Grant funder's data point: LLM reviewers already beat any single human
anshulkundaje · x · 2026-10-29
A first-hand data point from Impetus Grants: having tried both human and LLM reviewers, Martin Bjorn Jensen finds LLM reviewers already more capable than any individual human for many grants, since they spot issues across statistics, biology and epidemiology in one proposal. Humans still win on tacit knowledge — unpublished experiments, researcher motivations, ideas that only look good on paper. Quoted context: anshulkundaje adds that AI reviews beat the average reviewer, especially in esoteric cross-disciplinary areas, but the best workflow is to write your own review first, then augment it with AI.
More from AGI Musings
- Everyone talks about AI but they mean completely different things — almmaasoglu · 2026-10-29
- The printing press, not AI, was history's original slop cannon — RexDouglass · 2026-10-29
- New paper 'Training for Alignment' models how principals optimally train agents with endogenous values — Afinetheorem · 2026-10-29
- Google researcher: people posting these AI-work demo videos are 'doing nothing of consequence' — prajdabre · 2026-10-29
- AIxBio builder: people build the models they can, not the models they should — hardimanjames · 2026-10-29
- BallotBench pits 9 frontier AI agents against experts in live midterm forecasting — karinanguyen · 2026-10-29