AXRP - the AI X-risk Research Podcast cover image

4 - Risks from Learned Optimization with Evan Hubinger

AXRP - the AI X-risk Research Podcast

00:00

Is Optimization a Mechanistic Process?

Machine learning is notoriously difficult to defind. The process of gradiant i process of machine learning selects on the internals. It selects on implicit doctor biases, which might be pushing you in one direction or another. And it's very difficult to say that if you don't know what an authorizer looks like internally - how likely is it that i'll get something well described as an otiizer?

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