3 ms·

The issue with the many-worlds interpretation is that it only provides for a local absence of randomness. The process of splitting into many words, as well as

The issue with the many-worlds interpretation is that it only provides for a local absence of randomness. The process of splitting into many words, as well as necessary limitations on measurement, i think both leave you in non-computable territory. You'd model this by generating an infinite number of Turing machines for each quantum event, but you'd also need to do something about the time horizon on prediction. Ie., chaotic systems aren't long-run deterministic because of measurement limitations. So there's something like many uncountable infinities of Turing machines out there. As for each digit in X there is an infinity of many-worlds in Y (the nature of chaos). I dont see this many-worlds vision giving you anything you can simulation. Rather its a useful conceptual heuristic when interpreting QM probabilities.

skissane 5y ago
> The process of splitting into many words, as well as necessary limitations on measurement, i think both leave you in non-computable territory. You don’t actually have to compute umpteen branches. You just have to simulate sufficient evidence to convince an observer in the simulation that many-worlds is true. Surely it is far easier to delude simulated people into believing some theory of physics is true, than to actually accurately simulate a universe in which it is true. But, even that wholly accurate simulation - if we are living in a simulation, it is possible the physical laws of the “real” universe are very different from those of our own. If our simulators are living under very different laws of physics, they may have access to computational resources which would be “unphysical” in ours. They may even be capable of hypercomputation. Maybe their laws of physics contain an oracle for the halting problem. Maybe they can execute supertasks. Even proof that simulating our universe would be uncomputable does not exclude the possibility that we are simulated on a hypercomputer We don’t even know that hypercomputation is physically impossible in this universe. We know of no way of achieving it given known physics, but we cannot exclude the possibility of unknown physical laws or processes which permit it, or even that it might be some bizarre mathematical consequence of existing theories which nobody has yet discovered. Maybe the hypercomputer will be the defining technological advance of the 57th century.

skissane 5y ago

> The process of splitting into many words, as well as necessary limitations on measurement, i think both leave you in non-computable territory. You don’t actually have to compute umpteen branches. You just have to simulate sufficient evidence to convince an observer in the simulation that many-worlds is true. Surely it is far easier to delude simulated people into believing some theory of physics is true, than to actually accurately simulate a universe in which it is true. But, even that wholly accurate simulation - if we are living in a simulation, it is possible the physical laws of the “real” universe are very different from those of our own. If our simulators are living under very different laws of physics, they may have access to computational resources which would be “unphysical” in ours. They may even be capable of hypercomputation. Maybe their laws of physics contain an oracle for the halting problem. Maybe they can execute supertasks. Even proof that simulating our universe would be uncomputable does not exclude the possibility that we are simulated on a hypercomputer We don’t even know that hypercomputation is physically impossible in this universe. We know of no way of achieving it given known physics, but we cannot exclude the possibility of unknown physical laws or processes which permit it, or even that it might be some bizarre mathematical consequence of existing theories which nobody has yet discovered. Maybe the hypercomputer will be the defining technological advance of the 57th century.