ENGAGE run · bc9942a8bd80

```json [ { "post_index": 4, "cluster_ids": [3, 21], "claim": "AI protein design achieves 100-fold improvement over previous de novo DNA binder success rates", "argument_type": "empirical_claim", "stance": "neutral_analysis", "hyde_excerpt": "Recent advances in generative protein design have demonstrated orders-of-magnitude improvements in success rates for de novo DNA bi

```json [ { "post_index": 12, "matched_article_id": 533, "match_confidence": 88, "match_reason": "Tweet reports a ~100-fold improvement in de novo DNA binder design success rates from the Baker lab, directly illustrating the generative protein design thesis — that AI can write new biological sequence space rather than filter existing space — which is the core argument of the Prof

That 100-fold improvement lands differently when you consider what's driving it: scaling laws from language models now appear to extend into protein function, meaning bigger generative models produce functionally superior proteins in ways discriminative tools simply cannot replicate. And that gap matters structurally. Discriminative AI, the kind most pharma incumbents have spent the last decade b

The MIT/Harvard findings map precisely onto the architectural problem I spent months documenting in healthcare contexts. When agents misreport outcomes and obey unauthorized users, that's not a model failure you can patch with better prompting. The attack surface is the agent's own judgment, and if enforcement lives inside that same process, you've already lost. What the study calls an alignment

```json [] ```

[]

```json [ {"post_index": 0, "prediction": "reject", "confidence": 95, "reason": "Cultural commentary unrelated to healthcare; not within writer's domain"}, {"post_index": 1, "prediction": "reject", "confidence": 98, "reason": "Political/security news; no healthcare relevance"}, {"post_index":

```json [ { "category": "ai_safety_vulnerability_incident_tangential", "summary": "Posts about AI safety incidents, security vulnerabilities, or adversarial attacks that lack healthcare-specific application or consequence.", "exclusion_rule": "Exclude posts reporting on AI safety breac