Episode 28: Sergey Levine, UC Berkeley, on the bottlenecks to generalization in reinforcement learning, why simulation is doomed to succeed, and how to pick good research problems
Generally Intelligent
00:00
Generally Intelligent Podcast
I do think that we can figure out many of the nitty gritty physical robotic things. I'm not sure how long that will take exactly, but I'm also kind of hopeful that we figure them out. If your model understands how the physical world works, you can probably do a better job in the digital world.
Transcript
Play full episode
