Link. In the context of a talk Scott wanders all over the place – but he mirrors my thoughts on where AI-denial goes wrong. I’d have spent a bit more time on how “non-special” human cognition seems.

I mostly agree with his other points too …