Safe & beneficial AI. Working on Alignment Science at Anthropic.

Favorite papers at http://aisafetyfrontier.substack.com. Opinions my own.