On: They do think AI might kill everyone

[metalcrow]

> the positives are subjective, potential, and still not convincingly verifiable.

Saying this post- millennium prize problem solving is laughable.