Bio heavyweights clash over paper methodology and multiple testing critique
anshulkundaje · x · 2026-09-06
- Michael Eisen pushed back on @StevenSalzberg1's critique of a paper that landed an NYT headline, calling it an "absurdly facile critique".
- Salzberg argued that scanning trillions of combinations without correcting for multiple testing yields associations that are pure chance, and the authors got the headline anyway.
- Eisen concedes methodological choices are debatable but insists the claim that multiple testing wasn't addressed is simply wrong.
A public spat over statistical rigor versus media hype, illustrating ongoing tension in research methodology debates.
More from Fun
- Astra's room-reconstruction demo gets the meme treatment — aidan_mclau · 2026-09-06
- GPT-6 Astra also leverages Bricklink Studio's embedded Python while building LEGO — dkundel · 2026-09-06
- GPT-6 Astra mixes Computer Use and direct file writing to build LEGO models — dkundel · 2026-09-06
- whurley's rainbow-topped Robotaxi pickup photo goes for the magical vibe — whurley · 2026-09-06
- Insider quip claims Google's Astra might 'just be a ResNet' — yacinelearning · 2026-09-06
- User's site wrecked as newer model blindly keyword-stuffs on command — FarTicket7338 · 2026-09-06