4 ms·

The author is being slightly inconsistent, with respect to the other premises of the story, in saying there is a feedback loop: as it is set up, simulations hav

The author is being slightly inconsistent, with respect to the other premises of the story, in saying there is a feedback loop: as it is set up, simulations have no causal power in the world running them. If the top-level 'God' people choose to run the simulation faster than real-time, I think we can say that no change will be observable anywhere in the simulation stack, except that the people in them will feel as though they have chosen to speed up their simulation and yet not seen any change in it - which is another of the things that will tell that they are in a simulation and not at the top level. (Note that, as Diane has presumably already said in her paper, free will is definitely an illusion in the simulations. I guess it also is at the top-level, given that their universe is deterministic to the point that it can be simulated to the tiniest detail.) Now suppose the top-level people choose to run the simulation in reverse. Again, I think we can say that nothing will seem to change to those in any simulation, because, at any point in the reversal, everything, and in particular everyone, will be in the same state as they were in the forward pass, which for the people includes having the same memories, plans and expectations - it will be indistinguisable from an instant in forward-running time. They won't even remember that, at some point, the clock was reversed, as that was not something in the forward simulation's past. This is where it starts to get interesting: if the clock is reversed again and left to run, what will be observed in the simulations, and by the top-level people looking at the first-level simulation, when the simulations reach the time of the original reversal? Update: I'm leaning towards the view that the top-level simulation continues forward in time from this point, and observes its simulation go around the loop. In general, the Nth simulation goes round the loop N times before proceeding past the reversal point, but no simulation observes that it is going or has gone around the loop, and therefore cannot determine its depth by counting loops. Alternatively... The moment the 'camera' appears in a simulation, it has diverged from the 'real' world. Now some of its inhabitants know they are in a simulation, and so they are not replicating what happens in the real world. The butterfly effect will likely ensure that the divergence will become general. I think this could be undone by sending the simulations around a time loop, as then, if my supposition about how these loops play out is correct, each simulation will exit the loop in the same state as the real world, with no memory of the camera having once appeared.

Nition 6y ago
> Note that, as Diane has presumably already said in her paper, free will is definitely an illusion in the simulations. I guess it also is at the top-level, given that their universe is deterministic to the point that it can be simulated to the tiniest detail. The top level could be very different to the first level simulation. The only part that actually needs to be similar is that the real universe and the first-level simulated one both create their own simulated universe.
- mannykannot 6y ago
  I think you are saying that the top-level could be modeling a completely determinstic world that could be anywhere from slightly to completely different from their own? That's a good point, and it would mean that the first simulation need not see the viewport pop into existence. They could not then assume they are the top-level, however (which would be false), and I don't think they could conclude "either we are the top-level or the first simulation", as, if the latter of those possibilities is the case, they cannot rule out the level above them also being some sort of simulation, only that theirs is not an exact replication of it.
Chris2048 6y ago
> The author is being slightly inconsistent I agree, and I'm not sure the scenario given could happen. The basis of the possibility of the scenario is "the universe is deterministic" and "we have infinite computing power" - but then, is the infinite computer itself part of the universe? b/c as soon as the top layer interacts with the sims, it diverges, which means the sims are not the same as the top layer anymore, so are not equivalent to it even though they are otherwise deterministic - the first sim under the top is deterministic, but dependent on interaction with another deterministic (top level) universe. Hence, the first sim is no longer representative of the top level - the top level can still interact with the first sim, but it can no longer consider it to be a mirror - this is the answer to the paradox of "running the sim into the future" - once the top level observes the sim, it is altered in a way that might cause the sim to diverge. > when the simulations reach the time of the original reversal? The sum total time of the reversal is finite, so the top (level 0) level will observe the reversal in a finite time, the level 1 will therefor observe the same finite delay wrt the top. Hence each level N will be delayed by the same amount wrt the level above. The idea of a "loop" would be wrong, b/c when a sim reaches "the time of the original reversal" the result only affects the next sim down. But each sim as level N does go through the loop N times, because it shares the loop of each of its parents, hence if a loop delays by 10s, level N which goes through N loops is delayed wrt top by N*10s > The moment the 'camera' appears in a simulation, it has diverged from the 'real' world True, but now the "next" sim interference will not be replicated between level 0 and level 1, which in turn can cause level 1 to diverge, and so on. > with no memory of the camera having once appeared You could just stop all the sims after you interfered with them, and them restart the sim without interfering (or equivalent - like rewinding the sim and then playing it w/o interfering) - the same would happen in all sims too, so they would be, and have, the same pristine sims.

Nition 6y ago

> Note that, as Diane has presumably already said in her paper, free will is definitely an illusion in the simulations. I guess it also is at the top-level, given that their universe is deterministic to the point that it can be simulated to the tiniest detail. The top level could be very different to the first level simulation. The only part that actually needs to be similar is that the real universe and the first-level simulated one both create their own simulated universe.

mannykannot 6y ago
I think you are saying that the top-level could be modeling a completely determinstic world that could be anywhere from slightly to completely different from their own? That's a good point, and it would mean that the first simulation need not see the viewport pop into existence. They could not then assume they are the top-level, however (which would be false), and I don't think they could conclude "either we are the top-level or the first simulation", as, if the latter of those possibilities is the case, they cannot rule out the level above them also being some sort of simulation, only that theirs is not an exact replication of it.

Chris2048 6y ago

> The author is being slightly inconsistent I agree, and I'm not sure the scenario given could happen. The basis of the possibility of the scenario is "the universe is deterministic" and "we have infinite computing power" - but then, is the infinite computer itself part of the universe? b/c as soon as the top layer interacts with the sims, it diverges, which means the sims are not the same as the top layer anymore, so are not equivalent to it even though they are otherwise deterministic - the first sim under the top is deterministic, but dependent on interaction with another deterministic (top level) universe. Hence, the first sim is no longer representative of the top level - the top level can still interact with the first sim, but it can no longer consider it to be a mirror - this is the answer to the paradox of "running the sim into the future" - once the top level observes the sim, it is altered in a way that might cause the sim to diverge. > when the simulations reach the time of the original reversal? The sum total time of the reversal is finite, so the top (level 0) level will observe the reversal in a finite time, the level 1 will therefor observe the same finite delay wrt the top. Hence each level N will be delayed by the same amount wrt the level above. The idea of a "loop" would be wrong, b/c when a sim reaches "the time of the original reversal" the result only affects the next sim down. But each sim as level N does go through the loop N times, because it shares the loop of each of its parents, hence if a loop delays by 10s, level N which goes through N loops is delayed wrt top by N*10s > The moment the 'camera' appears in a simulation, it has diverged from the 'real' world True, but now the "next" sim interference will not be replicated between level 0 and level 1, which in turn can cause level 1 to diverge, and so on. > with no memory of the camera having once appeared You could just stop all the sims after you interfered with them, and them restart the sim without interfering (or equivalent - like rewinding the sim and then playing it w/o interfering) - the same would happen in all sims too, so they would be, and have, the same pristine sims.