The Simulation Argument goes like this: a simulated universe would necessarily feel absolutely real to the creatures simulated within it. Since a simulation can contain vastly more inhabitants than the number of people required to run the simulation, it is most likely that either we ourselves are living in such a simulation or that something prevents those simulations from being made (its theoretical impossibility, every civilization dying off before it gains the capability to create them, or something along those lines).
My question is, is it theoretically possible to simulate General Relativity on a grand scale? I’d imagine there’s only a few ways to go about it:
-You can simulate a privileged reference frame and apply relativity from there.
-You can simulate the entire universe from the reference frame of an arbitrarily large number of observers, and synchronize the universe across all these reference frames.
-Same as the second, but you don’t bother synchronizing the universe (there’s a glitch in the Matrix lmao)
Each of those seems less likely than the last, and as I understand it the first one is impossible by our current understanding of physics. Granted, it may be possible that math in the simulating universe doesn’t work like it does here, but when you go that route all bets are off and that universe is so alien it had might as well not exist.
I think your last point is key. The guys running the sim can make up any rules they want for inside it. Maybe Relativity applies in the outside world, and maybe it doesn’t. In either case, it’s observed here, so if we’re in a sim, they’re simming it.
And…why not a privileged frame of reference? Or maybe only a few. Or maybe one per each observer? If it’s just a sim, then they can cheat all they want.
The could simply hard-wire rules into our minds: “You perceive relativistic effects.” They don’t have to sim time-space; they just have to sim our telescopes.
You got it right when you said “All bets are off.”
I can’t actually see what the argument is that general relativity is impossible to use for a simulation. Any physical theory that was incapable for the use of simulating a physical situation wouldn’t really be a physical theory, besides which we do use general relativity to simulate certain physical situations and indeed cosmological simulations use general relativity as there base (e.g. the Bolshoi simulation).
Of course general relativity isn’t particularly easy and even describing a very simple physical situation using it can be complicated and also, not being a theory of everything, it doesn’t tell the whole story. But these are not unique properties of general relativity.
In practical terms of how you go about using general relativity to simulate the cosmos, the assumptions of cosmology provide a nice ‘privileged’ frame of reference. This frame of reference is ‘built’ from an infinite ‘network’ of ‘privileged’ (note to self don’t over-use ‘scare quotes’) observers. Though it’s not necessary that a meaningful global frame exists or that it is can be regarded as privileged in some way to use general relativity to simulate physical situations, it merely makes it easier.
I’m not trying to say that general relativity can’t be simulated at all, because as you said any physically meaningful theory can be simulated. That said, I would expect simulating a unique reference frame for every observer (person or machine that we can interact with) would be computationally prohibitive, or else there’s some observable artifact of the simulation in the form of a privileged reference frame. I’m going to ignore “we’re programmed to see relativity that isn’t there” as a cop-out because relativity has consistent and measurable effects that should require some kind of simulation.
I figure that giving every observer a unique reference frame would cause the cost of computation to increase factorially with the number of observers, as each would have to coordinate with every other reference frame. Right now we only have a few dozen observers in different reference frames that we know about, but there’s nothing suggesting that adding more is impossible.
If I’m reading your third paragraph correctly, Asympotically fat, you’re saying that our universe has something that could be considered a privileged reference frame? Would it be possible for someone inside that simulation to determine that that reference frame is unique among all others, or is it a computational simplification that doesn’t leave any artifacts?
In theory, if we’re all simulated, then computational time isn’t much of a factor. The computer could take as long as needed to run each calculation then execute a “clock” instruction that would make the calculated state the active state. From the simulated point of view time would flow the way we expect it to flow, but each clock cycle could take years to calculate the next state.
The simulation question depends only on one thing: “consciousness”.
My argument is that “consciousness” cannot be simulated and all your question is falling immediately.
Every relevant observer we know of is sharing the same reference frame, or reference frames so similar to each other that the differences between them are only minor, easily-calculated perturbations.
mendi80, your argument is equivalent to arguing that consciousness can’t exist at all. If it can exist in the world, then it can exist in a simulation of the world.
If I am understanding what is being argued in terms of “privileged” observers…
is it possible that I, as experiencing a simulation, am the only one in this Universe, and the rest of you are all quite complex facets of the simulation?
And then, perhaps I am the only person in “reality” left to experience the simulation, or perhaps there are an infinite number of people experiencing their own separate simulation.
Imagine there is a singular person keeping the machines running. Or an entire civilization. Or perhaps we are all, as a species, traveling away from our home planet of Earth, which once was, consenting to generation after generation of simulation as we wait to arrive at a new habitable planet, having destroyed the old one. Boggles the god-damned mind!
All right, that’s certainly good enough to square everything together. Thanks! I should have figured that out when I tried coming up with counterarguments and always ended up resorting to hypotheticals.
On a somewhat GD/IMHO note, I think the Simulation Argument is reasonable enough that it shouldn’t be discounted, even though I’m inclined to believe that we aren’t in a simulation. It’s an interesting extrapolation of transhumanist speculation/fantasies about an eternal simulated afterlife and what that would imply.
The speed of light limit has an important advantage - you can limit the things impacting each simulated entity based in its event horizon.
While it might be true that more people are being simulated than doing the simulation, the universe of the simulator could have a lot more people. I’ve written many simulators, and you try to keep the number of simulated objects to a minimum. What that minimum is we can’t say since we don’t know the purpose.
Indeed. Simulations of microprocessors in the process of being designed can take a day to simulate a few milliseconds worth of computation. And that is simulation at a relatively high level.
When simulating a 2-D surface do you specifically need to describe every imaginable curve that can be drawn on that surface and it’s relationship with every other curve on that surface? The analogy with general relativity is that spacetime is a 4-D surface and each observer is a curve
In our Universe there’s a set of observers for which the Universe appears to be isotropic and homogeneous (there must be because we assume the Universe is isotropic and homogeneous*, but it can’t appear this way to all observers), these observers can define a ‘privileged’ frame of reference.
*edited to add: minus the small peturbations which Chronos mentions.
Right, but you are assuming that the other observers are not constructs of the simulation itself. Of course, being a construct they would act as though they held the same view of reality. They would express it as being homogeneous, assuming that was their task, or better yet, they are artificial constructs unaware of their position.
I said everywhere, not for every reference frame. At every location, there is some velocity such that an observer at that location and velocity would see the universe as homogeneous and isotropic.