#131 Toby Ord - Will AI Destroy Humanity?
Within Reason
00:00
Training regimes: games versus language models
Toby contrasts reinforcement learning game agents with next-token pretraining and its human-data pull.
Play episode from 05:08
Transcript
