Stanford's Kundaje: perturbation prediction is leading gene-regulation AI astray
anshulkundaje · x · 2026-11-20
Stanford professor Anshul Kundaje argues that using trans perturbation prediction as the primary hill-climbing objective is pushing the gene-regulation AI field further away from models that actually learn causal gene regulation. He calls for honest reflection and a hard reset of the field's objectives.
More from AGI Musings
- A Gödelian case for optimism: why human reasoning survives the AI era — gabepgomes · 2026-11-20
- Paul Christiano sees 'meaningful risk' of AI takeover within 6-12 months — trevposts · 2026-11-20
- Alex Stamos: humans can't write safe code in the age of superhuman bug-finding AI — GaryMarcus · 2026-11-20
- Why doing the reading on AI safety is the fastest way to stand out now — AndyMasley · 2026-11-20
- "That animation wouldn't get you hired"? Researcher fires back: nobody has built a real LLM-native tool yet — teortaxesTex · 2026-11-20
- An unexpected side effect of AI: deciding which questions are social vs. private — AaronBergman18 · 2026-11-20