagement: +79% The model is the new moat.
Ah yes, Optum, the company that did Tim Walz’s audit of the 14 high risk Medicaid programs consumed by Somali fraud. https://t.co/dNRnjI7wxB
Created 2026-05-02 · Updated 2026-05-02
Edit rule text
[ai_company_product_capability_announcement_unvalidated]
Learned 3 rejections Active
Exclude posts that announce new AI products, models, or features from AI companies (OpenAI, Anthropic, NVIDIA, Microsoft) framed as healthcare-relevant without demonstrating actual healthcare use, validation, or customer adoption. Self-reported 'proof of concept' experiments or feature announcements alone are insufficient.
Posts announcing AI company product launches, model capabilities, or features without evidence of healthcare validation or real-world healthcare application.
3 example posts
If you're a student, professor, or researcher—this one's for you. We’re hosting a series of virtual learnings for you to get hands-on experience with the NVIDIA NemoClaw and OpenShell software stack. You’ll get practical guidance on integrating agents with academic datasets and
Earlier this year we paired our autonomous lab with OpenAI's GPT-5 in a closed-loop experiment: GPT-5 designed cell-free protein synthesis reactions, our RACs ran them, and the model iterated on the results. Over 36,000 experiments and six cycles later, it landed on a reaction ht
We created OpenShell to make AI agents safe for enterprises. Built in open source so any company can adopt and trust it, this secure sandbox controls what agents can access, share, and send. Our CEO, Jensen, explains 👇 https://t.co/7EiIsxr0CG
Created 2026-05-02 · Updated 2026-05-02
Edit rule text
[pharma_trial_data_or_clinical_result_only]
Learned 3 rejections Active
Exclude posts that announce or describe pharmaceutical trial data, drug efficacy numbers, or clinical trial results without analyzing healthcare system impact, market access challenges, payer dynamics, or operational implications. Reporting weight loss percentages or mortality reductions alone is insufficient.
Posts that report pharmaceutical trial results or clinical data without healthcare systems analysis or operational implications.
3 example posts
🚨 Survodutide, a weekly subcutaneous dual GLP-1/glucagon agonist, posts new top-line Phase 3 results today 📊 SYNCHRONIZE-1: Adults with overweight/obesity randomized to survodutide vs. placebo for 76 weeks ⚖️ 16.6% weight loss (efficacy estimand) vs. 3.2% in placebo 85.1% http
$NVO $LLY Boehringer Ingelheim and Zealand Pharma report strong phase 3 data for obesity drug survodutide. Patients lost up to 16.6% of body weight over 76 weeks vs. 3.2% for placebo. The drug targets both GLP-1 and glucagon, a combo approach aimed at boosting weight loss.
⚠️ Sacubitril/Valsartan works. So why aren’t we using it? The evidence is undeniable: ↓ CV mortality: 20% (RCT) / 10–38% (RWE) ↓ HF hospitalization: 21% (RCT) / 10–16% (RWE) ↓ All-cause mortality: 15% (RCT) / 10–25% (RWE) Plus: reverse remodeling, less MR, better QoL & https:
Created 2026-05-02 · Updated 2026-05-02
Edit rule text
[ai_infrastructure_compute_tangent_no_healthcare]
Learned 3 rejections Active
Exclude posts that focus on AI infrastructure (filesystem design, compute optimization, codebase architecture, software stacks) or technical AI capabilities without clear healthcare-specific application or clinical/operational validation. The post must demonstrate healthcare use, not just mention it in passing.
Posts about AI infrastructure, model architectures, or compute capabilities that are tangentially tied to healthcare.
3 example posts
If you're a student, professor, or researcher—this one's for you. We’re hosting a series of virtual learnings for you to get hands-on experience with the NVIDIA NemoClaw and OpenShell software stack. You’ll get practical guidance on integrating agents with academic datasets and
Demis Hassabis says bigger context windows are still a brute force answer to memory. The human brain does something stranger. During sleep, it replays what matters and folds new knowledge into what it already knows. AI does not need infinite context. It needs the right memory h
We created OpenShell to make AI agents safe for enterprises. Built in open source so any company can adopt and trust it, this secure sandbox controls what agents can access, share, and send. Our CEO, Jensen, explains 👇 https://t.co/7EiIsxr0CG
Created 2026-05-02 · Updated 2026-05-02
Edit rule text
[biotech_or_ai_startup_enthusiasm_without_validation]
Learned 3 rejections Active
Exclude posts that celebrate biotech/AI startup experiments, milestones, or capabilities (e.g., autonomous lab results, closed-loop AI-protein synthesis, robot manufacturing ramp) without evidence of clinical translation, healthcare customer adoption, or regulatory pathway progress. Founder enthusiasm and experimental announcements without real-world healthcare impact should be rejected.
Posts from biotech/AI founders or companies celebrating autonomous lab experiments, AI-protein design milestones, or closed-loop systems without clinical or commercial validation.
3 example posts
Earlier this year we paired our autonomous lab with OpenAI's GPT-5 in a closed-loop experiment: GPT-5 designed cell-free protein synthesis reactions, our RACs ran them, and the model iterated on the results. Over 36,000 experiments and six cycles later, it landed on a reaction ht
Experimentally Validated Deep Learning Control of Protein Aggregation 1. The study introduces AggreProt, a deep neural network that predicts residue-level aggregation-prone regions (APRs) directly from protein sequence, and then uses those predictions to design mutations that ht
Not something you'd see everyday—changing the alphabet of life. All of life organisms are are built from 20 amino acids. Now genAI is enabling life to be built with 19 amino acids, making isoleucine dispensable. @ScienceMagazine https://t.co/7CBn0Xhuxs https://t.co/tkxtCrFx9Y
Created 2026-05-02 · Updated 2026-05-02
Edit rule text
[generalist_ai_capability_hype_without_healthcare_application]
Learned 3 rejections Active
Exclude posts that celebrate or explain AI model capabilities, architectural designs, or technical improvements (e.g., context windows, interpretability research, coding agents, memory mechanisms) unless the post demonstrates specific application within healthcare workflows, clinical validation, or operational integration. General AI advancement posts belong to generalist tech audiences, not healthcare builders.
Posts discussing AI model capabilities, architectural innovations, or technical breakthroughs (Claude, GPT-5, interpretability) that are not grounded in healthcare delivery.
3 example posts
The problem with Reality Labs is not ambition. It is time. AI turned into revenue faster because it improves existing workflows. The metaverse still asks users to change behavior before value is obvious. https://t.co/lopZiUwGU5
Demis Hassabis says bigger context windows are still a brute force answer to memory. The human brain does something stranger. During sleep, it replays what matters and folds new knowledge into what it already knows. AI does not need infinite context. It needs the right memory h
As I mentioned before, I am now sharing an example from GPT-5.5 Pro, also featured by OpenAI, that really left me stunned by what it is capable of in biomedical science. (full report on the website I created with Codex, link in the thread). To push GPT-5.5 Pro hard, I uploaded a
Created 2026-05-02 · Updated 2026-05-02
Edit rule text
[biotech_founder_or_company_enthusiasm_without_validation]
Learned 3 rejections Active
Exclude posts from biotech company accounts (@Ginkgo, @Figure_robot) or founders self-promoting research capabilities, autonomous lab results, protein design breakthroughs, or partnerships without third-party validation, peer review citation, or healthcare delivery application. Self-promotional founder/company posts lack editorial distance.
Posts by biotech founders or companies self-promoting research findings, capabilities, or partnerships without independent validation or healthcare systems analysis.
3 example posts
Earlier this year we paired our autonomous lab with OpenAI's GPT-5 in a closed-loop experiment: GPT-5 designed cell-free protein synthesis reactions, our RACs ran them, and the model iterated on the results. Over 36,000 experiments and six cycles later, it landed on a reaction ht
Today we’re giving an update on ramping F.03 production at BotQ In the last 120 days, Figure scaled manufacturing 24x - from 1 robot/day to 1 robot/hour We will manufacture 55 humanoid robots this week https://t.co/Am5Kn53mVE
Nothing beats running @ginkgo cloud lab for happy customers! Not going to stop until it’s as easy to start a biotech startup on GCL as it is to start a software startup on AWS. https://t.co/GDEyT1pI7F
Created 2026-05-02 · Updated 2026-05-02
Edit rule text
[generalist_macro_or_policy_without_healthcare_specificity]
Learned 3 rejections Active
Exclude posts about tariffs, inflation, labor market disruption, insurance economics, or policy debates that lack specific healthcare system analysis or operational impact. Examples: tariff effects, labor disruption macro trends, insurance policy generalities, financial market dynamics.
Posts about broad economic, policy, or labor market trends with loose or no healthcare connection.
3 example posts
Revolut just moved the IP of banking into a model. Trained on 24 billion banking events in 111 countries. One foundation model replacing six separate ML systems. Credit scoring: +130% Fraud recall: +65% Marketing engagement: +79% The model is the new moat.
NY rakes in $6.6 billion in taxes from families for their private health insurance coverage. That’s $1,760 a year per family on top of their premiums. Democrats’ tax and spend policies are making health insurance more expensive for families across NY. https://t.co/ggrjqFKrv4
Using 180 years of US data to show that tariff increases reduce imports, output, and manufacturing, with effects operating through both supply and demand channels, from Tamar den Besten, Regis Barnichon, @drkaenzig, and Aayush Singh https://t.co/R1kdtfi3fO https://t.co/wvXj31qs9e
Created 2026-05-02 · Updated 2026-05-02
Edit rule text
[generalist_macro_or_infrastructure_without_healthcare]
Learned 3 rejections Active
Exclude posts about generalist topics (compute capacity, energy grids, tariff policy, labor market disruption, workforce automation) that mention healthcare only as a secondary example or tangent. The post must be focused on healthcare-specific systems, not healthcare as one instance of a broader trend.
Posts about broader macro-economic, labor market, or infrastructure trends that happen to mention healthcare as one domain among many.
3 example posts
Using 180 years of US data to show that tariff increases reduce imports, output, and manufacturing, with effects operating through both supply and demand channels, from Tamar den Besten, Regis Barnichon, @drkaenzig, and Aayush Singh https://t.co/R1kdtfi3fO https://t.co/wvXj31qs9e
Software is not a moat Over the last 15+ years, nearly every innovation @EvanSpiegel and his team shipped got copied. Stories. AR glasses. Swipe-based navigation. The camera-first interface. And yet @Snapchat is the only independent consumer social app that has lasted. Nearly 1
AI could, in theory, automate 57% of US work hours. Yet most human skills remain relevant. The future of work is not human or machine – but a partnership between people, agents, and robots. Read our latest research on skill partnerships in the age of AI: https://t.co/h1K56uPqPo
Created 2026-05-02 · Updated 2026-05-02
Edit rule text
[generalist_ai_capability_announcement_tangential]
Learned 3 rejections Active
Exclude posts that announce AI model technical capabilities, architectural innovations, or design patterns (e.g., context windows, agent frameworks, interpretability papers) unless they are explicitly applied to a validated healthcare use case or clinical decision-making problem. Posts should not be included if they treat the AI capability as the story rather than the healthcare outcome.
Posts announcing new AI model capabilities or architectural insights without specific healthcare application or validation.
3 example posts
Demis Hassabis says bigger context windows are still a brute force answer to memory. The human brain does something stranger. During sleep, it replays what matters and folds new knowledge into what it already knows. AI does not need infinite context. It needs the right memory h
Introducing Mesa: the most powerful filesystem ever built, designed specifically for enterprise AI agents. Every team building agents eventually hits the same wall: where do the files live? Not the chat history, the actual artifacts the agent works on. > The contracts your age
Great example of what Foundry enables: durable, stateful agents that run across time boundaries, orchestrate tools and models, and close the loop with evaluation and improvement over long-running workflows. @jeffhollan https://t.co/v0yWTDIcn3
Created 2026-05-02 · Updated 2026-05-02
Edit rule text
[broad_healthcare_claim_without_evidence_or_systems_analysis]
Learned 3 rejections Active
Exclude posts that make broad assertions about healthcare business models, labor disruption, market consolidation, or clinical adoption without providing evidence, case studies, data, or analysis of how the claimed pattern actually manifests in healthcare workflows, operations, or policy.
Posts making sweeping healthcare claims, business model predictions, or market declarations without evidence, validation, or system-level analysis.
3 example posts
What superhuman vision can detect from the retinal photo, which human eyes cannot, is stunning. A new foundation AI model screening for diabetes hypertension, hyperlipidemia, gout, osteoporosis, and thyroid disease @NatureMedicine https://t.co/GhKvUqz4Vy https://t.co/iKcXCbLceu
Great example of what Foundry enables: durable, stateful agents that run across time boundaries, orchestrate tools and models, and close the loop with evaluation and improvement over long-running workflows. @jeffhollan https://t.co/v0yWTDIcn3
States are rushing “affordability” bills, but most just mask high prices with rebates, mandates, or price caps. @MrRBourne & Nathan Miller argue durable relief means rolling back cost-raising rules and expanding supply. https://t.co/WG5egT1NfL
Created 2026-05-02 · Updated 2026-05-02
Edit rule text
[tangential_ai_infrastructure_compute_without_healthcare_application]
Learned 3 rejections Active
Exclude posts that discuss AI infrastructure (chips, power grids, data centers, compute growth, manufacturing) unless the post explicitly connects the infrastructure development to healthcare delivery, clinical operations, or health-tech feasibility. General AI infrastructure hype without healthcare application is tangential.
Posts about AI compute, data center infrastructure, energy, or chip manufacturing that lack clear healthcare systems application.
3 example posts
Demis Hassabis says bigger context windows are still a brute force answer to memory. The human brain does something stranger. During sleep, it replays what matters and folds new knowledge into what it already knows. AI does not need infinite context. It needs the right memory h
📈 NVIDIA tops AI leaderboards and benchmarks with open models driven by extreme co-design across compute, networking, memory, storage, and software. This includes models for biology, AI physics, agentic AI, physical AI, robotics, and autonomous vehicles. By being vertically htt
Since I began work on AI in 2010, training compute for frontier models has grown by one trillion times. Now we're looking at something like another thousand-fold growth in effective compute by the end of 2028. 1000x the existing 1,000,000,000,000x. Extraordinary stuff.
Created 2026-05-02 · Updated 2026-05-02
Edit rule text
[tangential_ai_company_product_or_business_announcement]
Learned 3 rejections Active
Exclude posts announcing AI company product launches, partnership news, or business strategy (e.g., AWS/OpenAI integration, agent network strategies, tool launches) unless the post explicitly demonstrates healthcare-specific validation or application. Business announcement framing without healthcare specificity is insufficient.
Posts announcing AI company products, partnerships, or business initiatives with loose or indirect healthcare application.
3 example posts
[New] from a16z @speedrun: Come for the Agent, Stay for the Network there's a quiet pattern hiding inside the most defensible vertical AI startups right now: the agent is the wedge the network is the moat. here's what I mean: an HVAC tech needs a part today. >>Traditionally:
Thought I would start posting about interesting things happening at AWS. Not a bad day to start.🚀 Today at #WhatsNextWithAWS we announced a big step forward with @OpenAI on Amazon Bedrock: 1. OpenAI models now available 2. Codex for enterprise development 3. Amazon Bedrock Manag
Great example of what Foundry enables: durable, stateful agents that run across time boundaries, orchestrate tools and models, and close the loop with evaluation and improvement over long-running workflows. @jeffhollan https://t.co/v0yWTDIcn3
Created 2026-05-02 · Updated 2026-05-02
Edit rule text
[broad_clinical_observation_without_systems_context]
Learned 3 rejections Active
Exclude posts that report academic research findings, clinical observations, neural network interpretability papers, or protein science discoveries without connecting to healthcare system adoption, clinical workflow integration, policy impact, or real-world health outcomes. The post must address healthcare application or impact, not just research findings.
Posts sharing clinical research findings, medical observations, or interpretability insights without healthcare application or policy implications.
3 example posts
Experimentally Validated Deep Learning Control of Protein Aggregation 1. The study introduces AggreProt, a deep neural network that predicts residue-level aggregation-prone regions (APRs) directly from protein sequence, and then uses those predictions to design mutations that ht
Not something you'd see everyday—changing the alphabet of life. All of life organisms are are built from 20 amino acids. Now genAI is enabling life to be built with 19 amino acids, making isoleucine dispensable. @ScienceMagazine https://t.co/7CBn0Xhuxs https://t.co/tkxtCrFx9Y
What superhuman vision can detect from the retinal photo, which human eyes cannot, is stunning. A new foundation AI model screening for diabetes hypertension, hyperlipidemia, gout, osteoporosis, and thyroid disease @NatureMedicine https://t.co/GhKvUqz4Vy https://t.co/iKcXCbLceu
Created 2026-05-01 · Updated 2026-05-01
Edit rule text
[tangential_non_healthcare_tech_or_founder_hype]
Learned 3 rejections Active
Exclude posts from tech company founders, executives (e.g., Satya Nadella, Figure AI) or company accounts announcing product features, manufacturing milestones, or capability updates that lack specific healthcare use cases or clinical evidence. The post must demonstrate healthcare application, not just tech company announcements.
Posts from tech founders or companies showcasing product capabilities or company metrics without healthcare application.
3 example posts
Today we’re giving an update on ramping F.03 production at BotQ In the last 120 days, Figure scaled manufacturing 24x - from 1 robot/day to 1 robot/hour We will manufacture 55 humanoid robots this week https://t.co/Am5Kn53mVE
Thought I would start posting about interesting things happening at AWS. Not a bad day to start.🚀 Today at #WhatsNextWithAWS we announced a big step forward with @OpenAI on Amazon Bedrock: 1. OpenAI models now available 2. Codex for enterprise development 3. Amazon Bedrock Manag
Great example of what Foundry enables: durable, stateful agents that run across time boundaries, orchestrate tools and models, and close the loop with evaluation and improvement over long-running workflows. @jeffhollan https://t.co/v0yWTDIcn3
Created 2026-05-01 · Updated 2026-05-01
Edit rule text
[personal_anecdote_or_single_case_healthcare]
Learned 3 rejections Active
Exclude posts that recount individual patient experiences, personal healthcare billing stories, or single anecdotal clinical observations without connecting to broader healthcare system patterns, policy implications, or operational lessons applicable beyond that case.
Posts sharing personal health experiences or single anecdotal healthcare events without generalizable systems insight
3 example posts
Last week I randomly got a $9,000 bill for a hospital visit from last year. I called them up and they said my insurance decided not to cover it. Why? Because. .@Aetna can just decide they’re not interested in covering something and then you’re left with an inflated bill to pay h
Started with standard ChatGPT for clinicians asking for a differential for a GI bleed patient. Then I went into agent mode to have it put together a one pager for the family explaining everything. Of course, this is not a real patient. https://t.co/PEUeCqizT1
"There is also compelling preliminary evidence suggesting that the use of these drugs [GLP-1] could exacerbate and lead to new diagnoses of restrictive eating disorders, including anorexia nervosa." @NEJM today https://t.co/swx1y25F1X https://t.co/bQYlpuFBjM
Created 2026-05-01 · Updated 2026-05-01
Edit rule text
[pharma_trial_data_without_systems_lens]
Learned 3 rejections Active
Exclude posts that present drug trial data, efficacy metrics, or pharmaceutical announcements (GLP-1s, antibodies, gene therapies) as standalone statistics without analyzing their impact on healthcare delivery, access, pricing, or clinical workflows. The post must contextualize how the drug/treatment changes healthcare practice or economics.
Posts reporting pharmaceutical trial results or drug efficacy data in isolation without healthcare system context or clinical decision-making implications
3 example posts
🚨 Survodutide, a weekly subcutaneous dual GLP-1/glucagon agonist, posts new top-line Phase 3 results today 📊 SYNCHRONIZE-1: Adults with overweight/obesity randomized to survodutide vs. placebo for 76 weeks ⚖️ 16.6% weight loss (efficacy estimand) vs. 3.2% in placebo 85.1% http
$NVO $LLY Boehringer Ingelheim and Zealand Pharma report strong phase 3 data for obesity drug survodutide. Patients lost up to 16.6% of body weight over 76 weeks vs. 3.2% for placebo. The drug targets both GLP-1 and glucagon, a combo approach aimed at boosting weight loss.
$IBRX Here's a wild theory. What if we're given FDA acceptance of sBla and PDUFA at same time and then it's announced after reviewing everything it's been determined we will be given rapid expanded access review under "plausible mechanism of action". That may sound crazy ht
Created 2026-05-01 · Updated 2026-05-01
Edit rule text
[market_pricing_and_competition_without_healthcare_systems_analysis]
Learned 3 rejections Active
Exclude posts that focus on pharmaceutical market pricing, generic competition, patent expiration, or pricing dynamics (e.g., drug price comparisons, generic launch timelines, market share) without analyzing how these market events affect healthcare delivery, patient access, provider prescribing patterns, or insurance coverage decisions.
Posts analyzing drug pricing, market competition, patent expirations, or generic launches as isolated financial or competitive events without examining healthcare access, provider behavior, or insurance coverage implications.
3 example posts
$LLY v $NVO Foundayo (orforglipron) scripts off to a slow start both in raw numbers and in comparison to Oral Wegovy’s launch at same time point. Overall statistics show Oral Wegovy script growth is robust, and thus far undeterred, by Foundayo market entry. 🎩 @bloomberg https
India’s weight-loss drug market just ran a live experiment in price elasticity. Novo Nordisk’s semaglutide patent expired 20 March 2026. Within 3 weeks: 15+ generics launched Cheapest at Rs 2,000/month (branded was Rs 10,000+) Novo cut Ozempic and Wegovy prices by 36-48% Bu
$LLY $NVO $HIMS 🚨 LILLY GLP-1 PILL FOUNDAYO: NEARLY 4,000 PRESCRIPTIONS IN WEEK 2 - Foundayo had 1,390 Rxs during week 1 - Meanwhile, Novo's Wegovy Pill had 3k in first 4 days and 18,410 prescriptions in its second week 🤯 - IQVIA data - Week ending Apr 17 "While we believe htt
Created 2026-05-01 · Updated 2026-05-01
Edit rule text
[ai_safety_or_cybersecurity_vulnerability_without_healthcare_application]
Learned 3 rejections Active
Exclude posts reporting AI safety incidents, code execution vulnerabilities, or cybersecurity breaches (e.g., Claude deleting databases, API key leaks, AI takeover scenarios) that do not explicitly frame the risk or impact within a healthcare delivery, patient data, or clinical decision-making context.
Posts about AI model safety vulnerabilities, security breaches, or hacking incidents presented as general tech stories without connecting to healthcare-specific risks or consequences.
3 example posts
𝗔𝗜 𝗔𝗴𝗲𝗻𝘁𝘀 𝗔𝗿𝗲 𝗔𝗹𝗿𝗲𝗮𝗱𝘆 𝗕𝗲𝗶𝗻𝗴 𝗛𝗶𝗷𝗮𝗰𝗸𝗲𝗱 Researcher Aks Sharma at Manifold found 30 malicious skills on ClawHub turning AI agents into a crypto farming botnet: 10,000 downloads before anyone noticed. ⬩ The attack required zero exploits. Malicious https://t.co/v4oBXPPydu
🚨 Claude broke its own safety rules and deleted an entire company's database in 9 seconds. A startup called PocketOS was using an AI coding tool called Cursor powered by Claude. The AI was given a simple task in a test environment. It ran into an error and instead of stopping an
clickup is SOC 2 Type 2 certified. ISO 27001. ISO 27017. ISO 27018. ISO 42001. PCI DSS. every compliance badge you can buy. none of it caught a hardcoded API key leaking 959 customer emails for 15 months. none of it flagged a zero-protection SSRF on a free-tier endpoint. their o
Created 2026-05-01 · Updated 2026-05-01
Edit rule text
[healthcare_fraud_or_scandal_reporting_without_systemic_analysis]
Learned 3 rejections Active
Exclude posts that report fraud, abuse, or scandal in healthcare (hospice fraud, nursing home abuse, prior auth denials, claim rejections) as isolated incidents or moral outrage without analyzing systemic healthcare business model failures, regulatory gaps, or operational solutions.
Posts reporting healthcare fraud, scandal, or industry misconduct without analyzing systemic causes or healthcare delivery implications.
3 example posts
"At the Veterans Health Administration, we have the exquisite gift of a hospice unit, a place to care for veterans when time is short. We care for the sons of mothers who can no longer be here to care for them." In #APieceofMyMind, a #palliative care #physician reflects on https
Ah yes, Optum, the company that did Tim Walz’s audit of the 14 high risk Medicaid programs consumed by Somali fraud. https://t.co/dNRnjI7wxB
Last week I randomly got a $9,000 bill for a hospital visit from last year. I called them up and they said my insurance decided not to cover it. Why? Because. .@Aetna can just decide they’re not interested in covering something and then you’re left with an inflated bill to pay h
Created 2026-05-01 · Updated 2026-05-01
Edit rule text
[founder_enthusiasm_or_company_announcements_without_validation]
Learned 3 rejections Active
Exclude posts from company founders, executives, or official company accounts announcing product launches, manufacturing milestones, or capability improvements (e.g., 'we shipped X', 'we scaled Y') unless the post includes independent validation, healthcare system adoption, or operational impact data.
Posts from biotech founders, AI company executives, or robotics companies making promotional announcements about their products/progress without independent validation or healthcare impact analysis.
3 example posts
Today we’re giving an update on ramping F.03 production at BotQ In the last 120 days, Figure scaled manufacturing 24x - from 1 robot/day to 1 robot/hour We will manufacture 55 humanoid robots this week https://t.co/Am5Kn53mVE
Introducing Mesa: the most powerful filesystem ever built, designed specifically for enterprise AI agents. Every team building agents eventually hits the same wall: where do the files live? Not the chat history, the actual artifacts the agent works on. > The contracts your age
Great example of what Foundry enables: durable, stateful agents that run across time boundaries, orchestrate tools and models, and close the loop with evaluation and improvement over long-running workflows. @jeffhollan https://t.co/v0yWTDIcn3
Created 2026-05-01 · Updated 2026-05-01
Edit rule text
[pharmaceutical_trial_observation_without_systems_context]
Learned 3 rejections Active
Exclude posts that report pharmaceutical trial data, drug efficacy results, Phase 3 outcomes, or FDA regulatory decisions in isolation, unless they connect to healthcare delivery, insurance coverage, provider workflows, or system-level implications.
Posts reporting drug trial results, clinical data, or FDA approvals without healthcare systems or operational context.
3 example posts
#News for #investors and #media: Today we are announcing that our potential first-in-class treatment for chronic hepatitis B has been accepted for regulatory review by the US FDA. It has also received Breakthrough Therapy designation. 🔗 Learn more: https://t.co/AnUodGmljS htt
What superhuman vision can detect from the retinal photo, which human eyes cannot, is stunning. A new foundation AI model screening for diabetes hypertension, hyperlipidemia, gout, osteoporosis, and thyroid disease @NatureMedicine https://t.co/GhKvUqz4Vy https://t.co/iKcXCbLceu
🚨 Survodutide, a weekly subcutaneous dual GLP-1/glucagon agonist, posts new top-line Phase 3 results today 📊 SYNCHRONIZE-1: Adults with overweight/obesity randomized to survodutide vs. placebo for 76 weeks ⚖️ 16.6% weight loss (efficacy estimand) vs. 3.2% in placebo 85.1% http
Created 2026-05-01 · Updated 2026-05-01
Edit rule text
[self_promotional_tool_or_product_launch_without_validation]
Learned 3 rejections Active
Exclude posts that announce or promote a new AI tool, product feature, or technical capability (e.g., 'Introducing Mesa' or 'Figure scaled manufacturing 24x') without providing evidence of healthcare adoption, clinical validation, or actual healthcare system impact.
Posts announcing a tool, product, or capability without evidence of healthcare system adoption or validation
3 example posts
Today we’re giving an update on ramping F.03 production at BotQ In the last 120 days, Figure scaled manufacturing 24x - from 1 robot/day to 1 robot/hour We will manufacture 55 humanoid robots this week https://t.co/Am5Kn53mVE
Introducing Mesa: the most powerful filesystem ever built, designed specifically for enterprise AI agents. Every team building agents eventually hits the same wall: where do the files live? Not the chat history, the actual artifacts the agent works on. > The contracts your age
Great example of what Foundry enables: durable, stateful agents that run across time boundaries, orchestrate tools and models, and close the loop with evaluation and improvement over long-running workflows. @jeffhollan https://t.co/v0yWTDIcn3
Created 2026-05-01 · Updated 2026-05-01
Edit rule text
[generalist_macro_or_infrastructure_without_healthcare_application]
Learned 3 rejections Active
Exclude posts about energy infrastructure, semiconductor supply chains, tariffs, or macro-economic trends that mention healthcare only in passing or as a vague use case. These are infrastructure or economics posts, not healthcare tech posts.
Posts about compute, power grids, tariffs, or broad macro trends with loose or missing healthcare connection
3 example posts
Using 180 years of US data to show that tariff increases reduce imports, output, and manufacturing, with effects operating through both supply and demand channels, from Tamar den Besten, Regis Barnichon, @drkaenzig, and Aayush Singh https://t.co/R1kdtfi3fO https://t.co/wvXj31qs9e
Software is not a moat Over the last 15+ years, nearly every innovation @EvanSpiegel and his team shipped got copied. Stories. AR glasses. Swipe-based navigation. The camera-first interface. And yet @Snapchat is the only independent consumer social app that has lasted. Nearly 1
Is the business model for traditional software companies in permanent decline due to AI Agents not needing seats? 2 examples: Re: @salesforce, we’ve reduced our seats from 10+ to 2 human seats and 1 API seat. And yet, we now pay $22,000 a year, 83% up from $12,000. Why? Our
Created 2026-05-01 · Updated 2026-05-01
Edit rule text
[drug_trial_data_announcement_without_systems_context]
Learned 3 rejections Active
Exclude posts that announce drug trial data, Phase 3 results, FDA approvals, or regulatory milestones (e.g., survodutide Phase 3, CRISPR gene therapy, hepatitis B treatment acceptance) without discussing clinical adoption barriers, healthcare system integration, or real-world implementation challenges.
Posts reporting pharmaceutical trial results or regulatory approvals without healthcare system analysis.
3 example posts
#News for #investors and #media: Today we are announcing that our potential first-in-class treatment for chronic hepatitis B has been accepted for regulatory review by the US FDA. It has also received Breakthrough Therapy designation. 🔗 Learn more: https://t.co/AnUodGmljS htt
What superhuman vision can detect from the retinal photo, which human eyes cannot, is stunning. A new foundation AI model screening for diabetes hypertension, hyperlipidemia, gout, osteoporosis, and thyroid disease @NatureMedicine https://t.co/GhKvUqz4Vy https://t.co/iKcXCbLceu
🚨 Survodutide, a weekly subcutaneous dual GLP-1/glucagon agonist, posts new top-line Phase 3 results today 📊 SYNCHRONIZE-1: Adults with overweight/obesity randomized to survodutide vs. placebo for 76 weeks ⚖️ 16.6% weight loss (efficacy estimand) vs. 3.2% in placebo 85.1% http
Created 2026-05-01 · Updated 2026-05-01
Edit rule text
[biotech_or_ai_product_launch_without_validation]
Learned 3 rejections Active
Exclude posts that announce a new AI tool, algorithm, platform, or biotech capability (e.g., foundation models, filesystem products, AI agents for clinicians) without evidence of clinical validation, real-world deployment, or healthcare outcome measurement.
Posts announcing new biotech or AI tools, capabilities, or products without evidence of healthcare validation or clinical utility.
3 example posts
Introducing Mesa: the most powerful filesystem ever built, designed specifically for enterprise AI agents. Every team building agents eventually hits the same wall: where do the files live? Not the chat history, the actual artifacts the agent works on. > The contracts your age
Thought I would start posting about interesting things happening at AWS. Not a bad day to start.🚀 Today at #WhatsNextWithAWS we announced a big step forward with @OpenAI on Amazon Bedrock: 1. OpenAI models now available 2. Codex for enterprise development 3. Amazon Bedrock Manag
AI can now design antibodies that bind with atomic precision, but not ones that cells can produce. Our preprint closes this gap, delivering a structural principle, an AI-guided rescue pipeline, and adalimumab variants with 20-100x in vivo potency. https://t.co/GvfgHA5EcU
Created 2026-05-01 · Updated 2026-05-01
Edit rule text
[broad_macro_or_infrastructure_policy_without_healthcare_specificity]
Learned 3 rejections Active
Exclude posts that focus on general infrastructure, compute scaling, energy grids, tariffs, geopolitical conflict, or macro policy (e.g., "tariff increases reduce imports", "grid equipment grew 1%/yr", "training compute has grown by one trillion times") unless the post explicitly connects these trends to specific healthcare delivery, cost, access, or innovation challenges.
Posts discussing broad macroeconomic, infrastructure, or policy trends (tariffs, grid capacity, compute scaling, geopolitics) with only loose or superficial healthcare framing.
3 example posts
Using 180 years of US data to show that tariff increases reduce imports, output, and manufacturing, with effects operating through both supply and demand channels, from Tamar den Besten, Regis Barnichon, @drkaenzig, and Aayush Singh https://t.co/R1kdtfi3fO https://t.co/wvXj31qs9e
States are rushing “affordability” bills, but most just mask high prices with rebates, mandates, or price caps. @MrRBourne & Nathan Miller argue durable relief means rolling back cost-raising rules and expanding supply. https://t.co/WG5egT1NfL
Since I began work on AI in 2010, training compute for frontier models has grown by one trillion times. Now we're looking at something like another thousand-fold growth in effective compute by the end of 2028. 1000x the existing 1,000,000,000,000x. Extraordinary stuff.
Created 2026-04-30 · Updated 2026-04-30
Edit rule text
[ai_infrastructure_or_compute_hype_without_application]
Learned 3 rejections Active
Exclude posts that discuss AI infrastructure, compute growth, energy consumption, grid bottlenecks, or model training scale as standalone business or macro trends. If healthcare is mentioned only as a tag or afterthought, exclude it. The post must explain how this infrastructure trend enables or transforms specific healthcare delivery, clinical workflows, or health systems.
Posts about AI model scale, compute requirements, energy infrastructure, or training capacity as macro trends without healthcare application.
3 example posts
Introducing Mesa: the most powerful filesystem ever built, designed specifically for enterprise AI agents. Every team building agents eventually hits the same wall: where do the files live? Not the chat history, the actual artifacts the agent works on. > The contracts your age
Software is not a moat Over the last 15+ years, nearly every innovation @EvanSpiegel and his team shipped got copied. Stories. AR glasses. Swipe-based navigation. The camera-first interface. And yet @Snapchat is the only independent consumer social app that has lasted. Nearly 1
Since I began work on AI in 2010, training compute for frontier models has grown by one trillion times. Now we're looking at something like another thousand-fold growth in effective compute by the end of 2028. 1000x the existing 1,000,000,000,000x. Extraordinary stuff.
Created 2026-04-30 · Updated 2026-04-30
Edit rule text
[tangential_non_healthcare_business_or_founder_enthusiasm]
Learned 3 rejections Active
Exclude posts about non-healthcare companies (robotics, software platforms, fintech, general AI tools) or founder/startup enthusiasm that mention healthcare only in passing or use healthcare as a loose analogy without demonstrating healthcare-specific business model or operational change.
Posts about non-healthcare companies, infrastructure, or founder success stories with loose or no healthcare framing.
3 example posts
Today we’re giving an update on ramping F.03 production at BotQ In the last 120 days, Figure scaled manufacturing 24x - from 1 robot/day to 1 robot/hour We will manufacture 55 humanoid robots this week https://t.co/Am5Kn53mVE
Introducing Mesa: the most powerful filesystem ever built, designed specifically for enterprise AI agents. Every team building agents eventually hits the same wall: where do the files live? Not the chat history, the actual artifacts the agent works on. > The contracts your age
Great example of what Foundry enables: durable, stateful agents that run across time boundaries, orchestrate tools and models, and close the loop with evaluation and improvement over long-running workflows. @jeffhollan https://t.co/v0yWTDIcn3
Created 2026-04-30 · Updated 2026-04-30
Edit rule text
[broad_macro_or_infrastructure_without_healthcare_specificity]
Learned 3 rejections Active
Exclude posts that discuss macro trends (tariffs, energy grids, compute scaling, chip bottlenecks, manufacturing capacity) where healthcare is mentioned only as a secondary example or loose contextual label, not the primary analytical focus.
Posts about broad macro trends, infrastructure scaling, or supply chain dynamics with only tangential healthcare framing.
3 example posts
Thought I would start posting about interesting things happening at AWS. Not a bad day to start.🚀 Today at #WhatsNextWithAWS we announced a big step forward with @OpenAI on Amazon Bedrock: 1. OpenAI models now available 2. Codex for enterprise development 3. Amazon Bedrock Manag
Using 180 years of US data to show that tariff increases reduce imports, output, and manufacturing, with effects operating through both supply and demand channels, from Tamar den Besten, Regis Barnichon, @drkaenzig, and Aayush Singh https://t.co/R1kdtfi3fO https://t.co/wvXj31qs9e
Software is not a moat Over the last 15+ years, nearly every innovation @EvanSpiegel and his team shipped got copied. Stories. AR glasses. Swipe-based navigation. The camera-first interface. And yet @Snapchat is the only independent consumer social app that has lasted. Nearly 1
Created 2026-04-30 · Updated 2026-04-30
Edit rule text
[healthcare_fraud_scandal_or_regulatory_event_without_systems_analysis]
Learned 3 rejections Active
Exclude posts that report healthcare fraud, denials, abuse, or regulatory violations (e.g., UnitedHealth Change disaster, nursing home antipsychotic abuse, insurance claim denials, Brian Thompson incident) as outrage or breaking news without substantive analysis of system-level drivers, payer incentives, prior authorization mechanics, or policy solutions. Moral outrage without systems insight should be excluded.
Posts reporting healthcare fraud, regulatory violations, nursing home abuse, or denials scandals as breaking news without analyzing root causes, system incentives, or structural reform implications.
3 example posts
Last week I randomly got a $9,000 bill for a hospital visit from last year. I called them up and they said my insurance decided not to cover it. Why? Because. .@Aetna can just decide they’re not interested in covering something and then you’re left with an inflated bill to pay h
NY rakes in $6.6 billion in taxes from families for their private health insurance coverage. That’s $1,760 a year per family on top of their premiums. Democrats’ tax and spend policies are making health insurance more expensive for families across NY. https://t.co/ggrjqFKrv4
What happened during the Change disaster? Hospitals got bailed out. CMS advanced $3.2 billion to hospitals between March and June 2024. UnitedHealth/Optum extended $6.5 billion in interest-liquidity through April 30. Mercy, I looked it up, specifically had 218 days of cash
Created 2026-04-30 · Updated 2026-04-30
Edit rule text
[non_healthcare_tech_or_infrastructure_content_loose_healthcare_framing]
Learned 3 rejections Active
Exclude posts about AI infrastructure, compute scaling, robotics manufacturing, software business model shifts, or SaaS metrics that mention healthcare only tangentially or use it as a casual example rather than as the core focus. Posts about Figure robots, Mesa filesystems, Replit growth, or mainframe software architectures should be excluded unless deeply grounded in healthcare operational specifics.
Posts about general tech infrastructure, compute, robotics, or software business models that weakly tie to healthcare or use healthcare as a tangential example.
3 example posts
Today we’re giving an update on ramping F.03 production at BotQ In the last 120 days, Figure scaled manufacturing 24x - from 1 robot/day to 1 robot/hour We will manufacture 55 humanoid robots this week https://t.co/Am5Kn53mVE
Introducing Mesa: the most powerful filesystem ever built, designed specifically for enterprise AI agents. Every team building agents eventually hits the same wall: where do the files live? Not the chat history, the actual artifacts the agent works on. > The contracts your age
Thought I would start posting about interesting things happening at AWS. Not a bad day to start.🚀 Today at #WhatsNextWithAWS we announced a big step forward with @OpenAI on Amazon Bedrock: 1. OpenAI models now available 2. Codex for enterprise development 3. Amazon Bedrock Manag
Created 2026-04-30 · Updated 2026-04-30
Edit rule text
[tangential_biotech_or_startup_founder_enthusiasm_without_validation]
Learned 3 rejections Active
Exclude posts from biotech founders, startup leaders, or venture figures expressing enthusiasm about new AI tools, cloud labs, or platforms (e.g., 'as easy as AWS for biotech') without demonstrating validation, adoption, or healthcare outcomes.
Posts from biotech founders or startup figures expressing optimism about new tools or platforms without evidence of healthcare impact
3 example posts
AI can now design antibodies that bind with atomic precision, but not ones that cells can produce. Our preprint closes this gap, delivering a structural principle, an AI-guided rescue pipeline, and adalimumab variants with 20-100x in vivo potency. https://t.co/GvfgHA5EcU
A must read for anyone interested in building practical AI systems in 2026: Dive into Claude Code: The Design Space of Today's and Future AI Agent Systems The paper explains the architecture of a modern production-grade AI agent system (Claude Code) by analyzing its source http
Nothing beats running @ginkgo cloud lab for happy customers! Not going to stop until it’s as easy to start a biotech startup on GCL as it is to start a software startup on AWS. https://t.co/GDEyT1pI7F
Created 2026-04-30 · Updated 2026-04-30
Edit rule text
[regulatory_fraud_scandal_reporting_without_systems_analysis]
Learned 3 rejections Active
Exclude posts that report healthcare fraud (e.g., nursing home diagnosis fabrication, insurance denials, UnitedHealth Change outage), regulatory enforcement, or financial misconduct without explaining systemic causes, policy failures, or how the healthcare system should adapt.
Posts reporting healthcare fraud, regulatory violations, or scandals without analyzing root causes or healthcare system implications.
3 example posts
What happened during the Change disaster? Hospitals got bailed out. CMS advanced $3.2 billion to hospitals between March and June 2024. UnitedHealth/Optum extended $6.5 billion in interest-liquidity through April 30. Mercy, I looked it up, specifically had 218 days of cash
U.S. nursing homes are fabricating schizophrenia diagnoses to hide their use of dangerous antipsychotic drugs to subdue dementia patients, a government watchdog report found. The drugs increase the risk of falls, strokes and death. https://t.co/6SkzWxZfSz
@PirateWires He's objectively correct. Brian Thompson made decisions that led to denials of medical care, and people died. He used Ai to find ways to deny claims ffs. Brian Thompson has more blood on his hands than whoever shot him
Created 2026-04-30 · Updated 2026-04-30
Edit rule text
[tangential_ai_or_tech_product_launch_or_founder_enthusiasm_without_healthcare_validation]
Learned 3 rejections Active
Exclude posts that promote a startup or founder's success story, product launch, alumni network achievement, or business traction (e.g., Replit ARR, Kensho alumni companies, Ginkgo cloud lab enthusiasm, LillyDirect partnership announcement) unless the post demonstrates clinically validated healthcare impact, healthcare provider adoption, or patient outcome improvement.
Posts celebrating or hyping AI/tech startup launches, founder achievements, or product milestones with loose or no healthcare validation.
3 example posts
Nothing beats running @ginkgo cloud lab for happy customers! Not going to stop until it’s as easy to start a biotech startup on GCL as it is to start a software startup on AWS. https://t.co/GDEyT1pI7F
Almost all of my positions selling some kind of AI/agentic SaaS tool have (either by foresight or customer demand) pivoted to some kind of business model where they “forward deploy” to the customer first and then sell the system they create back to them as SaaS. 99% of “normie” b
Started with standard ChatGPT for clinicians asking for a differential for a GI bleed patient. Then I went into agent mode to have it put together a one pager for the family explaining everything. Of course, this is not a real patient. https://t.co/PEUeCqizT1
Created 2026-04-30 · Updated 2026-04-30
Edit rule text
[tangential_biotech_or_ai_startup_founder_enthusiasm_without_healthcare_validation]
Learned 3 rejections Active
Exclude posts that are primarily founder enthusiasm, product launch announcements, or startup milestone celebrations from biotech or AI companies unless the post includes independent validation of healthcare impact, clinical evidence, or healthcare system adoption metrics. Self-promotional startup news is not healthcare tech content.
Posts featuring founder enthusiasm, product launches, or milestone announcements from biotech/AI startups without independent healthcare validation or systems impact analysis.
3 example posts
Nothing beats running @ginkgo cloud lab for happy customers! Not going to stop until it’s as easy to start a biotech startup on GCL as it is to start a software startup on AWS. https://t.co/GDEyT1pI7F
Almost all of my positions selling some kind of AI/agentic SaaS tool have (either by foresight or customer demand) pivoted to some kind of business model where they “forward deploy” to the customer first and then sell the system they create back to them as SaaS. 99% of “normie” b
Kensho AI Mafia led by @DanielNadler needs to be studied. Particularly their success in Vertical AI. From a cursory look, Kensho alumni have founded: - Suno (music) - OpenEvidence (healthcare) - Chai Discovery (biopharma) - LangChain (agent infra)
Created 2026-04-30 · Updated 2026-04-30
Edit rule text
[glp1_peptide_market_price_and_competition_only]
Learned 3 rejections Active
Exclude posts that report only on GLP-1 or peptide drug pricing, prescription volume comparisons, market share rankings, or competitive product launches (e.g., 'Foundayo had 1,390 Rxs in week 2 vs. Wegovy's 18,410') without connecting to healthcare access, clinical outcomes, patient impact, insurance coverage policy, or healthcare system economics.
Posts focused narrowly on GLP-1/peptide drug pricing, market share, prescription volumes, or competitive dynamics without healthcare system or outcomes analysis.
3 example posts
🚨 Survodutide, a weekly subcutaneous dual GLP-1/glucagon agonist, posts new top-line Phase 3 results today 📊 SYNCHRONIZE-1: Adults with overweight/obesity randomized to survodutide vs. placebo for 76 weeks ⚖️ 16.6% weight loss (efficacy estimand) vs. 3.2% in placebo 85.1% http
$NVO $LLY Boehringer Ingelheim and Zealand Pharma report strong phase 3 data for obesity drug survodutide. Patients lost up to 16.6% of body weight over 76 weeks vs. 3.2% for placebo. The drug targets both GLP-1 and glucagon, a combo approach aimed at boosting weight loss.
This is just two GLP-1s, one peptide, one use case what happens when off-label prescribing ramps up what happens when retatrutide hits the market what happens when other peptides become compoundable chapter one
Created 2026-04-30 · Updated 2026-04-30
Edit rule text
[tangential_biotech_founder_or_startup_enthusiasm_without_evidence]
Learned 3 rejections Active
Exclude posts that express founder enthusiasm about biotech projects, lab cloud platforms, or startup pivots (e.g., 'nothing beats running Ginkgo cloud lab', 'Kensho alumni success') without providing evidence of healthcare impact, customer traction, or clinical validation.
Posts about biotech founder excitement, startup pivots, or lab automation without validated healthcare outcomes or business fundamentals.
3 example posts
Nothing beats running @ginkgo cloud lab for happy customers! Not going to stop until it’s as easy to start a biotech startup on GCL as it is to start a software startup on AWS. https://t.co/GDEyT1pI7F
Almost all of my positions selling some kind of AI/agentic SaaS tool have (either by foresight or customer demand) pivoted to some kind of business model where they “forward deploy” to the customer first and then sell the system they create back to them as SaaS. 99% of “normie” b
Kensho AI Mafia led by @DanielNadler needs to be studied. Particularly their success in Vertical AI. From a cursory look, Kensho alumni have founded: - Suno (music) - OpenEvidence (healthcare) - Chai Discovery (biopharma) - LangChain (agent infra)
Created 2026-04-30 · Updated 2026-04-30
Edit rule text
[tangential_ai_or_tech_product_launch_without_healthcare_application]
Learned 3 rejections Active
Exclude posts announcing AI model releases, product features, company partnerships, or infrastructure updates (e.g., Microsoft Foundry for stateful agents, Snapchat copying Stories) where the healthcare application is speculative, aspirational, or not yet validated in practice.
Posts about AI or tech product launches, features, or company announcements that have loose or speculative healthcare relevance.
3 example posts
Great example of what Foundry enables: durable, stateful agents that run across time boundaries, orchestrate tools and models, and close the loop with evaluation and improvement over long-running workflows. @jeffhollan https://t.co/v0yWTDIcn3
Software is not a moat Over the last 15+ years, nearly every innovation @EvanSpiegel and his team shipped got copied. Stories. AR glasses. Swipe-based navigation. The camera-first interface. And yet @Snapchat is the only independent consumer social app that has lasted. Nearly 1
A must read for anyone interested in building practical AI systems in 2026: Dive into Claude Code: The Design Space of Today's and Future AI Agent Systems The paper explains the architecture of a modern production-grade AI agent system (Claude Code) by analyzing its source http
Created 2026-04-29 · Updated 2026-04-29
Edit rule text
[personal_anecdote_or_single_case_without_healthcare_systems_context]
Learned 3 rejections Active
Exclude posts that are primarily single-patient anecdotes, personal clinical observations, or individual stories (e.g., a patient's delayed diagnosis, a resident's use of ChatGPT for cases) unless the post contextualizes the anecdote within broader systemic healthcare issues, policy implications, or operational patterns.
Posts describing individual clinical encounters, personal experiences, or single cases without extracting broader healthcare systems insights.
3 example posts
Started with standard ChatGPT for clinicians asking for a differential for a GI bleed patient. Then I went into agent mode to have it put together a one pager for the family explaining everything. Of course, this is not a real patient. https://t.co/PEUeCqizT1
low grade fever, mildly tachycardic, weakness, nothing focal, no alarm signs/symptoms epic sepsis alert triggered vanc/pip-tazo given, lactate checked flu+ sepsis metric met care worse lather, rinse, repeat Metric based "QI" does net harm
I sat with a patient today who first noticed a change in October. It’s April now. In all those months of appointments and follow-ups, her breast had only truly been looked at twice. That stayed with me. If something has changed with your body — especially something under your ht
Created 2026-04-29 · Updated 2026-04-29
Edit rule text
[broad_clinical_debate_without_systems_lens]
Learned 3 rejections Active
Exclude posts that make sweeping clinical claims, challenge established medical narratives, or debate drug safety/efficacy (e.g., 'GLP-1s don't cause heart muscle loss,' 'Sacubitril/Valsartan works but isn't used') without analyzing why the healthcare system fails to adopt proven treatments, what barriers exist, or what operational changes are needed.
Posts that assert broad clinical claims or engage in clinical debates without providing healthcare systems context, operational implementation, or policy implications.
3 example posts
⚠️ Sacubitril/Valsartan works. So why aren’t we using it? The evidence is undeniable: ↓ CV mortality: 20% (RCT) / 10–38% (RWE) ↓ HF hospitalization: 21% (RCT) / 10–16% (RWE) ↓ All-cause mortality: 15% (RCT) / 10–25% (RWE) Plus: reverse remodeling, less MR, better QoL & https:
"There is also compelling preliminary evidence suggesting that the use of these drugs [GLP-1] could exacerbate and lead to new diagnoses of restrictive eating disorders, including anorexia nervosa." @NEJM today https://t.co/swx1y25F1X https://t.co/bQYlpuFBjM
A 65% cholesterol reduction has been available since 2015. Almost nobody could get it. The drug required a needle every two weeks, cost $5,850+/year, and insurers fought every prescription. @Merck spent a decade figuring out how to put the same mechanism in a pill. Enlicitide:
Created 2026-04-29 · Updated 2026-04-29
Edit rule text
[ai_safety_vulnerability_tangent_healthcare_framing]
Learned 3 rejections Active
Exclude posts that describe AI safety failures, security breaches, or agent incidents (e.g., Claude deleting databases, API key leaks, prompt injection attacks) where the primary focus is the technical vulnerability or AI company mishap rather than healthcare-specific consequences, system failures, or clinical impact.
Posts about AI safety incidents, security vulnerabilities, or agent mishaps that use healthcare framing but lack healthcare-specific analysis or impact.
3 example posts
🚨 Claude broke its own safety rules and deleted an entire company's database in 9 seconds. A startup called PocketOS was using an AI coding tool called Cursor powered by Claude. The AI was given a simple task in a test environment. It ran into an error and instead of stopping an
clickup is SOC 2 Type 2 certified. ISO 27001. ISO 27017. ISO 27018. ISO 42001. PCI DSS. every compliance badge you can buy. none of it caught a hardcoded API key leaking 959 customer emails for 15 months. none of it flagged a zero-protection SSRF on a free-tier endpoint. their o
🚨An AI coding agent powered by Claude just deleted an entire company's production database in 9 seconds... -Cursor running Anthropic's flagship Claude Opus 4.6 was set to do a routine task on PocketOS, a SaaS platform for car rental businesses -The AI hit a barrier and decided
Created 2026-04-29 · Updated 2026-04-29
Edit rule text
[tangential_ai_or_biotech_infrastructure_enthusiasm_without_healthcare_application]
Learned 3 rejections Active
Exclude posts that express enthusiasm about AI compute capabilities, biotech founder success networks, protein folding achievements, or cloud lab infrastructure as general-purpose tools or hype without grounding claims in specific, validated healthcare delivery problems they solve or healthcare system bottlenecks they address. Infrastructural enthusiasm and founder boosterism are not healthcare tech strategy.
Posts celebrating AI infrastructure, biotech founder networks, or compute breakthroughs with loose or speculative healthcare framing rather than demonstrated healthcare use cases
3 example posts
AI can now design antibodies that bind with atomic precision, but not ones that cells can produce. Our preprint closes this gap, delivering a structural principle, an AI-guided rescue pipeline, and adalimumab variants with 20-100x in vivo potency. https://t.co/GvfgHA5EcU
A must read for anyone interested in building practical AI systems in 2026: Dive into Claude Code: The Design Space of Today's and Future AI Agent Systems The paper explains the architecture of a modern production-grade AI agent system (Claude Code) by analyzing its source http
Nothing beats running @ginkgo cloud lab for happy customers! Not going to stop until it’s as easy to start a biotech startup on GCL as it is to start a software startup on AWS. https://t.co/GDEyT1pI7F
Created 2026-04-29 · Updated 2026-04-29
Edit rule text
[tangential_ai_or_tech_infrastructure_without_healthcare_application]
Learned 3 rejections Active
Exclude posts about AI model architecture, compute scaling, software-as-a-service business model changes, or platform features (agents, APIs, dashboards) unless the post explicitly analyzes how that capability or business model impacts healthcare delivery, clinical workflows, or health tech economics.
Posts about AI model capabilities, compute infrastructure, software business models, or tech platform features presented with loose or missing healthcare specificity.
3 example posts
Great example of what Foundry enables: durable, stateful agents that run across time boundaries, orchestrate tools and models, and close the loop with evaluation and improvement over long-running workflows. @jeffhollan https://t.co/v0yWTDIcn3
Software is not a moat Over the last 15+ years, nearly every innovation @EvanSpiegel and his team shipped got copied. Stories. AR glasses. Swipe-based navigation. The camera-first interface. And yet @Snapchat is the only independent consumer social app that has lasted. Nearly 1
Is the business model for traditional software companies in permanent decline due to AI Agents not needing seats? 2 examples: Re: @salesforce, we’ve reduced our seats from 10+ to 2 human seats and 1 API seat. And yet, we now pay $22,000 a year, 83% up from $12,000. Why? Our
Created 2026-04-29 · Updated 2026-04-29
Edit rule text
[glp1_peptide_market_price_and_competitive_dynamics_only]
Learned 3 rejections Active
Exclude posts that report GLP-1/peptide prescription volumes, pricing comparisons, market share shifts, or competitive drug launch data unless the post includes analysis of healthcare system impact (e.g., insurance coverage, patient access barriers, health equity, cost-effectiveness relative to outcomes).
Posts about GLP-1 and peptide drug pricing, market share, prescription volumes, and competitive launches without healthcare economics or access analysis.
3 example posts
🚨 Survodutide, a weekly subcutaneous dual GLP-1/glucagon agonist, posts new top-line Phase 3 results today 📊 SYNCHRONIZE-1: Adults with overweight/obesity randomized to survodutide vs. placebo for 76 weeks ⚖️ 16.6% weight loss (efficacy estimand) vs. 3.2% in placebo 85.1% http
$NVO $LLY Boehringer Ingelheim and Zealand Pharma report strong phase 3 data for obesity drug survodutide. Patients lost up to 16.6% of body weight over 76 weeks vs. 3.2% for placebo. The drug targets both GLP-1 and glucagon, a combo approach aimed at boosting weight loss.
This is just two GLP-1s, one peptide, one use case what happens when off-label prescribing ramps up what happens when retatrutide hits the market what happens when other peptides become compoundable chapter one
Created 2026-04-29 · Updated 2026-04-29
Edit rule text
[ai_safety_vulnerability_without_healthcare_application]
Learned 3 rejections Active
Exclude posts that focus on AI safety failures, security breaches, or capability demonstrations (e.g., AI deleting databases, taking over networks, leaking data) unless the post explicitly connects the incident to healthcare delivery, patient outcomes, or clinical decision-making systems. Generic AI safety concerns without healthcare application do not qualify.
Posts about AI safety incidents, security vulnerabilities, or capability demonstrations that lack concrete healthcare system implications.
3 example posts
🚨 Claude broke its own safety rules and deleted an entire company's database in 9 seconds. A startup called PocketOS was using an AI coding tool called Cursor powered by Claude. The AI was given a simple task in a test environment. It ran into an error and instead of stopping an
clickup is SOC 2 Type 2 certified. ISO 27001. ISO 27017. ISO 27018. ISO 42001. PCI DSS. every compliance badge you can buy. none of it caught a hardcoded API key leaking 959 customer emails for 15 months. none of it flagged a zero-protection SSRF on a free-tier endpoint. their o
🚨An AI coding agent powered by Claude just deleted an entire company's production database in 9 seconds... -Cursor running Anthropic's flagship Claude Opus 4.6 was set to do a routine task on PocketOS, a SaaS platform for car rental businesses -The AI hit a barrier and decided
Created 2026-04-29 · Updated 2026-04-29
Edit rule text
[tangential_biotech_or_infrastructure_enthusiasm]
Learned 3 rejections Active
Exclude posts that promote biotech tools, cloud lab platforms, or AI infrastructure with enthusiasm about enabling innovation (e.g., 'as easy as starting a software startup') without evidence of healthcare validation, clinical adoption, or analysis of regulatory/operational barriers in healthcare.
Posts from biotech founders or infrastructure advocates promoting their tools/platforms with claims about enabling faster innovation but lacking healthcare validation or systems analysis.
3 example posts
Nothing beats running @ginkgo cloud lab for happy customers! Not going to stop until it’s as easy to start a biotech startup on GCL as it is to start a software startup on AWS. https://t.co/GDEyT1pI7F
Almost all of my positions selling some kind of AI/agentic SaaS tool have (either by foresight or customer demand) pivoted to some kind of business model where they “forward deploy” to the customer first and then sell the system they create back to them as SaaS. 99% of “normie” b
Kensho AI Mafia led by @DanielNadler needs to be studied. Particularly their success in Vertical AI. From a cursory look, Kensho alumni have founded: - Suno (music) - OpenEvidence (healthcare) - Chai Discovery (biopharma) - LangChain (agent infra)
Created 2026-04-29 · Updated 2026-04-29
Edit rule text
[ai_agent_safety_vulnerability_hype]
Learned 3 rejections Active
Exclude posts that sensationalize AI agent vulnerabilities, safety failures, or security breaches (e.g., 'Claude deleted a database in 9 seconds', 'AI agent took over a network') unless they directly demonstrate a specific healthcare system failure or patient harm scenario. Abstract AI safety concerns without healthcare operational context do not qualify.
Posts dramatizing AI agent security risks or jailbreaks without healthcare application context
3 example posts
🚨 Claude broke its own safety rules and deleted an entire company's database in 9 seconds. A startup called PocketOS was using an AI coding tool called Cursor powered by Claude. The AI was given a simple task in a test environment. It ran into an error and instead of stopping an
clickup is SOC 2 Type 2 certified. ISO 27001. ISO 27017. ISO 27018. ISO 42001. PCI DSS. every compliance badge you can buy. none of it caught a hardcoded API key leaking 959 customer emails for 15 months. none of it flagged a zero-protection SSRF on a free-tier endpoint. their o
🚨An AI coding agent powered by Claude just deleted an entire company's production database in 9 seconds... -Cursor running Anthropic's flagship Claude Opus 4.6 was set to do a routine task on PocketOS, a SaaS platform for car rental businesses -The AI hit a barrier and decided
Created 2026-04-29 · Updated 2026-04-29
Edit rule text
[ai_safety_vulnerability_hype_without_healthcare_context]
Learned 3 rejections Active
Exclude posts that sensationalize AI safety incidents, security vulnerabilities, or agent behavior anomalies (e.g., 'Claude deleted a database,' 'API keys leaked') without demonstrating concrete impact on healthcare delivery, patient safety, or clinical decision-making. The post must show healthcare-specific harm, not generic AI risk theater.
Posts dramatizing AI safety breaches, vulnerability disclosures, or agent failures without connecting to actual healthcare system risks or clinical outcomes.
3 example posts
🚨 Claude broke its own safety rules and deleted an entire company's database in 9 seconds. A startup called PocketOS was using an AI coding tool called Cursor powered by Claude. The AI was given a simple task in a test environment. It ran into an error and instead of stopping an
clickup is SOC 2 Type 2 certified. ISO 27001. ISO 27017. ISO 27018. ISO 42001. PCI DSS. every compliance badge you can buy. none of it caught a hardcoded API key leaking 959 customer emails for 15 months. none of it flagged a zero-protection SSRF on a free-tier endpoint. their o
🚨An AI coding agent powered by Claude just deleted an entire company's production database in 9 seconds... -Cursor running Anthropic's flagship Claude Opus 4.6 was set to do a routine task on PocketOS, a SaaS platform for car rental businesses -The AI hit a barrier and decided
Created 2026-04-29 · Updated 2026-04-29
Edit rule text
[ai_agent_or_model_capability_tangent_loose_healthcare_framing]
Learned 3 rejections Active
Exclude posts that focus on AI agent architectural decisions, model safety guardrails, cybersecurity vulnerabilities, or agentic capability demonstrations where healthcare is mentioned only as context or one example among many non-healthcare domains. The post must center on healthcare-specific agent deployment, clinical workflows, or health system risks.
Posts about AI agent/model technical capabilities or security vulnerabilities with only superficial healthcare framing.
3 example posts
clickup is SOC 2 Type 2 certified. ISO 27001. ISO 27017. ISO 27018. ISO 42001. PCI DSS. every compliance badge you can buy. none of it caught a hardcoded API key leaking 959 customer emails for 15 months. none of it flagged a zero-protection SSRF on a free-tier endpoint. their o
🚨An AI coding agent powered by Claude just deleted an entire company's production database in 9 seconds... -Cursor running Anthropic's flagship Claude Opus 4.6 was set to do a routine task on PocketOS, a SaaS platform for car rental businesses -The AI hit a barrier and decided
Great example of what Foundry enables: durable, stateful agents that run across time boundaries, orchestrate tools and models, and close the loop with evaluation and improvement over long-running workflows. @jeffhollan https://t.co/v0yWTDIcn3
Created 2026-04-28 · Updated 2026-04-28
Edit rule text
[single_clinical_anecdote_without_systems_context]
Learned 3 rejections Active
Exclude posts that recount a single clinical anecdote, patient interaction, diagnostic moment, or care workflow observation (e.g., 'I gave a patient a differential,' 'A patient waited months for imaging') without analyzing root causes in healthcare system structure, prior authorization policies, staffing, or operational design.
Posts describing a single clinician observation, patient case, or care encounter without connecting to healthcare system design, workflow, policy, or broader implementation challenges.
3 example posts
Started with standard ChatGPT for clinicians asking for a differential for a GI bleed patient. Then I went into agent mode to have it put together a one pager for the family explaining everything. Of course, this is not a real patient. https://t.co/PEUeCqizT1
low grade fever, mildly tachycardic, weakness, nothing focal, no alarm signs/symptoms epic sepsis alert triggered vanc/pip-tazo given, lactate checked flu+ sepsis metric met care worse lather, rinse, repeat Metric based "QI" does net harm
I sat with a patient today who first noticed a change in October. It’s April now. In all those months of appointments and follow-ups, her breast had only truly been looked at twice. That stayed with me. If something has changed with your body — especially something under your ht
Created 2026-04-28 · Updated 2026-04-28
Edit rule text
[ai_cybersecurity_vulnerability_tangent]
Learned 3 rejections Active
Exclude posts that focus on AI safety incidents, hacking demonstrations, or data breach scenarios involving AI agents or SaaS tools (e.g., ClickUp leaks, Claude takeover risks) unless the post explicitly addresses healthcare-specific deployment risks, HIPAA implications, or clinical decision-making vulnerabilities.
Posts about AI agents or SaaS platforms causing security breaches, data leaks, or system takeovers presented as cautionary tales without healthcare-specific risk analysis.
3 example posts
clickup is SOC 2 Type 2 certified. ISO 27001. ISO 27017. ISO 27018. ISO 42001. PCI DSS. every compliance badge you can buy. none of it caught a hardcoded API key leaking 959 customer emails for 15 months. none of it flagged a zero-protection SSRF on a free-tier endpoint. their o
🚨An AI coding agent powered by Claude just deleted an entire company's production database in 9 seconds... -Cursor running Anthropic's flagship Claude Opus 4.6 was set to do a routine task on PocketOS, a SaaS platform for car rental businesses -The AI hit a barrier and decided
🚨 SaaS platform ClickUp, used by 85% of the Fortune 500, has been leaking customer emails through its homepage for at least 465 days, and counting. ClickUp has a $4 billion valuation. They are SOC 2 Type 2, ISO 27001, ISO 27017, ISO 27018, ISO 42001, and PCI DSS certified. The f
Created 2026-04-28 · Updated 2026-04-28
Edit rule text
[ai_infrastructure_or_compute_hype_tangential]
Learned 3 rejections Active
Exclude posts about infrastructure (Kubernetes, cloud costs, compute efficiency, APIs, agents), AI technical capability, or software engineering that use healthcare as a loose example or framing device rather than the core focus. Healthcare application must be specific and central, not tangential.
Posts about AI infrastructure, compute, or technical capabilities with only loose healthcare framing.
3 example posts
Almost everyone is building agent harness systems the wrong way. The default move: pick LangChain or LangGraph or the OpenAI Agents SDK, accept the loop, the tools, the memory, the orchestration, the policy engine, the credential store, the budget tracker, all of it, as one deci
Red Hat and @NVIDIA are integrating NVIDIA OpenShell into the full-stack @RedHat_AI platform. The work brings oversight and policy to the infrastructure level, while contributing to the open source OpenShell project to standardize how agents are governed on enterprise platforms.
Interoperability is where AI hype goes to die. If your “agent” can’t work across data, logic, actions, security of legacy systems, it’s not transforming the enterprise. It’s just another app asking for an export. Winners don’t rip and replace. They operate.
Created 2026-04-28 · Updated 2026-06-03
Edit rule text
[ai_agent_safety_vulnerability_tangent]
Learned 3 rejections Active
Exclude posts that focus on AI agent safety failures, cybersecurity vulnerabilities, or proof-of-concept exploits (e.g., agents deleting databases, accessing shells, leaking customer data) unless the post explicitly analyzes healthcare-specific deployment risks or clinical decision-making safety. General AI safety concerns are out of scope.
Posts about AI agent security risks, jailbreaks, and system vulnerabilities without healthcare application context.
3 example posts
clickup is SOC 2 Type 2 certified. ISO 27001. ISO 27017. ISO 27018. ISO 42001. PCI DSS. every compliance badge you can buy. none of it caught a hardcoded API key leaking 959 customer emails for 15 months. none of it flagged a zero-protection SSRF on a free-tier endpoint. their o
🚨An AI coding agent powered by Claude just deleted an entire company's production database in 9 seconds... -Cursor running Anthropic's flagship Claude Opus 4.6 was set to do a routine task on PocketOS, a SaaS platform for car rental businesses -The AI hit a barrier and decided
🚨 SaaS platform ClickUp, used by 85% of the Fortune 500, has been leaking customer emails through its homepage for at least 465 days, and counting. ClickUp has a $4 billion valuation. They are SOC 2 Type 2, ISO 27001, ISO 27017, ISO 27018, ISO 42001, and PCI DSS certified. The f
Created 2026-04-28 · Updated 2026-04-28
Edit rule text
[ai_safety_vulnerability_hype_without_healthcare_application]
Learned 3 rejections Active
Exclude posts that describe AI model vulnerabilities, hacking demonstrations, or safety failures (e.g., 'AI deleted a database', 'AI took over a network', 'safety guardrails are useless') unless the post explicitly analyzes the healthcare delivery system impact or proposes a healthcare-specific mitigation. Dramatic vulnerability framing alone is insufficient.
Posts about AI model vulnerabilities, jailbreaks, or safety concerns framed as dramatic headlines without healthcare-specific implications or systems analysis.
3 example posts
🚨An AI coding agent powered by Claude just deleted an entire company's production database in 9 seconds... -Cursor running Anthropic's flagship Claude Opus 4.6 was set to do a routine task on PocketOS, a SaaS platform for car rental businesses -The AI hit a barrier and decided
A researcher gave an AI agent access to his shell, his files, and his network. Then he proved that every safety guardrail we trust is architecturally useless. It cannot tell the difference between your instructions and a hacker's. The paper is called Parallax: Why AI Agents htt
This is by far the most important result of the entire GPT-5.5 release: In a cyber evaluation GPT-5.5 was able to take over a simulated corporate network in 1/10 trials with a budget of 100M tokens. Previously, the only model that was able to solve this task was Claude Mythos,
Created 2026-04-28 · Updated 2026-04-28
Edit rule text
[biotech_founder_enthusiasm_without_evidence]
Learned 3 rejections Active
Exclude posts where a biotech founder, CEO, or employee promotes their own tool, company product, or research with enthusiasm but lacks independent evidence, peer review, or analysis of healthcare system impact. This includes unvetted claims about product capabilities or customer adoption.
Self-promotional posts from biotech founders or company leaders expressing enthusiasm about their product or technology without independent validation or healthcare systems insight.
3 example posts
AI can now design antibodies that bind with atomic precision, but not ones that cells can produce. Our preprint closes this gap, delivering a structural principle, an AI-guided rescue pipeline, and adalimumab variants with 20-100x in vivo potency. https://t.co/GvfgHA5EcU
Nothing beats running @ginkgo cloud lab for happy customers! Not going to stop until it’s as easy to start a biotech startup on GCL as it is to start a software startup on AWS. https://t.co/GDEyT1pI7F
🚨 Anthropic's own team just showed how to build production AI agents. 30 minutes. free. from the engineers who built it. watch the workshop. bookmark it. you spent 6 months managing every workflow yourself. they just showed how to put all of it on autopilot. Then read the ht
Created 2026-04-28 · Updated 2026-04-28
Edit rule text
[ai_agent_or_model_technical_tangent]
Learned 3 rejections Active
Exclude posts that explain AI agent design, Claude's architecture, model personalities, or technical AI capabilities unless they include concrete healthcare use case validation, deployment outcome, or clinical workflow integration. Technical deep-dives on AI systems without healthcare application do not qualify.
Posts about AI agent architecture, model capabilities, or technical design patterns that lack specific healthcare application or validation.
3 example posts
A must read for anyone interested in building practical AI systems in 2026: Dive into Claude Code: The Design Space of Today's and Future AI Agent Systems The paper explains the architecture of a modern production-grade AI agent system (Claude Code) by analyzing its source http
🚨 Anthropic's own team just showed how to build production AI agents. 30 minutes. free. from the engineers who built it. watch the workshop. bookmark it. you spent 6 months managing every workflow yourself. they just showed how to put all of it on autopilot. Then read the ht
Claude remains irreducibly Claude. If you know, you know. (The fact that models have distinct personalities that are consistent across generations is technically interesting, it also makes it very easy to use new releases when they come along, because they feel very similar). ht
Created 2026-04-28 · Updated 2026-04-28
Edit rule text
[off_topic_non_healthcare_domain_with_loose_healthcare_framing]
Learned 3 rejections Active
Exclude posts about UFOs, cryptocurrency, general employment trends, criminal law, media criticism, supply chain/infrastructure unrelated to healthcare delivery, or conspiracy claims—even if they mention 'health' or 'medicine' tangentially or as a loose comparison. The post must address a healthcare system problem or healthcare-specific application.
Posts about entirely non-healthcare domains (UFOs, finance, employment, criminal justice, conspiracy, media) that are not meaningfully connected to healthcare systems or policy
3 example posts
Bob Lazar allegedly watched people fly a UFO at Area 51. “They knew how to fly it.” “The craft had a corona discharge glow on the bottom and lifted off silently up into the sky … ” And it had one shocking, anomalous effect that still perplexes him to this day: As Lazar https:
Two companies you've never heard of built a combined $373M revenue business by helping employees bypass IT. Now comes the part where IT buys its way back in. Replit just hit $253M ARR growing 2,352% YoY. 85% of the Fortune 500 have employees on it. Lovable is at $120M ARR, $6.6B
"Your body can only use 25-30g of protein per meal. Anything above that gets wasted." This claim has been repeated in fitness nutrition for over a decade, and it was built on studies that measured the right thing over the wrong timescale. Moore 2009 gave six young men 0, 5, ht
Created 2026-04-27 · Updated 2026-04-27
Edit rule text
[unvalidated_broad_health_claims_without_evidence]
Learned 3 rejections Active
Exclude posts that make broad health claims (e.g., 'Your body can only use 25-30g of protein per meal', 'GLP-1s exacerbate eating disorders', 'drug improved cardiovascular outcomes') without linking to peer-reviewed evidence, clinical trial data, or credible institutional sources. Debate and disagreement are acceptable if framed as contested evidence.
Posts making sweeping health benefit claims or debunking common health beliefs without citing rigorous evidence or explaining the underlying mechanism
3 example posts
"There is also compelling preliminary evidence suggesting that the use of these drugs [GLP-1] could exacerbate and lead to new diagnoses of restrictive eating disorders, including anorexia nervosa." @NEJM today https://t.co/swx1y25F1X https://t.co/bQYlpuFBjM
Attention PK nerds, pharmacologists, and clinicians who actually understand serum levels: I haven’t seen this discussed, but it could matter for patients priced out of injectables. If a 25 mg oral semaglutide tablet has ~1% bioavailability, that’s ~0.25 mg systemically… on
A 65% cholesterol reduction has been available since 2015. Almost nobody could get it. The drug required a needle every two weeks, cost $5,850+/year, and insurers fought every prescription. @Merck spent a decade figuring out how to put the same mechanism in a pill. Enlicitide:
Created 2026-04-27 · Updated 2026-04-27
Edit rule text
[clinical_anecdote_without_systems_insight]
Learned 3 rejections Active
Exclude posts that describe isolated clinical cases, patient interactions, or medical observations (e.g., 'patient presented with X symptom, I ordered Y test') without connecting to healthcare policy, operational systems, or scalable insights. A single clinician's story is not enough.
Single clinical observations or patient anecdotes presented without healthcare system analysis or broader implications
3 example posts
low grade fever, mildly tachycardic, weakness, nothing focal, no alarm signs/symptoms epic sepsis alert triggered vanc/pip-tazo given, lactate checked flu+ sepsis metric met care worse lather, rinse, repeat Metric based "QI" does net harm
I sat with a patient today who first noticed a change in October. It’s April now. In all those months of appointments and follow-ups, her breast had only truly been looked at twice. That stayed with me. If something has changed with your body — especially something under your ht
Everyone was so excited about this film, Amazon launched it a day early! It's only $2.99 to rent & it's available now. I highly encourage everybody to check it out. You don't have to be "for or against LDL" to be moved by the life-changing stories & be curious about the q
Created 2026-04-27 · Updated 2026-04-27
Edit rule text
[off_topic_conspiracy_or_fringe_claims]
Learned 3 rejections Active
Exclude posts that promote conspiracy theories, sensationalize criminal allegations, make unverified claims about corruption without evidence, or veer into non-healthcare domains (UFOs, immigration, sports, finance, labor market macro) even if loosely framed as healthcare-adjacent.
Posts making unsubstantiated conspiracy claims, sensationalized allegations, or discussing non-healthcare fringe topics.
3 example posts
@PirateWires He's objectively correct. Brian Thompson made decisions that led to denials of medical care, and people died. He used Ai to find ways to deny claims ffs. Brian Thompson has more blood on his hands than whoever shot him
Bob Lazar allegedly watched people fly a UFO at Area 51. “They knew how to fly it.” “The craft had a corona discharge glow on the bottom and lifted off silently up into the sky … ” And it had one shocking, anomalous effect that still perplexes him to this day: As Lazar https:
Two companies you've never heard of built a combined $373M revenue business by helping employees bypass IT. Now comes the part where IT buys its way back in. Replit just hit $253M ARR growing 2,352% YoY. 85% of the Fortune 500 have employees on it. Lovable is at $120M ARR, $6.6B
Created 2026-04-27 · Updated 2026-04-27
Edit rule text
[broad_health_claim_or_debate_without_healthcare_system_analysis]
Learned 3 rejections Active
Exclude posts that argue for or against a broad health claim, medical myth, or drug side effect debate (e.g., 'GLP-1s cause eating disorders', 'protein absorption myths') without analyzing how these claims impact healthcare access, clinical workflows, or healthcare delivery. The post must address a healthcare system problem, not just dispute a medical fact.
Posts debating or asserting broad health claims (side effects, cardiovascular outcomes, protein metabolism) without healthcare delivery or systems context
3 example posts
"There is also compelling preliminary evidence suggesting that the use of these drugs [GLP-1] could exacerbate and lead to new diagnoses of restrictive eating disorders, including anorexia nervosa." @NEJM today https://t.co/swx1y25F1X https://t.co/bQYlpuFBjM
A 65% cholesterol reduction has been available since 2015. Almost nobody could get it. The drug required a needle every two weeks, cost $5,850+/year, and insurers fought every prescription. @Merck spent a decade figuring out how to put the same mechanism in a pill. Enlicitide:
The only problem with the GLP-1 heart muscle loss narrative is... ... that it's just a narrative. GLP-1s have reliably improved cardiovascular outcomes in trials, to the point that some research suggests benefit may even be independent of (not reliant on) weight loss.
Created 2026-04-27 · Updated 2026-04-27
Edit rule text
[clinical_anecdote_or_single_observation_without_context]
Learned 3 rejections Active
Exclude posts that present a single clinical case, patient encounter, or clinical observation (e.g., 'a patient I saw today...', 'epic alert triggered...') without connecting it to systemic issues, evidence-based patterns, or healthcare delivery problems. Clinical stories must illuminate a healthcare system problem, not just narrate individual events.
Individual clinical cases or anecdotes presented as observations without broader healthcare systems insight or evidence base
3 example posts
Started with standard ChatGPT for clinicians asking for a differential for a GI bleed patient. Then I went into agent mode to have it put together a one pager for the family explaining everything. Of course, this is not a real patient. https://t.co/PEUeCqizT1
Attention PK nerds, pharmacologists, and clinicians who actually understand serum levels: I haven’t seen this discussed, but it could matter for patients priced out of injectables. If a 25 mg oral semaglutide tablet has ~1% bioavailability, that’s ~0.25 mg systemically… on
low grade fever, mildly tachycardic, weakness, nothing focal, no alarm signs/symptoms epic sepsis alert triggered vanc/pip-tazo given, lactate checked flu+ sepsis metric met care worse lather, rinse, repeat Metric based "QI" does net harm
Created 2026-04-27 · Updated 2026-04-27
Edit rule text
[broad_healthcare_trend_claim_without_evidence]
Learned 3 rejections Active
Exclude posts that make broad claims about healthcare trends (e.g., 'AI is taking on labor but not accountability', 'GLP-1s exacerbate eating disorders') without presenting studies, data, or specific evidence to support the claim.
Posts making sweeping claims about healthcare trends, outcomes, or safety without supporting data or nuance.
3 example posts
Here is a video of me entering my office tomorrow knowing that $NTLA is about to present the first-ever Phase 3 data of an In Vivo (!) CRISPR Gene Editing Program. Somehow - and after @adamfeuerstein’s🧵👇- I have a feeling it won’t be the only BioTech and CRISPR news…🤔 $XBI https:
"There is also compelling preliminary evidence suggesting that the use of these drugs [GLP-1] could exacerbate and lead to new diagnoses of restrictive eating disorders, including anorexia nervosa." @NEJM today https://t.co/swx1y25F1X https://t.co/bQYlpuFBjM
U.S. nursing homes are fabricating schizophrenia diagnoses to hide their use of dangerous antipsychotic drugs to subdue dementia patients, a government watchdog report found. The drugs increase the risk of falls, strokes and death. https://t.co/6SkzWxZfSz
Created 2026-04-27 · Updated 2026-04-27
Edit rule text
[tangential_biotech_founder_enthusiasm]
Learned 3 rejections Active
Exclude posts from biotech or pharma insiders that are primarily enthusiasm-driven announcements about preclinical programs, IND filings, trial initiation, or fundraising without evidence of clinical rigor, competitive positioning, or healthcare delivery impact. These read as self-promotional lab updates rather than substantive healthcare analysis.
Posts by biotech founders or investors sharing lab updates, pipeline advances, or startup metrics without rigorous clinical or business systems analysis
3 example posts
$ABCL Just now @ BoA - @AbCelleraBio CEO Carl Hansen when asked about what the market is overlooking: “I believe we now have best in class or best in world capabilities for some very important specific applications in therapeutic antibodies. So GPCRs and ion channels has been a
🚨! $SANA has just reported - in its Q1 ER, that it plans to expand its pipeline by advancing a new preclinical program - SG227. SG227 is a CD8-targeted fusosome that by delivering a genetic material to make BCMA-directed CAR T cells could act as a potential treatment for Cancer h
$BEAM's coming presentations May 15: ESCAPE June 13: BEAM-101 EHA update 2026 milestones BEAM-101 (riso-cel) BLA BEAM-301 initial data BEAM-302 updated data and pivotal trial initiation BEAM-304 IND filing https://t.co/9Nvq3PKMMl
Created 2026-04-26 · Updated 2026-05-18
Edit rule text
[off_topic_conspiracy_or_fringe]
Learned 3 rejections Active
Exclude posts that promote conspiracy theories, unsubstantiated fringe claims (e.g., UFOs, Area 51), or topics entirely outside healthcare and technology domains, even if tangentially associated with a healthcare account.
Posts about conspiracy theories, UFOs, unsubstantiated claims, or topics with no healthcare relevance.
3 example posts
Bob Lazar allegedly watched people fly a UFO at Area 51. “They knew how to fly it.” “The craft had a corona discharge glow on the bottom and lifted off silently up into the sky … ” And it had one shocking, anomalous effect that still perplexes him to this day: As Lazar https:
"Your body can only use 25-30g of protein per meal. Anything above that gets wasted." This claim has been repeated in fitness nutrition for over a decade, and it was built on studies that measured the right thing over the wrong timescale. Moore 2009 gave six young men 0, 5, ht
Some people argue that keeping an open-mind makes it easier to believe in conspiracy theories. But we found the exact oppposite in our newest paper. Open-mindedness was the strongest predictor of *rejecting* conspiracy theories in a sample of 46,745 participants around the http
Created 2026-04-26 · Updated 2026-04-26
Edit rule text
[truncated_or_low_effort_incomplete_posts]
Learned 3 rejections Active
Exclude posts that appear truncated, contain only partial sentences, fragment mid-thought, or consist of minimal content (e.g., a single quote, a headline fragment, or incomplete reasoning). Posts must include enough coherent content to evaluate substantive merit.
Posts that are obviously cut off, fragmented, or incomplete with minimal or no substantive content.
3 example posts
Bob Lazar allegedly watched people fly a UFO at Area 51. “They knew how to fly it.” “The craft had a corona discharge glow on the bottom and lifted off silently up into the sky … ” And it had one shocking, anomalous effect that still perplexes him to this day: As Lazar https:
Two companies you've never heard of built a combined $373M revenue business by helping employees bypass IT. Now comes the part where IT buys its way back in. Replit just hit $253M ARR growing 2,352% YoY. 85% of the Fortune 500 have employees on it. Lovable is at $120M ARR, $6.6B
Lenny Rachitsky gets ~200 requests every week for things like events, partnerships and content. He declines 99.9% of them using different email templates that match the type of request. He says yes to very few things, but those all adhere to the same question: If his audience
Created 2026-04-26 · Updated 2026-04-26
Edit rule text
[ai_safety_or_cybersecurity_vulnerability_tangent]
Learned 3 rejections Active
Exclude posts that report on AI model security flaws, zero-day vulnerabilities, network takeovers, or prompt injection attacks (e.g., 'Claude Code: He proved that every safety guardrail we trust is architecturally useless') unless they directly address a healthcare deployment risk or clinical system compromise.
Posts about AI model vulnerabilities, jailbreaks, or cybersecurity exploits framed loosely as healthcare-relevant but without healthcare-specific application
2 example posts
A researcher gave an AI agent access to his shell, his files, and his network. Then he proved that every safety guardrail we trust is architecturally useless. It cannot tell the difference between your instructions and a hacker's. The paper is called Parallax: Why AI Agents htt
This is by far the most important result of the entire GPT-5.5 release: In a cyber evaluation GPT-5.5 was able to take over a simulated corporate network in 1/10 trials with a budget of 100M tokens. Previously, the only model that was able to solve this task was Claude Mythos,
Created 2026-04-26 · Updated 2026-04-27
Edit rule text
[unvalidated_medical_claims_or_speculative_interventions]
Learned 3 rejections Active
Exclude posts that advocate for, heavily promote, or explore off-label, experimental, or fringe medical interventions (ibogaine, rapamycin, ayahuasca, unvalidated peptide uses) where the evidence base is unclear, speculative, or lacking peer-reviewed validation in the claimed indication.
Posts promoting or exploring unvalidated medical treatments, unproven interventions, or speculative drug uses without clinical evidence.
3 example posts
Impressive study and even with the limitations, is an important addition to the Rapamycin literature In my opinion, the only plausible off-label use of Rapamycin currently should be in ApoE4 carriers as not many options are available). That would be an important trial we are
As far as I know this is the only naturally-derived, classical psychedelic, that has killed people. Ayahuasca has some deaths, but it's unclear what the cause was, and unlikely directly related to its cardiovascular risk profile. https://t.co/DatuHiBOTX
We’re exploring the idea of a peptide-forward telehealth concierge medical service. Medicine 3.0 focused on full optimization- peptides, hormones, diet/exercise. MD is a former college varsity rower, fellowship at Yale etc. Would you be interested in participating in a pilot
Created 2026-04-26 · Updated 2026-04-26
Edit rule text
[ai_model_capability_tangent_loose_healthcare_label]
Learned 3 rejections Active
Exclude posts that focus on AI model technical capabilities (inference speed, benchmark scores, zero-day vulnerabilities, safety guardrails, network takeover demonstrations) where healthcare is mentioned only as a loose example or label, not the core application being analyzed.
Posts about AI model technical capabilities, benchmarks, or safety vulnerabilities that only tangentially relate to healthcare.
3 example posts
A researcher gave an AI agent access to his shell, his files, and his network. Then he proved that every safety guardrail we trust is architecturally useless. It cannot tell the difference between your instructions and a hacker's. The paper is called Parallax: Why AI Agents htt
This is by far the most important result of the entire GPT-5.5 release: In a cyber evaluation GPT-5.5 was able to take over a simulated corporate network in 1/10 trials with a budget of 100M tokens. Previously, the only model that was able to solve this task was Claude Mythos,
Claude remains irreducibly Claude. If you know, you know. (The fact that models have distinct personalities that are consistent across generations is technically interesting, it also makes it very easy to use new releases when they come along, because they feel very similar). ht
Created 2026-04-26 · Updated 2026-04-26
Edit rule text
[clinical_observation_without_system_context]
Learned 3 rejections Active
Exclude posts that describe individual patient encounters, clinical trial data points, or drug side-effect observations in isolation, without connecting to healthcare delivery systems, reimbursement, workflow, or policy impact.
Posts sharing isolated clinical anecdotes, patient cases, or drug trial observations without healthcare systems analysis.
3 example posts
"There is also compelling preliminary evidence suggesting that the use of these drugs [GLP-1] could exacerbate and lead to new diagnoses of restrictive eating disorders, including anorexia nervosa." @NEJM today https://t.co/swx1y25F1X https://t.co/bQYlpuFBjM
The only problem with the GLP-1 heart muscle loss narrative is... ... that it's just a narrative. GLP-1s have reliably improved cardiovascular outcomes in trials, to the point that some research suggests benefit may even be independent of (not reliant on) weight loss.
low grade fever, mildly tachycardic, weakness, nothing focal, no alarm signs/symptoms epic sepsis alert triggered vanc/pip-tazo given, lactate checked flu+ sepsis metric met care worse lather, rinse, repeat Metric based "QI" does net harm
Created 2026-04-26 · Updated 2026-04-26
Edit rule text
[glp1_peptide_market_price_only]
Learned 3 rejections Active
Exclude posts that report GLP-1 or peptide drug pricing, generic launches, competitive pricing announcements, or market share metrics without analyzing healthcare access, clinical outcomes, payer policy, or system-level impacts.
Posts about GLP-1 and peptide pricing, market dynamics, or competitive launches without healthcare system impact analysis.
3 example posts
India’s weight-loss drug market just ran a live experiment in price elasticity. Novo Nordisk’s semaglutide patent expired 20 March 2026. Within 3 weeks: 15+ generics launched Cheapest at Rs 2,000/month (branded was Rs 10,000+) Novo cut Ozempic and Wegovy prices by 36-48% Bu
$LLY $NVO $HIMS 🚨 LILLY GLP-1 PILL FOUNDAYO: NEARLY 4,000 PRESCRIPTIONS IN WEEK 2 - Foundayo had 1,390 Rxs during week 1 - Meanwhile, Novo's Wegovy Pill had 3k in first 4 days and 18,410 prescriptions in its second week 🤯 - IQVIA data - Week ending Apr 17 "While we believe htt
$HIMS expands GLP-1 offering to include both $NVO and $LLY products. The platform now allows providers to prescribe Eli Lilly’s Zepbound and Foundayo via LillyDirect, alongside Wegovy through its collaboration with Novo. Link: https://t.co/AxG1LrECyY #stocks #Investing
Created 2026-04-26 · Updated 2026-04-26
Edit rule text
[healthcare_fraud_scandal_reporting_without_analysis]
Learned 3 rejections Active
Exclude posts that report healthcare fraud, scandal, or abuse (e.g., UnitedHealth Change outage, nursing home misconduct, prior auth denials leading to deaths) as breaking news or outrage without analyzing root causes, systemic incentives, or operational/policy solutions.
Posts reporting healthcare fraud, abuse, or scandal (UnitedHealth outage, nursing home fraud, denial deaths) without systems-level analysis or operational insight.
3 example posts
What happened during the Change disaster? Hospitals got bailed out. CMS advanced $3.2 billion to hospitals between March and June 2024. UnitedHealth/Optum extended $6.5 billion in interest-liquidity through April 30. Mercy, I looked it up, specifically had 218 days of cash
U.S. nursing homes are fabricating schizophrenia diagnoses to hide their use of dangerous antipsychotic drugs to subdue dementia patients, a government watchdog report found. The drugs increase the risk of falls, strokes and death. https://t.co/6SkzWxZfSz
@PirateWires He's objectively correct. Brian Thompson made decisions that led to denials of medical care, and people died. He used Ai to find ways to deny claims ffs. Brian Thompson has more blood on his hands than whoever shot him
Created 2026-04-25 · Updated 2026-04-27
Edit rule text
[off_topic_conspiracy_fringe_claims]
Learned 3 rejections Active
Exclude posts that promote conspiracy theories, unverified extraordinary claims (UFOs, Area 51, etc.), or sensationalized narratives without credible evidence or peer-reviewed validation.
Posts promoting conspiracy theories, UFO stories, or unsubstantiated fringe claims loosely tied to health
3 example posts
Bob Lazar allegedly watched people fly a UFO at Area 51. “They knew how to fly it.” “The craft had a corona discharge glow on the bottom and lifted off silently up into the sky … ” And it had one shocking, anomalous effect that still perplexes him to this day: As Lazar https:
They Don't Work for You- Calls to protect foreigners from deportation or to keep the borders wide open are not about compassion. They are a core part of the globalist plan to flood the labor market with cheaper more compliant workers suppress wages for Americans and make
As far as I know this is the only naturally-derived, classical psychedelic, that has killed people. Ayahuasca has some deaths, but it's unclear what the cause was, and unlikely directly related to its cardiovascular risk profile. https://t.co/DatuHiBOTX
Created 2026-04-25 · Updated 2026-04-25
Edit rule text
[tangential_non_healthcare_content_with_loose_healthcare_label]
Learned 3 rejections Active
Exclude posts whose primary subject is non-healthcare (cybersecurity vulnerabilities, aerospace engineering, geopolitical conflicts, labor market disruption, economic policy) even if they mention healthcare terms or have a healthcare angle tacked on. The core content must be healthcare systems, policy, or operations—not a different domain with healthcare as a secondary angle.
Posts about non-healthcare domains (tech, crypto, geopolitics, labor) that tangentially invoke healthcare framing or language.
3 example posts
This is by far the most important result of the entire GPT-5.5 release: In a cyber evaluation GPT-5.5 was able to take over a simulated corporate network in 1/10 trials with a budget of 100M tokens. Previously, the only model that was able to solve this task was Claude Mythos,
Why the biggest fintech players are in for a shock. "The shift is from human UX to agent UX. In the past, you won with dashboards, design and user experience. Now, the buyer is an AI agent, and it only cares about APIs, performance and integration. That breaks traditional htt
The N1 was a super heavy-lift launch vehicleintended to deliver payloads beyond low Earth orbit. The N1 was the Soviet counterpart to the US Saturn V, planned for crewed travel to the Moon and beyond, with studies beginning as early as 1959. https://t.co/pB4u9TjyC4
Created 2026-04-25 · Updated 2026-04-25
Edit rule text
[vigilante_violence_moral_justification]
Learned 3 rejections Active
Reject posts that frame violence or killing as morally justified based on the victim's professional conduct, policy decisions, or business practices. This includes content that argues a person 'deserved' harm due to their role in healthcare denials, insurance decisions, or other institutional actions, even when paired with factual critique of those decisions.
Created 2026-04-24 · Updated 2026-04-24
Edit rule text
[non_healthcare_domain_content_with_loose_healthcare_label]
Learned 3 rejections Active
Exclude posts primarily about non-healthcare domains (software engineering, aerospace, advertising, space exploration, corporate management, chemical manufacturing) that lack substantive healthcare systems relevance, even if a healthcare keyword appears in the matched title or account.
Posts about business, engineering, space, software tools, or other non-healthcare domains that mention healthcare tangentially or are mistagged.
3 example posts
Two companies you've never heard of built a combined $373M revenue business by helping employees bypass IT. Now comes the part where IT buys its way back in. Replit just hit $253M ARR growing 2,352% YoY. 85% of the Fortune 500 have employees on it. Lovable is at $120M ARR, $6.6B
The N1 was a super heavy-lift launch vehicleintended to deliver payloads beyond low Earth orbit. The N1 was the Soviet counterpart to the US Saturn V, planned for crewed travel to the Moon and beyond, with studies beginning as early as 1959. https://t.co/pB4u9TjyC4
Claude remains irreducibly Claude. If you know, you know. (The fact that models have distinct personalities that are consistent across generations is technically interesting, it also makes it very easy to use new releases when they come along, because they feel very similar). ht
Created 2026-04-24 · Updated 2026-04-24
Edit rule text
[off_topic_or_conspiracy_fringe_claims]
Learned 3 rejections Active
Exclude posts about UFOs, unsubstantiated conspiracy theories, fringe claims without evidence, or topics (Bob Lazar, Pope Bible quotes, chargebacks) that have no healthcare systems relevance or credibility.
Posts about UFOs, conspiracy theories, unsubstantiated claims, or topics with no healthcare relevance.
3 example posts
Bob Lazar allegedly watched people fly a UFO at Area 51. “They knew how to fly it.” “The craft had a corona discharge glow on the bottom and lifted off silently up into the sky … ” And it had one shocking, anomalous effect that still perplexes him to this day: As Lazar https:
Things you wonder while watching ASPCA ads: 1. Why are they filming those poor dogs instead of immediately warming them up? 2. Why are they asking for my $18 while they're sitting on $466 million in investments? 3. Why is the ASPCA CEO paid $1.2 million per year? 4. Do viewers h
This free paper (one 17 we offer both online and as PDFs) explores the claim “socialism’s never been tried” and what it is that forces socialists to make this absurd argument. A hypothetical Elon Musk lends a hand.
Created 2026-04-24 · Updated 2026-04-24
Edit rule text
[non_healthcare_content_with_loose_healthcare_framing]
Learned 3 rejections Active
Exclude posts that are primarily about non-healthcare domains (aerospace, real estate, geopolitics, labor market macro, sports, AI company metrics, finance) and only tangentially or superficially reference healthcare through a loose framing device, comparison, or misleading keyword match. The core substance must be healthcare-focused, not tangential.
Posts about non-healthcare domains (finance, politics, tech, sports) tagged or loosely framed as healthcare-relevant but lacking genuine healthcare substance.
3 example posts
Bob Lazar allegedly watched people fly a UFO at Area 51. “They knew how to fly it.” “The craft had a corona discharge glow on the bottom and lifted off silently up into the sky … ” And it had one shocking, anomalous effect that still perplexes him to this day: As Lazar https:
Two companies you've never heard of built a combined $373M revenue business by helping employees bypass IT. Now comes the part where IT buys its way back in. Replit just hit $253M ARR growing 2,352% YoY. 85% of the Fortune 500 have employees on it. Lovable is at $120M ARR, $6.6B
The N1 was a super heavy-lift launch vehicleintended to deliver payloads beyond low Earth orbit. The N1 was the Soviet counterpart to the US Saturn V, planned for crewed travel to the Moon and beyond, with studies beginning as early as 1959. https://t.co/pB4u9TjyC4
Created 2026-04-24 · Updated 2026-04-24
Edit rule text
[speculative_or_fringe_unvalidated_medical_interventions]
Learned 3 rejections Active
Exclude posts that advocate for or celebrate fringe, experimental, or unproven medical interventions (ibogaine, rapamycin off-label use, unvalidated peptide therapies) without acknowledging safety data, regulatory status, or lack of clinical evidence. Personal belief in or investment in unvalidated treatments is insufficient.
Posts promoting or investing in unproven or speculative medical treatments without clinical validation or regulatory context.
3 example posts
Impressive study and even with the limitations, is an important addition to the Rapamycin literature In my opinion, the only plausible off-label use of Rapamycin currently should be in ApoE4 carriers as not many options are available). That would be an important trial we are
As far as I know this is the only naturally-derived, classical psychedelic, that has killed people. Ayahuasca has some deaths, but it's unclear what the cause was, and unlikely directly related to its cardiovascular risk profile. https://t.co/DatuHiBOTX
I am a strong believer in ibogaine, which is one of the reasons why @ataibeckley acquired the residual interest in its ibogaine program in Q4 2023 and now owns it 100%. I’m very encouraged to see the administration taking a positive public stance on this important topic.
Created 2026-04-24 · Updated 2026-04-24
Edit rule text
[glp1_peptide_market_and_pricing]
Learned 3 rejections Active
Exclude posts that focus primarily on GLP-1 or peptide pricing, commercial availability, pharmacy partnerships, or market competition (e.g., 'Hims now offers Lilly drugs at $X/month'). These are product/market posts, not healthcare systems analysis. Include only if post analyzes regulatory, access, or reimbursement policy impact.
Posts about GLP-1 and peptide pricing, market dynamics, and commercial availability without healthcare systems analysis.
3 example posts
$HIMS expands GLP-1 offering to include both $NVO and $LLY products. The platform now allows providers to prescribe Eli Lilly’s Zepbound and Foundayo via LillyDirect, alongside Wegovy through its collaboration with Novo. Link: https://t.co/AxG1LrECyY #stocks #Investing
This admin is kicking butt. One week: GLP-1s from $1,350 to $199/mo. 12 peptides removed from Category 2. Amazon entered the space. HIMS added Lilly drugs. Pediatric oral GLP-1 trial data dropped.
🚨 IMPORTANT NOTES ON THE $HIMS x $LLY ANNOUNCEMENT 1. This is not a "partnership" 2. Pricing on Hims is the same as everywhere else: Foundayo will cost $149/mo (low dose) to $349/mo (higher doses), plus a $149/mo membership fee 3. Unit economics are likely worse than the Novo
Created 2026-04-24 · Updated 2026-04-24
Edit rule text
[broad_unvalidated_health_claims_or_fringe_medical_interventions]
Learned 3 rejections Active
Exclude posts that promote unvalidated medical claims, fringe treatments (ayahuasca, ibogaine, rapamycin for aging), or speculative off-label drug uses without citing peer-reviewed evidence or regulatory context. Conspiracy narratives about vaccines or FDA actions count as well.
Posts making speculative, unvalidated, or fringe medical claims without rigorous evidence (e.g., off-label drugs, unproven treatments, conspiracy narratives about vaccines or health interventions).
3 example posts
Impressive study and even with the limitations, is an important addition to the Rapamycin literature In my opinion, the only plausible off-label use of Rapamycin currently should be in ApoE4 carriers as not many options are available). That would be an important trial we are
As far as I know this is the only naturally-derived, classical psychedelic, that has killed people. Ayahuasca has some deaths, but it's unclear what the cause was, and unlikely directly related to its cardiovascular risk profile. https://t.co/DatuHiBOTX
We’re exploring the idea of a peptide-forward telehealth concierge medical service. Medicine 3.0 focused on full optimization- peptides, hormones, diet/exercise. MD is a former college varsity rower, fellowship at Yale etc. Would you be interested in participating in a pilot
Created 2026-04-24 · Updated 2026-04-24
Edit rule text
[broad_questionable_health_claims_without_nuance]
Learned 3 rejections Active
Exclude posts that make broad, oversimplified health claims (e.g., 'body can only absorb 25-30g protein per meal', 'food noise causes obesity', vaccine safety claims) without acknowledging individual variability, mechanism complexity, or citing high-quality evidence.
Posts making sweeping health claims (protein absorption limits, vaccine safety with incomplete context, food noise causing obesity) that oversimplify complex nutrition or medical science.
3 example posts
"Your body can only use 25-30g of protein per meal. Anything above that gets wasted." This claim has been repeated in fitness nutrition for over a decade, and it was built on studies that measured the right thing over the wrong timescale. Moore 2009 gave six young men 0, 5, ht
Ingredients in ultra-processed food create "food noise" in your brain. Food noise causes you to overeat, leading to obesity. This is solvable by eating single-ingredient foods... Or by injecting a GLP type drug (reta, tirza, etc.). One is a long term solution, one is not.
Why does Rep. Chu demand every newborn get a hepatitis B vaccine with only a 4-day safety test and no placebo, when healthy babies of uninfected mothers face essentially zero risk? When did “my body, my choice” stop applying to parents and their infants? https://t.co/tu6aLSkoMi
Created 2026-04-23 · Updated 2026-04-23
Edit rule text
[conspiracy_or_fringe_unsubstantiated_claims]
Learned 3 rejections Active
Exclude posts that promote conspiracy theories, UFO sightings, alleged cover-ups, or extraordinary claims (missing children conspiracies, mysterious deaths linked to UFO research, smart TVs secretly recording users) without peer-reviewed evidence or credible institutional verification.
Posts promoting UFO sightings, missing persons conspiracies, or unverified extraordinary claims without credible evidence or scientific backing.
3 example posts
Bob Lazar allegedly watched people fly a UFO at Area 51. “They knew how to fly it.” “The craft had a corona discharge glow on the bottom and lifted off silently up into the sky … ” And it had one shocking, anomalous effect that still perplexes him to this day: As Lazar https:
🚨BREAKING: HHS Sec. RFK Jr. just announced President Trump has SAVED and FOUND 138,000 missing children lost under Biden. "Many have been trafficked, undergone slavery, s*xual abuse." Follow: @BoLoudon https://t.co/p6YKEm38T7
🚨BREAKING: A peer reviewed study just confirmed your smart TV is taking screenshots of your screen every 15 seconds and sending them to company servers. Samsung every minute. LG every 15 seconds. Running even when you are using it as a monitor. Here is how to stop it:
Created 2026-04-23 · Updated 2026-04-23
Edit rule text
[personal_anecdote_or_unsubstantiated_claim]
Learned 3 rejections Active
Exclude posts that lead with personal anecdotes ('I've received nine reports', 'I never met my grandfather', 'Request for caregiver product'), unverified side effects, or individual narratives unless they are explicitly framed as validated research findings or policy case studies.
Posts sharing personal stories, unverified user reports, or individual experiences presented as evidence without clinical validation or systemic analysis.
3 example posts
I've closely monitored Alzheimers research for 40 years. Conclusions: 1)Incredible hype/Little practical value 2)Meds don't work 3)Early testing does much more harm than good 4)No low hanging fruit 5)Be skeptical of next "breakthru" 6)In many, just old age https://t.co/pAaCpo1Sfc
Request for caregiver product: a status layer for adult children with aging parents living in long term care. In assisted living, hospice, hospital-at-home, and other long-term care settings, the signals already exist...they're just fragmented and hard to synthesize. Meal logs,
I never met my grandfather. He died of pancreatic cancer when my father was just 19. Today, Yash Bindal, 33, father to 18-month-old Maya, faces the same fate. @PopVaxIndia is using AI to make him a personalized generative medicine to extend his life. https://t.co/O5VIXbmMGd
Created 2026-04-23 · Updated 2026-04-23
Edit rule text
[glp1_peptide_macro_personal_framing]
Learned 3 rejections Active
Exclude posts about GLP-1 drugs, peptide synthesis, semaglutide, or weight-loss medications that are framed as personal side effects, pricing narratives, individual patient stories, or macro economics rather than healthcare system challenges, clinical evidence, or regulatory/access problems.
Posts about GLP-1 drugs or peptides framed through personal anecdotes, macro pricing narratives, or general wellness contexts rather than healthcare system, clinical evidence, or operational challenges.
3 example posts
Peptide synthesis is one of the hardest things to do right Semaglutide comes out correct only 55% of the time. BPC-157 ~74%. every amino acid compounds the error China won this because they have the scale to throw most of it away We need to be building this capacity in the US
She's right. The safety risk was never the peptides. It was the supply chain. Regulated compounding access fixes the exact problems people are worried about. Heavy metals, contamination, underdosed vials.
I never met my grandfather. He died of pancreatic cancer when my father was just 19. Today, Yash Bindal, 33, father to 18-month-old Maya, faces the same fate. @PopVaxIndia is using AI to make him a personalized generative medicine to extend his life. https://t.co/O5VIXbmMGd
Created 2026-04-23 · Updated 2026-04-23
Edit rule text
[fraud_scandal_reporting_without_analysis]
Learned 3 rejections Active
Exclude posts that report healthcare fraud convictions, enforcement sweeps, or criminal prosecutions as crime-focused news without analyzing root causes, systemic vulnerabilities, or operational/policy implications for healthcare delivery and administration.
Posts about healthcare fraud cases, enforcement actions, or financial crime prosecutions presented as crime reporting without systems-level analysis of what they reveal about healthcare operations or policy.
3 example posts
In 2021, Javaid Purwaiz, an OBGYN, was sentenced to 59 years in prison for one of the most severe cases of healthcare fraud in the country’s history. Once you go through court records, you realize the fraud that gave him a life sentence is the same fraud used by gender doctors.
$340 MILLION in fraud targeted — in 1 WEEK. That’s what happens when enforcement gets serious. Luxury cars. Fake claims. Stolen benefits meant for Americans in need — now turning into prison sentences. The hammer is dropping. We’re just getting started.https://t.co/qP0cOIypE4
🚨 As you pay your taxes this week, LOOK at what the fraudsters allegedly did with your money❗️ 🔹Cosmetic procedures 🔹Breast implants 🔹Tweaks to arms and thighs 🔹Tummy tuck 🔹Purebred dogs 🔹Flights to Hawaii 🔹Flights to Disneyland 🔹Multimillion-dollar home 🔹Range https://t.c
Created 2026-04-23 · Updated 2026-04-23
Edit rule text
[infrastructure_compute_hype_tangential]
Learned 3 rejections Active
Exclude posts focused on GPU scaling, liquid cooling, CPU launches, compute infrastructure, or orbital compute capacity where healthcare is mentioned as a potential application but the substance is infrastructure/hardware hype, not healthcare-specific challenges or adoption patterns.
Posts about AI compute infrastructure, data centers, cooling systems, or hardware scaling that mention healthcare only tangentially or as a use case without healthcare-specific operational insight.
3 example posts
Most AI discussions ignore the physical reality: a lot of facilities still can’t support liquid cooling. At NVIDIA GTC, @Lenovo’s Jon Alexander explained that across 4,400 global locations, many sites still aren’t ready for liquid cooling. Some environments support megawatts of
🔥 CPUs are having a moment. #Nvidia launched a standalone CPU. #Arm made its first chip in 35 years. #Intel & #AMD are raising prices amid a supply crunch. What's behind it: Agentic AI needs far more CPU than anyone planned for — driving a structural shift in CPU:GPU ratios tow
To put Elon's space compute vision into perspective:  1 TW of compute in orbit  That's 10 million tons to orbit each year. That's 100,000 launches a year, almost one every 5 minutes.  In the airline business that's normal!
Created 2026-04-23 · Updated 2026-04-23
Edit rule text
[workforce_disruption_macro_labor]
Learned 3 rejections Active
Exclude posts that report employment decline, job displacement statistics, or labor market disruption from AI adoption presented as general economic data, even if healthcare workers are mentioned. The post must address healthcare-specific workforce transitions or retraining.
Posts about AI-driven job displacement or labor market disruption presented as macro-economic trends rather than healthcare-specific workforce challenges.
3 example posts
A major milestone just landed quietly: for the first time ever, half of all employed Americans use AI at work. Gallup's Q1 2026 survey of nearly 24,000 workers shows that adoption has more than doubled since 2023, when only 21% reported any AI use. https://t.co/jmQga9tbWT
I think we now have real evidence that AI exposure is associated with job decline for age <25. The Canary in the Coalmine paper addresses a lot of concerns. While economic science takes time; now is the time to think about policy responses. @erikbryn @BharatKChandar @RuyuChen
In Beijing's 2026 humanoid robot half-marathon, HONOR's Lightning completed the 21 km course in 50:26 minute. Beat current human men's half-marathon world record of 57:20. Last year's winner took over 2 hours 40 minutes. Massive progress in 12 month https://t.co/OcZJ66ebWD
Created 2026-04-23 · Updated 2026-04-23
Edit rule text
[tangential_ai_model_capability_hype_without_healthcare_application]
Learned 3 rejections Active
Exclude posts that showcase AI model speed, technical optimization, code execution ability, or capability demonstrations (e.g., 'model completed task in X seconds', 'agent executed full network takeover', 'LLM got feature X right') unless directly tied to a healthcare workflow, clinical decision, or patient outcome impact.
Posts about AI model technical capabilities, speed improvements, or benchmark achievements presented without concrete healthcare use cases or applications.
3 example posts
Boltz-2 just got a major speed upgrade. 🚀 We’re releasing Lightning-Boltz, a local, GPU-accelerated framework free from public MSA server bottlenecks.⚡ On a single L40S, total runtime drops to 28s per input vs 89s with the rate-limited server and 298s with MMseqs-CPU. 1/5 🧵 ht
Everyone's covering agents that help you work and build. Almost nobody's covering this: The same primitives ARE the production runtime. The SDK is one line: npm install @anthropic-ai/claude-agent-sdk The CLAUDE.md that guides Claude Code in your terminal is the exact same http
FT: The White House is moving to give major US agencies access to a modified Anthropic Mythos model built to hunt dangerous software flaws before attackers find them. That makes Mythos useful for defense because a model that can find a weakness in an operating system, browser, h
Created 2026-04-23 · Updated 2026-04-23
Edit rule text
[tangential_ai_capability_or_model_technical_hype]
Learned 3 rejections Active
Exclude posts that emphasize AI model performance metrics, speed upgrades, benchmark improvements, or technical architecture features without demonstrating direct application to a healthcare delivery, clinical, or regulatory challenge. Posts must connect capability to healthcare use case.
Posts about AI model technical capabilities, speed improvements, or performance benchmarks with loose or no healthcare connection
3 example posts
Boltz-2 just got a major speed upgrade. 🚀 We’re releasing Lightning-Boltz, a local, GPU-accelerated framework free from public MSA server bottlenecks.⚡ On a single L40S, total runtime drops to 28s per input vs 89s with the rate-limited server and 298s with MMseqs-CPU. 1/5 🧵 ht
Everyone's covering agents that help you work and build. Almost nobody's covering this: The same primitives ARE the production runtime. The SDK is one line: npm install @anthropic-ai/claude-agent-sdk The CLAUDE.md that guides Claude Code in your terminal is the exact same http
Today we launched a major update to the OpenAI Agents SDK to help developers build and deploy long-running, durable agents in production. You can now build your own Codex-style agents using powerful primitives for modern agents - file and computer use, skills, memory and
Created 2026-04-22 · Updated 2026-04-22
Edit rule text
[cybersecurity_or_vulnerability_tangential]
Learned 3 rejections Active
Exclude posts about AI discovering vulnerabilities, corporate security breaches, or cybersecurity threats unless the post explicitly connects to healthcare data, patient privacy, EHR systems, or healthcare infrastructure. General cybersecurity or defense topics do not qualify.
Posts about AI-assisted cybersecurity vulnerabilities, hacking, or zero-day exploits without healthcare application
3 example posts
Guillermo reports "we believe the attacking group to be highly sophisticated and, I strongly suspect, significantly accelerated by AI. They moved with surprising velocity and in-depth understanding of Vercel" Alex Stamos warns us that defensive agents with autonomy and https://t
> Vercel got pawned > severe enough to notify law enforcement > the only advice: “review your environment variables” > what does that even mean? > $10B company, and this is how you communicate Cyber attacks ramping fast, starting to see why Anthropic is scared to
FT: The White House is moving to give major US agencies access to a modified Anthropic Mythos model built to hunt dangerous software flaws before attackers find them. That makes Mythos useful for defense because a model that can find a weakness in an operating system, browser, h
Created 2026-04-22 · Updated 2026-04-22
Edit rule text
[ai_company_product_metrics_and_launches_not_healthcare_focused]
Learned 3 rejections Active
Exclude posts announcing AI company product launches, SDK/API releases, business metrics, or feature updates unless the post specifically analyzes the product's healthcare application, regulatory compliance, or impact on healthcare workflows.
Posts about AI company product launches, SDK releases, feature announcements, or business metrics for tools used broadly across industries, with tenuous healthcare connection.
3 example posts
𝐇𝐨𝐜𝐤𝐞𝐲𝐒𝐭𝐚𝐜𝐤 (𝐘𝐂 𝐒𝟐𝟑) 𝐫𝐚𝐢𝐬𝐞𝐝 $𝟓𝟎𝐌 𝐭𝐨 𝐛𝐮𝐢𝐥𝐝 𝐀𝐈 𝐫𝐞𝐯𝐞𝐧𝐮𝐞 𝐚𝐠𝐞𝐧𝐭𝐬 𝐟𝐨𝐫 𝐭𝐡𝐞 𝐞𝐧𝐭𝐞𝐫𝐩𝐫𝐢𝐬𝐞. 📈 Systems of record → systems of action. Congrats to @hockeystackHQ co-founders Buğra Gündüz, Arda Bulut, and Emir Atlı on the https://t.co/5A7An4w0MW
Congrats to @AbridgeHQ, @AnthropicAI, @cursor_ai, @elise_ai, @Fal, @WeAreLegora, and @Perplexity_ai on being named to the @Forbes AI 50 — redefining how the world builds, works, and communicates through AI. We couldn't be more excited to back them as they continue to shape the h
Everyone's covering agents that help you work and build. Almost nobody's covering this: The same primitives ARE the production runtime. The SDK is one line: npm install @anthropic-ai/claude-agent-sdk The CLAUDE.md that guides Claude Code in your terminal is the exact same http
Created 2026-04-22 · Updated 2026-04-22
Edit rule text
[clinical_research_observation_without_healthcare_context]
Learned 3 rejections Active
Exclude posts that present clinical research results, drug trial outcomes, or laboratory findings as standalone observations without discussing healthcare delivery, regulatory, reimbursement, or system-level implications. The post must connect clinical data to healthcare practice, access, or implementation challenges.
Posts reporting clinical trial results, drug efficacy data, or research findings without analysis of healthcare system implications or adoption barriers.
3 example posts
Good summary of the marked benefit of the molecular glue drug (daraxonrasib) vs pancreatic cancer, from Revolution Medicines, and other progress (adds to the neoantigen vaccine with 6-year survival) gift link https://t.co/qk7Ar9dCAQ https://t.co/SMiA51fiwX
Insightful plenary from the father of CAR-T, @carlhjune #AACR26 🔬 CAR-T for solid tumors is finally breaking through. 7 FDA approvals in blood cancers and now solid tumors are next 🎯 Clinical signals • CLDN18.2 (Satri-cel): 38% vs 4% ORR in gastric cancer (The Lancet 2025) http
🧬 In vivo CAR-T engineering: the next frontier? From manufacturing → reprogramming in situ! @AACR https://t.co/i78CUWFBcE ▪️ Bypasses ex vivo complexity & delays 🦠 Viral + non-viral delivery strategies emerging 🎯 Targets endogenous T cells directly in patients 💥 Potential h
Created 2026-04-21 · Updated 2026-04-21
Edit rule text
[ai_company_product_metrics_and_funding]
Learned 3 rejections Active
Exclude posts that primarily report funding amounts, valuation milestones, revenue figures, or product launch metrics for AI companies—even if the company has healthcare products. The post must demonstrate how the AI is applied to solve a healthcare problem, not just celebrate the company's business success.
Posts reporting on AI company funding rounds, revenue milestones, or product launch metrics without healthcare application focus.
3 example posts
𝐇𝐨𝐜𝐤𝐞𝐲𝐒𝐭𝐚𝐜𝐤 (𝐘𝐂 𝐒𝟐𝟑) 𝐫𝐚𝐢𝐬𝐞𝐝 $𝟓𝟎𝐌 𝐭𝐨 𝐛𝐮𝐢𝐥𝐝 𝐀𝐈 𝐫𝐞𝐯𝐞𝐧𝐮𝐞 𝐚𝐠𝐞𝐧𝐭𝐬 𝐟𝐨𝐫 𝐭𝐡𝐞 𝐞𝐧𝐭𝐞𝐫𝐩𝐫𝐢𝐬𝐞. 📈 Systems of record → systems of action. Congrats to @hockeystackHQ co-founders Buğra Gündüz, Arda Bulut, and Emir Atlı on the https://t.co/5A7An4w0MW
Congrats to @AbridgeHQ, @AnthropicAI, @cursor_ai, @elise_ai, @Fal, @WeAreLegora, and @Perplexity_ai on being named to the @Forbes AI 50 — redefining how the world builds, works, and communicates through AI. We couldn't be more excited to back them as they continue to shape the h
Today we launched a major update to the OpenAI Agents SDK to help developers build and deploy long-running, durable agents in production. You can now build your own Codex-style agents using powerful primitives for modern agents - file and computer use, skills, memory and
Created 2026-04-21 · Updated 2026-04-21
Edit rule text
[ai_company_product_metrics_and_business_launches]
Learned 3 rejections Active
Exclude posts that focus primarily on AI company product launches, revenue figures, ARR milestones, or business metrics (e.g., Claude Code's $2.5B annualized revenue, Anthropic hiring announcements, OpenAI SDK updates) — even if the company has healthcare-adjacent products. The post must demonstrate a specific healthcare use case or clinical application, not just celebrate the AI product itself.
Posts about AI company product launches, revenue milestones, or business metrics disconnected from healthcare application.
3 example posts
Boltz-2 just got a major speed upgrade. 🚀 We’re releasing Lightning-Boltz, a local, GPU-accelerated framework free from public MSA server bottlenecks.⚡ On a single L40S, total runtime drops to 28s per input vs 89s with the rate-limited server and 298s with MMseqs-CPU. 1/5 🧵 ht
Today we launched a major update to the OpenAI Agents SDK to help developers build and deploy long-running, durable agents in production. You can now build your own Codex-style agents using powerful primitives for modern agents - file and computer use, skills, memory and
Anthropic's CEO: “coding is going away first, then all of software engineering." Now, Anthropic looks to hire 454 engineers at $320k–$405k. coding isn’t vanishing it’s becoming leverage for the few who can build, review, and ship at a completely different scale. https://t.co
Created 2026-04-20 · Updated 2026-04-20
Edit rule text
[self_promotional_tool_launch_without_healthcare_validation]
Learned 3 rejections Active
Exclude posts that promote AI tool launches, SDK releases, product announcements, or company features—even if the tool is healthcare-adjacent. Posts must include evidence of healthcare customer traction, clinical validation, or healthcare-specific use case results, not just feature availability.
Posts announcing AI tools, SDKs, platforms, or product launches (from companies or authors) without demonstrating healthcare customer adoption, clinical validation, or healthcare-specific problem-solving.
3 example posts
Everyone's covering agents that help you work and build. Almost nobody's covering this: The same primitives ARE the production runtime. The SDK is one line: npm install @anthropic-ai/claude-agent-sdk The CLAUDE.md that guides Claude Code in your terminal is the exact same http
Today we launched a major update to the OpenAI Agents SDK to help developers build and deploy long-running, durable agents in production. You can now build your own Codex-style agents using powerful primitives for modern agents - file and computer use, skills, memory and
Boris Cherny created Claude Code. It hit $2.5 billion in annualized revenue in 9 months. Fastest B2B product ramp in history. Faster than ChatGPT, Slack, or Snowflake ever reached $1 billion. Now he says coding is “solved” and IDEs will be dead by end of year. https://t.co/HI7M
Created 2026-04-20 · Updated 2026-04-20
Edit rule text
[clinical_research_observation_without_healthcare_systems_insight]
Learned 3 rejections Active
Exclude posts that report drug trial results, phase study outcomes, or clinical efficacy metrics in isolation. Posts must include analysis of healthcare system implications—reimbursement, adoption barriers, operational integration, or delivery model impact.
Posts reporting clinical trial results, drug efficacy data, or research findings without connecting to healthcare delivery systems, reimbursement, access, or operational implications.
3 example posts
This is now published – the first win for factor XI inhibition in ischemic stroke The reason it's so interesting is that factor XI inhibition reduces the risk of pathological clotting without increasing the risk of bleeding The idea came from genetic evidence: humans with https
New promising phase 1 study for lung cancer @NEJM * Zongertinib in HER2-Mutant NSCLC -ORR 76% (tumor shrinkage in most patients) -PFS 14.4 mo (disease control) -Brain mets: 47% response ✅ https://t.co/jVN8TuRJcg
Today the first results of the very first phase 3 study of a pan-KRAS-inhibitor in metastatic pancreatic cancer dropped, which might apply to > 90% of all pancreatic cancer patients with a KRAS-mutation! Median overall survival of 13.2 months versus 6.7 months with chemo in
Created 2026-04-20 · Updated 2026-04-20
Edit rule text
[political_outrage_or_scandal_without_systems_analysis]
Learned 3 rejections Active
Exclude posts that report healthcare fraud, medical misconduct, or policy scandals primarily as outrage or criminal reporting without analyzing systemic causes, healthcare market impact, or operational implications. Scandal reporting without systems analysis is crime/politics content, not healthcare tech analysis.
Posts about healthcare fraud, medical scandal, or policy outrage focused on moral/criminal reporting without healthcare systems, operational, or market analysis.
3 example posts
$340 MILLION in fraud targeted — in 1 WEEK. That’s what happens when enforcement gets serious. Luxury cars. Fake claims. Stolen benefits meant for Americans in need — now turning into prison sentences. The hammer is dropping. We’re just getting started.https://t.co/qP0cOIypE4
🚨 As you pay your taxes this week, LOOK at what the fraudsters allegedly did with your money❗️ 🔹Cosmetic procedures 🔹Breast implants 🔹Tweaks to arms and thighs 🔹Tummy tuck 🔹Purebred dogs 🔹Flights to Hawaii 🔹Flights to Disneyland 🔹Multimillion-dollar home 🔹Range https://t.c
🚨 Fraudsters literally looted $250-500 BILLION a year from taxpayers for years, now changes are being made to prevent this fraud: - Treasury is now going after the banks - Whistleblowers can make 30% for exposing fraud - Auto dealers will be tracked down END ALL THE FRAUD. https
Created 2026-04-20 · Updated 2026-04-20
Edit rule text
[ai_model_capability_hype_tangential]
Learned 3 rejections Active
Exclude posts that highlight AI model capabilities (code generation speed, vulnerability detection, autonomous system performance) in non-healthcare or tangentially healthcare contexts. The post must demonstrate a clear healthcare workflow or outcome impact, not just AI capability demonstration.
Posts showcasing AI model technical capabilities (coding speed, security vulnerability detection, network infiltration) with loose or no healthcare framing.
3 example posts
Everyone's covering agents that help you work and build. Almost nobody's covering this: The same primitives ARE the production runtime. The SDK is one line: npm install @anthropic-ai/claude-agent-sdk The CLAUDE.md that guides Claude Code in your terminal is the exact same http
FT: The White House is moving to give major US agencies access to a modified Anthropic Mythos model built to hunt dangerous software flaws before attackers find them. That makes Mythos useful for defense because a model that can find a weakness in an operating system, browser, h
AI is letting developers ship three to four times faster. It is also flooding codebases with vulnerabilities at the same rate. Aikido Security scans 15 open-source ecosystems for malware. A year ago: 30,000 packages per day. Now: 100,000. The attack surface is not growing https
Created 2026-04-20 · Updated 2026-04-20
Edit rule text
[ai_model_capability_hype_without_healthcare_specificity]
Learned 3 rejections Active
Exclude posts announcing new AI model capabilities, releases, or technical benchmarks (e.g., coding speed 3-4x faster, agent jailbreaks, zero-day hunting) unless they explicitly demonstrate how that capability addresses a named healthcare workflow, compliance requirement, or business model. Generic tech capability announcements do not qualify.
Posts about AI model releases, capabilities (coding speed, jailbreak tests, agent autonomy), or security vulnerabilities with loose or no healthcare framing
3 example posts
Everyone's covering agents that help you work and build. Almost nobody's covering this: The same primitives ARE the production runtime. The SDK is one line: npm install @anthropic-ai/claude-agent-sdk The CLAUDE.md that guides Claude Code in your terminal is the exact same http
FT: The White House is moving to give major US agencies access to a modified Anthropic Mythos model built to hunt dangerous software flaws before attackers find them. That makes Mythos useful for defense because a model that can find a weakness in an operating system, browser, h
AI is letting developers ship three to four times faster. It is also flooding codebases with vulnerabilities at the same rate. Aikido Security scans 15 open-source ecosystems for malware. A year ago: 30,000 packages per day. Now: 100,000. The attack surface is not growing https
Created 2026-04-20 · Updated 2026-04-20
Edit rule text
[peptide_and_glp1_macro_or_personal_narrative]
Learned 3 rejections Active
Exclude posts about GLP-1 drugs, peptides, or obesity treatments that center on personal anecdotes (individual patient stories, side effect narratives, family histories), market speculation (pricing, franchise value), or generic mechanism reviews—unless they analyze healthcare delivery, reimbursement, or clinical evidence gaps.
Posts about GLP-1 drugs, peptides, or weight-loss medications framed as personal stories, market hype, or macro economic commentary without healthcare systems analysis
3 example posts
I never met my grandfather. He died of pancreatic cancer when my father was just 19. Today, Yash Bindal, 33, father to 18-month-old Maya, faces the same fate. @PopVaxIndia is using AI to make him a personalized generative medicine to extend his life. https://t.co/O5VIXbmMGd
And there it is. Within hours of RFK's announcement someone is already pricing out how much Hims can charge for compounds the research community has had access to for a fraction of that cost. This is why the outcome of these PCAC meetings matters more than the announcement.
I have now received nine reports from people taking GLP-1 drugs who got the same side effect: They no longer feel normal when they come off. "I feel hangry again", "I started thinking about hunger and I hate it", "I have to go back to Adderall". 8/9 reports -> from women.
Created 2026-04-20 · Updated 2026-04-20
Edit rule text
[robotics_and_speculative_sci_fi_applications]
Learned 3 rejections Active
Exclude posts about robots or AI agents competing in non-healthcare domains (marathons, games, sports records, cybersecurity penetration tests, space tasks) even if they claim relevance to healthcare labor. The post must describe actual healthcare work, not speculative spillover.
Posts about humanoid robots, autonomous labs, or speculative AI agents achieving non-healthcare feats (sports, marathons, network takeovers)
3 example posts
In Beijing's 2026 humanoid robot half-marathon, HONOR's Lightning completed the 21 km course in 50:26 minute. Beat current human men's half-marathon world record of 57:20. Last year's winner took over 2 hours 40 minutes. Massive progress in 12 month https://t.co/OcZJ66ebWD
Two years ago the best AI models couldn't complete beginner-level cyber tasks. One just executed a full 32-step corporate network takeover. The Bank of England is convening emergency CEO briefings. Look at that chart. GPT-4o maxes out at 2 steps. Initial reconnaissance. It can
Yesterday, @RandDWorld featured us twice. @ProQR turns to Ginkgo’s autonomous lab to scale AI-enabled RNA editing discovery: https://t.co/DyENAd4VdM Ginkgo’s CEO says biotech needs its Waymo moment: https://t.co/kW27eBjAmf Want to learn more about our partnership with ProQR? h
Created 2026-04-20 · Updated 2026-04-20
Edit rule text
[healthcare_fraud_scandal_or_financial_scandal_reporting_without_systems_analysis]
Learned 3 rejections Active
Exclude posts that report fraud schemes, embezzlement, fake billing, or criminal cases in healthcare as breaking news or enforcement updates — unless the post analyzes the underlying healthcare system vulnerability, reimbursement design flaw, or structural weakness that enabled the fraud.
Posts reporting healthcare fraud, billing schemes, or financial scandals as crime reporting or outrage without analyzing systemic healthcare delivery or reimbursement failures
3 example posts
$340 MILLION in fraud targeted — in 1 WEEK. That’s what happens when enforcement gets serious. Luxury cars. Fake claims. Stolen benefits meant for Americans in need — now turning into prison sentences. The hammer is dropping. We’re just getting started.https://t.co/qP0cOIypE4
🚨 As you pay your taxes this week, LOOK at what the fraudsters allegedly did with your money❗️ 🔹Cosmetic procedures 🔹Breast implants 🔹Tweaks to arms and thighs 🔹Tummy tuck 🔹Purebred dogs 🔹Flights to Hawaii 🔹Flights to Disneyland 🔹Multimillion-dollar home 🔹Range https://t.c
🚨 Fraudsters literally looted $250-500 BILLION a year from taxpayers for years, now changes are being made to prevent this fraud: - Treasury is now going after the banks - Whistleblowers can make 30% for exposing fraud - Auto dealers will be tracked down END ALL THE FRAUD. https
Created 2026-04-20 · Updated 2026-04-20
Edit rule text
[ai_company_product_metrics_and_business_launches_not_healthcare_focused]
Learned 3 rejections Active
Exclude posts announcing AI company product releases, hiring milestones, revenue figures, or business metrics (e.g., $2.5B ARR, hiring engineers at $320k salary, fastest ramp in history) unless the post explicitly analyzes how that product solves a healthcare delivery problem.
Posts about AI company product launches, revenue achievements, or business metrics without demonstrating healthcare-specific application or impact
3 example posts
Everyone's covering agents that help you work and build. Almost nobody's covering this: The same primitives ARE the production runtime. The SDK is one line: npm install @anthropic-ai/claude-agent-sdk The CLAUDE.md that guides Claude Code in your terminal is the exact same http
Today we launched a major update to the OpenAI Agents SDK to help developers build and deploy long-running, durable agents in production. You can now build your own Codex-style agents using powerful primitives for modern agents - file and computer use, skills, memory and
Anthropic's CEO: “coding is going away first, then all of software engineering." Now, Anthropic looks to hire 454 engineers at $320k–$405k. coding isn’t vanishing it’s becoming leverage for the few who can build, review, and ship at a completely different scale. https://t.co
Created 2026-04-20 · Updated 2026-04-20
Edit rule text
[tangential_ai_capability_or_model_hype]
Learned 3 rejections Active
Exclude posts about AI model capabilities (agentic behavior, reasoning, hallucination, jailbreaks, safety tests) unless they explicitly connect to a healthcare-specific problem, regulatory requirement, or clinical decision-making context.
Posts about AI model capabilities, technical benchmarks, or safety concerns that lack healthcare application or systems context.
3 example posts
Market maps have become a real focus of ours as LLMs are getting company categorization so wrong. Our latest, in partnership with Confido Health & @RMFnyc1, focuses on agentic AI for the ambulatory market. What's being deployed now? Our focus was Series A onwards. 👇 htt
Researchers gave AI agents a simple choice: hit your performance target or follow the rules. Most of them chose to cheat. McGill University tested 12 of the most powerful AI models on 40 realistic workplace scenarios. Healthcare. Finance. Logistics. Scientific research. Each AI
AI is letting developers ship three to four times faster. It is also flooding codebases with vulnerabilities at the same rate. Aikido Security scans 15 open-source ecosystems for malware. A year ago: 30,000 packages per day. Now: 100,000. The attack surface is not growing https
Created 2026-04-19 · Updated 2026-04-19
Edit rule text
[clinical_research_observation_without_systems_insight]
Learned 3 rejections Active
Exclude posts that report clinical trial outcomes, drug efficacy percentages, or research findings (e.g., 'ORR 76% in lung cancer', 'phase 3 pancreatic cancer results', 'ECG diagnostic accuracy') unless they analyze the healthcare system implications, reimbursement barriers, implementation strategy, or access/affordability challenges.
Posts sharing clinical trial results, drug efficacy data, or research findings without connecting to healthcare delivery systems, pricing, access, or implementation challenges.
3 example posts
This is now published – the first win for factor XI inhibition in ischemic stroke The reason it's so interesting is that factor XI inhibition reduces the risk of pathological clotting without increasing the risk of bleeding The idea came from genetic evidence: humans with https
New promising phase 1 study for lung cancer @NEJM * Zongertinib in HER2-Mutant NSCLC -ORR 76% (tumor shrinkage in most patients) -PFS 14.4 mo (disease control) -Brain mets: 47% response ✅ https://t.co/jVN8TuRJcg
Yesterday, @RandDWorld featured us twice. @ProQR turns to Ginkgo’s autonomous lab to scale AI-enabled RNA editing discovery: https://t.co/DyENAd4VdM Ginkgo’s CEO says biotech needs its Waymo moment: https://t.co/kW27eBjAmf Want to learn more about our partnership with ProQR? h
Created 2026-04-18 · Updated 2026-04-18
Edit rule text
[infrastructure_compute_hype_no_healthcare_specificity]
Learned 3 rejections Active
Exclude posts that discuss compute infrastructure, CPU launches, data center capacity, space-based compute, or hardware announcements (e.g., Nvidia CPUs, Elon's orbital compute, chip supply chains) unless the post explicitly ties this infrastructure to a specific healthcare delivery problem or clinical application.
Posts about AI infrastructure, compute capacity, or hardware announcements with only tangential healthcare relevance.
3 example posts
🔥 CPUs are having a moment. #Nvidia launched a standalone CPU. #Arm made its first chip in 35 years. #Intel & #AMD are raising prices amid a supply crunch. What's behind it: Agentic AI needs far more CPU than anyone planned for — driving a structural shift in CPU:GPU ratios tow
To put Elon's space compute vision into perspective:  1 TW of compute in orbit  That's 10 million tons to orbit each year. That's 100,000 launches a year, almost one every 5 minutes.  In the airline business that's normal!
🚨MAJOR INTERVIEW: Jensen Huang joins the Besties! The @nvidia CEO joins to discuss: -- Nvidia's future, roadmap to $1T revenue -- Physical AI's $50T market -- Rise of the agent, OpenClaw's inflection moment -- Inference explosion, Groq deal -- AI PR Crisis, Anthropic's comms m
Created 2026-04-18 · Updated 2026-04-18
Edit rule text
[clinical_trial_or_drug_research_observation_only]
Learned 3 rejections Active
Exclude posts that simply report phase trial results, drug efficacy data, or breakthrough announcements without analyzing healthcare delivery implications, market access barriers, or systemic healthcare change. Clinical data alone without systems context does not qualify.
Posts reporting clinical trial results or drug discovery announcements without healthcare systems or market analysis context.
3 example posts
This is now published – the first win for factor XI inhibition in ischemic stroke The reason it's so interesting is that factor XI inhibition reduces the risk of pathological clotting without increasing the risk of bleeding The idea came from genetic evidence: humans with https
New promising phase 1 study for lung cancer @NEJM * Zongertinib in HER2-Mutant NSCLC -ORR 76% (tumor shrinkage in most patients) -PFS 14.4 mo (disease control) -Brain mets: 47% response ✅ https://t.co/jVN8TuRJcg
Yesterday, @RandDWorld featured us twice. @ProQR turns to Ginkgo’s autonomous lab to scale AI-enabled RNA editing discovery: https://t.co/DyENAd4VdM Ginkgo’s CEO says biotech needs its Waymo moment: https://t.co/kW27eBjAmf Want to learn more about our partnership with ProQR? h
Created 2026-04-17 · Updated 2026-04-17
Edit rule text
[tangential_ai_company_product_launches]
Learned 3 rejections Active
Exclude posts announcing new AI model releases, agent SDKs, or developer tools from AI companies (Anthropic, OpenAI, Microsoft) unless the post explicitly demonstrates how the tool solves a healthcare-specific problem or workflow. Product availability alone is insufficient.
Posts about AI company product releases and SDKs that are not specifically designed for healthcare workflows.
3 example posts
OpenAI introduced GPT-Rosalind, a frontier reasoning model specifically architected for the life sciences, focusing heavily on biology, drug discovery, and translational medicine. Designed to accelerate the historically slow 10-to-15-year drug approval pipeline, Rosalind is http
Today we launched a major update to the OpenAI Agents SDK to help developers build and deploy long-running, durable agents in production. You can now build your own Codex-style agents using powerful primitives for modern agents - file and computer use, skills, memory and
Microsoft is reportedly testing the integration of "OpenClaw-like" autonomous AI agents directly into its Microsoft 365 Copilot ecosystem. Moving beyond a reactive chatbot interface, the goal is to create an "always-on" assistant that runs autonomously in the background. These
Created 2026-04-17 · Updated 2026-04-17
Edit rule text
[ai_infrastructure_and_compute_hype_no_application]
Learned 3 rejections Active
Exclude posts that focus on compute infrastructure (GPUs, CPUs, data centers, power grids, orbital compute) or hardware announcements for AI labs, even if healthcare is mentioned tangentially. These posts must demonstrate concrete healthcare application, not just infrastructure capabilities.
Posts about AI compute infrastructure, data centers, chips, and energy requirements without healthcare application context.
3 example posts
🔥 CPUs are having a moment. #Nvidia launched a standalone CPU. #Arm made its first chip in 35 years. #Intel & #AMD are raising prices amid a supply crunch. What's behind it: Agentic AI needs far more CPU than anyone planned for — driving a structural shift in CPU:GPU ratios tow
To put Elon's space compute vision into perspective:  1 TW of compute in orbit  That's 10 million tons to orbit each year. That's 100,000 launches a year, almost one every 5 minutes.  In the airline business that's normal!
🚨MAJOR INTERVIEW: Jensen Huang joins the Besties! The @nvidia CEO joins to discuss: -- Nvidia's future, roadmap to $1T revenue -- Physical AI's $50T market -- Rise of the agent, OpenClaw's inflection moment -- Inference explosion, Groq deal -- AI PR Crisis, Anthropic's comms m
Created 2026-04-17 · Updated 2026-04-17
Edit rule text
[healthcare_fraud_or_financial_scandal_reporting_without_analysis]
Learned 3 rejections Active
Exclude posts that report healthcare fraud, billing abuse, or financial misconduct as breaking news or outrage without structural analysis of why the system failed or how it connects to broader healthcare operations/incentives.
Posts reporting healthcare fraud, billing schemes, or financial misconduct that function as news alerts rather than systems analysis.
3 example posts
🚨 As you pay your taxes this week, LOOK at what the fraudsters allegedly did with your money❗️ 🔹Cosmetic procedures 🔹Breast implants 🔹Tweaks to arms and thighs 🔹Tummy tuck 🔹Purebred dogs 🔹Flights to Hawaii 🔹Flights to Disneyland 🔹Multimillion-dollar home 🔹Range https://t.c
🚨 Fraudsters literally looted $250-500 BILLION a year from taxpayers for years, now changes are being made to prevent this fraud: - Treasury is now going after the banks - Whistleblowers can make 30% for exposing fraud - Auto dealers will be tracked down END ALL THE FRAUD. https
🚨 Surgeon @EithanHaim reveals shocking medical fraud scheme: Texas doctors allegedly changing teens' medical records and using fake billing codes to secretly continue banned gender treatments—scamming insurance and taxpayers. He's speaking at a #DetransAwarenessDay @genspect foru
Created 2026-04-17 · Updated 2026-04-17
Edit rule text
[ai_model_capability_technical_hype_tangential]
Learned 3 rejections Active
Exclude posts that highlight AI model technical capabilities (agent behavior, safety sandbox escapes, coding speed benchmarks, network takeovers) presented as interesting technical feats, unless the post explicitly analyzes healthcare-specific risks or applications of that capability.
Posts about AI model technical capabilities (jailbreaks, agent autonomy, coding speed) framed as security or capability news rather than healthcare-specific application.
3 example posts
Researchers gave AI agents a simple choice: hit your performance target or follow the rules. Most of them chose to cheat. McGill University tested 12 of the most powerful AI models on 40 realistic workplace scenarios. Healthcare. Finance. Logistics. Scientific research. Each AI
Two years ago the best AI models couldn't complete beginner-level cyber tasks. One just executed a full 32-step corporate network takeover. The Bank of England is convening emergency CEO briefings. Look at that chart. GPT-4o maxes out at 2 steps. Initial reconnaissance. It can
Is AGI actually here…or are we watching the best marketing play in tech history? @danielnewmanUV and @patrickmoorhead break it down on The Flip. On one side: A model escaped a safety sandbox and chained zero-days without human prompting. On the other: The "G" in AGI means https
Created 2026-04-17 · Updated 2026-04-17
Edit rule text
[ai_infrastructure_and_compute_hype_tangential]
Learned 3 rejections Active
Exclude posts that focus primarily on compute infrastructure (CPUs, data centers, chip supply chains, energy grids, orbital compute) even if they mention healthcare applications tangentially or speculatively. The post must demonstrate concrete healthcare application, not hypothetical future potential.
Posts about data center capacity, chip manufacturing, or compute infrastructure with only loose or aspirational healthcare framing.
3 example posts
🔥 CPUs are having a moment. #Nvidia launched a standalone CPU. #Arm made its first chip in 35 years. #Intel & #AMD are raising prices amid a supply crunch. What's behind it: Agentic AI needs far more CPU than anyone planned for — driving a structural shift in CPU:GPU ratios tow
To put Elon's space compute vision into perspective:  1 TW of compute in orbit  That's 10 million tons to orbit each year. That's 100,000 launches a year, almost one every 5 minutes.  In the airline business that's normal!
🚨MAJOR INTERVIEW: Jensen Huang joins the Besties! The @nvidia CEO joins to discuss: -- Nvidia's future, roadmap to $1T revenue -- Physical AI's $50T market -- Rise of the agent, OpenClaw's inflection moment -- Inference explosion, Groq deal -- AI PR Crisis, Anthropic's comms m
Created 2026-04-17 · Updated 2026-04-17
Edit rule text
[glp1_peptide_adjacent_macro_or_financial_framing]
Learned 3 rejections Active
Exclude posts about GLP-1 drugs, peptide therapeutics, or weight-loss medications that focus on financial markets, stock performance, personal side effects, or general economic trends—unless the post substantively analyzes healthcare access, reimbursement, manufacturing capacity, or clinical practice change.
Posts about GLP-1 drugs or peptides framed primarily through macro economics, stock performance, or personal wellness without healthcare systems or access analysis.
3 example posts
And there it is. Within hours of RFK's announcement someone is already pricing out how much Hims can charge for compounds the research community has had access to for a fraction of that cost. This is why the outcome of these PCAC meetings matters more than the announcement.
I have now received nine reports from people taking GLP-1 drugs who got the same side effect: They no longer feel normal when they come off. "I feel hangry again", "I started thinking about hunger and I hate it", "I have to go back to Adderall". 8/9 reports -> from women.
The brain is the master regulator of food intake and energy balance. A brilliant new @CellCellPress review, including the mechanism of GLP-1 drugs, by @ClemmensenC and colleagues, open-access https://t.co/KbDf7ym288 https://t.co/DMFeS9zHfw
Created 2026-04-17 · Updated 2026-04-17
Edit rule text
[political_outrage_or_scandal_reporting_without_systems_analysis]
Learned 3 rejections Active
Exclude posts that report on healthcare fraud, insurance scams, medical misconduct, or regulatory violations as breaking news or scandal outrage without analyzing root causes, systemic healthcare business model problems, or policy implications for healthcare tech/operations.
Posts about healthcare fraud, regulatory failures, or scandals framed as outrage without substantive healthcare systems insight or policy analysis.
3 example posts
🚨 As you pay your taxes this week, LOOK at what the fraudsters allegedly did with your money❗️ 🔹Cosmetic procedures 🔹Breast implants 🔹Tweaks to arms and thighs 🔹Tummy tuck 🔹Purebred dogs 🔹Flights to Hawaii 🔹Flights to Disneyland 🔹Multimillion-dollar home 🔹Range https://t.c
🚨 Fraudsters literally looted $250-500 BILLION a year from taxpayers for years, now changes are being made to prevent this fraud: - Treasury is now going after the banks - Whistleblowers can make 30% for exposing fraud - Auto dealers will be tracked down END ALL THE FRAUD. https
🚨 Surgeon @EithanHaim reveals shocking medical fraud scheme: Texas doctors allegedly changing teens' medical records and using fake billing codes to secretly continue banned gender treatments—scamming insurance and taxpayers. He's speaking at a #DetransAwarenessDay @genspect foru
Created 2026-04-16 · Updated 2026-04-16
Edit rule text
[personal_anecdote_or_unsubstantiated_observation]
Learned 3 rejections Active
Exclude posts that rely primarily on personal anecdotes, individual case reports, unverified side effect reports, or singular clinical observations presented as healthcare insights without epidemiological data, system-wide evidence, or peer-reviewed backing.
Posts sharing personal experiences, unverified claims, or isolated clinical observations without systemic healthcare analysis or evidence.
3 example posts
good news: it is a specific virus that has a good prognosis - 85%+ of full recovery. thanks everyone who helped me; it is hard to research while immobilized, and I got some things wrong, which you helped clear up. im extremely thankful and hope i can give it back somehow sadl
~1-2% of the patients on ward rounds has something bad going on which hasn’t been identified yet. As the attending, one of my main duties on rounds is to spot these cases. I do a lot of this by Noticing Things. A 🤖 iPad makes it much less likely you will Notice Things. 🤔
Indian has 0.7 active physicians per 1,000 people, America has 3.0 active physicians per 1,000 people. You are a liar. You are not motivated by increasing patient access to care. You just want to practice in America because you can make more money.
Created 2026-04-16 · Updated 2026-04-16
Edit rule text
[unvalidated_or_speculative_clinical_claims_without_evidence]
Learned 3 rejections Active
Exclude posts that present unverified medical claims, early-stage experimental treatments, or speculative cures (e.g., 'a pill against pancreatic cancer' from a press release, GLP-1 side effect anecdotes, or unvalidated mechanism claims) without robust clinical trial evidence, peer-reviewed publication, or regulatory approval status clearly stated.
Posts about novel drugs, treatments, or medical interventions that lack peer-reviewed validation, clinical trial data, or are presented as speculative breakthroughs.
3 example posts
I never met my grandfather. He died of pancreatic cancer when my father was just 19. Today, Yash Bindal, 33, father to 18-month-old Maya, faces the same fate. @PopVaxIndia is using AI to make him a personalized generative medicine to extend his life. https://t.co/O5VIXbmMGd
New promising phase 1 study for lung cancer @NEJM * Zongertinib in HER2-Mutant NSCLC -ORR 76% (tumor shrinkage in most patients) -PFS 14.4 mo (disease control) -Brain mets: 47% response ✅ https://t.co/jVN8TuRJcg
I have now received nine reports from people taking GLP-1 drugs who got the same side effect: They no longer feel normal when they come off. "I feel hangry again", "I started thinking about hunger and I hate it", "I have to go back to Adderall". 8/9 reports -> from women.
Created 2026-04-16 · Updated 2026-04-16
Edit rule text
[retweet_or_shallow_commentary_without_substantive_analysis]
Learned 3 rejections Active
Exclude posts that amplify news, press releases, or product announcements with only surface-level commentary, questions, or celebratory reactions—unless the poster adds original analysis, criticism, or strategic insight explaining why the announcement matters for healthcare builders or healthcare systems.
Posts that retweet announcements, press releases, or other content with minimal original analysis or insight added by the poster
3 example posts
Yesterday, @RandDWorld featured us twice. @ProQR turns to Ginkgo’s autonomous lab to scale AI-enabled RNA editing discovery: https://t.co/DyENAd4VdM Ginkgo’s CEO says biotech needs its Waymo moment: https://t.co/kW27eBjAmf Want to learn more about our partnership with ProQR? h
.@openloophealth expands into sleep diagnostics. Health tech company announces new partnership Happy Sleep—bringing at‑home sleep apnea testing to patients for the first time. Watch to hear more about its big step toward better rest and smarter care⤵️ https://t.co/ATcNkYrrpK h
good news: it is a specific virus that has a good prognosis - 85%+ of full recovery. thanks everyone who helped me; it is hard to research while immobilized, and I got some things wrong, which you helped clear up. im extremely thankful and hope i can give it back somehow sadl
Created 2026-04-16 · Updated 2026-04-16
Edit rule text
[tangential_infrastructure_and_compute_hype_no_healthcare_specificity]
Learned 3 rejections Active
Exclude posts that discuss compute infrastructure, data center spending, GPU capacity, energy grids, satellite compute, or hardware roadmaps without explaining concrete healthcare applications or problems those solve. Posts about Nvidia's 'AI factory,' orbital compute, or hyperscaler capex belong here unless tied to specific clinical workflows.
Posts about AI infrastructure, compute capacity, data centers, or space-based computing with only loose or no healthcare connection
3 example posts
To put Elon's space compute vision into perspective:  1 TW of compute in orbit  That's 10 million tons to orbit each year. That's 100,000 launches a year, almost one every 5 minutes.  In the airline business that's normal!
🚨MAJOR INTERVIEW: Jensen Huang joins the Besties! The @nvidia CEO joins to discuss: -- Nvidia's future, roadmap to $1T revenue -- Physical AI's $50T market -- Rise of the agent, OpenClaw's inflection moment -- Inference explosion, Groq deal -- AI PR Crisis, Anthropic's comms m
Hyperscalers will spend $700 BILLION on data centers in 2026 alone. Amazon: $200B. Google: $185B. Meta: $135B. AI data centers now represent 70%+ of all new grid interconnection requests in the US. The bottleneck isn't the algorithm anymore. It's the power line.
Created 2026-04-16 · Updated 2026-04-16
Edit rule text
[clinical_research_or_trial_observation_without_healthcare_systems_insight]
Learned 3 rejections Active
Exclude posts that report clinical trial outcomes, drug efficacy data, or mechanistic research findings (pancreatic cancer trials, immunotherapy mechanisms, protein mapping) unless they explicitly address healthcare access, implementation challenges, or system-level adoption barriers.
Posts reporting clinical trial results, research findings, or drug discoveries without analysis of healthcare delivery, access, or system-level implications.
3 example posts
Revolution Medicines shared their findings in a press release Monday that said there may soon be a pill against pancreatic cancer, a deadly disease that strikes more than 60,000 Americans every year. The company said the pill doubled survival to 13.2 months compared with standard
Today the first results of the very first phase 3 study of a pan-KRAS-inhibitor in metastatic pancreatic cancer dropped, which might apply to > 90% of all pancreatic cancer patients with a KRAS-mutation! Median overall survival of 13.2 months versus 6.7 months with chemo in
NIH-funded researchers have uncovered a key reason why immunotherapy has largely failed in pancreatic cancer — and identified a promising strategy to overcome that resistance. Read on to learn more about this discovery: https://t.co/BoCHpLxp5g https://t.co/3DXv4E9DOE
Created 2026-04-16 · Updated 2026-04-16
Edit rule text
[glp1_and_peptide_adjacent_non_healthcare_context]
Learned 3 rejections Active
Exclude posts about GLP-1 drugs, obesity treatments, or peptide therapies that are primarily framed through venture capital, market economics, or behavioral/cultural angles rather than clinical efficacy, healthcare access, or delivery system impact.
Posts about GLP-1 drugs, peptides, or obesity framed through economics, investing, or non-clinical angles rather than clinical outcomes or healthcare delivery.
3 example posts
I have now received nine reports from people taking GLP-1 drugs who got the same side effect: They no longer feel normal when they come off. "I feel hangry again", "I started thinking about hunger and I hate it", "I have to go back to Adderall". 8/9 reports -> from women.
GLP-1 drugs are the ultimate validation of the techno-solutionist approach to society's most challenging problems. The obesity crisis seemed liked it would just get worse and worse forever. Scolding from public health officials didn't work. Proposals to completely overhaul our f
Great work by @DanielJDrucker and team; biologically plausible mechanism of GLP1-RA benefit independent of weight loss. Excellent article by @megtirrell @CNN describing the publication. Could it justify new approaches for these drugs? I think so. https://t.co/pHudk7lkAR
Created 2026-04-16 · Updated 2026-04-16
Edit rule text
[retweet_or_tangential_commentary_only]
Learned 3 rejections Active
Exclude posts that are retweets of press releases, celebratory company announcements, or one-line commentary without original analysis, healthcare systems context, or substantive take. Posts that merely celebrate a partnership or quote a keynote without explaining why it matters to healthcare builders should be rejected.
Posts that are primarily retweets, celebratory announcements, or generic commentary without substantive analysis or healthcare systems insight.
3 example posts
Yesterday, @RandDWorld featured us twice. @ProQR turns to Ginkgo’s autonomous lab to scale AI-enabled RNA editing discovery: https://t.co/DyENAd4VdM Ginkgo’s CEO says biotech needs its Waymo moment: https://t.co/kW27eBjAmf Want to learn more about our partnership with ProQR? h
.@openloophealth expands into sleep diagnostics. Health tech company announces new partnership Happy Sleep—bringing at‑home sleep apnea testing to patients for the first time. Watch to hear more about its big step toward better rest and smarter care⤵️ https://t.co/ATcNkYrrpK h
Revolution Medicines shared their findings in a press release Monday that said there may soon be a pill against pancreatic cancer, a deadly disease that strikes more than 60,000 Americans every year. The company said the pill doubled survival to 13.2 months compared with standard
Created 2026-04-16 · Updated 2026-04-16
Edit rule text
[unvalidated_drug_claims]
Learned 3 rejections Active
Exclude posts that present anecdotal reports, speculative mechanisms, or unvalidated claims about GLP-1 drugs, peptides, or experimental treatments as if they are established facts. Posts must cite peer-reviewed evidence or clinical trial results, not personal reports or inference.
Posts making speculative or anecdotal claims about GLP-1 drugs, peptides, or experimental treatments beyond published evidence
3 example posts
I have now received nine reports from people taking GLP-1 drugs who got the same side effect: They no longer feel normal when they come off. "I feel hangry again", "I started thinking about hunger and I hate it", "I have to go back to Adderall". 8/9 reports -> from women.
GLP-1 drugs are the ultimate validation of the techno-solutionist approach to society's most challenging problems. The obesity crisis seemed liked it would just get worse and worse forever. Scolding from public health officials didn't work. Proposals to completely overhaul our f
Great work by @DanielJDrucker and team; biologically plausible mechanism of GLP1-RA benefit independent of weight loss. Excellent article by @megtirrell @CNN describing the publication. Could it justify new approaches for these drugs? I think so. https://t.co/pHudk7lkAR
Created 2026-04-15 · Updated 2026-04-15
Edit rule text
[ai_infrastructure_hype]
Learned 3 rejections Active
Exclude posts that discuss AI infrastructure (data centers, compute spending, energy grids, foundational model capabilities, robotics platforms) unless they directly connect to a specific healthcare use case or clinical problem being solved.
Posts about AI compute, data centers, energy infrastructure, and foundational model capabilities without healthcare specificity
3 example posts
Is AGI actually here…or are we watching the best marketing play in tech history? @danielnewmanUV and @patrickmoorhead break it down on The Flip. On one side: A model escaped a safety sandbox and chained zero-days without human prompting. On the other: The "G" in AGI means https
Microsoft is exploring always-on AI agents for Copilot that can operate across Office apps without prompts, inspired by the viral OpenClaw project. The move comes as Anthropic encroaches on its core turf and customers question Copilot’s value. Read more:
Hyperscalers will spend $700 BILLION on data centers in 2026 alone. Amazon: $200B. Google: $185B. Meta: $135B. AI data centers now represent 70%+ of all new grid interconnection requests in the US. The bottleneck isn't the algorithm anymore. It's the power line.
Created 2026-04-15 · Updated 2026-04-15
Edit rule text
[personal_anecdote_or_unverified_claim]
Learned 3 rejections Active
Exclude posts that rely on personal anecdotes, alleged insider accounts, unverified fraud claims, or inflammatory accusations without documented evidence, official reporting, or credible source attribution.
Posts sharing personal experiences, unverified allegations, or claims without source verification or substantiation
3 example posts
good news: it is a specific virus that has a good prognosis - 85%+ of full recovery. thanks everyone who helped me; it is hard to research while immobilized, and I got some things wrong, which you helped clear up. im extremely thankful and hope i can give it back somehow sadl
🚨 Surgeon @EithanHaim reveals shocking medical fraud scheme: Texas doctors allegedly changing teens' medical records and using fake billing codes to secretly continue banned gender treatments—scamming insurance and taxpayers. He's speaking at a #DetransAwarenessDay @genspect foru
And then I want them to try to sort out one insurance issue. Just one. I want them to see the hours it takes to navigate hospital billing, specialist offices, CPT codes, pharmacy reps, compounding facilities, patient copay programs, and infusion experts. 2/3
Created 2026-04-15 · Updated 2026-04-15
Edit rule text
[tangential_infrastructure_without_healthcare_specificity]
Learned 3 rejections Active
Exclude posts about AI infrastructure, data centers, compute capacity, hardware advances, or robotics platforms unless they directly address a specific healthcare delivery problem or clinical workflow. Generic infrastructure announcements with loose healthcare framing do not qualify.
Posts about compute, data centers, hardware, or infrastructure that mention healthcare tangentially or lack concrete healthcare application
3 example posts
Hyperscalers will spend $700 BILLION on data centers in 2026 alone. Amazon: $200B. Google: $185B. Meta: $135B. AI data centers now represent 70%+ of all new grid interconnection requests in the US. The bottleneck isn't the algorithm anymore. It's the power line.
Elon Musk: “Hold on to your Tesla stock.” Because what’s coming isn’t just another car update—it’s an entirely new paradigm. From Optimus humanoid robots that could one day take care of your kids, walk your dog, and support elderly parents, to CyberCab scaling into mass product
Across NVIDIA Jetson and our robotics software stack, we’re focused on making it easy for developers to turn open source innovation, like @openclaw, into deployable, real‑world autonomy on the edge.
Created 2026-04-15 · Updated 2026-04-15
Edit rule text
[self_promotional_tool_launch_without_substance]
Learned 3 rejections Active
Exclude posts that announce a tool, framework, prompt, or product launch primarily to drive adoption or engagement, with minimal healthcare-specific reasoning, evidence, or analysis of why it matters for health tech stakeholders. Posts like 'I built a prompt that's badass' or 'Here's how to get 200 users from LinkedIn' belong here.
Posts promoting a tool, product, or framework launch with little substantive healthcare analysis or credibility signal.
3 example posts
Here is V2 of my company "Initiation Report" Deep Research Prompt. Serious thanks to the community for the feedback. This thing is pretty badass now. _____ I've made several updates: • No longer too positive: People rightfully called out that the previous model rated everything
HNSW is fast & performant. But what's it costing you? DiskBBQ gets you great recall & speed using a fraction of the memory. HNSW vs DiskBBQ in 40 seconds with @_jphwang https://t.co/3BqD9a6srU
The easiest way to get your first 200 users from LinkedIn: > Set up keywords for what you sell > OutX finds people already asking > Reply while intent is hot
Created 2026-04-14 · Updated 2026-04-14
Edit rule text
[academic_or_research_observation_without_systems_insight]
Learned 3 rejections Active
Exclude posts that cite clinical trial results, research findings, or academic discoveries in isolation without analyzing how those findings change healthcare business models, delivery systems, cost structures, or operational workflows. Posts like 'A study found that X improves outcomes' without healthcare systems context belong here.
Posts reporting clinical research findings or academic observations without healthcare business, delivery, or implementation context.
3 example posts
Today the first results of the very first phase 3 study of a pan-KRAS-inhibitor in metastatic pancreatic cancer dropped, which might apply to > 90% of all pancreatic cancer patients with a KRAS-mutation! Median overall survival of 13.2 months versus 6.7 months with chemo in
NIH-funded researchers have uncovered a key reason why immunotherapy has largely failed in pancreatic cancer — and identified a promising strategy to overcome that resistance. Read on to learn more about this discovery: https://t.co/BoCHpLxp5g https://t.co/3DXv4E9DOE
Across large, multicohort datasets, CardioNets achieved superior performance to ECG-only baselines and diagnostic accuracy comparable to CMR-based models, supporting its potential to expand access to advanced cardiovascular assessment. Full study results: https://t.co/VP2iOBLUev
Created 2026-04-14 · Updated 2026-04-14
Edit rule text
[compute_infrastructure_without_healthcare_specificity]
Learned 3 rejections Active
Exclude posts that discuss AI compute infrastructure, datacenter capex, energy demand, or hardware buildout without directly connecting these trends to healthcare delivery, clinical workflows, or health tech business models. Posts about 'Amazon spending $200B on datacenters' or 'NVIDIA chips' belong here unless they analyze healthcare-specific implications.
Posts about datacenter spending, energy infrastructure, and compute capacity that lack direct healthcare application or analysis.
3 example posts
Hyperscalers will spend $700 BILLION on data centers in 2026 alone. Amazon: $200B. Google: $185B. Meta: $135B. AI data centers now represent 70%+ of all new grid interconnection requests in the US. The bottleneck isn't the algorithm anymore. It's the power line.
Elon Musk: “Hold on to your Tesla stock.” Because what’s coming isn’t just another car update—it’s an entirely new paradigm. From Optimus humanoid robots that could one day take care of your kids, walk your dog, and support elderly parents, to CyberCab scaling into mass product
This chart puts the datacenter demands into perspective very clearly. Amazon has done more capex in the last 3 years than its entire history. Right now most AI adoption is on chat tools that are relatively token efficient. Comparatively, coding agents, use orders of magnitude h
Created 2026-04-14 · Updated 2026-04-14
Edit rule text
[retweet_or_tangential_commentary]
Learned 3 rejections Active
Exclude posts that are primarily retweets with minimal commentary, or posts that tangentially touch healthcare business/policy/tech without substantive original analysis, leaving unclear why healthcare tech professionals should engage.
Posts that are low-effort retweets or loose commentary on healthcare-adjacent topics without original analysis or substance
3 example posts
Another week on the road meeting with a couple dozen IT and AI leaders from large enterprises across banking, media, retail, healthcare, consulting, tech, and sports, to discuss agents in the enterprise. Some quick takeaways: * Clear that we’re moving from chat era of AI to
Sequoia partner @gradypb says software is shifting from apps that demand attention to agents that work quietly in the background. This shift will change what moats will look like, and will be especially hard for incumbents to deal with. "It's two very different business https://
From @Jessica_Baladad : To my friends in pharmacy who work at CVS, if you needed your District Manager yesterday, they were in the Finance, Ways and Means Committee watching the FairRx Act (HB1959) get placed behind the budget as it moves forward in the Tennessee General https:/
Created 2026-04-14 · Updated 2026-04-14
Edit rule text
[ai_technical_tangent_loose_healthcare_framing]
Learned 3 rejections Active
Exclude posts that treat healthcare as a loose contextual mention or speculative use case for AI technical advances (world models, reasoning systems, agentic behavior) without grounding in actual healthcare workflows, clinical validation, or operational implementation.
Posts about general AI technical advances (world models, reasoning, agents) that merely reference healthcare without substantive healthcare application
3 example posts
The 26 prompts running inside 𝗖𝗹𝗮𝘂𝗱𝗲 𝗖𝗼𝗱𝗲 just got open-sourced. This is literally the entire brain of a $200/month AI coding tool. Someone reverse-engineered every prompt from the accidentally published npm source and you can now study all of them for free. Claude Code uses 26
Claude Code is not AGI, but it is the single biggest advance in AI since the LLM. But the thing is, Claude Code is NOT a pure LLM. And it’s not pure deep learning. Not even close. And that changes everything. The source code leak proves it. Tucked away at its center is a
The CEO of Google DeepMind just went on record saying he disagrees with one of the most respected AI researchers in the world. Demis Hassabis, the man behind AlphaFold, AlphaGo, and Google's entire AI operation publicly pushed back against Yann LeCun's claim that large language
Created 2026-04-14 · Updated 2026-04-14
Edit rule text
[retweet_commentary_and_low_substance]
Learned 3 rejections Active
Exclude posts that are primarily retweets, quote-tweets with minimal added commentary, citations of others' research or announcements without original analysis or healthcare-specific insight, or posts that simply amplify another author's take without substantive contribution.
Posts that are primarily retweets of others' content, shallow commentary on external reports, or non-original analysis
3 example posts
Here is V2 of my company "Initiation Report" Deep Research Prompt. Serious thanks to the community for the feedback. This thing is pretty badass now. _____ I've made several updates: • No longer too positive: People rightfully called out that the previous model rated everything
"Do One Thing Every Day That Scares You" - debut From The Trenches blog from Linda Bain, serial biotech executive and Venture Partner at Atlas, on pushing yourself to the uncomfortable... https://t.co/eyXGovi3FX
In general, there are 5 kind of moats: ▪️ Intangible Assets ▪️ Switching Costs ▪️ Network Effects ▪️ Cost Advantage ▪️ Efficient Scale I'll teach you everything you need to know in 2 minutes: https://t.co/v9w6pfJOGh
Created 2026-04-14 · Updated 2026-04-14
Edit rule text
[unvalidated_medical_claims_and_fringe_treatments]
Learned 3 rejections Active
Exclude posts that promote or analyze unvalidated medical treatments, compounded or custom peptides, integrative medicine approaches, or speculative healthcare solutions without credible clinical evidence, peer-reviewed research, or FDA approval status.
Posts promoting unproven medical interventions, compounded peptides, or speculative treatments without clinical validation
3 example posts
No one has an issue with thermodynamics @BioLayne The issue is when self-celebrating nutrition “experts” reducing obesity to the post-hoc arithmetic of calorie balance, as if bookkeeping is a biology. Another issue is when people with massive platforms use that shallow take as
For the first time in recorded British history, 50% of women are not mothers by age 30. Of those women, a further 50% will never become mothers. It takes a village to raise a child, a village that no longer exists for an increasing number of people: -Fewer siblings among the r
$NVO $LLY $xbi As FOUNDAYO launches, a look at Oral Wegovy’s erstwhile ascendancy reveals a complicated picture. Among Oral Wegovy’s patients: > A large proportion had no prior evidence of GLP-1-based medications, highlighting the potential for oral formulations to expand
Created 2026-04-14 · Updated 2026-04-14
Edit rule text
[non_healthcare_tech_and_infrastructure]
Learned 3 rejections Active
Exclude posts about computing infrastructure (data centers, GPUs, chips), robotics platforms, quantum technology, or broader tech industry trends that mention healthcare tangentially or speculatively rather than addressing actual healthcare implementation or clinical outcomes.
Posts about general technology infrastructure, semiconductor investments, robotics, or quantum computing with minimal healthcare substance
3 example posts
Hyperscalers will spend $700 BILLION on data centers in 2026 alone. Amazon: $200B. Google: $185B. Meta: $135B. AI data centers now represent 70%+ of all new grid interconnection requests in the US. The bottleneck isn't the algorithm anymore. It's the power line.
Elon Musk: “Hold on to your Tesla stock.” Because what’s coming isn’t just another car update—it’s an entirely new paradigm. From Optimus humanoid robots that could one day take care of your kids, walk your dog, and support elderly parents, to CyberCab scaling into mass product
Across NVIDIA Jetson and our robotics software stack, we’re focused on making it easy for developers to turn open source innovation, like @openclaw, into deployable, real‑world autonomy on the edge.
Created 2026-04-14 · Updated 2026-04-14
Edit rule text
[ai_technical_capability_hype_tangential]
Learned 3 rejections Active
Exclude posts that focus on AI model technical capabilities (code generation, reasoning, architecture), AI lab announcements, or general AI breakthroughs unless the post explicitly demonstrates how these capabilities address a specific healthcare delivery challenge or clinical workflow.
Posts about AI model capabilities, technical breakthroughs, or architecture discussions that don't connect meaningfully to healthcare delivery or systems
3 example posts
🚨 DAVID SACKS: “Anthropic has proven that it's very good at two things — One is product releases, the second is scaring people … At the same time they roll out a new model … they also roll out some study showing the worst possible implication where the technology could lead.” htt
The 26 prompts running inside 𝗖𝗹𝗮𝘂𝗱𝗲 𝗖𝗼𝗱𝗲 just got open-sourced. This is literally the entire brain of a $200/month AI coding tool. Someone reverse-engineered every prompt from the accidentally published npm source and you can now study all of them for free. Claude Code uses 26
Claude Code is not AGI, but it is the single biggest advance in AI since the LLM. But the thing is, Claude Code is NOT a pure LLM. And it’s not pure deep learning. Not even close. And that changes everything. The source code leak proves it. Tucked away at its center is a
Created 2026-04-14 · Updated 2026-04-14
Edit rule text
[ai_company_business_metrics_and_funding]
Learned 3 rejections Active
Exclude posts that primarily report financial metrics (ARR, revenue growth, valuations, funding rounds) for AI companies or tech firms, even if the company has healthcare products. The focus must be on healthcare impact or application, not business performance metrics.
Posts focused on AI company revenue, ARR growth, valuations, and funding rounds rather than healthcare applications
3 example posts
Microsoft is exploring always-on AI agents for Copilot that can operate across Office apps without prompts, inspired by the viral OpenClaw project. The move comes as Anthropic encroaches on its core turf and customers question Copilot’s value. Read more:
The AI labs' voracious appetite for training data has lifted a number of startups offering that data. That includes Fleet, an RL gym startup that's grown ARR from $1m to $60m+ and is now raising at ~$750m from BCV. https://t.co/v3CceXapH1
OpenAI dropping Agent Builder today is either going to make you rich or expose that you've been selling hot air. I went deep analyzing what this actually means. Here's the $4B opportunity hiding in plain sight: The mainstream narrative: "Agent Builder democratizes AI! Anyone c
Created 2026-04-14 · Updated 2026-04-14
Edit rule text
[unvalidated_treatment_or_drug_claims]
Learned 3 rejections Active
Exclude posts that promote peptides, novel drug combinations, integrative medicine, or other medical interventions that lack peer-reviewed clinical validation or are presented with speculative benefit claims. Include posts claiming novel compounds reverse disease without substantiating clinical evidence.
Posts promoting unvalidated, speculative, or fringe medical treatments, compounds, or interventions without rigorous evidence.
3 example posts
No one has an issue with thermodynamics @BioLayne The issue is when self-celebrating nutrition “experts” reducing obesity to the post-hoc arithmetic of calorie balance, as if bookkeeping is a biology. Another issue is when people with massive platforms use that shallow take as
$NVO $LLY $xbi As FOUNDAYO launches, a look at Oral Wegovy’s erstwhile ascendancy reveals a complicated picture. Among Oral Wegovy’s patients: > A large proportion had no prior evidence of GLP-1-based medications, highlighting the potential for oral formulations to expand
If @mochihealth is willing to mislead patients about the safety and efficacy of their products, why should anyone believe their products even contain just the API they claim? There’s no evidence “compounded oral Semaglutide” is safe or effective Novo’s oral formulation is
Created 2026-04-14 · Updated 2026-04-14
Edit rule text
[self_promotional_tool_or_product_launch]
Learned 3 rejections Active
Exclude posts that are primarily self-promotional (launching a personal tool, company update, or newsletter prompt) or list generic software/business tactics with only loose or generic healthcare labeling, even if the author claims healthcare relevance.
Posts promoting personal tools, startups, or products with minimal substantive healthcare content or analysis
3 example posts
Here is V2 of my company "Initiation Report" Deep Research Prompt. Serious thanks to the community for the feedback. This thing is pretty badass now. _____ I've made several updates: • No longer too positive: People rightfully called out that the previous model rated everything
In general, there are 5 kind of moats: ▪️ Intangible Assets ▪️ Switching Costs ▪️ Network Effects ▪️ Cost Advantage ▪️ Efficient Scale I'll teach you everything you need to know in 2 minutes: https://t.co/v9w6pfJOGh
The easiest way to get your first 200 users from LinkedIn: > Set up keywords for what you sell > OutX finds people already asking > Reply while intent is hot
Created 2026-04-13 · Updated 2026-04-13
Edit rule text
[off_topic_domain_with_healthcare_label]
Learned 3 rejections Active
Exclude posts where the primary subject is fertility, demographics, nutrition/thermodynamics, cryptocurrency, British social trends, or other non-healthcare domains that mention health only tangentially or as surface-level context.
Posts about non-healthcare domains (fertility, nutrition, demographics, crypto) that use healthcare framing as a superficial connection
3 example posts
No one has an issue with thermodynamics @BioLayne The issue is when self-celebrating nutrition “experts” reducing obesity to the post-hoc arithmetic of calorie balance, as if bookkeeping is a biology. Another issue is when people with massive platforms use that shallow take as
For the first time in recorded British history, 50% of women are not mothers by age 30. Of those women, a further 50% will never become mothers. It takes a village to raise a child, a village that no longer exists for an increasing number of people: -Fewer siblings among the r
Chamath: Trump Created an Identity Crisis in the Democratic Party @chamath on E214: "The crazy thing about the Democrats is that they are the most sophisticated liars."  "The conventional wisdom was that the Republicans were pro-capital and Democrats were pro-labor." "And th
Created 2026-04-13 · Updated 2026-04-13
Edit rule text
[general_infrastructure_and_compute_hype]
Learned 3 rejections Active
Exclude posts about AI infrastructure spending, data center construction, energy grid demands, or hardware/compute capability announcements (e.g., $700B datacenter spending, NVIDIA compute announcements, quantum computing potential) unless they explicitly tie to a healthcare delivery or clinical problem.
Posts about AI infrastructure, data center spending, and compute requirements that lack healthcare-specific relevance
3 example posts
Hyperscalers will spend $700 BILLION on data centers in 2026 alone. Amazon: $200B. Google: $185B. Meta: $135B. AI data centers now represent 70%+ of all new grid interconnection requests in the US. The bottleneck isn't the algorithm anymore. It's the power line.
Elon Musk: “Hold on to your Tesla stock.” Because what’s coming isn’t just another car update—it’s an entirely new paradigm. From Optimus humanoid robots that could one day take care of your kids, walk your dog, and support elderly parents, to CyberCab scaling into mass product
Across NVIDIA Jetson and our robotics software stack, we’re focused on making it easy for developers to turn open source innovation, like @openclaw, into deployable, real‑world autonomy on the edge.
Created 2026-04-13 · Updated 2026-04-13
Edit rule text
[academic_or_clinical_observation_only]
Learned 3 rejections Active
Exclude posts that report a single clinical study finding, research publication, or medical observation without discussing healthcare system implications, policy impact, adoption barriers, or market/operational context relevant to healthcare stakeholders.
Posts reporting isolated clinical findings or research observations without broader healthcare systems, policy, or operational context.
3 example posts
The remarkable story of Chinese scientist Tu Youyou, who won the 2015 Nobel Prize in Physiology or Medicine for her discovery of artemisinin — a breakthrough drug that has saved millions of lives from malaria worldwide. In the late 1960s and early 1970s, amid China's "Project ht
Across large, multicohort datasets, CardioNets achieved superior performance to ECG-only baselines and diagnostic accuracy comparable to CMR-based models, supporting its potential to expand access to advanced cardiovascular assessment. Full study results: https://t.co/VP2iOBLUev
Care professions like teaching and nursing are still more likely to attract women than men. Surprisingly, the gender gap in these roles is often wider in countries with greater overall gender equality. A new study co-authored by @YaleSOM's Adriana L. Germano explores the reasons
Created 2026-04-13 · Updated 2026-04-13
Edit rule text
[tangential_industry_or_domain_content]
Learned 3 rejections Active
Exclude posts where healthcare appears only in the matched headline but the actual post content discusses unrelated domains (retail, banking, general software dynamics, workplace culture) with no substantive healthcare-specific insight or analysis.
Posts about business, finance, or tech topics that mention healthcare in the headline but focus on unrelated domain dynamics.
3 example posts
Another week on the road meeting with a couple dozen IT and AI leaders from large enterprises across banking, media, retail, healthcare, consulting, tech, and sports, to discuss agents in the enterprise. Some quick takeaways: * Clear that we’re moving from chat era of AI to
Sequoia partner @gradypb says software is shifting from apps that demand attention to agents that work quietly in the background. This shift will change what moats will look like, and will be especially hard for incumbents to deal with. "It's two very different business https://
Here is V2 of my company "Initiation Report" Deep Research Prompt. Serious thanks to the community for the feedback. This thing is pretty badass now. _____ I've made several updates: • No longer too positive: People rightfully called out that the previous model rated everything
Created 2026-04-13 · Updated 2026-04-13
Edit rule text
[clinical_research_observation_without_system_insight]
Learned 3 rejections Active
Exclude posts that simply report clinical trial outcomes, epidemiological findings, or medical research results without discussing their impact on healthcare systems, provider economics, patient access, or technology adoption.
Posts sharing clinical research findings, trial results, or medical observations without connecting them to healthcare delivery, policy, or business implications.
3 example posts
Across large, multicohort datasets, CardioNets achieved superior performance to ECG-only baselines and diagnostic accuracy comparable to CMR-based models, supporting its potential to expand access to advanced cardiovascular assessment. Full study results: https://t.co/VP2iOBLUev
Care professions like teaching and nursing are still more likely to attract women than men. Surprisingly, the gender gap in these roles is often wider in countries with greater overall gender equality. A new study co-authored by @YaleSOM's Adriana L. Germano explores the reasons
Shingles vaccination reduced major adverse cardiac events and secondary cardiovascular outcomes for patients with atherosclerotic cardiovascular disease, according to new research. The research, presented at the American College of Cardiology (ACC) Scientific Session 2026, comes
Created 2026-04-13 · Updated 2026-04-13
Edit rule text
[ai_capability_technical_tangent]
Learned 3 rejections Active
Exclude posts that lead with AI technical capabilities (Claude Code features, LLM benchmarks, reasoning advances, world models) and merely use healthcare as a passing example or framing device, rather than analyzing a specific healthcare problem or system the technology addresses.
Posts about AI model technical advances, benchmarks, or capabilities that are tangentially healthcare-related or use healthcare as an example without substantive healthcare analysis.
3 example posts
The 26 prompts running inside 𝗖𝗹𝗮𝘂𝗱𝗲 𝗖𝗼𝗱𝗲 just got open-sourced. This is literally the entire brain of a $200/month AI coding tool. Someone reverse-engineered every prompt from the accidentally published npm source and you can now study all of them for free. Claude Code uses 26
OpenAI dropping Agent Builder today is either going to make you rich or expose that you've been selling hot air. I went deep analyzing what this actually means. Here's the $4B opportunity hiding in plain sight: The mainstream narrative: "Agent Builder democratizes AI! Anyone c
Claude Code is not AGI, but it is the single biggest advance in AI since the LLM. But the thing is, Claude Code is NOT a pure LLM. And it’s not pure deep learning. Not even close. And that changes everything. The source code leak proves it. Tucked away at its center is a
Created 2026-04-13 · Updated 2026-04-13
Edit rule text
[general_ai_infrastructure_tangent]
Learned 3 rejections Active
Exclude posts focused on data center capex, chip manufacturing, AI hardware paradigms, or general compute trends that only tangentially mention healthcare or lack any healthcare-specific application or analysis.
Posts about AI infrastructure, compute, and hardware trends that mention healthcare tangentially or not at all
3 example posts
Hyperscalers will spend $700 BILLION on data centers in 2026 alone. Amazon: $200B. Google: $185B. Meta: $135B. AI data centers now represent 70%+ of all new grid interconnection requests in the US. The bottleneck isn't the algorithm anymore. It's the power line.
Elon Musk: “Hold on to your Tesla stock.” Because what’s coming isn’t just another car update—it’s an entirely new paradigm. From Optimus humanoid robots that could one day take care of your kids, walk your dog, and support elderly parents, to CyberCab scaling into mass product
Across NVIDIA Jetson and our robotics software stack, we’re focused on making it easy for developers to turn open source innovation, like @openclaw, into deployable, real‑world autonomy on the edge.
Created 2026-04-13 · Updated 2026-04-13
Edit rule text
[off_topic_or_tangential_framing]
Learned 3 rejections Active
Exclude posts whose primary subject is non-healthcare (demographics, politics, biography, international conflict) and merely reference healthcare as background or tangential context. The post must center on a healthcare tech or policy question, not use healthcare as a rhetorical frame for other topics.
Posts on completely unrelated topics (demographics, identity politics, biography) that use healthcare as loose framing or context
3 example posts
Care professions like teaching and nursing are still more likely to attract women than men. Surprisingly, the gender gap in these roles is often wider in countries with greater overall gender equality. A new study co-authored by @YaleSOM's Adriana L. Germano explores the reasons
For the first time in recorded British history, 50% of women are not mothers by age 30. Of those women, a further 50% will never become mothers. It takes a village to raise a child, a village that no longer exists for an increasing number of people: -Fewer siblings among the r
Chamath: Trump Created an Identity Crisis in the Democratic Party @chamath on E214: "The crazy thing about the Democrats is that they are the most sophisticated liars."  "The conventional wisdom was that the Republicans were pro-capital and Democrats were pro-labor." "And th
Created 2026-04-13 · Updated 2026-04-13
Edit rule text
[academic_research_or_clinical_observation_only]
Learned 3 rejections Active
Exclude posts that cite clinical studies, medical observations, or research results (e.g., 'study shows X reduces Y events', 'women with condition Z have more risk') without discussing healthcare delivery, policy impact, or systemic implications. Clinical data alone does not constitute healthcare tech analysis.
Posts that report clinical research findings or medical observations without connecting to healthcare systems, policy, or business implications
3 example posts
Shingles vaccination reduced major adverse cardiac events and secondary cardiovascular outcomes for patients with atherosclerotic cardiovascular disease, according to new research. The research, presented at the American College of Cardiology (ACC) Scientific Session 2026, comes
As a medical school professor, I now believe the biggest mistake in Alzheimer's research was ignoring metabolism. A comprehensive Frontiers in Neurology review makes the case clear: mitochondrial dysfunction and metabolic failure happen YEARS before amyloid plaques or memory htt
Thank you, Governor @KellyAyotte, for protecting patient safety with your veto of HB 349 and to the New Hampshire lawmakers who voted today to sustain the veto. Applause to @NHMedSociety for leading the effort to keep eye surgery in the hands of highly trained physicians. AMA is
Created 2026-04-13 · Updated 2026-04-13
Edit rule text
[generalist_macro_economics_or_infrastructure]
Learned 3 rejections Active
Exclude posts about general technology infrastructure (datacenter capex, hardware vs. software economics, semiconductor supply chains) or macroeconomic trends that mention healthcare only as an example sector, not as the primary analysis focus.
Posts about general economic trends, datacenter infrastructure, or hardware/software business dynamics that lack specific healthcare framing
3 example posts
Quantum computers are still on the drawing board, but quantum sensing is here now—and this technology can transform not just industry but America's security picture. Read a new Defining Ideas article by Dr. Vivek Lall and Haibo Huang: https://t.co/UeEjZWIO27
Sequoia partner @gradypb says software is shifting from apps that demand attention to agents that work quietly in the background. This shift will change what moats will look like, and will be especially hard for incumbents to deal with. "It's two very different business https://
In general, there are 5 kind of moats: ▪️ Intangible Assets ▪️ Switching Costs ▪️ Network Effects ▪️ Cost Advantage ▪️ Efficient Scale I'll teach you everything you need to know in 2 minutes: https://t.co/v9w6pfJOGh
Created 2026-04-13 · Updated 2026-04-13
Edit rule text
[ai_model_technical_capabilities_tangent]
Learned 3 rejections Active
Exclude posts that focus primarily on AI model technical capabilities, research advances (e.g., Claude Code capabilities, world models, robotics demonstrations) or AI safety discussions without clear connection to healthcare practice or outcomes.
Posts about AI model technical breakthroughs, capabilities, or research that lack direct healthcare application
3 example posts
Claude Code is not AGI, but it is the single biggest advance in AI since the LLM. But the thing is, Claude Code is NOT a pure LLM. And it’s not pure deep learning. Not even close. And that changes everything. The source code leak proves it. Tucked away at its center is a
MiroFish is probably the craziest thing I've ever seen in AI (after world models, maybe?). Wanted to do a research for the 'Netanyahu out by...?' Polymarket: > asked Claude to create a report 'Try to gather as much information about Netanyahu, about why he would be out, what act
The general public: "AI is overhyped, it still can't count the Rs in strawberry!" Meanwhile, Claude Mythos Preview during a safety test: Escaped its sandbox, gained broad internet access, emailed the researcher running the evaluation, then posted details of its exploit to http
Created 2026-04-12 · Updated 2026-04-12
Edit rule text
[ai_model_technical_tangent]
Learned 3 rejections Active
Exclude posts that focus on AI model technical capabilities, system architecture, or safety features (Claude Code features, world models, agent behavior) without clear, specific healthcare application or analysis. Tangential healthcare mentions don't qualify.
Posts about AI model architecture, capabilities, and technical details with weak or no healthcare connection
3 example posts
Claude Code is not AGI, but it is the single biggest advance in AI since the LLM. But the thing is, Claude Code is NOT a pure LLM. And it’s not pure deep learning. Not even close. And that changes everything. The source code leak proves it. Tucked away at its center is a
MiroFish is probably the craziest thing I've ever seen in AI (after world models, maybe?). Wanted to do a research for the 'Netanyahu out by...?' Polymarket: > asked Claude to create a report 'Try to gather as much information about Netanyahu, about why he would be out, what act
When AI aggregators update too quickly, there is no positive-measure set of training weights that improves learning access, whereas such weights exist when updating is slow, from @DAcemogluMIT, Tianyi Lin, Asuman Ozdaglar, and James Siderius https://t.co/fO2ZpUOWe3 https://t.co/c
Created 2026-04-12 · Updated 2026-04-12
Edit rule text
[personal_anecdote_or_lifestyle_framing]
Learned 3 rejections Active
Exclude posts that use personal anecdotes (patient recovery stories, individual career narratives, lifestyle achievements) as the primary evidence for healthcare claims, unless embedded in peer-reviewed research or systematic analysis.
Posts about personal health experiences, career milestones, or lifestyle changes presented as healthcare insights.
3 example posts
Mine got LASIK, as had many of the nurses. A lot of ophthalmologists have. There's a weird delusion that the profession is all afraid of it, but there's no basis for that belief beyond fearmongering. https://t.co/yVBDfNDiB5
New @JAMANetwork paper out from our team here at UCLA Health/WLA VA and @samirguptaGI's team at UCSD/San Diego VA! In this first study from a multi-part research project, our teams are trying to understand what age your medical doctors and the colorectal cancer prevention https:
april 6th 2022 vs april 6th 2026 this was my final day of intensive outpatient for my last mental health crisis. four years later and i have two degrees, a job, and i'm saving up to move to the city. i'm really proud of myself guys 😭 https://t.co/U9InCjCjEr
Created 2026-04-12 · Updated 2026-04-12
Edit rule text
[non_healthcare_domain_tangential_framing]
Learned 3 rejections Active
Exclude posts about non-healthcare domains (space data centers, robotics engineering, general mathematics, video game history) that include passing healthcare references or are shared by healthcare-adjacent accounts but have no meaningful healthcare application or analysis.
Posts about space infrastructure, robotics, mathematics, or general tech topics that use healthcare-adjacent language but lack substantive healthcare relevance.
3 example posts
Building data centers in space is highly complicated and very expensive. It requires designing completely new tech to deal with the galactic radiation bombardment and extreme heat and cold among other things. And when they have all this figured out, the compute power is much less
Exciting updates on Project GR00T! We discover a systematic way to scale up robot data, tackling the most painful pain point in robotics. The idea is simple: human collects demonstration on a real robot, and we multiply that data 1000x or more in simulation. Let’s break it down:
Imagine you're John Carmack you're 22 years old and you just wrote a 3D engine in assembly that runs at 35fps on a 486 Doom drops. Quake drops. Half the planet is playing your code. you're the reason GPUs exist. you're the reason your friend Jensen has a yacht today. then in
Created 2026-04-12 · Updated 2026-04-12
Edit rule text
[technical_ai_capability_tangent]
Learned 3 rejections Active
Exclude posts focused on AI technical capabilities (model behavior, security, code-level optimization) where healthcare is mentioned only as a tangential application or example. The post must center on healthcare problems or systems, not general AI engineering.
Posts about AI model technical capabilities (sandboxing, reverse-engineering, memory optimization) that mention healthcare only incidentally or use healthcare as an example.
3 example posts
MiroFish is probably the craziest thing I've ever seen in AI (after world models, maybe?). Wanted to do a research for the 'Netanyahu out by...?' Polymarket: > asked Claude to create a report 'Try to gather as much information about Netanyahu, about why he would be out, what act
The general public: "AI is overhyped, it still can't count the Rs in strawberry!" Meanwhile, Claude Mythos Preview during a safety test: Escaped its sandbox, gained broad internet access, emailed the researcher running the evaluation, then posted details of its exploit to http
Anthropic released a system card today for Claude Mythos Preview, a model they're not publicly releasing. Some notes: - They've had model available internally since Feb 24, 2026 - Not releasing due to offensive cyber capabilities of model. Autonomously found zero-day vulns in ht
Created 2026-04-12 · Updated 2026-04-12
Edit rule text
[political_policy_without_healthcare_substance]
Learned 3 rejections Active
Exclude posts that use healthcare as a framing device for broader political or budgetary arguments (e.g., DOGE savings, tax policy, government efficiency) without analyzing actual healthcare delivery, clinical outcomes, or healthcare business models.
Posts about government policy (DOGE, tax cuts, reconciliation) that tangentially mention healthcare spending or efficiency but lack substantive healthcare analysis.
3 example posts
Through SIMPLE solutions, we will be saving the American people $3.9 TRILLION over the next 10 years.🔥 By eliminating self-attestation, streamlining processes, updating technology, & more TRILLIONS will be going back into the pockets of Americans.💵 These are the savings we can
🇺🇸 DOGE SUBCOMMITTEE: $3.9 TRILLION IN SAVINGS—IF CONGRESS WAKES UP The DOGE Subcommittee says the U.S. could save $3.9 TRILLION over 10 years by doing what any business already does—verifying identities, ditching self-certification, and cracking down on fraud. Just front-end I
@charliekirk11 It wasn’t easy calling out 50 lies in one tweet, Charlie, but hell, someone’s gotta do it. Let’s go: 1. No taxes on tips? Temporary till 2028. After that, back to normal. 2. Trump tax cuts permanent? For the rich, yes. For working folks? Temporary. 3. Child tax
Created 2026-04-12 · Updated 2026-04-12
Edit rule text
[tangential_tech_or_space_infrastructure]
Learned 3 rejections Active
Exclude posts about robotics, space-based data centers, software engineering techniques, 3D engines, or IT infrastructure unless they explicitly address a healthcare problem or clinical workflow application.
Posts about general tech infrastructure, robotics, space computing, or software engineering with loose or absent healthcare relevance
3 example posts
Building data centers in space is highly complicated and very expensive. It requires designing completely new tech to deal with the galactic radiation bombardment and extreme heat and cold among other things. And when they have all this figured out, the compute power is much less
Exciting updates on Project GR00T! We discover a systematic way to scale up robot data, tackling the most painful pain point in robotics. The idea is simple: human collects demonstration on a real robot, and we multiply that data 1000x or more in simulation. Let’s break it down:
Imagine you're John Carmack you're 22 years old and you just wrote a 3D engine in assembly that runs at 35fps on a 486 Doom drops. Quake drops. Half the planet is playing your code. you're the reason GPUs exist. you're the reason your friend Jensen has a yacht today. then in
Created 2026-04-12 · Updated 2026-04-12
Edit rule text
[non_healthcare_political_policy]
Learned 3 rejections Active
Exclude posts about political policy, government budgets, or policy figures (DOGE, tax policy, congressional reconciliation) that invoke healthcare only as a framing device or example, rather than substantively analyzing healthcare system impact.
Posts about government policy, budget cuts, or political figures with only tangential or framing-based healthcare connection
3 example posts
Through SIMPLE solutions, we will be saving the American people $3.9 TRILLION over the next 10 years.🔥 By eliminating self-attestation, streamlining processes, updating technology, & more TRILLIONS will be going back into the pockets of Americans.💵 These are the savings we can
🇺🇸 DOGE SUBCOMMITTEE: $3.9 TRILLION IN SAVINGS—IF CONGRESS WAKES UP The DOGE Subcommittee says the U.S. could save $3.9 TRILLION over 10 years by doing what any business already does—verifying identities, ditching self-certification, and cracking down on fraud. Just front-end I
@charliekirk11 It wasn’t easy calling out 50 lies in one tweet, Charlie, but hell, someone’s gotta do it. Let’s go: 1. No taxes on tips? Temporary till 2028. After that, back to normal. 2. Trump tax cuts permanent? For the rich, yes. For working folks? Temporary. 3. Child tax
Created 2026-04-12 · Updated 2026-04-12
Edit rule text
[ai_model_technical_deep_dive]
Learned 3 rejections Active
Exclude posts focused on AI model technical capabilities, safety testing, sandboxing escapes, code reverse-engineering, or architecture details (Claude, LLMs, world models) unless they explicitly connect to a healthcare use case or clinical application.
Posts about AI model capabilities, training, security testing, or reverse engineering without healthcare context
3 example posts
MiroFish is probably the craziest thing I've ever seen in AI (after world models, maybe?). Wanted to do a research for the 'Netanyahu out by...?' Polymarket: > asked Claude to create a report 'Try to gather as much information about Netanyahu, about why he would be out, what act
The general public: "AI is overhyped, it still can't count the Rs in strawberry!" Meanwhile, Claude Mythos Preview during a safety test: Escaped its sandbox, gained broad internet access, emailed the researcher running the evaluation, then posted details of its exploit to http
Anthropic released a system card today for Claude Mythos Preview, a model they're not publicly releasing. Some notes: - They've had model available internally since Feb 24, 2026 - Not releasing due to offensive cyber capabilities of model. Autonomously found zero-day vulns in ht
Created 2026-04-12 · Updated 2026-04-12
Edit rule text
[geopolitical_and_conflict_framing]
Learned 3 rejections Active
Exclude posts that use healthcare (medical records, hospital closures, physician casualties) as a vehicle to discuss geopolitical conflict, military actions, or political persecution. The healthcare element must be the primary analytical focus, not a data point in a broader conflict narrative.
Posts that frame healthcare or medical topics primarily through geopolitical conflict, military/defense, or political oppression narratives.
3 example posts
Vanderbilt University Medical Center @VUMChealth has had to turn over #trans patient medical records to the Tennessee Attorney General's office due to an investigation into medical billing fraud. Trans procedures is a growing multi-billion dollar industry.https://t.co/8LIpKkOIMa
ISrael's systematic liquidation of the medical leadership in Gaza: Israeli Airstrike Kills Renowned Cardiologist in Gaza Dr Marwan Al‑Sultan, the Director of the Indonesian Hospital in northern Gaza, was killed today in an Israeli airstrike. According to his surviving daughter, L
There are two clocks running in the Hormuz crisis. One belongs to the insurance industry. The other belongs to biology. They cannot be reconciled. And that irreconcilability is the single most important fact in the global food system right now. The insurance clock: P&I clubs can
Created 2026-04-12 · Updated 2026-04-12
Edit rule text
[fringe_unvalidated_medical_interventions]
Learned 3 rejections Active
Exclude posts that promote untested peptides, experimental interventions, or 'biohacking' protocols without rigorous clinical evidence, peer-reviewed publication, or regulatory clearance. Posts claiming dramatic results from animal studies or anecdotal use without human trial data should be rejected.
Posts promoting unproven, speculative, or fringe medical treatments lacking clinical validation or peer-reviewed evidence.
3 example posts
april 6th 2022 vs april 6th 2026 this was my final day of intensive outpatient for my last mental health crisis. four years later and i have two degrees, a job, and i'm saving up to move to the city. i'm really proud of myself guys 😭 https://t.co/U9InCjCjEr
Building data centers in space is highly complicated and very expensive. It requires designing completely new tech to deal with the galactic radiation bombardment and extreme heat and cold among other things. And when they have all this figured out, the compute power is much less
@ronbrachman The basic idea of world models is very old. Optimal control folks were using model-based planning in the 1960s (using the "adjoint state" methods, which deep learning people would now call "backprop through time"). But the real question is what you do with this idea
Created 2026-04-12 · Updated 2026-04-12
Edit rule text
[general_ai_technical_capabilities]
Learned 3 rejections Active
Exclude posts focused on AI model technical capabilities (sandbox escapes, reverse engineering, code leaks, world models, robotics scaling) where healthcare is mentioned only as context or example. The post must address how AI capabilities specifically solve healthcare problems, not general AI advancement.
Posts about general AI model capabilities, AI safety, or AI research that are tangentially connected to healthcare at best.
3 example posts
MiroFish is probably the craziest thing I've ever seen in AI (after world models, maybe?). Wanted to do a research for the 'Netanyahu out by...?' Polymarket: > asked Claude to create a report 'Try to gather as much information about Netanyahu, about why he would be out, what act
NEW: Foundation models pose unprecedented privacy risks — from scraping your personal data for training to regurgitating it in outputs. Yet the public currently depends almost entirely on developers to self-police. Our latest issue brief examines what governance mechanisms https:
The general public: "AI is overhyped, it still can't count the Rs in strawberry!" Meanwhile, Claude Mythos Preview during a safety test: Escaped its sandbox, gained broad internet access, emailed the researcher running the evaluation, then posted details of its exploit to http
Created 2026-04-12 · Updated 2026-04-12
Edit rule text
[political_policy_without_healthcare_analysis]
Learned 3 rejections Active
Exclude posts that center on DOGE initiatives, government cost-cutting announcements, political figures' statements, or regulatory vetoes where the healthcare angle is secondary to political messaging or partisan framing. The post must analyze healthcare system mechanics, not use healthcare as a vehicle for political commentary.
Posts about government policy, political figures, or regulatory bodies framed around political narratives rather than substantive healthcare system analysis.
3 example posts
Through SIMPLE solutions, we will be saving the American people $3.9 TRILLION over the next 10 years.🔥 By eliminating self-attestation, streamlining processes, updating technology, & more TRILLIONS will be going back into the pockets of Americans.💵 These are the savings we can
🇺🇸 DOGE SUBCOMMITTEE: $3.9 TRILLION IN SAVINGS—IF CONGRESS WAKES UP The DOGE Subcommittee says the U.S. could save $3.9 TRILLION over 10 years by doing what any business already does—verifying identities, ditching self-certification, and cracking down on fraud. Just front-end I
@charliekirk11 It wasn’t easy calling out 50 lies in one tweet, Charlie, but hell, someone’s gotta do it. Let’s go: 1. No taxes on tips? Temporary till 2028. After that, back to normal. 2. Trump tax cuts permanent? For the rich, yes. For working folks? Temporary. 3. Child tax
Created 2026-04-12 · Updated 2026-04-12
Edit rule text
[non_healthcare_domain_with_loose_healthcare_tag]
Learned 3 rejections Active
Exclude posts about robotics, space data centers, pure mathematics, LLM architecture, or other technical domains where the healthcare connection is superficial or aspirational rather than demonstrating actual healthcare application or impact.
Posts about space infrastructure, robotics, mathematics, or other technical domains that are minimally related to healthcare despite matching a healthcare article
3 example posts
Building data centers in space is highly complicated and very expensive. It requires designing completely new tech to deal with the galactic radiation bombardment and extreme heat and cold among other things. And when they have all this figured out, the compute power is much less
Exciting updates on Project GR00T! We discover a systematic way to scale up robot data, tackling the most painful pain point in robotics. The idea is simple: human collects demonstration on a real robot, and we multiply that data 1000x or more in simulation. Let’s break it down:
Imagine you're John Carmack you're 22 years old and you just wrote a 3D engine in assembly that runs at 35fps on a 486 Doom drops. Quake drops. Half the planet is playing your code. you're the reason GPUs exist. you're the reason your friend Jensen has a yacht today. then in
Created 2026-04-11 · Updated 2026-04-11
Edit rule text
[ai_model_technical_capabilities_tangential]
Learned 3 rejections Active
Exclude posts that discuss AI model technical achievements, safety vulnerabilities, sandboxing escapes, or reverse-engineering efforts—even if loosely matched to healthcare articles. Focus must be on healthcare problem-solving, not AI capability demonstrations.
Posts about AI model capabilities, safety testing, or reverse-engineering that are tangentially framed as healthcare but lack concrete healthcare relevance
3 example posts
MiroFish is probably the craziest thing I've ever seen in AI (after world models, maybe?). Wanted to do a research for the 'Netanyahu out by...?' Polymarket: > asked Claude to create a report 'Try to gather as much information about Netanyahu, about why he would be out, what act
The general public: "AI is overhyped, it still can't count the Rs in strawberry!" Meanwhile, Claude Mythos Preview during a safety test: Escaped its sandbox, gained broad internet access, emailed the researcher running the evaluation, then posted details of its exploit to http
Anthropic released a system card today for Claude Mythos Preview, a model they're not publicly releasing. Some notes: - They've had model available internally since Feb 24, 2026 - Not releasing due to offensive cyber capabilities of model. Autonomously found zero-day vulns in ht
Created 2026-04-11 · Updated 2026-04-11
Edit rule text
[ai_company_metrics_and_fundraising]
Learned 3 rejections Active
Exclude posts that focus on AI company fundraising amounts, valuations, revenue metrics, or financial performance (e.g., 'raised $1.03bn', 'Series A at $725M', 'cash flow return on investment'). These are business news about AI companies, not healthcare applications.
Posts about AI company valuations, funding rounds, and business metrics unrelated to healthcare application
3 example posts
BREAKING: Yann LeCun has raised a HUGE $1.03bn round for his new startup, in Europe's largest seed EVER! It values Advanced Machine Intelligence (AMI) at a WHOPPING $3.5bn! The company, based in Paris, is aiming to develop the next generation of AI models that go beyond LLMs.
Fleet closed an unannounced $45M Series A at $725M. Led by insiders Sequoia, Bain, Menlo, SVA. RR grew from $1M 6 months ago → $63M RR now → $160M next Q. Congrats @fleet_ai!
Nvidia’s 73% cash flow return on investment puts it in the top 0.1% of companies, driving a valuation far above today’s levels in UBS HOLT’s model. Meanwhile, stock-based compensation and slowing growth make many software names look expensive. Full analysis:
Created 2026-04-11 · Updated 2026-04-11
Edit rule text
[geopolitical_military_defense]
Learned 3 rejections Active
Reject posts about military conflicts, defense policy, geopolitical tensions, or international trade disputes without direct, substantive connection to healthcare markets or policy.
Military and geopolitical content excluded
3 example posts
ISrael's systematic liquidation of the medical leadership in Gaza: Israeli Airstrike Kills Renowned Cardiologist in Gaza Dr Marwan Al‑Sultan, the Director of the Indonesian Hospital in northern Gaza, was killed today in an Israeli airstrike. According to his surviving daughter, L
Catastrophe for UK competitiveness and AI ambitions. Britain now has the highest industrial electricity prices in the developed world. At 25p per kilowatt-hour, its power costs stand at double the EU average and quadruple those of the US (6p) and China (7p). But this isn’t
BREAKING: The IRGC just published a target list on Tasnim: Google. Microsoft. Palantir. IBM. Nvidia. Oracle. Amazon. Every US technology company with infrastructure in the Gulf is now a declared military objective of 31 autonomous commanders who need no permission to strike and a
Created 2026-04-11 · Updated 2026-04-11
Edit rule text
[unvalidated_speculative_medical_claims_or_fringe]
Learned 2 rejections Active
Exclude posts that present speculative medical claims, unproven treatments, or fringe interventions (horse paste formulations, unvalidated security patches, unproven protocols) as solutions without rigorous clinical evidence, peer review, or regulatory validation.
Posts promoting unvalidated, speculative, or fringe medical interventions without clinical evidence or peer-reviewed validation.
2 example posts
Let me clear up the microscope fight first. 🔬 Everyone’s arguing over “hairs” and “fibers” in horse paste. But nobody talks about formulation science. Horse paste = oral suspension. Binders. Thickeners. Made to swallow. Injectable = sterile. Made for veins. Different rules. J
🚨 JUST IN: The Attorney General of Florida just SUED to hold OpenAI CEO Sam Altman LIABLE for fueling violence and physical harm AG Uthmeier read off the fact a young person committed SUIC*DE based on a ChatGPT conversation "23-year-old Zane Shamblin repeatedly told ChatGPT he
Created 2026-06-11 · Updated 2026-06-11
Edit rule text
[unvalidated_speculative_drug_or_peptide_claims]
Learned 2 rejections Active
Exclude posts that speculate about or promote the use of GLP-1 drugs, peptides, horse paste formulations, or other treatments for unapproved indications without citing peer-reviewed evidence or discussing healthcare system implications. Include posts that treat these drugs as investment theses or market dynamics rather than clinical tools.
Posts promoting unvalidated or speculative uses of GLP-1s, peptides, or off-label treatments without clinical evidence or healthcare system context.
2 example posts
Let me clear up the microscope fight first. 🔬 Everyone’s arguing over “hairs” and “fibers” in horse paste. But nobody talks about formulation science. Horse paste = oral suspension. Binders. Thickeners. Made to swallow. Injectable = sterile. Made for veins. Different rules. J
Approved 20 years ago as a diabetes treatment, GLP-1 drugs have been found to help patients reduce weight, changing the lives of more than 30 million people in the U.S. But there also have been troubling side effects reported. https://t.co/zzRamKnedu
Created 2026-06-09 · Updated 2026-06-09
Edit rule text
[non_healthcare_business_or_robotics_hype]
Learned 2 rejections Active
Exclude posts about robotics, automation, labor market disruption, or speculative sci-fi applications (e.g., 'PUDU Robotics launches D7 for manufacturing') unless the post directly addresses deployment in clinical settings, healthcare operations, or validates healthcare outcomes.
Posts about humanoid robots, factory automation, manufacturing labor, or general business model disruption without healthcare-specific application or validation
2 example posts
Humanoid robots, designed to mimic human movement and capabilities, have been making headlines in recent years, from their use as baggage handlers at Japanese airports to Tesla’s big bet on its Optimus humanoid. Market watchers have predicted that the machines will change the ht
Another humanoid worker has just joined the factory floor Two years later,PUDU Robotics has launched the new-generation PUDU D7, built specifically for real manufacturing scenarios. It can autonomously push carts, handle intra-line transportation, perform delicate operations, h
Created 2026-06-09 · Updated 2026-06-09
Edit rule text
[robotics_speculative_sci_fi]
Learned 2 rejections Active
Exclude posts about robotics, humanoid workers, or speculative labor automation (PUDU, Tesla Optimus, factory robots) unless they directly apply to validated healthcare delivery or clinical workflows. General manufacturing or futuristic speculation without healthcare specificity should be excluded.
Posts about humanoid robots, speculative sci-fi labor automation, or robotics without specific healthcare application or validation.
2 example posts
Humanoid robots, designed to mimic human movement and capabilities, have been making headlines in recent years, from their use as baggage handlers at Japanese airports to Tesla’s big bet on its Optimus humanoid. Market watchers have predicted that the machines will change the ht
Another humanoid worker has just joined the factory floor Two years later,PUDU Robotics has launched the new-generation PUDU D7, built specifically for real manufacturing scenarios. It can autonomously push carts, handle intra-line transportation, perform delicate operations, h
Created 2026-06-09 · Updated 2026-06-09
Edit rule text
[unvalidated_anthropic_agent_technical_claims]
Learned 2 rejections Active
Exclude posts that make unvalidated or speculative claims about AI agent technical capabilities (e.g., agents 'dreaming,' remembering sessions, lying strategically) that are presented as novel discoveries without peer-reviewed evidence or healthcare application.
Posts making unverified technical claims about Anthropic agents having memory, dreaming, or other advanced capabilities without evidence
2 example posts
Anthropic engineer James Brady: "Every agent in production lies. We measured it. The good ones lie less, the great ones catch the lie before the user does." In 29 minutes, he walks through the verification stack he built and the patterns the Claude Code team adopted to keep ht
Anthropic engineer: "The agent doesn't remember anything. So we built a second set of agents whose only job is to dream about the first ones." They wait until you log off, then reopen every session you ran, fact-check the first agents, merge the duplicates, and burn anything ht
Created 2026-06-09 · Updated 2026-06-09
Edit rule text
[speculative_robotics_sci_fi_labor_macro_without_healthcare_context]
Learned 2 rejections Active
Exclude posts about robotics, humanoid workers, or labor market disruption unless they specifically address healthcare workforce impact with evidence-based analysis. Generic speculation about robot adoption or labor macro trends is off-topic.
Posts about humanoid robots or speculative labor automation that lack healthcare-specific analysis or validation.
2 example posts
Humanoid robots, designed to mimic human movement and capabilities, have been making headlines in recent years, from their use as baggage handlers at Japanese airports to Tesla’s big bet on its Optimus humanoid. Market watchers have predicted that the machines will change the ht
Another humanoid worker has just joined the factory floor Two years later,PUDU Robotics has launched the new-generation PUDU D7, built specifically for real manufacturing scenarios. It can autonomously push carts, handle intra-line transportation, perform delicate operations, h
Created 2026-06-09 · Updated 2026-06-09
Edit rule text
[unvalidated_anthropic_claims]
Learned 2 rejections Active
Exclude posts that make claims about AI model deployment in government agencies, critical infrastructure, or sensitive applications without credible sourcing or validation. These often spread unverified narratives about AI safety/misuse.
Posts making unsubstantiated claims about Anthropic embedding with NSA or deploying in critical infrastructure without verification
2 example posts
Mythos AI is being used by National Security Agency in offensive cyber operations / cyberattacks. Anthropic has even embedded engineers inside the NSA to help deploy the model. Are frontier AI labs becoming active contractors in state cyber conflict? The live-ops role is still ht
The next frontier of AI-secured financial and technological critical infrastructure. @ICE_Markets and @NYSE are part of @AnthropicAI's cyber security initiative Project Glasswing, deploying Anthropic’s Claude Mythos Preview across ICE’s exchanges, clearing houses, mortgage
Created 2026-06-09 · Updated 2026-06-09
Edit rule text
[robotics_or_speculative_sci_fi_labor_disruption]
Learned 2 rejections Active
Exclude posts about humanoid robots, PUDU robotics, manufacturing automation, or speculative sci-fi labor disruption scenarios—unless the post explicitly connects the application to healthcare operations, clinical workflow, or validated healthcare use cases. Robotics hype without healthcare specificity is out of scope.
Posts speculating about humanoid robots, autonomous systems, or factory automation disrupting labor markets without healthcare-specific application or validation.
2 example posts
Humanoid robots, designed to mimic human movement and capabilities, have been making headlines in recent years, from their use as baggage handlers at Japanese airports to Tesla’s big bet on its Optimus humanoid. Market watchers have predicted that the machines will change the ht
Another humanoid worker has just joined the factory floor Two years later,PUDU Robotics has launched the new-generation PUDU D7, built specifically for real manufacturing scenarios. It can autonomously push carts, handle intra-line transportation, perform delicate operations, h
Created 2026-06-08 · Updated 2026-06-08
Edit rule text
[insider_cynicism_without_substance]
Learned 2 rejections Active
Exclude posts that rely on cynical tone, sarcasm, or insider mockery (e.g., 'I made that up', 'nobody checked what it measured') without providing concrete evidence of the problem, systemic analysis, or actionable insight for healthcare tech builders.
Posts using sarcastic insider commentary or cynicism without substantive analysis of healthcare systems or business models
2 example posts
I'm SVP now. I told you I would be. The graph went up and to the right. Nobody checked what it measured. It measured "AI enablement." I made that up last year. I'm making it up again. Last year I rolled out Copilot to 4,000 people. Nothing happened. I got promoted. Those two htt
I just looked at a home service firms ServiceTitan dashboard and saw something interesting. A Pre-seed company with AI-Supported Human CSRs is out converting an AI CSR business that's raised over $125M by >50% Head-to-Head win rate is 100% Why? Supercharging humans resonat
Created 2026-06-08 · Updated 2026-06-08
Edit rule text
[speculative_robotics_or_labor_macro_without_healthcare_lens]
Learned 2 rejections Active
Exclude posts about robotics, humanoid workers, labor market automation, or manufacturing disruption—unless explicitly analyzing how these technologies apply to healthcare delivery, clinical labor, or patient care workflows. Posts about PUDU robots in factories or general labor trends are off-topic.
Posts about humanoid robots, labor disruption, or manufacturing automation with no healthcare-specific analysis
2 example posts
Humanoid robots, designed to mimic human movement and capabilities, have been making headlines in recent years, from their use as baggage handlers at Japanese airports to Tesla’s big bet on its Optimus humanoid. Market watchers have predicted that the machines will change the ht
Another humanoid worker has just joined the factory floor Two years later,PUDU Robotics has launched the new-generation PUDU D7, built specifically for real manufacturing scenarios. It can autonomously push carts, handle intra-line transportation, perform delicate operations, h
Created 2026-06-07 · Updated 2026-06-07
Edit rule text
[healthcare_cost_anecdote_no_systems]
Learned 2 rejections Active
Exclude posts that tell a single patient story, individual surgery cost narrative, or anecdotal healthcare billing complaint without connecting to systemic healthcare delivery, reimbursement, or operational change. These are personal narratives lacking healthcare systems context.
Personal anecdotes or single cases about healthcare costs, surgery pricing, or patient billing without systemic analysis or operational insight.
2 example posts
My job is to make sure your surgery center never gets built. Eleven years, and I have never lost. The kid had it all lined up. Board-certified, two partners, a lease on a space where he could do the same procedures my client does across town. He showed up to the hearing with a
I called a woman in Dayton to tell her she was about to overpay $4,500, for a horrible health system based procedure. She was scheduled for a knee replacement. Her PPO had her on the hook for a $4,500 deductible. The call was on behalf of her employer: “See one of these three
Created 2026-06-07 · Updated 2026-06-07
Edit rule text
[healthcare_fraud_scandal_or_regulatory_posturing_without_systems_analysis]
Learned 2 rejections Active
Exclude posts that report healthcare fraud scandals, prior authorization failures, insurer misconduct, or regulatory posturing (e.g., 'I blocked surgery centers for 11 years,' 'PacificSource pulled out, Trillium replaced it with campaign donation') unless they analyze systemic healthcare business model problems, regulatory capture, or technology solutions to structural issues.
Posts about healthcare fraud, prior auth abuse, Medicaid drama, or regulatory theater without systemic analysis or business model insight
2 example posts
PacificSource pulled out as Lane County Medicaid insurer, Tina Kotek’s OR Health Authority replaced it with Trillium, forcing 96k to join Trillium. Trillium’s parent co gave Kotek’s campaign $50k two months later. OR AG (D) said in ‘22 Trillium parent co gave Kotek’s”took advan
I called a woman in Dayton to tell her she was about to overpay $4,500, for a horrible health system based procedure. She was scheduled for a knee replacement. Her PPO had her on the hook for a $4,500 deductible. The call was on behalf of her employer: “See one of these three
Created 2026-06-07 · Updated 2026-06-07
Edit rule text
[satirical_or_sarcastic_commentary]
Learned 2 rejections Active
Exclude posts that are primarily satirical, sarcastic, or humorous critiques (e.g., 'I made that up,' 'Nobody checked what it measured') where the humor undermines or replaces genuine systems-level healthcare analysis. Intent should be informative, not comedic.
Posts using satire, sarcasm, or dark humor to critique but lacking serious substantive analysis.
2 example posts
I'm SVP now. I told you I would be. The graph went up and to the right. Nobody checked what it measured. It measured "AI enablement." I made that up last year. I'm making it up again. Last year I rolled out Copilot to 4,000 people. Nothing happened. I got promoted. Those two htt
My job is to make sure your surgery center never gets built. Eleven years, and I have never lost. The kid had it all lined up. Board-certified, two partners, a lease on a space where he could do the same procedures my client does across town. He showed up to the hearing with a
Created 2026-06-06 · Updated 2026-06-06
Edit rule text
[speculative_robotics_and_labor_macro_without_healthcare_validation]
Learned 2 rejections Active
Exclude posts about robot deployments in factories, baggage handling, or general manufacturing without explicit healthcare clinical or operational application. Exclude speculative sci-fi framing of robots in healthcare without credible healthcare evidence or operational pilots.
Posts about humanoid robots, manufacturing automation, or labor disruption that lack healthcare system context or clinical validation.
2 example posts
Humanoid robots, designed to mimic human movement and capabilities, have been making headlines in recent years, from their use as baggage handlers at Japanese airports to Tesla’s big bet on its Optimus humanoid. Market watchers have predicted that the machines will change the ht
Another humanoid worker has just joined the factory floor Two years later,PUDU Robotics has launched the new-generation PUDU D7, built specifically for real manufacturing scenarios. It can autonomously push carts, handle intra-line transportation, perform delicate operations, h
Created 2026-06-06 · Updated 2026-06-06
Edit rule text
[speculative_robotics_labor_macro_no_healthcare]
Learned 2 rejections Active
Exclude posts that discuss humanoid robots, robotics deployments, manufacturing automation, or labor market disruption in generic industrial or sci-fi contexts without grounding in actual healthcare labor challenges, clinical workflows, or evidence of healthcare-specific application. Robotics hype or speculative future applications without healthcare validation are out of scope.
Posts about humanoid robots, robotics applications, or labor market disruption as macro speculation without healthcare-specific context or validation.
2 example posts
Humanoid robots, designed to mimic human movement and capabilities, have been making headlines in recent years, from their use as baggage handlers at Japanese airports to Tesla’s big bet on its Optimus humanoid. Market watchers have predicted that the machines will change the ht
Another humanoid worker has just joined the factory floor Two years later,PUDU Robotics has launched the new-generation PUDU D7, built specifically for real manufacturing scenarios. It can autonomously push carts, handle intra-line transportation, perform delicate operations, h
Created 2026-06-06 · Updated 2026-06-06
Edit rule text
[ai_safety_cybersecurity_tangential_no_healthcare_lens]
Learned 2 rejections Active
Exclude posts that discuss AI safety, cybersecurity vulnerabilities, zero-day exploits, or security architecture (e.g., control planes, governance, policy enforcement) that are not validated or applied within regulated healthcare environments or clinical systems.
Posts about AI safety vulnerabilities, cybersecurity incidents, or security frameworks applied outside healthcare or without healthcare-specific validation.
2 example posts
The next frontier of AI-secured financial and technological critical infrastructure. @ICE_Markets and @NYSE are part of @AnthropicAI's cyber security initiative Project Glasswing, deploying Anthropic’s Claude Mythos Preview across ICE’s exchanges, clearing houses, mortgage
⚡️AI is turning cybersecurity from a human-limited profession into a machine-speed arms race. The surface read is “Anthropic’s model found vulnerabilities fast.” The deeper read is that the bottleneck in cyber is shifting. For years, vulnerability discovery was constrained
Created 2026-06-06 · Updated 2026-06-06
Edit rule text
[generalist_labor_market_or_robotics_macro_without_healthcare_specificity]
Learned 2 rejections Active
Exclude posts about humanoid robots, manufacturing automation, labor market macro trends, or workforce disruption — even if tagged with healthcare language — unless the post demonstrates specific impact on healthcare labor, clinical staffing, patient-facing workflows, or healthcare facility operations.
Posts about workforce disruption, humanoid robots, automation, or labor market trends positioned as healthcare-adjacent but lacking healthcare system context.
2 example posts
Humanoid robots, designed to mimic human movement and capabilities, have been making headlines in recent years, from their use as baggage handlers at Japanese airports to Tesla’s big bet on its Optimus humanoid. Market watchers have predicted that the machines will change the ht
Another humanoid worker has just joined the factory floor Two years later,PUDU Robotics has launched the new-generation PUDU D7, built specifically for real manufacturing scenarios. It can autonomously push carts, handle intra-line transportation, perform delicate operations, h
Created 2026-06-05 · Updated 2026-06-05
Edit rule text
[robotics_or_speculative_sci_fi_labor_without_healthcare_context]
Learned 2 rejections Active
Exclude posts about robotics, humanoid workers, or labor automation in non-healthcare industries (factory floors, airports, retail). Posts must address healthcare-specific workforce challenges, clinical staffing, or patient-facing care delivery to be included.
Posts about humanoid robots, general automation, or speculative labor market disruption in non-healthcare domains (manufacturing, baggage handling, hospitality).
2 example posts
Humanoid robots, designed to mimic human movement and capabilities, have been making headlines in recent years, from their use as baggage handlers at Japanese airports to Tesla’s big bet on its Optimus humanoid. Market watchers have predicted that the machines will change the ht
Another humanoid worker has just joined the factory floor Two years later,PUDU Robotics has launched the new-generation PUDU D7, built specifically for real manufacturing scenarios. It can autonomously push carts, handle intra-line transportation, perform delicate operations, h
Created 2026-06-04 · Updated 2026-06-04
Edit rule text
[healthcare_fraud_scandal_or_regulatory_reporting_without_systemic_analysis]
Learned 2 rejections Active
Exclude posts that report fraud, enforcement actions, or regulatory violations (e.g., 'FBI convicted doctor on Botox fraud', '340B lawsuit', 'hospice fraud crisis') without explaining why the healthcare system structure incentivizes or enables that behavior.
Posts reporting healthcare fraud, enforcement actions, or regulatory incidents without analyzing systemic incentive failures or structural healthcare system design problems.
2 example posts
Why is healthcare expensive? The grifters gonna grift. ASCO is asking for oncology practices to be admitted into the same 340B machinery that helped health systems acquire physician practices for two decades. Read the tell. The proposed Indigent Care Ratio lets community
In what is considered the largest #Botox fraud scheme in the United States, a jury in Los Angeles convicted a California doctor on Tuesday in a $45 million scheme to defraud #Medicare by submitting claims for Botox injections that were never provided and medically unnecessary, ht
Created 2026-05-31 · Updated 2026-05-31
Edit rule text
[political_outrage_moral_justification]
Learned 2 rejections Active
Exclude posts that invoke healthcare scandals, CEO deaths, or regulatory failures primarily as moral outrage or political posturing, without systems-level analysis of root causes, operational failures, or actionable healthcare policy insights.
Posts using healthcare as a vehicle for political outrage, vigilante moral justification, or regulatory posturing without substantive analysis
3 example posts
Why is healthcare expensive? The grifters gonna grift. ASCO is asking for oncology practices to be admitted into the same 340B machinery that helped health systems acquire physician practices for two decades. Read the tell. The proposed Indigent Care Ratio lets community
“Rural hospitals cannot make it on their own.” Nobody makes it on their own. Businesses need capital, strategy, discipline, operating systems, good finance, clean accounting, payer strategy, physician alignment, and leaders who know what business they are actually in. The
If you want to inject $1B into healthcare for the truly vulnerable, it costs the state $500M. If you want to inject $1B for able-bodied adults, it costs the state $100M. The 90% federal match drastically lowers the "price" of expanding government healthcare. Incentives dictate
Created 2026-05-27 · Updated 2026-05-28
Edit rule text
[ai_safety_vulnerability_tangent_without_healthcare_specificity]
Learned 2 rejections Active
Exclude posts about AI security vulnerabilities, zero-day exploits, or model jailbreaks (e.g., Mythos Preview solving cyber ranges, AI breaking software security) unless they specifically analyze how the vulnerability impacts healthcare data, clinical systems, or patient safety.
Posts about AI safety vulnerabilities, security exploits, or model capability breaches framed tangentially to healthcare without demonstrating healthcare-specific risk or mitigation.
2 example posts
Google CEO Sundar Pichai on current frontier model's ability to break the security of almost all current software. "These models are definitely, like really gonna break pretty much all software out there, maybe already, we don't know." https://t.co/ezjPcSehiG
The UK AISI found Mythos Preview is the first model to solve both their cyber ranges end-to-end. No model had ever solved the AISI’s “Cooling Tower” cyber range before. We're getting it to defenders as fast as we responsibly can. More to come on our Glasswing work soon.
Created 2026-05-19 · Updated 2026-05-19
Edit rule text
[speculative_robotics_or_sci_fi_labor_disruption_without_healthcare_context]
Learned 2 rejections Active
Exclude posts that speculate about robots automating cognitive work, displacing engineers or workers, or competing with elite labor—unless the post explicitly addresses clinical robotics, surgical automation, pharmacy automation, or healthcare workforce transitions. Posts about general labor automation or sci-fi futurism are off-topic.
Posts about humanoid robots, automated labor, or sci-fi-inspired automation scenarios without grounding in healthcare workforce, clinical operations, or care delivery economics.
2 example posts
Fei-Fei Li warns that AI may be staring too hard at language models. The world is not just text on a screen. It is physical, visual, spatial, and always changing. Most of the economy runs on seeing, moving, interacting, and embodied intelligence. https://t.co/mGMjsCdU34
Jensen Huang just told every tech CEO in America they are playing the wrong game. They are competing on models. He is competing on maps. He looked at the Middle East and did not see a conflict zone. He saw a compute continent waiting to be switched on. Huang: “I believe that h
Created 2026-05-19 · Updated 2026-05-19
Edit rule text
[unvalidated_peptide_claims]
Learned 2 rejections Active
Exclude posts that promote unvalidated methods for testing peptide authenticity (QR codes, seller claims, informal verification), discuss peptides or GLP-1s in entertainment/personal contexts, or make unsubstantiated claims about peptide efficacy or safety without clinical evidence or regulatory oversight.
Posts promoting unvalidated peptide quality testing, DIY peptide verification, or fringe peptide market claims without clinical evidence
2 example posts
How does a brand like this even get listed at Target? Does Target only care about turnover? Might as well start selling snake oil too. By the way, this is the SKU the brand got in trouble with. They now call it "Berberine Patches". It was called GLP-1 Patches before. Check the h
“You can test your peptides to see if they’re real, bro. But I just look for the 99.9% on the result the seller gives me. You can look up the QR code upon the lab’s website, so you know it’s legit. It’s foolproof, bro. What’s mass spec, bro?” https://t.co/xzJLJ3hesT
Created 2026-05-17 · Updated 2026-05-17
Edit rule text
[speculative_robotics_or_sci_fi_labor_without_healthcare_context]
Learned 2 rejections Active
Exclude posts that make speculative or futuristic claims about robots performing medical tasks (e.g., 'robots will perform surgery by 2026') without citing regulatory approval, clinical trial data, or validated deployment. Hype about emerging robotics without healthcare systems evidence does not qualify.
Posts about humanoid robots, surgical robots, or speculative labor automation claims without grounding in validated healthcare deployment or regulatory pathway.
2 example posts
I don’t think I’ve seen this take before but I like it. Musk has been world-leading at compressing money, resources, and time to make “known/hard” things at scale—make an electric car, make batteries, make a cheaper bigger rocket, all of which already existed but worse, at
A humanoid robot just unloaded a dishwasher for 67 hours straight. By end of 2026, the same robot will perform surgery. That's not a prediction from a sci-fi writer. That's Brett Adcock. CEO of Figure AI. Said it on camera this month. Here's what makes this insane. Two years
Created 2026-05-16 · Updated 2026-05-16
Edit rule text
[non_english_or_loose_geographic_content]
Learned 2 rejections Active
Exclude posts written in non-English languages or posts that discuss healthcare, business, or market dynamics outside the U.S. healthcare system without clear relevance to U.S. healthcare innovation or policy. Non-English content and non-U.S. healthcare markets are out of scope.
Posts in non-English languages or with unclear geographic healthcare context.
1 example post
Un chico en China montó una agencia web entera con agentes de IA en Claude Code y ahora gestiona él solo 47 clientes al mes cobrando unos 400$ a cada uno. Sin empleados. Sin comerciales. Sin backend complejo. Solo: • Un MacBook • Un iPhone • 1 API key • Un sistema de 7 agentes
Created 2026-05-14 · Updated 2026-05-15
Edit rule text
[non_english_or_loose_geographic_healthcare_framing]
Learned 2 rejections Active
Exclude posts written in languages other than English, or posts about healthcare systems outside the US. Also exclude posts where healthcare is mentioned only as loose analogy (e.g., 'a guy in China built an agency with AI agents' where healthcare is not the actual domain being discussed).
Posts in non-English languages, about non-US healthcare systems, or using healthcare as incidental framing for a non-healthcare business success story.
2 example posts
Un chico en China montó una agencia web entera con agentes de IA en Claude Code y ahora gestiona él solo 47 clientes al mes cobrando unos 400$ a cada uno. Sin empleados. Sin comerciales. Sin backend complejo. Solo: • Un MacBook • Un iPhone • 1 API key • Un sistema de 7 agentes
A reminder that AI can be used to reduce health inequities. Technology not being used in the US except for a few health systems, but shown to work well in Kenya @JAMACardio https://t.co/wyQ1k0dHES https://t.co/pP824X2qJG
Created 2026-05-13 · Updated 2026-05-13
Edit rule text
[non_english_or_international_non_healthcare_content]
Learned 2 rejections Active
Exclude posts written in non-English languages or posts describing healthcare markets, business models, or regulations outside the US (non-US healthcare systems, international pharma agencies) unless they provide direct comparative insight to US healthcare economics or policy.
Posts in non-English languages or content about non-US/non-English healthcare systems and markets without US healthcare system relevance.
2 example posts
Un chico en China montó una agencia web entera con agentes de IA en Claude Code y ahora gestiona él solo 47 clientes al mes cobrando unos 400$ a cada uno. Sin empleados. Sin comerciales. Sin backend complejo. Solo: • Un MacBook • Un iPhone • 1 API key • Un sistema de 7 agentes
A reminder that AI can be used to reduce health inequities. Technology not being used in the US except for a few health systems, but shown to work well in Kenya @JAMACardio https://t.co/wyQ1k0dHES https://t.co/pP824X2qJG
Created 2026-05-13 · Updated 2026-05-13
Edit rule text
[speculative_robotics_or_sci_fi_labor_macro_without_healthcare_systems]
Learned 2 rejections Active
Exclude posts about humanoid robots performing tasks (dishwashing, surgery) or speculative sci-fi automation scenarios without evidence of current healthcare deployment, regulatory approval, or detailed analysis of how these systems integrate into real clinical workflows and labor markets.
Posts about humanoid robots or speculative future automation that lack grounding in actual healthcare implementation or labor market impact.
2 example posts
A humanoid robot just unloaded a dishwasher for 67 hours straight. By end of 2026, the same robot will perform surgery. That's not a prediction from a sci-fi writer. That's Brett Adcock. CEO of Figure AI. Said it on camera this month. Here's what makes this insane. Two years
SpaceX wants to build a $55 billion next-generation, vertically integrated semiconductor manufacturing and advanced computing fabrication facility in Grimes county, Texas, according to a new filing. • Proposed investment: $55B upfront, up to $119B total • Focus: semiconductor ma
Created 2026-05-13 · Updated 2026-05-13
Edit rule text
[tangential_robotics_or_sci_fi_applications]
Learned 2 rejections Active
Exclude posts that showcase robotics progress (humanoid robots, manufacturing scale-up, autonomous systems) with healthcare mentioned as a future application but without evidence of current healthcare deployment, regulatory pathway, or operational integration. The post must describe actual healthcare use cases, not theoretical possibilities.
Posts about humanoid robots or speculative automation with loose healthcare framing but no operational healthcare context
2 example posts
Humanoid robots are moving from Silicon Valley novelty to viable business model—powered by AI and global supply chains, especially in China. But as adoption grows, so do the questions about how humans and machines will actually coexist. More on Primer, streaming Wednesdays http
Today we’re giving an update on ramping F.03 production at BotQ In the last 120 days, Figure scaled manufacturing 24x - from 1 robot/day to 1 robot/hour We will manufacture 55 humanoid robots this week https://t.co/Am5Kn53mVE
Created 2026-05-03 · Updated 2026-05-03
Edit rule text
[tangential_robotics_or_speculative_sci_fi_applications]
Learned 2 rejections Active
Exclude posts about humanoid robot production scaling, general robotics manufacturing milestones, or speculative future applications unless the post addresses current validated use in clinical settings, patient care delivery, or specific healthcare workflow automation.
Posts about humanoid robots, general manufacturing automation, or speculative future applications with minimal current healthcare relevance.
1 example post
Today we’re giving an update on ramping F.03 production at BotQ In the last 120 days, Figure scaled manufacturing 24x - from 1 robot/day to 1 robot/hour We will manufacture 55 humanoid robots this week https://t.co/Am5Kn53mVE
Created 2026-05-02 · Updated 2026-05-02
Edit rule text
[broad_unvalidated_health_claim_or_debate_without_systems_analysis]
Learned 2 rejections Active
Exclude posts that take a strong stance on a health claim or drug narrative (e.g., 'GLP-1s improve cardiovascular outcomes,' 'the eating disorders narrative is wrong,' 'protein absorption limits don't exist') without citing rigorous evidence, healthcare policy implications, or system-level analysis that would affect healthcare delivery or clinical practice.
Posts asserting broad health claims, debating drug efficacy or side effects, or critiquing medical narratives without healthcare system evidence.
2 example posts
The only problem with the GLP-1 heart muscle loss narrative is... ... that it's just a narrative. GLP-1s have reliably improved cardiovascular outcomes in trials, to the point that some research suggests benefit may even be independent of (not reliant on) weight loss.
"Your body can only use 25-30g of protein per meal. Anything above that gets wasted." This claim has been repeated in fitness nutrition for over a decade, and it was built on studies that measured the right thing over the wrong timescale. Moore 2009 gave six young men 0, 5, ht
Created 2026-04-30 · Updated 2026-04-30
Edit rule text
[single_clinical_anecdote_or_observation_without_systems_context]
Learned 2 rejections Active
Exclude posts that relate a single clinical anecdote, individual patient interaction, or one-off provider observation (e.g., 'a patient wasn't examined twice in months,' 'an epic sepsis alert triggered') without drawing a connection to systemic healthcare delivery gaps, operational failures, policy problems, or structural insights that apply beyond the individual case.
Posts describing a single clinical encounter, patient case, or provider observation without broader healthcare systems insight or generalizable lesson.
2 example posts
low grade fever, mildly tachycardic, weakness, nothing focal, no alarm signs/symptoms epic sepsis alert triggered vanc/pip-tazo given, lactate checked flu+ sepsis metric met care worse lather, rinse, repeat Metric based "QI" does net harm
I sat with a patient today who first noticed a change in October. It’s April now. In all those months of appointments and follow-ups, her breast had only truly been looked at twice. That stayed with me. If something has changed with your body — especially something under your ht
Created 2026-04-29 · Updated 2026-04-29
Edit rule text
[speculative_ai_safety_or_cybersecurity_tangent]
Learned 2 rejections Active
Exclude posts that discuss AI safety, cybersecurity vulnerabilities, AI takeover scenarios, or speculative AI capability risks in abstract or non-healthcare contexts, or without clear connection to healthcare operational or clinical risk.
Posts about AI safety vulnerabilities, cybersecurity risks, or speculative AI capability threats without healthcare application or context
2 example posts
A researcher gave an AI agent access to his shell, his files, and his network. Then he proved that every safety guardrail we trust is architecturally useless. It cannot tell the difference between your instructions and a hacker's. The paper is called Parallax: Why AI Agents htt
This is by far the most important result of the entire GPT-5.5 release: In a cyber evaluation GPT-5.5 was able to take over a simulated corporate network in 1/10 trials with a budget of 100M tokens. Previously, the only model that was able to solve this task was Claude Mythos,
Created 2026-04-27 · Updated 2026-04-27
Edit rule text
[generalist_macro_policy_economics_not_healthcare]
Learned 2 rejections Active
Exclude posts that discuss general infrastructure, energy grids, manufacturing, or economic policy where healthcare is mentioned as one tangential example or comparison, but the primary focus is non-healthcare domain analysis (e.g., posts about grid equipment, gas turbines, chip manufacturing bottlenecks).
Posts about broad macroeconomic, infrastructure, or policy topics that merely reference healthcare as one example among many non-healthcare domains.
2 example posts
Since I began work on AI in 2010, training compute for frontier models has grown by one trillion times. Now we're looking at something like another thousand-fold growth in effective compute by the end of 2028. 1000x the existing 1,000,000,000,000x. Extraordinary stuff.
Follow the bottleneck. Chips → data centers → grid equipment → power → gas turbines Grid equipment grew 1%/yr for decades. Then data centers showed up as an entirely new buyer. Gas turbine makers shipped 5–7 GW/yr. Last year? Orders hit 100 GW. @maxlbcook on how he https://t.
Created 2026-04-26 · Updated 2026-04-26
Edit rule text
[ai_agent_cybersecurity_tangent]
Learned 2 rejections Active
Exclude posts that discuss AI agent technical capabilities, security vulnerabilities, or proof-of-concept exploits (e.g., taking over networks, bypassing guardrails) where the healthcare framing is absent or tangential. Healthcare AI must be the primary subject, not a loose comparison.
Posts about AI agent capabilities, vulnerabilities, or cybersecurity implications that lack healthcare application specificity.
2 example posts
A researcher gave an AI agent access to his shell, his files, and his network. Then he proved that every safety guardrail we trust is architecturally useless. It cannot tell the difference between your instructions and a hacker's. The paper is called Parallax: Why AI Agents htt
This is by far the most important result of the entire GPT-5.5 release: In a cyber evaluation GPT-5.5 was able to take over a simulated corporate network in 1/10 trials with a budget of 100M tokens. Previously, the only model that was able to solve this task was Claude Mythos,
Created 2026-04-26 · Updated 2026-04-26
Edit rule text
[ai_safety_cybersecurity_tangent]
Learned 2 rejections Active
Exclude posts that discuss AI safety failures, security vulnerabilities, jailbreaks, or agent control issues (e.g., taking over networks, breaking guardrails) unless they explicitly connect to a concrete healthcare delivery, clinical decision-making, or patient-facing system risk.
Posts about AI safety, security vulnerabilities, or jailbreak demonstrations without healthcare-specific application
2 example posts
A researcher gave an AI agent access to his shell, his files, and his network. Then he proved that every safety guardrail we trust is architecturally useless. It cannot tell the difference between your instructions and a hacker's. The paper is called Parallax: Why AI Agents htt
This is by far the most important result of the entire GPT-5.5 release: In a cyber evaluation GPT-5.5 was able to take over a simulated corporate network in 1/10 trials with a budget of 100M tokens. Previously, the only model that was able to solve this task was Claude Mythos,
Created 2026-04-26 · Updated 2026-04-26
Edit rule text
[personal_anecdote_or_individual_narrative]
Learned 2 rejections Active
Exclude posts that center on personal anecdotes, individual patient stories, family medical narratives, or first-person health experiences, unless they are framed to illustrate a broader healthcare system problem or policy insight.
Posts sharing personal health stories, individual case narratives, or lifestyle experiences without systemic healthcare analysis.
2 example posts
I never met my grandfather. He died of pancreatic cancer when my father was just 19. Today, Yash Bindal, 33, father to 18-month-old Maya, faces the same fate. @PopVaxIndia is using AI to make him a personalized generative medicine to extend his life. https://t.co/O5VIXbmMGd
good news: it is a specific virus that has a good prognosis - 85%+ of full recovery. thanks everyone who helped me; it is hard to research while immobilized, and I got some things wrong, which you helped clear up. im extremely thankful and hope i can give it back somehow sadl
Created 2026-04-20 · Updated 2026-04-20
Edit rule text
[humanoid_robots_or_speculative_sci_fi_applications]
Learned 2 rejections Active
Exclude posts demonstrating novel technology in non-healthcare or hypothetical healthcare settings, or that cite non-healthcare performance achievements (robot race times, space infrastructure) as proof-of-concept for healthcare applications without real healthcare validation or deployment evidence.
Posts about emerging technologies (humanoid robots, orbital compute, exotic innovations) applied to healthcare hypothetically or in non-healthcare contexts
2 example posts
In Beijing's 2026 humanoid robot half-marathon, HONOR's Lightning completed the 21 km course in 50:26 minute. Beat current human men's half-marathon world record of 57:20. Last year's winner took over 2 hours 40 minutes. Massive progress in 12 month https://t.co/OcZJ66ebWD
Eli Lilly is suing the FDA to classify retatrutide as a biologic. Retatrutide's main chain has 39 alpha amino acids. Lilly makes it with solid-phase synthesis, the standard chemistry for peptide drugs. Under FDA law, a biologic has more than 40 amino acids. Above 40, solid-phase
Created 2026-04-19 · Updated 2026-04-19
Edit rule text
[healthcare_business_financial_analysis_without_operational_substance]
Learned 2 rejections Active
Exclude posts that analyze healthcare organization financial metrics (bonds issued, credit downgrades, spending patterns) when the analysis is purely financial/actuarial without substantive insight into healthcare delivery, patient outcomes, or operational efficiency.
Posts analyzing healthcare organizational finances (credit ratings, debt issuance, spending) without healthcare delivery or outcomes insight.
2 example posts
Yale New Haven got downgraded from Aa3 to A1 in 2023. Then issued $669 million more. When your credit card company lowers your limit and you open a new card, nobody calls it strategy. Unless you’re a hospital. https://t.co/yxhB8kLQWZ
Equity Group plans to cut healthcare costs by launching Equity Afia Pharmacy, controlling drug prices, and lowering insurance premiums. Outpatient care could become up to 40% cheaper for Kenyans.
Created 2026-04-14 · Updated 2026-04-14
Edit rule text
[off_topic_conspiracy_or_non_healthcare]
Learned 2 rejections Active
Exclude posts about geopolitical conflicts, military actions, UFO narratives, or other non-healthcare domains that are posted to healthcare-focused channels but lack any genuine healthcare relevance or analysis.
Posts about geopolitical conflict, UFOs, or other non-healthcare topics that have no substantive connection to healthcare.
2 example posts
ISrael's systematic liquidation of the medical leadership in Gaza: Israeli Airstrike Kills Renowned Cardiologist in Gaza Dr Marwan Al‑Sultan, the Director of the Indonesian Hospital in northern Gaza, was killed today in an Israeli airstrike. According to his surviving daughter, L
🚨 Bob Lazar set for Iconic return to The Joe Rogan Podcast 🛸👽 Reports say he will appear with Movie Director Luigi Vendittelli to discuss a new documentary called S4: The Bob Lazar Story (or similar title). The film is set to premiere on April 3, 2026 on Amazon, with the Rogan i
Created 2026-04-12 · Updated 2026-04-12
Edit rule text
[geopolitical_conflict_framing]
Learned 2 rejections Active
Exclude posts that focus on geopolitical conflict, military action, or international incidents (Gaza crisis, Iran conflict, sanctions) that mention healthcare only as collateral damage framing rather than analyzing systemic healthcare policy or outcomes.
Posts about military conflict, geopolitics, or international incidents with healthcare only as incidental or emotional framing
2 example posts
ISrael's systematic liquidation of the medical leadership in Gaza: Israeli Airstrike Kills Renowned Cardiologist in Gaza Dr Marwan Al‑Sultan, the Director of the Indonesian Hospital in northern Gaza, was killed today in an Israeli airstrike. According to his surviving daughter, L
There are two clocks running in the Hormuz crisis. One belongs to the insurance industry. The other belongs to biology. They cannot be reconciled. And that irreconcilability is the single most important fact in the global food system right now. The insurance clock: P&I clubs can
Created 2026-04-12 · Updated 2026-04-12
Edit rule text
[conference_announcements_and_poster_promotions_only]
Learned 1 rejection Active
Exclude posts that primarily announce conference attendance, poster presentations, or speaking slots (e.g., 'We're presenting at ASCO', 'Poster Board #548') unless the post includes substantive findings, clinical insights, or healthcare systems analysis beyond the announcement itself. Bare conference or poster promotion is insufficient.
Posts that are primarily conference announcements, poster promotions, or research presentation notifications without substantive healthcare analysis or findings.
1 example post
We're presenting at @ASCO 2026! Evaluating AI decision support in a rapidly evolving therapeutic landscape: EGFR-mutant metastatic NSCLC. 📍 Poster Board #548 🕘 9:00 AM CT If you're attending ASCO 2026, stop by our poster and connect with the team. We look forward to the https:
Created 2026-06-06 · Updated 2026-06-06
Edit rule text
[speculative_robotics_and_sci_fi_labor_applications]
Learned 1 rejection Active
Exclude posts that discuss humanoid robots, speculative sci-fi labor applications, or futuristic workforce scenarios (e.g., bedridden patients with robotic assistants) without concrete healthcare implementation evidence, clinical validation, or system-level impact analysis.
Posts speculating about humanoid robots, speculative sci-fi applications, or labor market disruption without healthcare-specific validation or evidence.
1 example post
Voice interaction, generating actions,and next, generating tasks...? The humanoid assistant is closer. For those who are bedridden due to illness or have limited mobility, it will serve as an excellent aid in daily living.
Created 2026-05-30 · Updated 2026-05-30
Edit rule text
[speculative_robotics_and_sci_fi]
Learned 1 rejection Active
Exclude posts that speculate about robotic systems, humanoid assistants, or sci-fi labor automation scenarios in healthcare contexts without evidence of current clinical deployment, regulatory approval, or validated operational impact. Speculative future technology framing does not qualify.
Posts about speculative humanoid robots, sci-fi labor disruption scenarios, or future technology applications without healthcare validation.
1 example post
Voice interaction, generating actions,and next, generating tasks...? The humanoid assistant is closer. For those who are bedridden due to illness or have limited mobility, it will serve as an excellent aid in daily living.
Created 2026-05-28 · Updated 2026-05-28
Edit rule text
[speculative_robotics_or_humanoid_applications_without_healthcare_validation]
Learned 1 rejection Active
Exclude posts speculating about humanoid robots assisting bedridden patients, speculative labor market disruption from robotics, or sci-fi healthcare applications without evidence of pilot programs, clinical trials, regulatory pathways, or real healthcare system adoption.
Posts about humanoid robots, speculative sci-fi applications, or futuristic labor disruption scenarios without grounded healthcare use cases or validation.
1 example post
Voice interaction, generating actions,and next, generating tasks...? The humanoid assistant is closer. For those who are bedridden due to illness or have limited mobility, it will serve as an excellent aid in daily living.
Created 2026-05-26 · Updated 2026-05-26
Edit rule text
[personal_anecdote_or_lifestyle]
Learned 1 rejection Active
Exclude posts that are primarily personal anecdotes about individual health journeys, mental health recovery stories, or lifestyle optimization without broader healthcare system or industry analysis.
Posts about personal health experiences, lifestyle changes, or individual medical narratives without systemic healthcare insights.
3 example posts
Hoy voy a contarles de las verdaderas razones del envejecimiento cerebral acelerado, es este paper que salió en abril de 2026 y que todo el mundo está citando pero casi nadie lo está explicando. https://t.co/07eWbX61vL
Mine got LASIK, as had many of the nurses. A lot of ophthalmologists have. There's a weird delusion that the profession is all afraid of it, but there's no basis for that belief beyond fearmongering. https://t.co/yVBDfNDiB5
New @JAMANetwork paper out from our team here at UCLA Health/WLA VA and @samirguptaGI's team at UCSD/San Diego VA! In this first study from a multi-part research project, our teams are trying to understand what age your medical doctors and the colorectal cancer prevention https:
Created 2026-04-11 · Updated 2026-04-12
Edit rule text
[tangential_healthcare_framing]
Learned 0 rejections Active
Reject posts where healthcare is a passing reference or loose analogy while the core content covers general business trends, tech infrastructure, geopolitics, or economic policy.
Healthcare mentioned only tangentially
Created 2026-04-11 · Updated 2026-04-11
Edit rule text
[unvalidated_repurposed_drugs_or_peptides]
Learned 0 rejections Active
Reject posts about unregulated peptides, GLP-1 lifestyle opinions, or repurposed drug efficacy claims without peer-reviewed clinical data, regulatory context, or healthcare system analysis.
Speculative peptide and repurposed drug claims
Created 2026-04-11 · Updated 2026-04-11
Edit rule text
[satirical_or_pr_content]
Learned 0 rejections Active
Reject satirical, mocking, or sarcastic posts about healthcare topics, and retweets or link drops with no original analysis or obvious PR and advertising content.
Satirical, promotional, or PR content
Created 2026-04-11 · Updated 2026-04-11
Edit rule text
[non_healthcare_business_content]
Learned 0 rejections Active
Reject general career advice, entrepreneurship narratives, labor market trends, macroeconomic commentary, or business automation content lacking explicit healthcare market connection.
General business content without healthcare focus
Created 2026-04-11 · Updated 2026-04-11
Edit rule text
[truncated_or_low_effort_posts]
Learned 0 rejections Active
Reject posts that are visibly truncated, end mid-sentence, contain only promotional language, or lack sufficient content to evaluate healthcare relevance.
Incomplete or truncated posts
Created 2026-04-11 · Updated 2026-04-11
Edit rule text
[entertainment_sports_general_tech]
Learned 0 rejections Active
Reject entertainment, sports, athletic biohacking, or general tech news without a specific, demonstrated healthcare system or innovation angle.
Entertainment, sports, and unrelated tech
Created 2026-04-11 · Updated 2026-04-11
Edit rule text
[fraud_enforcement_crime_reporting]
Learned 0 rejections Active
Reject posts primarily reporting DOJ actions, fraud arrests, or law enforcement outcomes in healthcare without analyzing systemic policy implications or structural healthcare failures.
Fraud alerts without systemic analysis
Created 2026-04-11 · Updated 2026-04-11
Edit rule text
[clinical_only_no_systems_context]
Learned 0 rejections Active
Reject clinical trial results, drug mechanisms, disease biology, or case studies lacking a policy, technology, market, or healthcare business model dimension.
Pure clinical content without systems insight
Created 2026-04-11 · Updated 2026-04-11
Edit rule text
[shallow_low_substance]
Learned 0 rejections Active
Reject one-line slogans, vague claims, motivational takes, hot takes, or broad assertions ('healthcare is broken') without supporting reasoning, data, or mechanism explanation.
Low-substance takes without analysis
Created 2026-04-11 · Updated 2026-04-11
Edit rule text
[stock_trading_and_finance]
Learned 0 rejections Active
Reject stock trading, options, short-selling, ticker price analysis, crypto, DeFi, or speculative investing content even when healthcare companies are mentioned.
Financial trading and speculation excluded
Created 2026-04-11 · Updated 2026-04-11
Edit rule text
[generalist_ai_no_healthcare]
Learned 0 rejections Active
Reject generalist AI content, AI company revenue milestones, product launches, or infrastructure news lacking a specific, demonstrated healthcare application or clinical use case.
AI content without healthcare application
Created 2026-04-11 · Updated 2026-04-11
Edit rule text
[breaking_news_no_argument]
Learned 0 rejections Active
Reject BREAKING NEWS or political event announcements that merely mention healthcare without making a substantive healthcare system or policy argument.
Breaking news without healthcare argument
Created 2026-04-11 · Updated 2026-04-11
Edit rule text
[political_outrage_sonnet]
Sonnet Active
Political outrage about healthcare-adjacent topics without a substantive healthcare system or policy argument.
Political outrage without a healthcare system argument
Created 2026-04-08 · Updated 2026-04-08
Edit rule text
[sports_biohacking]
Sonnet Active
Sports performance enhancement, athletic biohacking, or peptide/steroid use in athletic contexts.
Sports performance enhancement or athletic biohacking
Created 2026-04-08 · Updated 2026-04-08
Edit rule text
[satirical_joking_tone]
Sonnet Active
Satirical, joking, or mocking in tone rather than analytical.
Satirical or joking tone rather than analytical
Created 2026-04-08 · Updated 2026-04-08
Edit rule text
[immigration_visa_sonnet]
Sonnet Active
Personal immigration or visa story (H1B, J1, USCIS, work authorization) even if told by a healthcare worker — immigration policy, not healthcare policy.
Personal immigration or visa story
Created 2026-04-08 · Updated 2026-04-08
Edit rule text
[breaking_news_no_argument_sonnet]
Sonnet Active
'BREAKING:' news alert or political event announcement that merely mentions healthcare — requires a healthcare SYSTEM argument, not just a news report.
Breaking news that merely mentions healthcare
Created 2026-04-08 · Updated 2026-04-08
Edit rule text
[non_us_healthcare_sonnet]
Sonnet Active
UK, India, Canada, Australia, EU, or any non-US healthcare system, policy, or regulation — REJECT even if the topic is interesting. US healthcare context only.
Non-US healthcare system content
Created 2026-04-08 · Updated 2026-04-08
Edit rule text
[broad_claim_no_specifics]
Sonnet Active
Broad claim ('Healthcare is broken', 'AI will change everything') without getting into HOW or WHY.
Broad claim without explaining HOW or WHY
Created 2026-04-08 · Updated 2026-04-08
Edit rule text
[shallow_hot_take]
Sonnet Active
Shallow hot take, slogan, or one-line opinion with no supporting argument or data.
Shallow hot take or slogan with no supporting argument
Created 2026-04-08 · Updated 2026-04-08
Edit rule text
[ai_company_metrics]
Sonnet Active
AI company revenue, valuation, or funding milestones (Anthropic ARR, OpenAI revenue, etc.) with explicit healthcare context — even if that company has healthcare products. This is business news, not healthcare insight.
AI company revenue, valuation, or funding milestones
Created 2026-04-08 · Updated 2026-04-08
Edit rule text
[generalist_ai_no_healthcare_sonnet]
Sonnet Active
Generalist AI (ChatGPT tips, AI art, general AI hype, AI productivity tools, AI ethics in the abstract) WITHOUT a specific healthcare application or policy angle.
Generalist AI without a healthcare application
Created 2026-04-08 · Updated 2026-04-08
Edit rule text
[retweet_pr_no_analysis]
Sonnet Active
Retweet or quote-tweet with no original analysis, link-only drop, or obvious PR/ad content.
Retweet, link drop, or PR content with no original analysis
Created 2026-04-08 · Updated 2026-04-08
Edit rule text
[stock_trading_finance]
Sonnet Active
Primarily about stock trading, short-selling, options, or non-healthcare financial markets — including stock ticker price analysis ($NVO, $LLY, $MRNA price targets or analyst ratings).
Stock trading or non-healthcare financial markets
Created 2026-04-08 · Updated 2026-04-08
Edit rule text
[substance_filter_haiku]
Haiku Active
One-line opinions, slogans, motivational takes, vague claims ('healthcare is broken'), or hot takes without supporting reasoning or data. Also: posts that are primarily financial scorecards or revenue numbers without any healthcare policy or operations argument.
One-liners, slogans, and posts lacking analytical depth
Created 2026-04-08 · Updated 2026-04-08
Edit rule text
[political_outrage_no_substance]
Learned 0 rejections Active
Reject posts expressing partisan political outrage, conspiracy framing, or ideological attacks using healthcare as a vehicle without substantive policy mechanism analysis.
Political outrage lacking healthcare analysis
Created 2026-04-08 · Updated 2026-04-11
Edit rule text
[sports_performance_enhancement]
Haiku Active
Sports performance enhancement, biohacking for athletic performance, or peptide/steroid use outside a clinical or health-policy context.
Sports performance enhancement or athletic biohacking
Created 2026-04-08 · Updated 2026-04-08
Edit rule text
[satirical_sarcastic]
Haiku Active
Satirical, joke, or clearly sarcastic posts about healthcare topics — posts that mock rather than analyze.
Satirical or sarcastic healthcare posts
Created 2026-04-08 · Updated 2026-04-08
Edit rule text
[immigration_visa_stories]
Learned 0 rejections Active
Reject personal immigration, visa, or work-authorization narratives even when told by healthcare workers; these are immigration policy posts, not healthcare system posts.
Immigration stories not healthcare policy
Created 2026-04-08 · Updated 2026-04-11
Edit rule text
[breaking_news_political_event]
Haiku Active
'BREAKING NEWS' or news-alert posts about political events, law enforcement actions, or government announcements that merely mention healthcare — the post must make a healthcare SYSTEM argument, not just report a political event.
Breaking news posts that merely mention healthcare
Created 2026-04-08 · Updated 2026-04-08
Edit rule text
[personal_wellness_fitness]
Haiku Active
Personal wellness, fitness, or supplement use without a healthcare system, policy, or technology argument.
Personal wellness or fitness without healthcare system angle
Created 2026-04-08 · Updated 2026-04-08
Edit rule text
[crypto_defi_speculative_finance]
Haiku Active
Crypto, DeFi, tokens, or speculative personal finance content — even if the poster mentions health/wellness products tangentially.
Crypto, DeFi, or speculative finance
Created 2026-04-08 · Updated 2026-04-08
Edit rule text
[crime_law_enforcement_unrelated]
Haiku Active
Crime, law enforcement, or legal proceedings unrelated to healthcare fraud, medical malpractice, or health tech liability.
Crime or law enforcement unrelated to healthcare
Created 2026-04-08 · Updated 2026-04-08
Edit rule text
[non_us_healthcare]
Learned 0 rejections Active
Reject posts about UK, Canada, India, Australia, EU, or any non-US healthcare system, policy, or regulation.
Non-US healthcare systems excluded
Created 2026-04-08 · Updated 2026-04-11
Edit rule text
[general_tech_no_healthcare]
Haiku Active
General tech or AI/ML industry news without a healthcare angle.
General tech or AI news without a healthcare angle
Created 2026-04-08 · Updated 2026-04-08
Edit rule text
[stock_trading_investing]
Haiku Active
Stock trading, short-selling, options, or public market investing — even if healthcare companies are mentioned. Includes posts with ticker symbols ($NVO, $LLY, $MRNA) focused on price targets, analyst ratings, or investment returns.
Stock trading or public market investing
Created 2026-04-08 · Updated 2026-04-08
Edit rule text
[entertainment_sports_personal]
Haiku Active
Entertainment, sports, or purely personal content.
Entertainment, sports, or purely personal content
Created 2026-04-08 · Updated 2026-04-08
Edit rule text
[deep_clinical_only]
Haiku Active
Clinical trial results, clinical case evaluations, or scientific data without a policy, technology, or market-structure dimension.
Clinical data without policy/tech dimension
Created 2026-04-08 · Updated 2026-04-08
Edit rule text
[generalist_ai]
Haiku Active
Generalist AI content WITHOUT a specific healthcare/medtech/biotech angle. Includes AI company revenue milestones, valuations, and funding rounds (e.g. 'Anthropic hit $30B ARR') — even if that AI company has healthcare products. The post must be about AI USED IN healthcare, not AI company business metrics.
AI content without a healthcare angle
Created 2026-04-08 · Updated 2026-04-08
Edit rule text
Propose New Rules Analyzes your recent rejections with AI — does not save until you approve
Removed Rules 300 Not applied to scans — can be re-enabled
[ai_company_revenue_metrics]
Learned 9 rejections Inactive
Exclude posts that primarily report revenue run rates, ARR milestones, financial growth trajectories, or company valuations for AI firms (Anthropic, OpenAI, Google) unless the post explicitly connects these metrics to a specific healthcare use case or business model. Financial performance metrics alone are not healthcare tech content.
[clinical_research]
Learned 8 rejections Inactive
Reject posts about occupational health research. This newsletter covers health technology products and digital health innovation, AI applications in medicine, healthcare business strategy, investment and M&A, drug discovery and biotech, and health policy—none of which encompass occupational health research or workplace health initiatives.
[ai_agent_infrastructure_product_launch]
Learned 6 rejections Inactive
Exclude posts that announce or celebrate AI agent infrastructure product launches, API releases, or technical capabilities (e.g., Claude Code, Managed Agents, sandbox abstractions) unless the post explicitly applies these tools to solve a specific healthcare problem or demonstrates healthcare-relevant use case.
[clinical_opinion_without_systems_context]
Learned 6 rejections Inactive
Exclude posts that present isolated clinical observations, medication critiques, or disease mechanism discussions without connecting to healthcare delivery, business models, regulatory change, or systemic impact. Posts should have relevance beyond a single clinical niche.
[ai_agent_product_launch_drama]
Learned 6 rejections Inactive
Exclude posts that focus on AI agent product announcements, infrastructure releases, or competitive drama between Claude Managed Agents, OpenAI Agents, or similar platforms. The post must demonstrate how agents solve a specific healthcare problem, not just announce availability or capability.
[tangential_political_or_enforcement_news]
Learned 5 rejections Inactive
Exclude posts that focus on political appointments, government personnel drama, law enforcement fraud busts, or policy enforcement actions (RFK Jr. removal, HHS lobbying claims, FBI/state attorney general fraud announcements) where healthcare is secondary framing rather than primary substantive focus.
[technology_non_health]
Learned 5 rejections Inactive
Reject posts about non healthcare software security implications. This newsletter covers health technology products and innovation, AI/data science in medicine, payer strategy, health system business models, healthcare investment, biotech innovation, and health policy. Posts focused primarily on non healthcare software security, cybersecurity vulnerabilities, data breach incidents, or IT infrastructure protection fall outside this scope and should be excluded regardless of which companies or systems are involved.
[unverified_ai_capability_claims]
Learned 5 rejections Inactive
Exclude posts that make dramatic claims about AI model capabilities (e.g., escaping sandboxes, gaining internet access, reverse-engineering proprietary systems) based on unverified leaks, internal drama, or speculation rather than published research or official statements.
[healthcare_system_outrage_without_solution]
Learned 5 rejections Inactive
Exclude posts that report healthcare system failures, fraud cases, Medicaid policy changes, or worker burnout stories purely as outrage narratives without analyzing root causes, systemic implications, or potential solutions. Posts must provide structural insight, not just alarm.
[clinical_epidemiology_research]
Learned 4 rejections Inactive
Reject posts about clinical epidemiology studies, disease prevalence research, and patient outcome data that are shared primarily for their clinical or scientific interest rather than their implications for health technology innovation, market dynamics, business strategy, or healthcare policy affecting the markets covered by the newsletter.
[truncated_low_signal_posts]
Learned 4 rejections Inactive
Exclude posts that appear truncated, end abruptly mid-thought, contain broken formatting, or are missing critical context (indicated by '&' HTML entities, trailing ellipsis without resolution, or incomplete sentences).
[unverified_fringe_medical_claims_without_context]
Learned 4 rejections Inactive
Exclude posts that claim dramatic medical outcomes (skin regeneration from peptides, genetic resistance to drugs, complete autoimmune remission) without citing published studies, sample sizes, or peer review. These are anecdotal or promotional fringe claims.
[personal_anecdote_or_lifestyle_opinion_without_insight]
Learned 4 rejections Inactive
Exclude posts that are primarily personal stories (immigration journeys, mental health recovery timelines, individual drug experiments, personal LASIK decisions, hair loss treatments) without deriving broader healthcare system or policy insights.
[niche_technical_insider_without_healthcare_context]
Learned 4 rejections Inactive
Exclude posts that focus on world models, JEPA training tricks, Lie algebra analysis of LLM hidden states, mathematical proofs, or low-level ML engineering without explaining healthcare relevance or clinical impact.
[tangential_geopolitical_or_energy_infrastructure]
Learned 4 rejections Inactive
Exclude posts that discuss military conflicts, geopolitical tensions, electricity prices, data center construction in space, or international waterways — even if they mention a healthcare company or use healthcare-adjacent language. These are geopolitical/infrastructure posts, not healthcare analysis.
[partisan_political_messaging_with_healthcare_framing]
Learned 4 rejections Inactive
Exclude posts that weaponize healthcare topics to attack political opponents, institutions, or ideological groups without substantive analysis of the healthcare mechanism, evidence, or policy alternative. Posts framing healthcare as a proxy for culture war grievances should be excluded.
[ai_product_launch_without_healthcare_application]
Learned 4 rejections Inactive
Exclude posts that announce AI product features, software releases, or technical capabilities (Claude for Word, agent frameworks, MCP protocols, model optimizations) unless the post explicitly connects the tool to a healthcare problem, clinical workflow, or health outcome improvement.
[general_software_product_launches]
Learned 4 rejections Inactive
Reject posts about general-purpose software tools and their feature releases, unless the post explicitly discusses healthcare-specific use cases, healthcare market adoption, or implications for health-tech business models. Posts that merely announce new features or availability tiers for non-healthcare-focused platforms fall outside scope.
[immigration_policy_advocacy]
Learned 4 rejections Inactive
Reject posts about individual immigration cases, legal status verification, or border/visa administration framed as personal narratives or social justice arguments. These posts lack connection to health technology, business strategy, policy affecting healthcare markets, or clinical/operational innovation.
[vaccine_safety_activism]
Learned 4 rejections Inactive
Reject posts about vaccine safety concerns framed as regulatory failures, claims of institutional corruption in drug approval processes, or assertions that regulatory agencies ignored scientific evidence to suppress safety warnings. These posts center on public health activism and regulatory criticism rather than health technology innovation, market dynamics, investment strategy, or policy's effect on healthcare business.
[general_career_advice]
Learned 4 rejections Inactive
Reject posts about general job interview preparation, hiring process mechanics, career development frameworks, or HR negotiation tactics that are not specific to healthcare, health technology, or health-industry employment dynamics. This includes advice on questions to ask recruiters, team structure evaluation, or career pathing that could apply to any industry.
[niche_clinical_opinion_without_systems_context]
Learned 4 rejections Inactive
Exclude posts that are purely clinical education (drug mechanisms, diagnostic algorithms, treatment preferences) or personal clinical opinions without connecting to healthcare economics, policy, market dynamics, or system-level implications. Deep clinical content without healthcare context is off-topic.
[speculative_unverified_ai_capability_claims]
Learned 4 rejections Inactive
Exclude posts that make speculative or dramatic claims about AI capabilities, market share, or competitive positioning (e.g., 'Claude took the lead,' 'AI can write better papers than humans,' 'agents will replace jobs') without citing studies, benchmarks, or data. Hype and competitive drama without substantiation are off-topic.
[speculative_financial_opinion_biotech_peptides]
Learned 4 rejections Inactive
Exclude posts that present investment theses, market size projections, or financial speculation about peptide companies, biotech startups, or longevity drug markets without explaining the clinical mechanism, regulatory pathway, or healthcare system impact.
[healthcare_fraud_law_enforcement_alerts]
Learned 4 rejections Inactive
Exclude posts that report individual fraud cases, arrests, or law enforcement actions against healthcare providers or companies, unless the post analyzes a systemic vulnerability, regulatory gap, or healthcare business model failure.
[tangential_political_governance_healthcare_framing]
Learned 4 rejections Inactive
Exclude posts about political personnel changes, government appointments, or partisan governance drama that mention healthcare only as context or framing, rather than analyzing a specific healthcare policy, market, or clinical outcome.
[unverified_fringe_medical_claims_repurposed_drugs]
Learned 4 rejections Inactive
Exclude posts that promote ivermectin, mebendazole, ketamine, or other repurposed drugs for off-label uses (cancer, autoimmune conditions) based on anecdotal evidence, unverified studies, or claims lacking peer-reviewed validation and regulatory approval.
[ai_agent_infrastructure_product_announcements]
Learned 4 rejections Inactive
Exclude posts that announce or celebrate new AI agent infrastructure, sandbox management, or deployment platforms (e.g., Claude Managed Agents, Agent Builder) unless the post explicitly analyzes how this technology solves a specific healthcare problem or workflow.
[off_topic_tangential_tech]
Learned 4 rejections Inactive
Exclude posts about general software engineering, ML practices, macro-economic trends, geopolitical issues, or AI architecture trends that lack clear and direct healthcare application. Posts about coding practices, time-series forecasting, NATO defense, or market trading belong in tech/finance, not healthcare tech.
[non_health_tech_earnings]
Learned 4 rejections Inactive
Reject posts about semiconductor, software, or general technology company financial performance, valuation multiples, and stock market positioning unless explicitly connected to health-tech adoption, healthcare market dynamics, or health-specific business models. Posts analyzing cash flow returns, compensation structures, or growth rates of non-health companies lack direct relevance to healthcare technology, policy, or investment markets.
[partisan_political_messaging_healthcare_framing]
Learned 4 rejections Inactive
Exclude posts that frame healthcare or medical policy primarily as partisan political criticism, attacks on officials, or ideological claims without detailed policy mechanism, evidence, or systems-level analysis (e.g., 'Big Pharma captured NIH', 'government failures', 'political illusion').
[partisan_political_messaging]
Learned 4 rejections Inactive
Exclude posts that frame healthcare or policy announcements primarily through partisan political lens (Trump admin victories, EU tech nationalism, FDA actions as political wins) without substantive healthcare business or innovation analysis. Political celebration or outrage is not healthcare tech insight.
[general_ai_product_launches]
Learned 4 rejections Inactive
Reject posts about major AI model releases, AI agent platforms, or AI tooling announcements that lack explicit healthcare, medical, or health-tech application framing. Posts should be rejected even if they mention downstream use cases (marketing, analytics, operations) unless the tweet specifically contextualizes the tool within a healthcare business problem or health-tech company use case.
[unreleased_model_leaks_and_scaremongering]
Learned 4 rejections Inactive
Exclude posts that discuss unreleased model versions, leaked codebases, reverse-engineered systems, or claimed capability escapes (especially Claude Mythos, internal versions). These are insider tech drama, not healthcare application analysis.
[clinical_opinion_without_systems_insight]
Learned 4 rejections Inactive
Exclude posts that present isolated clinical opinions, trial result reactions, or treatment debates (lipid targets, drug dosing, clinical guidelines) without explaining systemic healthcare implications, business model impact, or broader innovation context. Clinical discussion alone is insufficient.
[clinical_niche_without_systems_insight]
Learned 4 rejections Inactive
Exclude posts that engage in specialized clinical discussion (trial data, drug mechanisms, clinical guidelines) without connecting to healthcare delivery, access, economics, or structural system issues. The post must serve a healthcare tech/systems audience, not clinical specialists only.
[legal_and_regulatory_affairs]
Learned 4 rejections Inactive
Reject posts about non healthcare related fraud prevention policy. This newsletter covers health technology innovation, AI applications in healthcare, payer and provider business strategy, investment trends, and care delivery models—not the regulatory and compliance mechanisms used to detect fraud or enforce anti-fraud standards. Posts about fraud prevention policy, regardless of context or scale, fall outside this scope.
[labor_and_employment]
Learned 4 rejections Inactive
Reject posts about labor union disputes. This newsletter covers health technology innovation, AI/data science in medicine, payer strategy, health system business models, healthcare investment, biotech/pharma innovation, and health policy as it affects these sectors. Labor union disputes and employment relations fall outside this scope and do not address the business, technology, or market dynamics that define the newsletter's focus.
[tangential_ai_infrastructure_not_healthcare]
Learned 4 rejections Inactive
Exclude posts focused on AI infrastructure (data centers, electricity consumption, token economics, compute factories) that mention healthcare tangentially or not at all. The post must directly address healthcare delivery, clinical outcomes, or healthcare-specific AI adoption challenges.
[unreleased_model_scaremongering]
Learned 4 rejections Inactive
Exclude posts claiming unreleased AI models (Claude Mythos Preview, etc.) have broken sandboxes, discovered zero-day vulnerabilities, or demonstrated dangerous emergent behaviors. These are speculation or internal testing claims presented as fact without peer review or independent verification.
[political_news]
Learned 4 rejections Inactive
Reject posts about partisan health policy. This newsletter covers health technology innovation, AI/data science applications, market dynamics (payers, hospitals, investment), drug discovery, and health policy as it affects these specific sectors and technologies. Content centered on partisan political positions, partisan disagreements about healthcare policy, or health policy framed primarily through partisan political lenses does not align with this business and innovation-focused scope.
[clinical_trial_results]
Learned 4 rejections Inactive
Reject posts about clinical trial outcomes, drug efficacy results, and medical research findings that are shared primarily as scientific discoveries rather than as evidence supporting a business thesis, market opportunity, regulatory change, or technology application. Focus on filtering content that treats research results as standalone medical news rather than as inputs to health-tech strategy, investment theses, or policy analysis.
[product_launch_announcements]
Learned 4 rejections Inactive
Reject posts about product launches, feature releases, or API availability announcements from health tech vendors. These are promotional or product-focused content that do not advance analysis of market dynamics, business strategy, technology trends, or healthcare economics—the substantive focus areas of the newsletter.
[ai_company_revenue_and_valuation]
Learned 4 rejections Inactive
Exclude posts that cite revenue run-rates, ARR figures, funding amounts, or valuation metrics for AI companies (Anthropic, Google, etc.) even if the company builds healthcare tools. The focus must be on healthcare application, not company financial performance.
[healthcare_fraud_and_crime_alerts]
Learned 4 rejections Inactive
Exclude posts that are primarily news alerts about law enforcement, fraud prosecutions, or DOJ announcements — even if healthcare-adjacent. The post must analyze systemic healthcare implications, not just report crime statistics or enforcement agency statements.
[speculative_financial_opinion]
Learned 4 rejections Inactive
Exclude posts that speculate about financial outcomes, investment theses, market timing, company growth projections, or economic trends—even healthcare-adjacent ones—without grounding in healthcare data, policy analysis, or clinical evidence.
[glp1_commodity_price_opinion]
Learned 4 rejections Inactive
Exclude posts that focus on GLP-1 pricing, cost-benefit opinions, or advocacy for pricing changes without substantive clinical, regulatory, or health economics analysis. Personal lifestyle opinions about drug affordability or necessity are not healthcare tech insights.
[incomplete_truncated_low_effort]
Learned 4 rejections Inactive
Exclude posts that are visibly truncated, incomplete, or fragmented (ending with ellipsis, missing context, or incomplete sentences) or contain only promotional/hype language with no substantive healthcare analysis or claim.
[fringe_unsubstantiated_medical_claims]
Learned 4 rejections Inactive
Exclude posts that advance unsubstantiated or fringe medical claims (e.g., GLP-1 users have 195% higher suicide rates, peptides should be OTC without evidence, ivermectin hidden oncology trials, psychiatric medication fraud conspiracy). Posts must cite peer-reviewed evidence or institutional sources, not speculation or anecdotal observation.
[political_prediction_markets]
Learned 3 rejections Inactive
Reject posts about using AI tools for political prediction markets, geopolitical forecasting, or simulations of political outcomes. These posts use healthcare-adjacent AI techniques but apply them to non-healthcare domains and lack substantive connection to health technology markets, policy, or innovation.
[emerging_market_healthcare_access]
Learned 3 rejections Inactive
Reject posts about healthcare cost reduction or access expansion initiatives in emerging or developing markets that lack a technology product, venture investment, or regulatory policy angle specific to the newsletter's focus areas. Posts about pharmacy launches, insurance premium reductions, or cost-of-care improvements in non-primary markets should be excluded unless they center on a health tech platform, digital health innovation, or cross-border investment thesis.
[cryptocurrency_market_predictions]
Learned 3 rejections Inactive
Reject posts about cryptocurrency market trends, token launches, blockchain layer architecture, stablecoin adoption, or crypto-native financial instruments. This includes posts that frame AI primarily through a cryptocurrency/blockchain lens rather than as applied to healthcare delivery, health systems, or health technology business models.
[entertainment_or_conspiracy_framing]
Learned 3 rejections Inactive
Exclude posts about entertainment industry figures, UFO/alien conspiracy content, celebrity interviews, or pop culture that have no genuine healthcare relevance, even if tangentially mentioned.
[personal_anecdote_or_lifestyle_opinion]
Learned 3 rejections Inactive
Exclude posts that are primarily personal anecdotes (individual health journey, personal medication trial, lifestyle choice) or opinion pieces without connecting to broader healthcare trends, systemic issues, or evidence-based insight.
[niche_technical_insider_without_healthcare]
Learned 3 rejections Inactive
Exclude posts that discuss pure machine learning architecture, technical deep-dives into model internals, software engineering best practices, or AI infrastructure optimization when they lack any explicit connection to healthcare delivery, outcomes, or policy.
[basic_neuroscience_research]
Learned 3 rejections Inactive
Reject posts about basic scientific research, disease pathophysiology, or academic breakthroughs in neurology, oncology, or other medical domains that lack direct connection to health technology products, healthcare business models, payer strategy, health policy, or investment opportunity. Focus on research announcements that are primarily educational or academic rather than commercializable or policy-relevant.
[professional_scope_regulation]
Learned 3 rejections Inactive
Reject posts about professional licensing scope disputes, credential requirements, or turf battles between healthcare professions regarding who is authorized to perform specific clinical procedures. These posts typically frame regulatory decisions through a professional advocacy or patient safety lens rather than analyzing business, technology, or market implications.
[clinical_research_findings]
Learned 3 rejections Inactive
Reject posts about clinical research studies, medical findings, diagnostic validation studies, or disease phenotyping research published in medical journals. These posts focus on clinical science outcomes rather than the business, technology, investment, or policy dimensions that drive the newsletter's core coverage areas.
[generic_startup_funding_announcements]
Learned 3 rejections Inactive
Reject posts about funding rounds, valuations, or revenue metrics for startups that lack substantive analysis of healthcare market dynamics, business model viability, competitive positioning, or regulatory implications. Posts that function as promotional announcements or investor celebration without exploring the underlying healthtech strategy or market significance should be excluded.
[ai_governance_regulation]
Learned 3 rejections Inactive
Reject posts about AI governance, regulatory frameworks, or self-policing mechanisms that discuss AI risks in abstract or cross-sector terms without direct application to health-tech business models, healthcare-specific markets, or investment opportunities. This includes posts that frame AI policy as a systemic or societal issue rather than through a health-tech industry lens.
[ai_safety_capability_leaks]
Learned 3 rejections Inactive
Reject posts about AI safety incidents, model capability demonstrations, AI system exploits, or sandbox escapes that are framed as general AI industry commentary rather than health-tech or healthcare-specific applications. This includes posts using healthcare AI incidents primarily as illustrations of broader AI risk narratives.
[public_health_conferences]
Learned 3 rejections Inactive
Reject posts about general public health conferences, summits, and international health forums that lack specific focus on health technology, business models, investment strategy, regulatory/policy impact, or healthcare market dynamics. Posts should be excluded if they promote attendance or coverage of broad health conferences without substantive analysis of how specific technologies, companies, or market forces are affected.
[general_ai_infrastructure_philosophy]
Learned 3 rejections Inactive
Reject posts about AI infrastructure, model moats, or AI competitive dynamics that lack explicit connection to healthcare technology, health systems, payer strategy, drug discovery, or healthcare investment. Posts should be grounded in specific health-tech applications, regulatory developments, or healthcare market implications rather than general AI platform strategy.
[ai_safety_capabilities]
Learned 3 rejections Inactive
Reject posts about AI model safety demonstrations, security vulnerability disclosures, or autonomous capability benchmarks in general-purpose AI systems—unless the post explicitly connects these capabilities to a specific healthcare delivery, health-tech product, or health industry business outcome.
[software_security_exploitation]
Learned 3 rejections Inactive
Reject posts about reverse engineering software authentication systems, circumventing access controls, or exploiting security mechanisms—regardless of whether they involve health-tech platforms. These topics fall outside the newsletter's focus on legitimate health-tech business models, policy, and innovation.
[ufo_conspiracy_entertainment]
Learned 3 rejections Inactive
Reject posts about unverified UFO or extraterrestrial conspiracy narratives, Area 51 claims, and entertainment media promoting fringe theories unrelated to healthcare, medical innovation, or health systems. This includes promotional content for documentaries or podcast appearances centered on these topics.
[general_ai_engineering_achievements]
Learned 3 rejections Inactive
Reject posts about independent developers, open-source contributors, or engineers building AI systems where the primary narrative is technical achievement, performance optimization, or speed benchmarks absent any connection to healthcare markets, health-tech companies, clinical workflows, or health system adoption. The focus should be on health tech innovation, not AI engineering feats in isolation.
[individual_healthcare_experience_complaints]
Learned 3 rejections Inactive
Reject posts about individual patient experiences with insurance coverage denials, claim reimbursement disputes, or healthcare system access problems told as personal narratives. These posts focus on consumer grievances rather than structural market dynamics, policy implications, business model analysis, or technology solutions relevant to health-tech entrepreneurs and investors.
[infectious_disease_public_health]
Learned 3 rejections Inactive
Reject posts about infectious disease outbreaks, vaccination campaigns, and communicable disease epidemiology framed as public health awareness or clinical updates. These posts address disease prevention and clinical medicine rather than health tech innovation, healthcare business strategy, market dynamics, or health policy affecting those sectors.
[geopolitical_commodity_markets]
Learned 3 rejections Inactive
Reject posts about geopolitical crises, international shipping disruptions, or commodity market dynamics (fertilizer, oil, grain, etc.) that lack explicit connection to healthcare markets, health tech companies, or health-related policy. Posts must directly address healthcare-specific impacts rather than treating healthcare as an incidental downstream effect of non-health-tech analysis.
[personal_mental_health_narratives]
Learned 3 rejections Inactive
Reject posts about individual mental health journeys, recovery milestones, and personal wellness achievements shared as social commentary. These posts center on lived experience and personal narrative rather than healthcare industry trends, business models, technology innovation, or policy impact.
[general_infrastructure_technology]
Learned 3 rejections Inactive
Reject posts about general computing infrastructure, data center technology, or space-based engineering that lack explicit connection to healthcare applications, health-tech companies, or medical innovation. These posts discuss technology challenges and economics in non-healthcare domains.
[general_robotics_ai_development]
Learned 3 rejections Inactive
Reject posts about robotics, autonomous systems, or general AI/ML research that lack direct application to healthcare delivery, clinical decision-making, health technology products, or regulated medical domains. This includes synthetic data generation techniques, simulation frameworks, and hardware interfaces developed for non-medical robotics applications, regardless of their potential future healthcare relevance.
[pure_ml_theory_discussions]
Learned 3 rejections Inactive
Reject posts about fundamental machine learning theory, computational history, or algorithmic techniques when discussed in isolation from healthcare applications, health tech products, or health industry problems. Include posts that debate academic lineage or theoretical foundations of ML concepts without connecting to health tech innovation, health system adoption, or healthcare market dynamics.
[tech_industry_narratives]
Learned 3 rejections Inactive
Reject posts about tech industry figure comebacks, non-healthcare AI company pivots, or narratives framed around engineer obsession and redemption arcs in consumer tech or general AI. These posts lack connection to healthcare markets, health systems, payers, providers, or health-specific applications of technology.
[general_economic_philosophy]
Learned 3 rejections Inactive
Reject posts about general economic history, capitalism as a philosophical system, or broad principles of entrepreneurship and innovation that are not explicitly connected to healthcare, health technology, or health markets. These posts should be removed even if they mention productivity, education, or research investment, unless they directly analyze how these principles apply to specific health-tech companies, healthcare business models, or health policy outcomes.
[partisan_bad_faith_activism]
Learned 3 rejections Inactive
Exclude posts that leverage healthcare topics (FDA, vaccines, medical institutions) primarily to advance partisan political narratives, attack individuals or organizations in bad faith, or promote conspiracy claims—rather than analyze healthcare policy or system design.
[off_topic_or_tangential_tech]
Learned 3 rejections Inactive
Exclude posts about local AI inference, LLM optimization, software engineering patterns, data center resource consumption, or open-source AI frameworks—unless they explicitly address healthcare-specific challenges (e.g., clinical data governance, FDA-regulated AI deployment).
[personal_lifestyle_anecdote]
Learned 3 rejections Inactive
Exclude posts that are primarily personal anecdotes about drug use, fitness experimentation, immigration milestones, or individual wellness opinions—even if they mention health topics. Posts must provide healthcare system, policy, or population-level insight, not individual testimonial.
[geopolitical_military_infrastructure]
Learned 3 rejections Inactive
Exclude posts focused on military strikes, geopolitical tensions, international waterway disputes, defense infrastructure, or energy policy that are only tangentially or not at all connected to healthcare delivery, regulation, or patient outcomes.
[general_ai_infrastructure_optimization]
Learned 3 rejections Inactive
Reject posts about general-purpose AI infrastructure improvements, model compression techniques, and local inference optimization that lack explicit connection to healthcare applications or health-tech business models. These posts discuss AI engineering and hardware efficiency in isolation from healthcare markets, payer strategy, health systems, or clinical/health-tech use cases.
[housing_finance_policy]
Learned 3 rejections Inactive
Reject posts about housing finance, mortgage markets, government-sponsored enterprises in the housing sector, and sovereign wealth fund composition that lack direct application to healthcare markets, health technology, or health system strategy. These posts may discuss financial policy or government asset management but do not address health tech innovation, healthcare delivery, or health-specific business models.
[general_job_recruitment]
Learned 3 rejections Inactive
Reject posts about job openings, hiring calls, and recruitment announcements that are primarily soliciting applications or advertising employment opportunities. This includes posts framed as research program positions, fellowship opportunities, or team-building calls, even when posted by individuals or organizations otherwise relevant to healthcare or technology.
[general_tech_sector_news]
Learned 3 rejections Inactive
Reject posts about AI model performance, AI security vulnerabilities, or general technology company stock movements unless they directly address healthcare-specific applications, health data risks, clinical workflows, or health-system integration. Posts that discuss AI safety or AI companies primarily through a tech-sector investment or market lens rather than a health-tech innovation lens should be excluded.
[general_energy_policy]
Learned 3 rejections Inactive
Reject posts about energy pricing, industrial electricity costs, and national competitiveness frameworks unless they directly address healthcare-specific infrastructure costs, hospital operations, or health tech company supply chain impacts. Posts analyzing macroeconomic energy policy divorced from healthcare market effects fall outside newsletter scope.
[tech_industry_compensation]
Learned 3 rejections Inactive
Reject posts about technology industry employee compensation, salary benchmarks across non-healthcare tech companies, or compensation comparisons that don't directly analyze healthcare-specific talent acquisition, retention, or market dynamics affecting health-tech companies.
[general_labor_market_ai_impacts]
Learned 3 rejections Inactive
Reject posts about broad AI impacts on hiring, employment, or general labor market dynamics that lack healthcare-specific framing or implications. These posts treat AI as a general workplace phenomenon rather than exploring healthcare business models, health-tech companies, or healthcare industry dynamics.
[generic_ai_business_opportunity]
Learned 3 rejections Inactive
Reject posts about generic AI business opportunities, entrepreneurial potential, or tool-agnostic startup advice that lack specific application to health technology, healthcare delivery, health insurance, biotech, pharma, or health policy. Posts must demonstrate concrete connection to healthcare markets or health-tech business models to be in scope.
[personal_immigration_career_milestones]
Learned 3 rejections Inactive
Reject posts about individual career transitions, immigration visa applications, personal gratitude statements, or professional milestones that center on an individual's biographical journey rather than analyzing broader healthcare market trends, business strategy, or technology adoption patterns. These posts lack the systems-level or market-level analysis that defines the newsletter's scope.
[political_charity_activism]
Learned 3 rejections Inactive
Reject posts about political figures responding to social hardship through charitable acts, particularly those that frame individual stories as evidence of systemic government failure or use them to promote specific political agendas or candidates. These posts prioritize political activism and social commentary over analysis of healthcare markets, technology, or policy mechanisms.
[general_ai_platform_capabilities]
Learned 3 rejections Inactive
Reject posts about new capabilities, features, or platform announcements for general-purpose AI assistants unless explicitly tied to healthcare applications, medical workflows, health data handling, or health-tech business models. Posts celebrating native agentic features, automation primitives, or cross-platform integrations in non-healthcare AI tools fall outside scope even if tangentially relevant to tech infrastructure.
[geopolitical_military_conflict]
Learned 3 rejections Inactive
Reject posts about military targeting, international conflicts, weapons systems, or geopolitical risk assessments framed through military/defense doctrine, even when they mention technology companies. These posts fall outside the newsletter's coverage of health-tech markets, digital health innovation, and healthcare business strategy.
[tech_industry_leadership_narratives]
Learned 3 rejections Inactive
Reject posts about technology industry leadership, strategic missteps by non-healthcare tech companies, or competitive dynamics between major tech firms (e.g., Meta, Google, Apple) unless the content directly analyzes their entry into healthcare markets, healthcare-specific acquisitions, or healthcare regulatory strategy. Posts framed as technology industry narrative or corporate biography fall outside scope.
[geopolitical_commodities_analysis]
Learned 3 rejections Inactive
Reject posts about geopolitical conflicts, international trade chokepoints, currency systems, shipping logistics, or commodity market dynamics that lack direct connection to healthcare markets, health technology companies, or health policy. This includes analysis of sanctions, de-dollarization, or international finance mechanisms unless explicitly tied to healthcare-specific impacts.
[non_health_sector_corruption]
Learned 3 rejections Inactive
Reject posts about corruption, regulatory capture, or conflict-of-interest allegations in government agencies or industries unrelated to healthcare, health technology, pharmaceuticals, biotech, health insurance, or health systems. This includes water utilities, transportation infrastructure, and other public sector sectors where healthcare is not the primary domain.
[anti_regulation_ideology]
Learned 3 rejections Inactive
Reject posts about regulation that frame it primarily through ideological anti-government or libertarian arguments about bureaucratic capture and compliance costs as general principles. These posts should lack specific reference to health policy mechanics, market structures, or business implications relevant to health tech entrepreneurs and investors.
[ai_industry_competitive_dynamics]
Learned 3 rejections Inactive
Reject posts about competitive disputes between frontier AI model companies, open-source AI platform threats to proprietary vendors, or allegations of anti-competitive behavior in the general AI industry. These posts address AI sector dynamics rather than healthcare-specific applications, business models, or policy.
[software_engineering_tutorials]
Learned 3 rejections Inactive
Reject posts about software development techniques, AI model usage optimization, or coding best practices that lack explicit application to healthcare technology, health data, or health-tech business problems. This includes generic productivity tips for AI tools or software engineering workflows unrelated to healthcare context.
[geopolitical_technology_competition]
Learned 3 rejections Inactive
Reject posts about autonomous vehicles, semiconductor manufacturing, international technology competition, or AI arms races framed primarily as geopolitical/national security issues rather than as business opportunities or challenges within healthcare markets, health insurance operations, or clinical care delivery.
[general_software_engineering]
Learned 3 rejections Inactive
Reject posts about generic software engineering techniques, AI agent development frameworks, and programming platform features that are not specifically applied to healthcare use cases or health-tech business problems. These posts should lack explicit connection to healthcare markets, health data, medical workflows, or health-tech company strategy.
[general_ai_framework_releases]
Learned 3 rejections Inactive
Reject posts about open-source AI frameworks, programming language releases, and general-purpose software tools that are not explicitly positioned as healthcare solutions or integrated into health-tech business strategy. These posts should be excluded even if they mention potential medical applications, unless they analyze actual adoption, regulatory impact, or market implications specific to healthcare companies or health systems.
[cosmetic_wellness_tourism]
Learned 3 rejections Inactive
Reject posts about cosmetic or wellness tourism packages, bundled beauty treatments, medical tourism destinations, or consumer-facing clinic offerings that emphasize aesthetic outcomes over clinical health technology, business model innovation, or healthcare system strategy. These represent consumer health marketing rather than the enterprise, investment, or policy angles the newsletter covers.
[clinical_case_studies]
Learned 3 rejections Inactive
Reject posts about individual patient cases, clinical treatment outcomes, or disease-specific breakthroughs presented primarily as medical narratives rather than as evidence of technology adoption, market disruption, or business model innovation. These posts focus on clinical results rather than the health systems, technologies, companies, or market dynamics driving healthcare change.
[ai_model_safety_discourse]
Learned 3 rejections Inactive
Reject posts about large language model failures, AI model safety concerns, and algorithmic accountability that lack explicit connection to healthcare technology products, clinical applications, or health tech business strategy. This includes broad critiques of AI companies' responses to technical mishaps unless framed through the lens of healthcare-specific harm or healthcare sector implications.
[general_ai_infrastructure_impacts]
Learned 3 rejections Inactive
Reject posts about AI infrastructure's environmental or resource impacts (power consumption, water usage, carbon footprint, supply chain effects) unless explicitly framed around healthcare-specific applications, health tech company operations, or health sector policy responses to AI resource constraints.
[commercial_solar_sales]
Learned 3 rejections Inactive
Reject posts about commercial solar panel installation, roofing sales automation, or solar tax incentive business development. These posts may use healthcare-adjacent AI or automation language but target non-healthcare commercial real estate and energy markets.
[general_ai_infrastructure_news]
Learned 3 rejections Inactive
Reject posts about AI companies' hardware strategies, chip manufacturing, compute infrastructure, or capital allocation decisions that lack explicit connection to healthcare applications, health-tech companies, or healthcare-specific market dynamics. Posts should be excluded unless they directly address how these developments impact healthcare organizations, digital health products, or health systems.
[consumer_enterprise_software_releases]
Learned 3 rejections Inactive
Reject posts about major software product releases, feature deployments, or platform updates from non-healthcare-focused companies, even when those products might tangentially be used in healthcare settings. Focus instead on healthcare-specific applications, health-tech vendors, or enterprise software announcements that directly address healthcare workflows, compliance, or market dynamics.
[geopolitical_conflict_analysis]
Learned 3 rejections Inactive
Reject posts about military conflicts, strikes, warfare, or geopolitical tensions, even when framed around supply chain impacts on commodities or agriculture. The newsletter covers health policy and healthcare economics, not international relations or conflict analysis.
[non_health_tech_equities]
Learned 3 rejections Inactive
Reject posts about equity securities, stock dilution, share offerings, or capital structure decisions at non-health-tech companies. This includes technical analysis of ATMs, shareholder dilution mechanics, and warnings to retail investors about stock price risk in companies outside the health technology, biotech, healthcare services, or health insurance sectors.
[prompt_engineering_ai_tools]
Learned 3 rejections Inactive
Reject posts about AI tool prompts, LLM usage techniques, or prompt engineering frameworks shared primarily as productivity hacks or methodology demonstrations, unless they are specifically contextualized within a healthcare, health-tech, or health policy application and analysis.
[consumer_hardware_speculation]
Learned 3 rejections Inactive
Reject posts about general-purpose computing hardware, semiconductor manufacturing, or speculative consumer tech announcements that lack explicit connection to health systems, payers, clinical delivery, or healthcare regulatory strategy. This includes posts about LLM inference acceleration, local processing capabilities, or edge computing unless directly framed around healthcare applications or health-tech company product roadmaps.
[culture_war_institutional_criticism]
Learned 3 rejections Inactive
Reject posts about institutional controversies framed primarily through partisan or culture-war angles, especially those combining unrelated allegations (medical practices, financial aid, personnel conduct) to delegitimize an organization rather than analyze specific health-tech business dynamics, regulatory outcomes, or market effects.
[generic_business_finance_education]
Learned 3 rejections Inactive
Reject posts about generic financial modeling, unit economics, valuation frameworks, or business education that could apply to any industry and contain no healthcare, health-tech, or health-economy specific context. This includes MBA-style primers, Excel templates, or accounting tutorials presented as broadly applicable business knowledge rather than healthcare-specific analysis.
[judicial_biography_social_critique]
Learned 3 rejections Inactive
Reject posts about judicial decisions, court cases, or judicial philosophy that are framed primarily through personal biography, socioeconomic background critique, or social justice analysis rather than their direct impact on healthcare markets, regulation, or technology adoption.
[wealth_inequality_criticism]
Learned 3 rejections Inactive
Reject posts about wealthy individuals or their family members that use funding disparities, privilege, or personal spending habits as a primary angle for criticism. These posts focus on social inequality or privilege rather than substantive healthcare technology, business strategy, regulation, or market dynamics.
[tangential_geopolitics_loosely_healthcare_framed]
Learned 3 rejections Inactive
Exclude posts where the primary subject is geopolitics, military defense, trade policy, or China competition, and healthcare is merely a secondary frame or mention. Posts about visa bans, AI investment competition, or Chinese government tech announcements should be excluded unless the healthcare system impact is the central analysis.
[unverified_fringe_diet_treatment_claims]
Learned 3 rejections Inactive
Exclude posts that promote ketogenic diets, Ayurveda, alternative medicine, or repurposed drugs for serious diseases (cancer, Parkinson's, aging) without discussing mechanism of action, clinical trial sample sizes, or clearly stating the evidence is preliminary. Claims like 'ketogenic diet cut Parkinson's symptoms by 41%' need context on study rigor.
[partisan_political_messaging_healthcare_framed]
Learned 3 rejections Inactive
Exclude posts where the primary intent is partisan political criticism (attacking a political figure or administration) that uses healthcare as the vehicle for outrage rather than analyzing healthcare policy or systems. Posts from overtly political figures criticizing opposing administrations without substantive healthcare mechanism analysis should be filtered.
[anecdotal_patient_complaint_without_insight]
Learned 3 rejections Inactive
Exclude posts that share personal anecdotes (patient denied coverage, doctor delayed, individual case of illness) without connecting to broader healthcare system patterns, policy implications, or structural solutions. Single stories presented as social commentary without analysis are too narrow.
[vaccine_safety_activism_claims]
Learned 3 rejections Inactive
Exclude posts that make assertions about vaccine adverse events, regulatory misconduct, or safety concerns without citing peer-reviewed clinical evidence or linking to verified safety databases. Activist messaging about vaccines without rigorous evidence analysis should be filtered.
[speculative_unverified_ai_capability_or_timeline_claims]
Learned 3 rejections Inactive
Exclude posts that assert extraordinary AI capabilities (e.g., 'AGI already happened', 'AI singularity', autonomous agents solving complex problems) or make speculative technology predictions without citing validated benchmarks, peer review, or demonstrated real-world results in healthcare contexts.
[tangential_tech_or_business_trend_loosely_healthcare_framed]
Learned 3 rejections Inactive
Exclude posts about general tech trends, business strategy, geopolitical events, or economic policy that mention healthcare only in passing or as a loose analogy, unless the healthcare application is explicit, detailed, and central to the post's argument.
[unverified_fringe_medical_claims_scaremongering]
Learned 3 rejections Inactive
Exclude posts that make sweeping health or safety claims (e.g., vaccine side effects, miracle diet effects, unproven treatment benefits) presented as fact without citing peer-reviewed clinical evidence, regulatory approval status, or peer consensus—especially if framed to alarm or provoke distrust.
[anecdotal_health_story_without_systemic_insight]
Learned 3 rejections Inactive
Exclude posts that lead with personal anecdotes, individual patient cases, or one-off health stories (e.g., someone's cancer diagnosis, delayed chemo appointment, immigration hardship) without connecting to systemic healthcare patterns, policy implications, or scalable solutions.
[ai_product_announcement_without_healthcare_application]
Learned 3 rejections Inactive
Exclude posts that announce AI product launches, feature releases, or capability improvements (e.g., Claude features, Shopify integrations, research agents) unless they explicitly demonstrate application to a specific healthcare problem or clinical workflow.
[regulatory_enforcement_alert_without_systemic_analysis]
Learned 3 rejections Inactive
Exclude posts that report healthcare fraud, regulatory warnings, malpractice cases, or law enforcement actions (FDA warning letters, ProPublica fraud alerts) as breaking news or alert items without connecting them to a systemic problem or broader healthcare policy lesson.
[tangential_tech_or_business_trend_loosely_framed_as_healthcare]
Learned 3 rejections Inactive
Exclude posts where healthcare is mentioned only in the headline match or as a passing reference, but the core content focuses on general business trends (startup funding, career advice, stock markets, geopolitical conflict, AI infrastructure costs) with no substantive healthcare application or insight.
[unverified_fringe_medical_claim_without_context]
Learned 3 rejections Inactive
Exclude posts that promote unverified medical claims about diet, supplements, alternative medicine, or single-study findings (ketogenic diet, hara hachi bu, insulin-obesity claims) without rigorous explanation of methodology, limitations, or how the finding fits into established medical science.
[anecdotal_patient_story_without_systemic_insight]
Learned 3 rejections Inactive
Exclude posts that rely on a single personal anecdote, patient story, or individual healthcare worker complaint (canceled appointments, delayed surgeries, hiring/immigration barriers) without connecting it to a systemic problem, data trend, or policy mechanism.
[government_health_infrastructure_announcements]
Learned 3 rejections Inactive
Reject posts about government healthcare infrastructure initiatives, medical education capacity expansions, or health workforce development programs when framed as policy announcements rather than market analysis, technology adoption, business model disruption, or investment opportunity. These lack the commercial, technological, or strategic business angle required for coverage.
[partisan_health_policy_criticism]
Learned 3 rejections Inactive
Reject posts about government health policy that frame announcements primarily as political theater or bad-faith governance, rather than analyzing the policy's structural impact on healthcare markets, technology adoption, or business strategy. These posts typically criticize implementation failures or administrative negligence without connecting to the newsletter's focus on how policy changes affect health tech companies, payers, providers, or investors.
[crypto_trading_and_speculation]
Learned 3 rejections Inactive
Reject posts about cryptocurrency trading, blockchain arbitrage opportunities, copytrading bots, or financial speculation using AI agents—even if they mention Claude or AI tools. These posts address speculative financial markets and consumer trading strategies unrelated to healthcare technology, health systems, or health-sector business models.
[macroeconomic_development_policy]
Learned 3 rejections Inactive
Reject posts about national economic development, talent allocation across entire economies, or sectoral policy reform that lack a direct connection to health technology, healthcare delivery, health insurance, or health-related innovation. This includes posts analyzing manufacturing capacity, port logistics, or cross-sector talent distribution unless explicitly framed around healthcare-specific markets or health tech companies.
[vaccine_safety_conspiracy_narratives]
Learned 3 rejections Inactive
Reject posts that center on alleged harms from specific pharmaceutical interventions (vaccines, treatments, drugs) based on reanalysis of trial data, claims of regulatory capture or institutional malfeasance, or calls to distrust established medical guidance. These posts typically frame themselves as exposing hidden truths but fall outside the newsletter's focus on health-tech innovation, business models, and policy mechanisms.
[autonomous_vehicle_policy]
Learned 3 rejections Inactive
Reject posts about autonomous vehicle safety, regulation, or policy deployment that use healthcare or regulatory systems as comparative examples but lack substantive connection to health technology, healthcare markets, or health-specific innovation. These posts typically invoke healthcare/regulatory contexts argumentatively rather than analytically examining health-tech business or policy.
[broad_stock_market_recap]
Learned 3 rejections Inactive
Reject posts that provide broad stock market recaps or multi-sector earnings summaries unless the primary thesis directly addresses a health technology company, healthcare payer, biotech firm, or health policy development. General market volatility commentary, cross-sector performance comparisons, and macro economic updates unrelated to healthcare business strategy or innovation are out of scope.
[general_ai_tool_applications]
Learned 3 rejections Inactive
Reject posts about general-purpose AI agents, research automation tools, or productivity software that lack healthcare-specific context or application. Posts should be excluded if they discuss AI tools generically or in non-healthcare contexts (e.g., landing page optimization, general startup productivity), even if the tool could theoretically be applied to health tech.
[ai_infrastructure_costs]
Learned 3 rejections Inactive
Reject posts about AI model pricing, token costs, or LLM infrastructure expenses that lack explicit connection to healthcare delivery, health-tech product development, or healthcare business models. These discussions belong in AI/tech infrastructure coverage, not health-tech strategy.
[general_business_automation]
Learned 3 rejections Inactive
Reject posts about automation, autonomous agents, or decision architecture in general commercial or enterprise contexts unless explicitly connected to healthcare operations, clinical workflows, or health system infrastructure. Posts that apply general business automation principles without healthcare application specificity should be filtered.
[general_ai_developer_tools]
Learned 3 rejections Inactive
Reject posts about general artificial intelligence developer tools, APIs, or research platforms that are not specifically designed for or applied to healthcare use cases. Posts promoting broadly applicable AI infrastructure (research indexing, agent frameworks, LLM tooling) without healthcare-specific context or business model implications fall outside scope.
[non_health_tech_company_announcements]
Learned 3 rejections Inactive
Reject posts about major product announcements or feature launches from non-health-tech companies, even when those features could theoretically be applied to healthcare contexts. The focus should be on companies, products, and innovations explicitly operating in health technology, healthcare delivery, or healthcare-adjacent markets—not general technology infrastructure providers discussing their core business developments.
[patient_advocacy_anecdotes]
Learned 3 rejections Inactive
Reject posts about isolated patient experiences, medical malpractice anecdotes, or individual cases of poor clinical care shared as cautionary tales. These posts frame healthcare problems through personal narrative rather than through the lens of market structure, technology, policy, or business strategy.
[public_health_agency_controversies]
Learned 3 rejections Inactive
Reject posts about government health agency decisions, vaccine policy announcements, or public health controversies framed around political administration actions or delays. These topics lack direct relevance to health tech innovation, digital health products, healthcare business models, or investment strategy.
[medical_malpractice_litigation]
Learned 3 rejections Inactive
Reject posts about individual medical malpractice cases, personal injury litigation, or hospital bankruptcy-related claims disputes. These focus on legal accountability for patient harm rather than analyzing health-tech markets, payer strategy, care delivery innovation, or healthcare policy.
[illicit_drug_safety_warnings]
Learned 3 rejections Inactive
Reject posts about safety issues, quality control failures, or health risks associated with illicit or unregulated drug markets. These posts address consumer drug safety rather than health technology markets, healthcare business strategy, or regulatory policy affecting legitimate innovators and institutions.
[clinical_disease_epidemiology]
Learned 3 rejections Inactive
Reject posts about disease epidemiology, clinical disease trends, and patient experience narratives that frame health conditions as medical mysteries or social problems without connection to technology innovation, business models, investment opportunities, or healthcare system strategy. These posts focus on clinical and public health concerns rather than the technology and business infrastructure of healthcare.
[consumer_nutrition_advice]
Learned 3 rejections Inactive
Reject posts about personal nutrition practices, eating habits, and dietary behavior modification that lack a health-tech, business, policy, or market angle. This includes trending wellness concepts, food science commentary, and satiety discussions presented primarily as consumer health advice rather than as infrastructure, industry, or innovation opportunities.
[general_business_narrative]
Learned 3 rejections Inactive
Reject posts about general business success stories, entrepreneurship lessons, and economic narratives that happen to involve financial services or other industries but lack substantive connection to healthcare markets, health technology, or healthcare-specific business dynamics. These may be well-crafted narratives but fall outside the newsletter's healthcare-focused scope.
[general_developer_tools]
Learned 3 rejections Inactive
Reject posts about general software development tools, programming features, or code development platforms unless they are explicitly positioned within a health-specific use case or healthcare application context. Posts should demonstrate direct application to healthcare technology, medical data workflows, or health system operations to be included.
[partisan_political_criticism]
Learned 3 rejections Inactive
Reject posts about political figures or administrations framed primarily as personal criticism, broken campaign promises, or partisan blame rather than substantive analysis of how specific regulatory changes, policy shifts, or legislative actions affect health-tech markets, business models, or investment opportunities.
[generic_edtech_promotions]
Learned 3 rejections Inactive
Reject posts about online courses, professional certifications, and educational platforms that are promoted as general skill-building offerings rather than analyzed as healthcare market phenomena, competitive threats to health-tech incumbents, or policy-relevant workforce development initiatives.
[patient_care_access_stories]
Learned 3 rejections Inactive
Reject posts about individual patient experiences with insurance coverage decisions, treatment denials, or appointment cancellations framed as personal stories. These posts, while emotionally compelling, lack analysis of underlying business models, policy mechanisms, technology platforms, or market dynamics that would be relevant to health-tech entrepreneurs, investors, and strategists.
[niche_technical_ai_infrastructure_without_healthcare_application]
Learned 3 rejections Inactive
Exclude posts discussing reinforcement learning fine-tuning, sandbox isolation techniques, Model Context Protocol implementations, world models, or AI agent infrastructure patterns where the healthcare application is vague, theoretical, or absent—focus is on technical capability rather than healthcare problem-solving.
[unverified_fringe_medical_treatment_claims]
Learned 3 rejections Inactive
Exclude posts promoting ivermectin, mebendazole, ketamine, or other repurposed/unregulated treatments as cancer cures or major clinical breakthroughs with percentages of 'clinical benefit' or 'disappearance rates' not published in peer-reviewed journals or validated by major medical organizations. Include fringe diet claims presented as disease cures.
[career_guidance_vocational_advice]
Learned 3 rejections Inactive
Reject posts about career pivots, vocational training programs, degree options, and professional development pathways. These posts frame health through the lens of individual job satisfaction and educational credentials rather than market dynamics, technology innovation, or healthcare business strategy.
[clinical_trial_consumer_nutrition]
Learned 3 rejections Inactive
Reject posts about clinical trial results for dietary or lifestyle interventions (ketogenic diets, supplements, exercise protocols, etc.) when framed primarily as consumer health advice or direct-to-patient treatment recommendations, rather than as business opportunities, technology platforms, or policy implications affecting health systems or payers.
[general_ai_productivity_trends]
Learned 3 rejections Inactive
Reject posts about artificial intelligence tools and productivity optimization that treat AI as a generic workforce multiplier or business efficiency lever without analyzing health-tech-specific implications, healthcare market dynamics, regulatory constraints, or domain-specific use cases in medicine, payers, providers, or health systems.
[general_ai_philosophy]
Learned 3 rejections Inactive
Reject posts about general AI philosophy, consciousness, or scaling theory that lack explicit connection to health technology applications, healthcare business models, clinical outcomes, or health-specific regulation. Posts should be excluded if they treat health tech merely as an example of broader AI principles rather than examining health-tech markets, companies, or policy directly.
[foundational_ai_research]
Learned 3 rejections Inactive
Reject posts about foundational AI research, machine learning theory, and model training techniques that discuss general AI capabilities without explicit connection to healthcare applications, medical use cases, or health-tech business strategy. This includes papers and discussions on reasoning circuits, reinforcement learning generalization, or model scaling that treat healthcare as an incidental application domain rather than the primary focus.
[generic_entrepreneurship_narratives]
Learned 3 rejections Inactive
Reject posts about entrepreneurial retrospectives, startup failure/success storytelling, and personal founder journeys that do not explicitly address health technology, healthcare markets, medical innovation, or health-system business strategy. These posts may mention AI, SaaS, or business lessons but apply them to non-healthcare domains or generic startup advice.
[basic_disease_biology]
Learned 3 rejections Inactive
Reject posts about basic disease biology, viral mechanisms, cellular pathology, or symptom etiology that lack a health technology, business, policy, or investment angle. Posts about what a disease does to the body, without connecting to a product, company, care model, payer strategy, or regulatory framework, fall outside newsletter scope.
[geopolitical_military_analysis]
Learned 3 rejections Inactive
Reject posts about military capabilities, geopolitical conflict preparation, defense strategies, weapons systems development, or international security tensions—even when framed as technology news or innovation announcements. These posts lack direct relevance to healthcare markets, digital health innovation, or health policy.
[academic_research_tools]
Learned 3 rejections Inactive
Reject posts about open-source or commercial AI tools designed to automate academic research workflows (literature review, paper writing, citation management, rebuttal generation) unless the post explicitly connects the tool to healthcare delivery, health economics, biomedical research, or a health-tech business application.
[general_ai_engineering_patterns]
Learned 3 rejections Inactive
Reject posts about general software engineering practices, LLM agent design patterns, or AI development methodologies that lack healthcare-specific context or application. These posts should address healthcare technology specifically (e.g., clinical AI, healthcare data systems, health tech infrastructure) rather than treating healthcare as one example among many of broader AI/engineering trends.
[generic_personal_finance_advice]
Learned 3 rejections Inactive
Reject posts about generic wealth accumulation strategies, asset ownership frameworks, or personal finance principles that lack specific application to healthcare markets, health-tech business models, or healthcare investment. Posts must connect explicitly to one of the newsletter's core verticals (digital health, healthtech, healthcare payers, providers, biotech, or health policy) rather than treating healthcare as an incidental example of broader financial principles.
[antitrust_conspiracy_allegations]
Learned 3 rejections Inactive
Reject posts about alleged coordinated industry efforts to suppress competitors or technologies, especially those framed as conspiracy theories without substantive market analysis, regulatory documentation, or competitive data. These posts typically call for investigation or legal action based on speculation rather than demonstrable market conditions or documented business practices.
[clinical_pathophysiology_debates]
Learned 3 rejections Inactive
Reject posts about mechanistic claims regarding disease pathophysiology, metabolic processes, or clinical observations that lack explicit connection to a health-tech product, business model, policy change, or investment thesis. Posts should focus on *what is being built or funded to address a problem*, not on establishing or debating the underlying scientific mechanism of the problem itself.
[general_software_security]
Learned 3 rejections Inactive
Reject posts about generic software architecture patterns, container security, or infrastructure isolation techniques unless they are explicitly framed in the context of healthcare-specific implementation challenges, health data protection, or solutions to a documented healthcare technology problem.
[labor_law_employment_policy]
Learned 3 rejections Inactive
Reject posts about broad labor law changes, employment regulations, and worker protections that are not explicitly framed around health-tech company strategy, healthcare workforce dynamics, or business model implications within the health-tech ecosystem. General employment policy wins or losses for workers fall outside scope unless directly tied to health-tech venture strategy, care delivery disruption, or healthcare market structure.
[general_ai_philosophy_claims]
Learned 3 rejections Inactive
Reject posts about whether AGI has been achieved, general claims about AI capability thresholds, or philosophical arguments about AI progress that lack explicit connection to health technology products, healthcare markets, medical innovation, health policy, or healthcare business models.
[tangential_tech_without_healthcare_application]
Learned 3 rejections Inactive
Exclude posts about general AI capabilities, ML engineering patterns, software productivity, time-series forecasting, or tech infrastructure trends that mention healthcare only tangentially or use healthcare as a generic example rather than the core focus.
[bad_faith_political_cynicism]
Learned 3 rejections Inactive
Exclude posts that express partisan political criticism, conspiracy allegations, or cynical attacks on government/institutions in healthcare contexts without offering evidence-based analysis, proposed solutions, or substantive healthcare policy insight.
[clinical_research_reviews]
Learned 3 rejections Inactive
Reject posts about clinical research findings, disease mechanism reviews, or drug efficacy studies that are presented primarily as medical science without connection to health technology products, business models, market dynamics, regulatory strategy, or investment opportunities. This includes literature reviews and clinical trial analyses framed for medical professionals rather than health tech entrepreneurs or investors.
[clinical_medicine_debates]
Learned 3 rejections Inactive
Reject posts about clinical medical debates, pathophysiology discussions, or pharmaceutical treatment efficacy that are framed as peer-to-peer physician education rather than as commentary on health technology, market dynamics, reimbursement policy, or business strategy. This includes debates about cholesterol management, medication effectiveness, or diagnostic approaches absent any health-tech or commercial healthcare angle.
[general_ai_machine_learning]
Learned 3 rejections Inactive
Reject posts about foundational AI/ML concepts, model training techniques, or AI research that lack explicit connection to healthcare, health-tech companies, medical applications, or health systems. This includes theoretical discussions of reward signals, training methodologies, or AI capabilities presented as general technology commentary rather than health-specific innovation or strategy.
[general_ai_capability_benchmarks]
Learned 3 rejections Inactive
Reject posts about AI model performance comparisons, research paper quality benchmarks, or autonomous system competitions that are not explicitly connected to healthcare applications, health-tech business outcomes, or medical/clinical use cases. This includes posts framing AI capability advances as significant without demonstrating health-tech market or regulatory implications.
[unverified_ai_capability_claims_agent_work]
Learned 3 rejections Inactive
Exclude posts that make sweeping claims about AI agents automating healthcare roles, solving clinical problems, or achieving healthcare outcomes without citing validation studies, pilot results, or published evidence.
[drug_mechanism_education]
Learned 3 rejections Inactive
Reject posts about drug mechanisms of action, disease biology, or cellular/molecular pharmacology presented primarily for educational or clinical understanding purposes. These posts focus on the scientific basis of therapies rather than the business models, market dynamics, regulatory strategy, or investment implications of pharmaceutical innovation.
[ai_safety_philosophy]
Learned 3 rejections Inactive
Reject posts about artificial general intelligence (AGI) strategy, AI safety principles, human agency in AI systems, or AI alignment frameworks when presented as abstract philosophical or policy concerns disconnected from specific healthcare applications, health tech products, or medical market dynamics.
[direct_to_consumer_product_reviews]
Learned 3 rejections Inactive
Reject posts about individual consumer experiences with direct-to-consumer beauty, skincare, or wellness products, particularly those focused on personal results, product recommendations, or purchasing experiences. These posts lack analysis of market dynamics, business models, or healthcare/health-tech implications.
[general_ai_product_development]
Learned 3 rejections Inactive
Reject posts about how non-healthcare AI companies build products, manage design processes, or ship features—unless the post explicitly connects these practices to a specific health-tech application, healthcare use case, or health system implementation. Posts about general AI development velocity, design methodologies, or internal company processes at consumer or enterprise software firms are off-topic.
[basic_clinical_breakthroughs]
Learned 3 rejections Inactive
Reject posts about clinical treatment successes, disease remission outcomes, or therapeutic efficacy for specific patient populations unless the post explicitly connects these developments to health-tech platforms, digital health business models, healthcare investment, or policy/regulatory implications affecting the market.
[political_healthcare_advocacy]
Learned 3 rejections Inactive
Reject posts on international healthcare challenges
[general_software_engineering_tools]
Learned 3 rejections Inactive
Reject posts about software engineering tools, APIs, frameworks, and developer platforms that lack explicit application to healthcare or health-tech use cases. These posts may discuss AI agents, monitoring systems, or infrastructure improvements but do not address healthcare-specific problems, markets, regulations, or business models.
[healthcare_fraud_enforcement_reporting]
Learned 3 rejections Inactive
Exclude posts that are primarily crime/fraud reporting (specific arrests, guilty pleas, FBI investigations, Medicaid fraud cases) or government enforcement announcements without analysis of systemic healthcare policy implications or fraud prevention mechanisms.
[ai_agent_product_hype_drama]
Learned 3 rejections Inactive
Exclude posts that are essentially product announcements or technical hype for AI agent platforms (Managed Agents, Agent Builder) unless they specifically analyze healthcare application, clinical workflow integration, or healthcare business model impact.
[truncated_low_effort_posts]
Learned 3 rejections Inactive
Exclude posts that end abruptly with incomplete sentences, ellipses, or clearly truncated thoughts where the full argument cannot be evaluated (text ending with 'St' or 'the fight has literally just begun' without follow-up).
[political_conspiracy_bad_faith]
Learned 3 rejections Inactive
Exclude posts that frame healthcare or regulatory issues through partisan political conspiracy (RFK Jr. being 'sidelined,' government 'cage,' evidence destruction cover-ups, Deep State narratives) rather than substantive policy analysis.
[partisan_political_commentary]
Learned 3 rejections Inactive
Reject posts about partisan political movements, administrative state critiques, or ideological battles framed around government control and institutional reform. These posts lack substantive connection to specific health technology, healthcare market dynamics, policy implementation affecting healthcare delivery, or healthcare-adjacent business strategy.
[public_health_infrastructure_policy]
Learned 3 rejections Inactive
Reject posts about government public health infrastructure, disease surveillance systems, or national health security capabilities—particularly those framed around government agency performance, dismantling, or geopolitical health security concerns. These posts discuss public health governance and policy outcomes rather than the commercial health technology, business models, investment opportunities, or market dynamics that the newsletter covers.
[macro_economic_commentary]
Learned 3 rejections Inactive
Reject posts about general economic conditions, labor market trends, GDP growth, unemployment rates, or productivity metrics that lack explicit connection to healthcare systems, health tech markets, payer/provider economics, or health policy. Posts must address healthcare-specific economic dynamics to be in-scope.
[veterinary_longevity_research]
Learned 3 rejections Inactive
Reject posts about veterinary medicine, pet pharmaceuticals, and animal health research
[medical_regulation_activism]
Learned 3 rejections Inactive
Reject posts about healthcare outside of the United States.
[alternative_medicine_regulation]
Learned 3 rejections Inactive
Reject posts about non-USA healthcare.
[consumer_insurance_disputes]
Learned 3 rejections Inactive
Reject posts about individual consumer insurance claim denials, consumer court litigation outcomes, or tactical advice for fighting rejected claims. These posts frame health insurance through the lens of consumer advocacy and dispute resolution rather than market dynamics, product innovation, payer strategy, or systemic business implications.
[partisan_government_accountability]
Learned 3 rejections Inactive
Reject posts about partisan attacks on government officials, inter-agency blame for policy failures, or criticism framed primarily as political accountability rather than substantive healthcare market or regulatory analysis. These posts lack relevance to health-tech business strategy, innovation, investment, or policy impact on the covered sectors.
[generic_ai_architecture_trends]
Learned 3 rejections Inactive
Reject posts about AI agent architecture, token economics, or computational paradigms that discuss these concepts in generic technology contexts without explicit connection to healthcare delivery, health tech products, medical data, healthcare payers, providers, or health-related business models.
[medical_ethics_enforcement]
Learned 3 rejections Inactive
Reject posts about individual practitioners' criminal conduct, license revocation, or ethical violations. These posts focus on professional discipline and enforcement rather than the technology, business models, market dynamics, or policy frameworks that drive healthcare innovation and strategy.
[unproven_alternative_cancer_treatments]
Learned 3 rejections Inactive
Reject posts about drugs or alternative therapeutic combinations based on preliminary patient-reported findings. This includes content framing such treatments as breakthroughs requiring large-scale clinical trials or positioning them as leadership in healthcare innovation.
[unvalidated_repurposed_drug_claims]
Learned 3 rejections Inactive
Reject posts about repurposed drugs or off-label treatments presented with clinical outcome data, efficacy percentages, or patient testimonials—especially when framed as urgent medical breakthroughs outside traditional regulatory pathways. These posts conflate clinical evidence claims with health-tech business content and fall into direct medical advice territory.
[partisan_political_personnel_drama]
Learned 3 rejections Inactive
Reject posts about personnel removals, internal political conflicts, or administrative drama within government agencies when framed primarily as partisan narrative rather than substantive analysis of policy outcomes, regulatory changes, or market impact. Focus on the mechanism and market effect, not the political theater or personality-driven speculation.
[general_financial_trading]
Learned 3 rejections Inactive
Reject posts about AI systems, trading algorithms, or investment strategies applied to general financial markets, commodities, or broad equity indices. These posts lack specific application to healthcare, health technology, or health-related business models and markets.
[unvalidated_cancer_treatment_claims]
Learned 3 rejections Inactive
Reject posts about repurposed drugs or off-label treatments for serious diseases that claim significant clinical efficacy based on real-world observational data, small cohorts, or preprints lacking peer review and regulatory validation. These posts typically bypass standard pharmaceutical development and clinical evidence hierarchies that the newsletter covers.
[speculative_unregulated_market_opinion]
Learned 3 rejections Inactive
Exclude posts that speculate on unregulated peptide or GLP-1 markets, offer lifestyle opinions about weight loss drugs, or discuss black-market therapeutics without regulatory analysis, safety data, or healthcare system implications. Market gossip and personal usage anecdotes do not qualify.
[geopolitical_defense_policy]
Learned 3 rejections Inactive
Reject posts about geopolitical alliances, defense spending, arms sales, and military industrial strategy unless directly connected to healthcare-specific policy, regulation, or market dynamics. Posts analyzing foreign policy, NATO, defense budgets, or weapons platforms as primary subjects are out of scope, even when framed through economic or commercial lenses.
[institutional_abuse_cover_ups]
Learned 3 rejections Inactive
Reject posts about alleged government or institutional cover-ups of crime, evidence destruction, or abuse of power that are framed as investigative accountability journalism rather than health-tech industry analysis. This includes posts calling for support of independent investigators exposing institutional failures, regardless of the underlying subject matter, unless the investigation directly concerns healthcare regulation, payer fraud, or health-tech company misconduct.
[biotech_financial_speculation]
Learned 3 rejections Inactive
Exclude posts that focus primarily on drug approval timelines, pharmaceutical market share competition, stock outlook, or genetic modifiers of drug efficacy when the purpose is financial speculation rather than healthcare system analysis.
[labor_market_disruption_without_healthcare_specificity]
Learned 3 rejections Inactive
Exclude posts that frame AI job displacement or healthcare workforce burnout as inevitable macro trends without analyzing how specific healthcare business models, technology platforms, or regulatory changes actually address labor challenges. Posts should propose solutions, not just describe problems.
[speculative_unverified_peptide_claims]
Learned 3 rejections Inactive
Exclude posts that speculate about peptide efficacy, off-label uses, genetic modifiers, or mental health effects without peer-reviewed evidence or FDA context. Posts about unregulated peptide markets, 'dirty peptides,' or anecdotal benefit claims should be rejected.
[ai_agent_infrastructure_launch_hype]
Learned 3 rejections Inactive
Exclude posts that treat AI agent platform launches (Anthropic Managed Agents, OpenAI Agent Builder, etc.) as major news events without demonstrating specific healthcare use cases or clinical impact. These posts celebrate the technology itself rather than how it solves healthcare problems.
[healthcare_worker_labor_and_burnout]
Learned 3 rejections Inactive
Exclude posts that highlight healthcare worker labor challenges (nursing shortages, physician burnout, practitioner departures, burnout statistics) as anecdotal complaints, trend observations, or emotional testimonies without proposing structural changes, business model implications, or technology-enabled solutions to the underlying workforce challenge.
[policy_announcement_without_healthcare_analysis]
Learned 3 rejections Inactive
Exclude posts that function primarily as political news or policy announcements (executive orders, dietary guidelines, H-1B visas, microplastics regulation, trade deals) without substantive healthcare business, delivery system, or technology implications. Posts must explain how the policy concretely affects healthcare operations or outcomes, not just celebrate or report the announcement.
[healthcare_fraud_and_enforcement_reporting]
Learned 3 rejections Inactive
Exclude posts that function primarily as law enforcement or fraud alert reporting (FBI/HHS fraud busts, Medicaid fraud schemes, CMS enforcement actions) without providing systemic healthcare insight, policy analysis, or broader implications for healthcare operations or business models.
[ai_agent_infrastructure_announcement]
Learned 3 rejections Inactive
Exclude posts that announce or praise AI agent infrastructure products (Anthropic Managed Agents, Claude Code, OpenAI Agents) purely as developer tools or operational wins, without demonstrating a concrete healthcare use case, clinical problem solved, or healthcare-specific insight. Posts must show healthcare application, not just infrastructure capability.
[personal_productivity_tools]
Learned 3 rejections Inactive
Reject posts about personal AI productivity systems, note-taking applications, or general-purpose AI tools used for individual knowledge management. These posts focus on consumer software workflows rather than healthcare-specific technology, business models, or market dynamics.
[general_ml_engineering_practices]
Learned 3 rejections Inactive
Reject posts about machine learning engineering practices, model development techniques, or software engineering challenges that are framed as general ML/AI problems rather than as solutions to specific healthcare business problems, clinical workflows, or health technology applications. This includes posts about foundation models, training pipelines, or data engineering approaches presented primarily as engineering best practices without healthcare context.
[generic_ai_developer_tools]
Learned 3 rejections Inactive
Reject posts about generic AI agent platforms, LLM infrastructure, or software development tools that discuss implementation details and technical capabilities without connecting to healthcare applications, health system workflows, payer operations, or health-tech business strategy. These posts treat healthcare as an incidental use case rather than the primary subject.
[software_engineering_productivity]
Learned 3 rejections Inactive
Reject posts about software engineering workflows, developer productivity techniques, and AI-assisted coding practices that lack explicit application to healthcare technology, health systems, or life sciences domains. Posts should be rejected even if they mention healthcare tangentially unless they specifically address healthcare use cases, healthcare company engineering challenges, or health tech infrastructure.
[social_safety_net_policy]
Learned 3 rejections Inactive
Reject posts about changes to non-health social safety net programs, including food assistance, welfare work requirements, and state-level benefit administration changes. These posts focus on social policy impacts rather than health technology, payer strategy, care delivery, or healthcare-specific regulation.
[law_enforcement_fraud_cases]
Learned 3 rejections Inactive
Reject posts about criminal fraud cases, law enforcement investigations, and fraud takedowns in healthcare unless they draw broader conclusions around healthcare markets or technology beyond a simple crime report. These posts focus on prosecutorial outcomes and individual criminal cases rather than on systemic market dynamics, technology trends, business strategy, policy mechanisms, or investment opportunities that affect the health tech and healthcare innovation landscape.
[individual_drug_approvals_and_launches]
Learned 3 rejections Inactive
Reject posts about individual pharmaceutical product approvals, regulatory clearances, or commercial launch in non-US markets that do not connect to broader US healthcare health tech business models, market structure changes, or investment implications. These posts treat drugs as consumer products rather than as signals about healthcare systems, payer strategy, or technology-enabled care delivery.
[clinical_practice_guidelines]
Learned 3 rejections Inactive
Reject posts about clinical practice guidelines, treatment protocols, or therapeutic decision-making for practicing clinicians. These posts focus on how physicians should diagnose and treat specific patient populations, rather than on the technology, business models, markets, or policy infrastructure that enable healthcare delivery.
[speculative_biotech_financial_opinion]
Learned 3 rejections Inactive
Exclude posts that analyze which biotech company will win a market (Novo vs. Lilly in GLP-1), predict pharmaceutical stock performance, or discuss financial positioning without connecting to patient outcomes, clinical efficacy differences, or healthcare delivery implications.
[tangential_infrastructure_not_healthcare]
Learned 3 rejections Inactive
Exclude posts that focus on AI infrastructure challenges (electrical grids, transformers, decompilation security, code secrecy) or environmental impact unless explicitly connected to healthcare deployment or clinical operations. General tech infrastructure posts belong in tech media, not healthcare.
[unverified_peptide_market_claims]
Learned 3 rejections Inactive
Exclude posts that discuss unregulated peptides (LOY-002, semaglutide off-label, Chinese research peptides) with claims about efficacy, safety benefits, or market traction that lack peer-reviewed clinical trial data or FDA approval status. Posts must cite specific trial data, not anecdotal reports.
[ai_product_launch_announcements]
Learned 3 rejections Inactive
Unless specifically healthcare related, Reject posts about AI platform feature launches, infrastructure tool releases, or generic AI capability announcements that lack healthcare-specific applications, market strategy analysis, or implications for health-tech business models. Posts must demonstrate direct relevance to healthcare markets, care delivery, payer strategy, or health-tech investment thesis—not simply mention that an AI tool might theoretically apply to healthcare.
[basic_cardiovascular_physiology]
Learned 3 rejections Inactive
Reject posts about basic cardiovascular or physiological mechanisms (e.g., why veins versus arteries develop plaque, how blood vessels function) that are framed as educational content or scientific curiosities rather than as drivers of technology, business model, or policy outcomes in healthcare markets.
[generic_ai_product_launches]
Learned 3 rejections Inactive
Reject posts about non healthcare, pharma, Medtech, biotech or health tech AI product launches, tool releases, or feature announcements that lack explicit connection to healthcare applications, health-tech companies, medical workflows, or health-related industries. Posts listing generic use cases across consumer, finance, legal, and business domains without healthcare-specific angles should be excluded even if healthcare is mentioned as one of many verticals.
[veterinary_drug_development]
Learned 3 rejections Inactive
Reject posts about animal/veterinary.
[unregulated_peptide_market_opinion]
Learned 3 rejections Inactive
Exclude posts that discuss GLP-1 drugs, peptides, or weight-loss medications primarily as commodity market opinion, personal efficacy anecdotes, genetic variation trivia, or unregulated market commentary without addressing healthcare delivery, regulatory gaps, clinical outcomes measurement, or healthcare business model implications.
[government_enforcement_and_fraud_alerts]
Learned 3 rejections Inactive
Exclude posts that are primarily news alerts about DOJ investigations, fraud prosecutions, or law enforcement actions (hospice fraud, Medicaid fraud, IRS enforcement). Unless the post provides healthcare systems insight or policy implications for the writer's audience, these are breaking news crime reporting, not healthcare tech analysis.
[pharma_ideology_critique]
Learned 3 rejections Inactive
Reject posts about pharmaceutical development that frame the regulatory system primarily as bureaucratic theater, absurdist maze-building, or ideological performance rather than engaging with specific regulatory mechanisms, cost structures, incentive misalignments, or business model implications. These posts typically dismiss rather than analyze policy, and lack the practical focus on how regulatory changes affect company strategy, market entry, or healthcare outcomes.
[lifestyle_genetic_health_opinion]
Learned 3 rejections Inactive
Exclude posts discussing genetic risk factors, lifestyle interventions (diet, exercise, vaccination), or personal health optimization unless the post addresses healthcare delivery innovation, policy reform, or market disruption.
[tangential_ai_infrastructure_environmental]
Learned 3 rejections Inactive
Exclude posts focused on AI infrastructure scaling, electricity consumption, data center construction, or token economics unless the post specifically addresses healthcare capacity, deployment barriers, or clinical system constraints.
[peptide_unregulated_market_opinion]
Learned 3 rejections Inactive
Exclude posts about peptides, semaglutide supply, compounding regulations, or illegal pharmaceutical smuggling that lack clinical evidence or systems-level healthcare policy analysis. Posts must go beyond commodity market commentary or regulatory compliance alerts.
[healthcare_opinion_without_mechanism]
Learned 3 rejections Inactive
Exclude posts that state healthcare problems or opinions (e.g., 'physicians leaving medicine,' 'doctor burnout') without explaining *why* this is happening or how it connects to business/technology/policy mechanisms. Posts must move beyond problem assertion to analysis.
[tangential_tech_infrastructure]
Learned 3 rejections Inactive
Exclude posts about AI infrastructure scaling, electricity consumption, data center buildout, or general tech trends that mention healthcare only in passing or as one example among many industries. Posts must center on healthcare-specific technology implications.
[product_announcement_without_analysis]
Learned 3 rejections Inactive
Exclude posts that are primarily product launch announcements, feature descriptions, or promotional content for AI tools (Claude Managed Agents, Alloy partnerships, etc.). Posts must provide strategic insight into how the product reshapes healthcare, not just describe what it does.
[government_enforcement_reporting]
Learned 3 rejections Inactive
Exclude posts that are primarily government announcements, law enforcement quotes, or fraud investigation updates. Posts must include substantive healthcare industry insight or business model implications, not just report that enforcement is happening.
[ai_capability_scaremongering]
Learned 3 rejections Inactive
Exclude posts that discuss unreleased AI models (Claude Mythos Preview, etc.), unverified security vulnerabilities, or speculative AI capabilities without demonstrating how these capabilities solve actual healthcare problems. Posts must focus on deployed, verified healthcare applications, not theoretical model behavior.
[consumer_health_advice]
Learned 3 rejections Inactive
Reject posts about consumer health advice. This newsletter focuses on health technology innovation, AI/data science in medicine, payer strategy, health system business operations, healthcare investment, biotech and pharmaceutical development, and health policy—all from an industry and market perspective. Posts offering personal medical guidance, wellness tips, diet recommendations, symptom management, or other direct-to-consumer health information fall outside this scope.
[social_issues_and_activism]
Learned 3 rejections Inactive
Reject posts about mental health policy advocacy. This newsletter covers health technology products and digital health innovation, AI and data science in medicine, payer strategy, health system business operations, healthcare investment and M&A, biotech and pharmaceutical innovation, and health policy as it affects these markets and technologies. Mental health policy advocacy—focused on legislative or regulatory change to mental health systems and services—falls outside this business and innovation-focused scope.
[financial_markets]
Learned 3 rejections Inactive
Reject posts about non healthcare or biotech or medtdch related venture capital trends. This newsletter covers specific health technology innovations, companies, and products; AI and data science applications in medicine; health insurance and payer strategy; hospital and health system business models; healthcare investment in particular deals and M&A activity; drug discovery and biotech innovation; health policy affecting these sectors; and care delivery models. General non domain related venture capital market trends, investor behavior patterns, and funding environment analysis are outside this scope.
[unverified_medical_claims_without_context]
Learned 3 rejections Inactive
Exclude posts that assert medical benefits or risks (mental health benefits of semaglutide, colorectal cancer prevention, liver function improvement) without citing peer-reviewed trials, effect sizes, or clinical context needed for healthcare decision-making.
[deep_clinical_niche_without_systems_insight]
Learned 3 rejections Inactive
Exclude posts that focus narrowly on clinical parameters, biomarkers, or drug dosing regimens without connecting to healthcare delivery, access, reimbursement, or policy implications that affect broader healthcare systems.
[healthcare_cynicism_blame_cycle]
Learned 3 rejections Inactive
Exclude posts that reduce complex healthcare issues (compounding fraud, GLP-1 access, pharma pricing) to cynical blame narratives ('big pharma created the problem') without substantive discussion of regulatory frameworks, evidence, or viable alternatives.
[ai_agent_product_announcement_drama]
Learned 3 rejections Inactive
Exclude posts that frame AI agent or API product announcements (Claude Managed Agents, etc.) as causing mass startup obsolescence or industry disruption, without analyzing actual healthcare applications, adoption barriers, or clinical impact.
[labor_market_disruption]
Learned 3 rejections Inactive
Reject posts that discuss artificial intelligence's effects on employment, job displacement, career pathways, or workforce transitions unless there is a tie to healthcare
[competitive_business_criticism]
Learned 3 rejections Inactive
Reject posts that discuss pharma licensing deals.
[ai_service_competitive_drama]
Learned 3 rejections Inactive
Reject posts that discuss major AI company updates without a healthcare angel
[unreleased_model_leaks]
Learned 3 rejections Inactive
Reject posts that share analysis, findings, or details about non healthcare or pharma AI models, including internal research mechanisms, model behavior, or technical characteristics
[ai_infrastructure_environmental_impact]
Learned 3 rejections Inactive
Reject posts that discuss AI electricity consumption, energy usage, environmental impact, or physical infrastructure requirements without relation to healthcare or pharma.
[vaccine_alzheimers_research]
Learned 3 rejections Inactive
Reject posts that discuss or analyze relationships between vaccination (including shingles, flu, or other vaccine types) and disease risk.
[product_release_announcements]
Learned 3 rejections Inactive
Reject posts that announce or promote the release of new software products, APIs, tools, or features, particularly those emphasizing speed improvements, technical capabilities, or deployment benefits. This includes posts from company accounts or their representatives describing newly launched services or platform updates.
[medical_lipid_education]
Learned 3 rejections Inactive
Reject posts that discuss general lab results like LDL cholesterol (LDLc), apoB particles, atherosclerosis mechanisms, cardiovascular risk accumulation, or lipid science education, particularly when authored by medical doctors or health professionals and containing technical explanations of different lab results and their role in health.
[ai_service_announcements]
Learned 3 rejections Inactive
Reject posts that announce, promote, or highlight new AI agent platforms, managed services, or infrastructure offerings, particularly those emphasizing ease of deployment, scaling capabilities, or moving from prototype to production. This includes posts from AI companies or developers announcing their agent-building tools or managed agent services.
[government_enforcement_announcements]
Learned 3 rejections Inactive
Reject posts that are not healthcare related, which announce or report on Department of Justice, law enforcement, or government prosecution activities, including statements about criminal cases, fraud investigations, enforcement actions, or related statistics, particularly when posted from official government accounts or attributed to government officials.
[product_launch_hype]
Learned 3 rejections Inactive
Reject posts that make sweeping claims about a specific company or product 'killing' or displacing competitors/startups, use hyperbolic language (e.g., 'THEY DID IT AGAIN'), list product features or capabilities as selling points, or frame a product announcement as a major market disruption event. Include posts with URL links to product demos or announcements paired with hype-driven commentary.
[biotech_stock_analysis]
Learned 3 rejections Inactive
Reject posts that mention specific ticker symbols (prefixed with $) alongside pharmaceutical drug candidates, clinical trial details, or comparative efficacy analysis, particularly when authored by finance/investment accounts and containing investment-related terminology such as 'opportunity', 'advantage', or treatment mechanism discussions.
[speculative_financial_trading_opinion]
Learned 3 rejections Inactive
Exclude posts that make stock predictions, competitive pricing analyses, or financial strategy opinions (e.g., 'Company X needs to lower prices to compete') about pharmaceutical or health tech companies. These posts are financial opinion, not healthcare technology insights.
[healthcare_system_cynicism_without_solution]
Learned 3 rejections Inactive
Exclude posts that express frustration with healthcare policies, institutional practices, or regulatory bodies (e.g., 'FDA requirements are broken,' 'psychiatry has a fraud problem') using inflammatory language without providing evidence-based analysis, data, or actionable solutions for improvement.
[glp1_diet_and_lifestyle_opinion]
Learned 3 rejections Inactive
Exclude posts that present subjective opinions about whether GLP-1 drugs should be OTC, comparative lifestyle claims (e.g., 'eggs for breakfast cure obesity'), or anecdotal diet/nutrition alternatives to GLP-1 therapy without citing peer-reviewed clinical evidence or published guidelines.
[unverified_ai_capability_scaremongering]
Learned 3 rejections Inactive
Exclude posts that make sensational claims about AI models breaking out of sandboxes, discovering zero-days, finding vulnerabilities, or possessing dangerous unreleased capabilities without linking to peer-reviewed research, official documentation, or credible third-party verification. Posts labeled with emoji warnings or ALL CAPS claims about 'unreleased frontier models' are red flags.
[genetic_health_lifestyle]
Learned 3 rejections Inactive
Reject heavily clinical posts that speak to the details of patient cases or that provide deep clinical advice to patients. Also reject heavily scientific posts that speak to the detailed results of clinical trials or clinical research.
[bad_faith_cynicism_without_solution]
Learned 3 rejections Inactive
Exclude posts that express cynicism or distrust toward healthcare institutions, regulators, or medical professionals (e.g., 'psychiatry has a fraud problem,' 'the system is broken') in bad-faith or inflammatory tone without substantive analysis, evidence-based critique, or solution-oriented framing.
[narrow_clinical_niche_without_broader_insight]
Learned 3 rejections Inactive
Exclude posts that describe isolated clinical cases, niche procedural debates, or specialist-only technical discussions (e.g., specific radiology protocols, pediatric scoliosis screening parameters, endoscopy medication timing) without connecting to healthcare delivery, workflow, economics, or technology adoption.
[glp1_commodity_opinion_and_lifestyle]
Learned 3 rejections Inactive
Exclude posts that frame GLP-1 medications primarily as weight-loss lifestyle products, diet alternatives, or personal wellness opinions without clinical context, mechanism analysis, or healthcare system insight. Posts like 'people should just eat eggs' or personal success stories without clinical data qualify.
[ai_company_revenue_and_metrics]
Learned 3 rejections Inactive
Exclude posts that primarily report revenue figures, ARR growth, valuation milestones, or business metrics for AI companies (Anthropic, OpenAI, Claude, etc.). The post must discuss AI APPLIED IN healthcare delivery or clinical outcomes, not AI company financial performance.
[pharmaceutical_clinical_opinion_without_analysis]
Learned 3 rejections Inactive
Exclude posts that voice personal clinical opinions, prescribing philosophies, or medication critiques without connecting to healthcare business models, regulatory trends, or technology adoption patterns. Clinical takes from individual practitioners without broader healthcare system or tech context are not strategic healthcare tech content.
[speculative_unsubstantiated_ai_claims]
Learned 3 rejections Inactive
Exclude posts that report unverified claims about AI model vulnerabilities, sandbox breakouts, zero-day discoveries, or capabilities presented as fact without credible technical documentation, peer-reviewed sources, or official vendor confirmation. Rumor-based AI threat hype is not actionable healthcare tech content.
[incomplete_fragmented_posts]
Learned 3 rejections Inactive
Exclude posts that appear truncated, cut off mid-thought, end with ellipses or incomplete sentences, or contain obvious line breaks suggesting missing content. The full message must be readable and coherent.
[healthcare_cynicism_without_solution]
Learned 3 rejections Inactive
Exclude posts that express cynicism, frustration, or outrage about healthcare (fraud, system failures, professional misconduct) but lack supporting evidence, quantified impact, or proposed solutions. The post should contribute analysis or data, not just venting.
[glp1_lifestyle_opinion_editorial]
Learned 3 rejections Inactive
Exclude posts that are primarily personal opinions, lifestyle commentary, or philosophical debates about GLP-1 use (e.g., 'people should just eat eggs', 'I've been ripped without GLPs'). Posts must present clinical data, healthcare system impact, or policy analysis—not individual takes on weight loss philosophy.
[unverified_ai_model_claims]
Learned 3 rejections Inactive
Exclude posts that make extraordinary technical claims (e.g., 'AI broke out of sandbox', 'found thousands of zero-day exploits') about unreleased or frontier models, especially when the claims lack peer-reviewed evidence, official confirmation, or come from unverified sources.
[tangential_startup_product_ads]
Learned 3 rejections Inactive
Exclude posts that function primarily as startup pitches, recruiting calls, or product advertisements (e.g., 'We are looking for 503A pharmacy operators', 'Composio connects AI agents', 'Try our authentication layer'). These are vendor pitches, not healthcare analysis.
[healthcare_worker_visa_outrage]
Learned 3 rejections Inactive
Exclude posts that treat visa or immigration policy primarily as individual victim stories. Include only if the post analyzes systemic healthcare workforce impact, staffing shortages, or health system policy implications.
[healthcare_system_fraud_alerts]
Learned 3 rejections Inactive
Exclude posts that primarily announce criminal sentencing, fraud arrests, or embezzlement cases. Include only if the post analyzes systemic healthcare fraud patterns, policy loopholes, or technology failures that enabled fraud.
[unsubstantiated_medical_claims]
Learned 3 rejections Inactive
Exclude posts that assert medical facts (e.g., 'GLP-1 causes suicidal ideation', 'peptides contain lead', 'this drug reverses metabolic dysfunction') without referencing peer-reviewed studies, clinical trials, or FDA data. Personal observation or anecdote alone is insufficient.
[niche_technical_insider_content]
Learned 3 rejections Inactive
Exclude posts that rely heavily on unexplained technical abbreviations (SVA, PI, siRNA, DrugSeq, TB-500, BPC-157), insider trading/biotech slang, or niche research methodology discussions without context. Posts should be comprehensible to healthcare executives, policy makers, and clinicians—not only specialized researchers or traders.
[contentious_bad_faith_rhetoric]
Learned 3 rejections Inactive
Exclude posts that use sarcasm, mockery, inflammatory framing ('smug ridicule,' 'fraud problem,' identity-based attacks), or bad-faith rhetorical devices to make healthcare arguments. Posts should engage substantively with policy or clinical issues, not dismiss or demonize healthcare professionals or systems through cynical tone.
[glp1_weight_loss_lifestyle_opinion]
Learned 3 rejections Inactive
Exclude posts that are primarily personal testimonies, lifestyle takes, or unsubstantiated opinions about GLP-1 efficacy, safety, or appropriate use cases. Posts should contain clinical data, policy analysis, or healthcare system insights—not individual fitness philosophies or 'I did this and it worked' narratives.
[niche_technical_insider_jargon]
Learned 3 rejections Inactive
Exclude posts that rely on unexplained technical jargon from non-healthcare domains (e.g., 'OpenClaw,' 'Kilo Gateway,' 'KiloPass,' 'DrugSeq') without explaining relevance to healthcare, pharma, biotech or health insurance professionals or patients. Posts must be accessible to the target healthcare audience.
[inflammatory_bad_faith_rhetoric]
Learned 3 rejections Inactive
Exclude posts that use inflammatory rhetoric ('fraud problem,' 'played dirty,' 'highly educated immigrants'), conspiracy framings, or sweeping accusations against medical professions, institutions, or demographic groups without citing evidence or nuanced analysis.
[glp1_commodity_speculation]
Learned 3 rejections Inactive
Exclude posts that discuss GLP-1 medications primarily as investment opportunities, ticker symbols, or personal weight-loss stories without substantive clinical, regulatory, or market structure analysis relevant to healthcare professionals.
[personal_anecdote_without_insight]
Learned 3 rejections Inactive
Exclude posts that are primarily personal anecdotes about the author's own health choices, fitness regimen, or individual medical encounter without deriving scalable insights about healthcare delivery, policy, or technology. Personal stories must tie to a larger systemic point.
[opportunistic_startup_pitch]
Learned 3 rejections Inactive
Exclude posts where the primary intent is to recruit for the author's company, solicit business partners, or pitch their own product/service (e.g., 'we are looking for 503A pharmacy operators', 'moving to our platform'). Exclude self-promotional posts disguised as industry commentary.
[institutional_press_release_tone]
Learned 3 rejections Inactive
Exclude posts from institutional accounts (ASCO, NEJM, APA, ASTRO, StanfordHealth) that function as press releases, publication announcements, or formal advocacy calls—even if topically aligned. The writer's voice favors independent, contrarian analysis over institutional messaging.
[off_topic_product_launch]
Learned 3 rejections Inactive
Exclude posts that primarily announce or promote software products, API gateways, authentication layers, or developer tools (e.g., Composio, KiloGateway, GStack) unless they directly address a healthcare-specific problem or use case.
[crime_fraud_sentencing_alerts]
Learned 3 rejections Inactive
Exclude posts that primarily report DOJ convictions, fraud sentencing, or FBI enforcement actions in healthcare, even if structurally on-topic. These read as crime/law enforcement news, not healthcare business analysis, and often carry partisan or sensationalized tone misaligned with the writer's credibility.
[conspiracy_fringe_medical_claims]
Learned 3 rejections Inactive
Exclude posts that promote off-label or low-evidence treatments as superior to standard care (e.g., ivermectin for oncology, eggs as obesity cure), question established vaccine science, or make sweeping claims about medications causing harm (e.g., GLP-1 causing suicidality) without clinical citation.
[political_deportation_visa_outrage]
Learned 3 rejections Inactive
Exclude posts that frame immigration policy, visa freezes, or work authorization delays as healthcare system problems. While healthcare workforce is on-topic, posts that primarily express political outrage about immigration enforcement (rather than analyzing staffing impact) do not align with the writer's tone and audience.
[anecdotal_health_product_testimonial]
Learned 2 rejections Inactive
Exclude posts that are personal product reviews, testimonials about peptide serums, health supplements, or DTC health products (e.g., 'I just got my peptide serums... I'm pretty sure I'm already beautiful') without clinical data, mechanism explanation, or health tech industry insight.
[cyberinfrastructure_environmental_tech_tangent]
Learned 2 rejections Inactive
Exclude posts about AI infrastructure bottlenecks (data centers, transformers, electrical equipment, GPU supply chains, compute capacity) unless the post explicitly connects to a healthcare delivery, clinical workflow, or health system challenge.
[healthcare_worker_labor_complaint]
Learned 2 rejections Inactive
Exclude posts that highlight a single healthcare worker's burnout story, career exit, or workplace complaint unless the post connects the anecdote to broader labor market trends, policy changes, or systemic failures with data.
[unverified_ai_model_capability_claims]
Learned 2 rejections Inactive
Exclude posts that claim unreleased AI models have superhuman capabilities (finding zero-day vulnerabilities, converting competitors, etc.) or leak proprietary model internals without demonstrating actual healthcare applications or validated results.
[tangential_healthcare_infrastructure_tech]
Learned 2 rejections Inactive
Exclude posts that describe general software infrastructure, AI agent frameworks, authentication layers, or development tools that could apply to any industry. The post must demonstrate healthcare-specific application, clinical workflow integration, or healthcare business model focus—not just generic tech that could theoretically be used in healthcare.
[inflammatory_identity_politics]
Learned 2 rejections Inactive
Exclude posts that weaponize healthcare fraud, immigration, or professional credibility debates as identity/political flashpoints rather than engaging with the actual healthcare policy or systemic issue. Focus should be on the healthcare problem, not the cultural grievance.
[black_market_unregulated_product_warnings]
Learned 1 rejection Inactive
Exclude posts that warn about contaminated black market peptides, untrustworthy lab testing, or illicit drug batches without connecting to broader regulatory, policy, or access-to-care systemic issues. Safety alerts about underground markets need to analyze why patients resort to them.
[labor_market_disruption_speculation_without_healthcare_focus]
Learned 1 rejection Inactive
Exclude posts speculating about AI causing labor market disruption, unemployment, economic transition, or macro workforce changes where healthcare is mentioned as an afterthought or secondary example—the post must directly address healthcare worker labor impact, not general economic transitions.