Chipmonk comments on How I turned doing therapy into object-level AI safety research

[Chipmonk]

Where do you think the boundary is here?