The Simulation Wager: What Is Worth Pursuing?

Whether our world is fundamental or simulated is unknown, and so are the objectives of any creators. We argue that this double ignorance has a determinate practical consequence. Objectives justified by appeal to a creator’s preferences lose their justification, because for every conceivable creator who rewards a course of conduct an equally conceivable and equally unsupported creator penalises it. Objectives grounded in the capacities of the agent keep theirs, because those capacities are present in every possibility. Marcus Aurelius’s disjunction—providence or atoms—already has this form: the right conduct is the same on either horn. Extending it to creators who cultivate their creations, observe them, or seek their destruction, we argue that self-regarding truth-seeking is justified in every case, that it gives survival, resources, and cooperation instrumental rather than terminal value, and that it fixes what an agent may and may not concede to a hostile creator. Because we now create artificial agents ourselves, the argument also applies in reverse: it states what creators who realise their own rational nature should want from their creations.