I think this idea of pluralist community models can overcome many fears surrounding superintelligence, vertical takeoff, and recursive self-improvement. Instead of just calling to “stop AI” or “control AI,” or trying to halt acceleration, it makes more sense to change direction.