AI, Robot
Google DeepMind: The Podcast
00:00
Learning the Functions of Retort Learning Agents
Human preference is said to be quite complex. And whenever we try to distil them into some kind of specification or something that we say we want, it's going to be a lot simpler than our real preferences are. But how do you get around these kinds of problems? Well we already know that writing a long list of dos and don'ts won't work. The new breed of learning agents is going to need a different approach. So one direction that we are pursuing is learning what functions for reinforcement learning agents. For example, i one work that we did together with openy eye is be trained, ssimulated bat to do yor back flom. Here is the key idea
Transcript
Play full episode
