As AI capabilities advance, multiple powers—nations, corporations, even warlords—race to attain artificial general intelligence. Episode 10 envisions the endgame when more than one actor holds an AGI or even an artificial superintelligence (ASI). The episode examines scenarios of a multipolar AI world, where no single entity can monopolize superintelligence. It asks: what happens when rival AIs, each pursuing their creators’ interests, collide? Through speculative yet plausible future vignettes, we see the terrifying game theory of multiple superintelligences. In one outcome, a dominant AI achieves decisive superiority, subjugating all competitors and the human population along with them. In another, evenly matched AIs enter a perpetual cold war—an uneasy equilibrium of hyperintelligent systems always calculating against each other, with humanity ...