Orpheus16 comments on How the AI safety technical landscape has changed in the last year, according to some practitioners

[Orpheus16]

I’d be interested in more information about how Anthropic and GDM are thinking about safety cases (or more details from people who are excited or unexcited about this work).