Generally Intelligent cover image

Dylan Hadfield-Menell, UC Berkeley/MIT: The value alignment problem in AI

Generally Intelligent

00:00

Is There a Loop of System Behavior?

I think that a lot of the ways in which systems are brittle and they don't work is because we rely on m l researchers or practitioners to fix them. The people wo've been doing m el work are really smart people who are really good at pattern matching. I think you should really be studying this interaction loop of system behavior, system designer and some sort of actual feedback mechanism - almost as a control system of its own.

Play episode from 12:07
Transcript

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