AXRP - the AI X-risk Research Podcast cover image

2 - Learning Human Biases with Rohin Shah

AXRP - the AI X-risk Research Podcast

00:00

Using a Neural Network to Make Optimum Predictions?

We just use a single neral network, a nera networks tend to be bias towards simplicity. What we do as we take our nerol net that corresponds to this planning maji, and we train it to produce the same things that valle etration produce. An value trationism is analgithm that produces optimal policies for the small environments that we consider. So by training, basically we're just training or nea met to make optimal predictions anyour initializing at thi.

Transcript
Play full episode

The AI-powered Podcast Player

Save insights by tapping your headphones, chat with episodes, discover the best highlights - and more!
App store bannerPlay store banner
Get the app