
Episode 22: Archit Sharma, Stanford, on unsupervised and autonomous reinforcement learning
Generally Intelligent
Do You Have Any Controversial Research Opinions?
I always think comfort has a effect in some sense. If the computer was a person, they would change their opinion based on what the initial conditions are. And for any generally intelligent agent, you want them to have like a consistent personality. Maybe this is a controversial opinion, but like broadly is something that makes like a lot of these things work. But it's a time to use the same system in like, I don't know, job parts which are possibly like dealing with people or counseling them.
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.