Cambridge AI safety professor pushes back: deep expertise does matter in AI debates

DavidSKrueger · x · 2026-09-26

Cambridge professor and AI safety researcher David Krueger responds to the claim that many AI questions have no real experts, arguing there's a huge difference between people with over a decade at the frontier of deep learning and AI safety and those unfamiliar with out-of-distribution generalization and instrumental convergence. He doesn't claim expertise makes him right about everything, but says basic mistakes and bad arguments recur constantly and their makers shouldn't be treated as credible voices.

Original post →

More from AGI Musings

AGI Musings channel →