In the past, I’ve spent some time agonizing about the question of why I should bother exercising my free will in a multiverse that already contains all possible outcomes. Putting aside for now the fact that free will is a pointless construct anyways, the real realization that hit me one day is that the entire concept of “existence” as being something confined to a singular branch of the multiverse makes no sense. If I exist in one universe, and I believe in a multiverse, then I (obviously) exist in all universes. If my existence gives rise to subjective consciousness, and my subjective consciousness gives rise to moral status, then logically all branches of my wave function must contribute to my moral utility function. A quick utilitarian argument then shows that focusing on only one branch’s happiness (at the expense of the others) presents far less utility than trying to take paths that maximize my happiness in all (future) universes.

Essentially, another way of putting the argument: If I can identity with my feelings in this universe enough to care about my future (which I do), then I can also identify with my feelings in all other diverging branches of the wavefunction enough to care about their future! There’s no reason to unnecessarily consider “identity” as something confined to a single timeline. The mistake is even considering them separate “timelines” to begin with. Really, what I mean by “identity” is some sufficient conscious causal influence, i.e. something that will emerge from my current observer-state in such a way that I still agree it represents something I can consider “myself”, and therefore have an innate reason to care about its reward function.¹ But this applies to all future branches of the wavefunction, even if they end up diverging into separate timelines. They’re all still forever causally connected, namely to this moment!

What strikes me as fascinating is that this way of viewing the multiverse actually deconstructs the entire concept of a multiverse, at least in terms of any impact it can have on our decision-making. By the subjective immortality argument, and the above paragraph, I should only conclude that the current branch of my wavefunction is worth culling from the “affect integral” of the multiverse if and only if I believe that all streams of consciousness emanating from my current observer-moment, weighed by their relative abundance in the space of possibilities (i.e. their probability), sum to below zero. Put less fancily, this is literally just saying “I should kill myself if I believe my average future not to be worth living.” Which is exactly what any reasonable agent would conclude even in the absence of a multiverse, by simple Bayesian reasoning alone! Indeed, by this line of equivalence, Tegmark’s Mathematical Universe Hypothesis ends up effectively equal to Solomonoff induction.

And, just like that, not only does the subjective immortality paradox disappear, indeed the very question of whether or not we live in a multiverse becomes completely irrelevant. Leading me to conclude I should stop wasting my time writing about it, and your time reading about it.