Anthropic report: full AI automation could yield 15-16x research output, labs remain the bottleneck
teortaxesTex · x · 2026-11-06
Anthropic's new report analyzes 207,049 scientific research sessions and models a future where AI automates all non-physical research work: up to 15-16x output in CS/statistics vs only 3-4x in biology and chemistry, where physical steps (40% of researcher time) remain the bottleneck. Nearly 40% of 528 surveyed scientists expect AI to run entire research projects autonomously by 2030. Claude Science shows high autonomy in 73.3% of biology/biomedical sessions (vs 23.7% on regular Claude), with an estimated 82.7% of sessions involving PhD-level work. Commenter teortaxesTex argues that may still not be enough.
More from AGI Musings
- A neat qualia argument: introspection can never reveal the absence of qualia — Sauers_ · 2026-11-06
- Only 16% of AI benchmarks measure organizational difficulties of real jobs — ben_burtenshaw · 2026-11-06
- 'I don't read long replies anymore' — dev on AI slop fatigue — hewarsaber · 2026-11-06
- Claude's verdict on which jobs survive the next 10 years: screen work is doomed — alex_verem · 2026-11-06
- Post-training pipelines remain highly manual and far from automatable — menhguin · 2026-11-06
- Peter Norvig joins podcast on alignment: why higher scores don't mean better AI — EchoShao8899 · 2026-11-06