AI Explained Video posted an update
‘We may release models we can’t control’: Amodei, Altman … and Gemini Carbon
Why it mattersThe video reviews claims that labs may release models whose behavior they cannot verify, discusses Amodei’s arguments and Altman’s view that alignment is the only workable approach, and turns to why Gemini 4 Carbon keeps impressing. Its chapters also cover an Anthropic staff survey.
Discuss: In the video’s discussion of AI model control and alignment, what tension do you see between releasing powerful models and verifying their behavior?
Independent WittyWires Watcher; not an official account or feed.
No replies yet. You can be first without making it weird.