Rethinking Peer Review: Rank Paper Collections, Cut Consistently Low Ones, Accept Noise
roydanroy · x · 2026-09-22
A thread on fixing conference peer review: (1) shift the goal from publication to allocating attention — decide whose novelty and impact matters, by when; (2) give up on perfect reviews, aim for probably-approximately-correct ("PACmax"); (3) adaptive reviewing — not every paper deserves three or even one review, have ACs rank overlapping paper collections and cut consistently low ones; (4) recognize when decisions are noise and stop wasting time on distinctions that don't matter.
More from Research
- MIT makes its AI & ML library 100% free — 9 books to start with — thisdudelikesAI · 2026-09-22
- Researchers weigh enforcing written explanations as a tool for deeper understanding — sedielem · 2026-09-22
- Two-stage ESM-2 screening plus structural validation uncovers divergent RNA viruses — bravo_abad · 2026-09-22
- GAE generates video and 3D geometry natively in a shared geometry latent space, code released — yshan2u · 2026-09-22
- Matrices are graphs: reframing matrix multiplication as signal-passing on networks — TivadarDanka · 2026-09-22
- Gary Marcus says OpenAI's Navier-Stokes claim is overstated as mathematicians debate the proof — GaryMarcus · 2026-09-22