Noosphere89 comments on Alignment Faking in Large Language Models

[Noosphere89]

Tagging @gwern here for your take on the situation.