[tomp]

Wait what? What this article is arguing is totally absurd - just because we can't model certain physical objects, they cannot exist?! That's like saying that since we can't model three gravitational objects interacting (i.e. the numerical solutions diverge, therefore to properly model the system, we would require increasing amounts of memory and time), therefore they cannot exist.

I'm not saying that the physical claim is wrong - I'm just saying that the explanation in the article is severely lacking/logically inconsistent.

[SnacksOnAPlane]

I'm a simulationist. I believe that the universe we're experiencing is a simulation made by a far-advanced civilization. So in my reality, if something can't be modeled, it can't exist.

Not saying that this is the truth, just the way I choose to understand things.

[mdxn]

I think you would have to be assuming that the far-advanced civilization's simulator has the same complexity as the models of computation we can construct. I think that's a huge leap to make. We might not be able to model something due to issues with, let's say, Turing decidability. Unlike us, the advanced civilization's might be able to because their constructable models of computation are strictly more powerful.

[drdeca]

I think you described what I wanted to in a much better and more concise way.

Thank you for doing that.

[drdeca]

Is a reality where computers can store and process e.g. Actual real numbers, inherently inconsistent?

I'm not sure, but from what I remember from Wikipedia, doesn't that allow for computations that can not be done with turing machines?

Iirc, there are models of computation which are self consistent, and can simulate Turing machines, but are such that it seems impossible to simulate in our universe (or on a Turing machine).

If I am correct in remembering this, and if these machines can simulate things with values from a set with cardinality greater than aleph null, wouldn't the argument that we are likely to be simulations also argue that we are likely to be simulations on a machine in a reality capable of simulating with a more powerful type of computation.

Ok I'm going to say that there are many potential hole in my argument.

But I think it might be possible that machines capable of hyper computation could simulate an infinite number of Turing machines in parallel.

And if it could do that, then it should be able to simulate an infinite number of universes like our own (not like it's own probably though).

And as such, shouldn't any argument that our universe is probably a simulation due to a universe like our own containing a large (but finite) number of simulations of universes like our own, Equally validly argue that our universe is almost certainly a simulation in a universe with greater computational ability than our own?

I acknowledge that this argument has many assumptions behind it, and that many of them could be wrong. I hope this argument doesn't come off as nonsense (preferably just Misinformed)

Also I suppose the large number of simulations might not be your line of reasoning for your beliefs.

Also, I think there's supposed to be a hiarchy(spelling?) of hyper computation, so maybe the surreal numbers would be a better basis for the argument than the reals?

I'm not sure if my line of reasoning makes sense, but I thought I would mention it anyway.

[tomp]

Still, until your belief is able to make physical predictions more accurate than other physical theories, it's just that - a belief, which has absolutely no physical validity.

[mikeash]

What says that the universe containing our simulation is unable to solve NP problems in polynomial time?

[bcbrown]

If something can neither be modeled nor observed, in what sense does it exist?

[scott_s]

That's not the question. Macroscopic behavior as described by Schrodinger's equation has never been observed. Because of having no observations, we assume it does not exist.

The linked paper is arguing, "Because solving Schrodinger's equation is non-polynomial, and such solutions are infeasible when N is large, then they can't exist." The linked paper is stating that the phenomenon has not been observed, assuming it does not exist, and then trying to explain why.

tomp's point is that we observe systems in our Universe that we can't model efficiently all the time. Hence, the line of argument in the paper doesn't hold up.

[yid]

What does modeling have to do with observability?

[jerf]

We observe n-body systems continuously. We are all participants in n-body systems continuously.