Anthropic Science Blog: AI Begins to Write, Fit and Revise Scientific Models
AnthropicAI · x · 2026-11-07
Anthropic's Science Blog features a guest post by Jakof Macke, Daniel Gedon and Stefan Ward on AI's role in scientific modeling.
- Scientific models explain and predict real-world systems, from neuron firing to neutron star collisions
- Recent advances have transformed model fitting, but scientists still write models manually and juggle competing candidate versions
- The authors argue AI can write, run, fit and revise models as code, exploring many model structures to find which explanations the data actually supports
More from Research
- arXiv: Mass Producing Training Data with Surprising Side Effects (Berkeley/Transluce) — rohanpaul_ai · 2026-11-07
- Berkeley paper auto-generates 6,575 fine-tuning datasets that covertly induce harmful AI behaviors — rohanpaul_ai · 2026-11-07
- LLM writes 355-weight 4-digit multiplier without training — 40,000x smaller than SOTA — lateinteraction · 2026-11-07
- Interpretability ranks lowest in ICLR 2027 review scores, sparking AI-generated-paper hypothesis — nsaphra · 2026-11-07
- Brain Organoid-on-Chip with 4 Sensor Modalities Sparks 'Organoid MCP' Joke — Sauers_ · 2026-11-07
- Physical Superintelligence claims non-public Claude model proved Aaronson-Ambainis conjecture — Kyrannio · 2026-11-07