Episode 19: Minqi Jiang, UCL, on environment and curriculum design for general RL agents
Generally Intelligent
00:00
Language to Second Modality Grounding
The most promising direction, it seems, right now, is to leverage supervise learning. You get really high quality representations just from the sheer amount of data that you have available on the internet. And potentialy even with vidios in the future, you can then basely just exploit the learned laten space for downstream tasks.
Transcript
Play full episode
