Toby Ord clashes with Jan Leike over whether human control of AI is a consensus

tobyordoxford · x · 2026-11-20

Former DeepMind VP of research Jan Leike argued that the stance "humans should fundamentally stay in charge of their destiny, even as we hand off to AI" is not a consensus everywhere. Philosopher Toby Ord agreed the position is a bad and dangerous idea but criticized Leike's rhetoric: claiming an undefined powerful group silently holds a view few have heard defended inflates its support to cast oneself as bravely bucking a consensus that doesn't really exist. A notable exchange among leading AI-safety figures over how AI risk debates are framed.

Original post →

More from AGI Musings

AGI Musings channel →