LessWrong (Curated & Popular) cover image

"Prologue to Terrified Comments on Claude’s Constitution" by Zack_M_Davis

LessWrong (Curated & Popular)

00:00

How Character-Based Training Generalizes

Zack explains the hope that models' latent knowledge of humans lets character training produce broadly safe behavior.

Play episode from 08:31
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