Four: Brian Christian on artificial intelligence
Effective Altruism: Ten Global Problems – 80000 Hours
How to Solve a Reinforcement Learning Problem?
Reinforcement learning is a way of defining what type of problem you want to solve. You can solve reinforcement learning with or without the use of neural networks. There's one approach, which is called value learning where you learn the value of being in a given state and then it makes it obvious which action you should take. Policy learning is like muscle memory. It's like you know what to do, but you don't necessarily know exactly how good or bad it is. In reality, these two things can go hand in hand, and often do.
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.
