
Episode 19: Minqi Jiang, UCL, on environment and curriculum design for general RL agents
Generally Intelligent
00:00
Getting Used to a Different Day Night Cycle of Life, Right?
Gepet three type of us set up where ou train on so much data. And now, yu ave this new data, but it's such a tiny fraction. Even finding it feels like a little doesn't quite feel lik what we're doing as people when we land on mars,. Like we're sort of learning about these things in a different way. But i think some of those questions are interesting to me, also kind of coming back to, like reinforce the larning stuff. Do we want them to copy behaviours from other agents that they see, or just like traint all the behaviours people have ever done? It's hard for me to exactly place my finger on
Transcript
Play full episode