Episode 25: Nicklas Hansen, UCSD, on long-horizon planning and why algorithms don't drive research progress
Generally Intelligent
00:00
Is RL a Good Idea for Local Motion Robots?
I'm very impressed by all the research in local motion robots recently. I was really surprised that other people had not already done that. And with RL, it's actually working pretty well. But still a lot of the RL policies are sort of focusing on how can we do simp to reel and overcome these generalization gap in dynamics.
Transcript
Play full episode
