Jakob Foerster
TalkRL: The Reinforcement Learning Podcast
The Future of Deep Learning and Multi-Agent Reinforcement Learning
In your paper from 2016, you mentioned end-to-end learning of protocols in complex environments. I think that was a groundbreaking paper in terms of figuring out how to get deep RL agents to communicate and to invent languages. Can we now ask those same questions again about having agents that can develop protocols starting from what is already seeded with all of human knowledge from a large-scale language model? It's an exciting line of work here which is picking up those initial ideas again but combining them with what we have in 2023. And I'll be very excited to see how we can combine these two now.
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.
