Chemist Critiques Biology Millennium Problems: All Engineering, No Discovery or Mechanism
burny_tech · x · 2026-09-21
Reacting to FutureHouse's "Millennium Problems for Biology," Jason DeKoning argues that while some are solid, the list is all engineering challenges. Biology isn't just "impressive — but I think I can do better!" Discovery and mechanism come first, and that's what motivates most biologists.
The critique targets a core design tradeoff: the authors deliberately excluded basic-science mysteries that are hard to grade objectively (e.g., the Vault protein's function), at the cost of shutting out discovery-driven problems. A worthwhile debate for anyone designing AI-for-Science evals.
More from Research
- Anthropic Sets Up Bay Area Wet Lab, Plans to Have Claude Run Experiment Robots — FinanceYF5 · 2026-09-21
- KLPO: a critic-free, single-rollout RL method for agentic LLMs, framed as 'Q* solved' — inductionheads · 2026-09-21
- FutureHouse publishes 'Millennium Problems for Biology' as wetlab-verifiable evals for AI — burny_tech · 2026-09-21
- No scientific test for consciousness exists even in humans — how can we declare AI unconscious? — burny_tech · 2026-09-21
- Formal Reasoning in AI: Song guest-lectures Virginia Tech course, slides public — wellecks · 2026-09-21
- Biologist: FutureHouse's Bio Millennium List Is Too Narrow — Really Just Synbio — burny_tech · 2026-09-21