cubefox comments on AI existential risk probabilities are too unreliable to inform policy

[cubefox]

This is an unusually well-written piece from AI doom skeptics. It deserves a response/​rebuttal.