
Pieter Abbeel: Deep Reinforcement Learning
Lex Fridman Podcast
Are You Interested in Self Play?
Self play is a form of imitation learning that can be used to teach robots new skills. In self play, you're on both sides - one succeeds and the other fails. The challenge in reinforcement learning is getting signal; if you don't never succeed, you don't get any signal. So I think if somehow we can turn more reinforcement learning problems into self play formulations, that would go really, really far.
00:00
Transcript
Play full episode
Remember Everything You Learn from Podcasts
Save insights instantly, chat with episodes, and build lasting knowledge - all powered by AI.