A game is a perfectly valid reason to do this. We're running millions of crude, low level simulations today.
> I’d like to see a more credible and compelling proposal than hand waved ‘history simulation’
Civilization is a crude, low level "history simulation" that doesn't try to be accurate.
Our world might be equally low level and crude by the standard of a future with vastly more computational capability.
Also consider that if you're in a simulation, then you have no basis for saying anything about the size or scope of the simulation: it's not a given many people are simulated at full fidelity, for example.
Hardly. They’re nowhere near a simulation of a “reality.”
How does a game character “see” something?
By a global agent checking the position of that character, along with the position of other characters, and a bunch of other variables and conditions, then tells the character if it sees anything.
This is a far, far, far cry from everything being made up of particles, emitting countless photons every “tick”, and those photons being absorbed by certain materials, emitting electrons and chemicals that travel to a neuron....
Nothing inside a game actually generates sound waves, or infrared and other parts of the EM spectrum that can only be perceived by certain creatures and tools..
It’s not even a simulation of that, just an extremely simplified and filtered representation of the end result as perceived by us, that we find convincing enough for the context.
Misses the point. The point is that each such instance is an example of something where the fidelity of the simulation is restricted by technical ability, not desire. It answers the "why" by giving a concrete example of an area where we keep pushing the boundaries of simulation, because more fidelity gives more believable interactions.
We went from SimCity and Civilization with really dumb AIs that never seemed remotely real, to games like Sid Meyer's Alpha Centauri where the faction leaders had distinct personalities, to Sims with large sets of competing traits. And yes, they're still crude. But you have game designers out there today toying with individual neural nets for the NPCs to augment the generic AIs.
And keep in mind it's not a question of whether or not we can do this tomorrow, but whether or not *our descendants from now until the heat death of the universe - or the end of civilization - will do this. If they do it even once, it's 50/50 whether you're in the simulation. Every additional run reduces the chance that we're in the "real world".
A simulation like GTA is only useful because it is staggeringly crude, but the goals of the simulation are commensurately staggeringly modest. I just don’t see how running a simulation of our universe, in our universe, at high enough fidelity to appear real even under detailed scientific analysis, could be useful or worthwhile.
I’d like to see a more credible and compelling proposal than hand waved ‘history simulation’. What’s the point of a history simulation that runs many times slower than real time? If it’s low fidelity, it will also be low accuracy, so why simulating at the physics level at all? I don’t see how it would give useful results.