RobinHanson comments on Some disjunctive reasons for urgency on AI risk