TalkRL: The Reinforcement Learning Podcast cover image

Rohin Shah

TalkRL: The Reinforcement Learning Podcast

CHAPTER

The Holy Grail for Ai Systems Training

The holy grail is to have a sedure for training ai systems at particular tasks. We can apply arbitrary, human understandable constraints to how the a system achieves those tasks. For example, we can buildn ai sistant that scheduled yor meetings, but ensure is that it's always very respectful when it's talking to other people in order to schedule your emales and is never like discriminating based on sex. Or you can just deploy it on an entirely new multiplaer server that includes both humans and ai systems. And then you can say, hey, you should just go help such and such player with whatever it ies they want to do. The agent just does that

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.
App store bannerPlay store banner