I originally wrote this for the Meta Coordination Forum. The organizers were interested in a memo on topics other than alignment that might be increasingly important as AI capabilities rapidly grow — in order to inform the degree to which community-building resources should go towards AI safety community building vs.