We hear from many quarters that "AI will destroy the world," but everyone's got a different scenario for what that means. The most sensational perspectives come from theorists like Geoffrey Hinton, the so-called "godfather of AI" or industrialists like Elon Musk. They warn us that one day,  a superintelligent AI could replace human beings in controlling the planet.  But we don't need to conjecture into the future. The scorched-earth destruction is already happening now. Jobs are hemorrhaging with no sign of return; data centers are turning U.S. farmlands into barren industrial gulags while consuming record levels of power, spewing volumes of carbon and using up our last freshwater sources; fusillades of deep-fake videos politically paralyze the public; an AI surveillance infrastructure is being constructed that will lock in fascism; and algorithms are telling ICE and the IDF who lives and who dies. It seems all too overwhelming. However, by tracing AI's lineage to the development of the atom bomb -- with the same ideologies and twisted logic -- it becomes apparent that solutions to the AI dilemma can also be found in nuclear history, in its disarmament successes.