
Episode 22: Archit Sharma, Stanford, on unsupervised and autonomous reinforcement learning
Generally Intelligent
Is There a Better Structure for Research?
The real world is super high dimensional, very continuous. And so language is somewhat of a compression that's like maybe more structured. computers are good at dealing with very structured things. So it could be the case that if we get general intelligences and they're really good at like these digital tasks, but not the physical ones? We robot resetter. Terrible stuff. That's why we need to make progress on RL. If you need some information or your stuff. I am more and more motivated than ever do about doing search right now.
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.