LessWrong (30+ Karma) cover image

“AI Corrigibility Debate: Max Harms vs. Jeremy Gillen” by Liron, Max Harms, Jeremy Gillen

LessWrong (30+ Karma)

00:00

Inductive biases and the limits of black-box training

Jeremy argues corrigibility's simplicity must align with training inductive biases, otherwise black-box methods won't reliably hit it.

Play episode from 01:57:45
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