2

Please note this question isn't about "simulation" as such. It is cast in this way to illustrate a particular sub-to-super ontology relationship:

Given that all we see or seem, are the product of arcane computations on an unimaginably sophisticated machine. We may be tempted to wonder if we could learn something of the outside, after all we too make computers and run simulations...

Problem though is that our most basic physical laws are not "real". Our thoughts may not notice if the simulation stops and starts. All of the information at our disposal is at once as solid as our physical reality and as fleeting as a forgotten dream. How can anything from here tell us anything about what is really real?

Maybe our computers can tell us how the Grand simulator works? We can imagine that it must have a clock cycle. Or if we could generate enough activity we might notice the processor struggling a bit? Alas that with our limited view we could hardly hope to imagine what every possible kind of computer could be.

Yet if there is a simulation, there must be simulators. And they, they must have put, like us, their knowledge into the machines and software. They would have certain aims and expectations when designing a simulation. We should be able to connect with that sentience, that "aspect of design" in our world.

"Question: Has there been any attempts to make ontological distinction based, on objects of our Reality that would be necessarily "inherited" from an upper ontology? Or objects that we could in principle not place in a simulation?

Are there things that would necessarily propagate right down a simulation hierarchy, essentially binding all levels into a particular kind of reality?

**To be fair I was looking into the question from @tidymonkey81 when this question came up.

11
  • It sounds like a copy of kantian noumenon-phenomen distiction. Commented Jul 15, 2019 at 7:34
  • @MauroALLEGRANZA As far as I can see both noumena and phenomena are candidates for objects necessarily shared between upper and lower ontologies. A relation between the noumenal world and an upper ontology world seem enticing. However one may just as easily assert it is the phenomenal that is inheritable to the sentience inside a simulation.
    – christo183
    Commented Jul 15, 2019 at 9:34
  • 1
    The answer is no (unless our designers let us know in a way that induces us to trust them), which is why simulation speculations are not taken very seriously outside of pop-culture. Their authors simply assume the laws of physics like ours, perhaps with minor modifications, and would have nothing to go on and talk about otherwise. Related If we live in a simulated world, doesn't there have to be a first world that's real?
    – Conifold
    Commented Jul 15, 2019 at 23:40
  • @Conifold Given your answer to the related question I assume your are looking at this as seeking for physical explanation, and in that context I'm more or less fully in agreement with your answer. But foregoing physical(ism) reality's solid footing, what are we left with? Could we think what thoughts the simulators must have had when creating our world? - Yes, I know this comes back to: "What's the purpose of life, the Universe, etc." But maybe there are some interesting steps in between.
    – christo183
    Commented Jul 16, 2019 at 6:06
  • 1
    You’re assuming that simulation is a philosophically coherent category. It’s not. Few philosophers have taken up Bostroms notion of a simulation as a philosophically coherent thought. It’s science-fiction dressed up as philosophy, and for we know, that’s where Bostrom got the idea from. Commented Jul 17, 2019 at 8:00

7 Answers 7

2

As you point out, a computation can be sped up, slowed down, or temporarily halted without affecting how the computation proceeds at all. Similarly, if everything is properly implemented, other computations could be taking place in parallel without affecting the computation in question at all. So if our experience is nothing more than a computation, it should also be unaffected by such things. I.e., if everything is properly implemented we should not expect to see any Matrix-style glitches as a result of what is going on outside.

On the other hand, if the computation is intentionally altered due to hardware considerations -- e.g., things are only simulated / executed at a given resolution / discretization -- then we might in principle observe this and infer something about the hardware. (See also here: Are We Living in a Computer Simulation?) In principle, one might also infer something about the intent of the programmer (or more generally the nature of the higher-level system) purely from the content of the computation itself, but of course it may be difficult to do so.

1

Maybe the question is what do you mean by "our simulation". Lets say I designed a very immersive computer game that I spend a lot of time playing. At certain times in my game it is just so immersive that I forget I am the designer of the game and become simply the gamer. Is the designer-me from before the same me as the gamer-me who plays so immersed? Are you the same "you" as you were yesterday? What about those moments you can no longer remember?

If there are these simulators on the "outside" I suppose they may be struggling with the same questions of who is simulating them from their outside and so on...but maybe the outside simulator is really the game designer who becomes just very immersed in his own game as a player on the inside.

What do you suppose it would feel like to be this higher reality simulator being? How do you suppose would it feel different than you feel now about reality? How can you be sure that your are at the highest reality and nothing is inherited from higher still?

5
  • I have rather ambiguous "feeling" toward reality, to an extent because of questions you raise. Thus it is that I'm looking for epistemological answers here, that is "How can we know?", is there particular knowledge pertaining to the relationship between simulator and simulation. E.g. if we found some object of our experience that, in principle, could not be placed in a simulation by us, that would be reasonable grounds to conclude that we are at a top level Reality. - Your notion that a single consciousness can traverse from simulation-designer, to user, to simulated entity is intriguing.
    – christo183
    Commented Aug 1, 2019 at 6:33
  • @christo183 - The epistemological issue is that we cannot know that the claim of mysticism regarding the ultimate unreality of space-time is false Thus the plot-line of the film 'The Matrix' largely works. You could see this as an argument for simulation, but the film borrows its ideas from Buddhism and this is not a simulation theory. The extended universe would not be a simulation but a conceptual imputation. The scientific evidence cannot decide which it is but logical considerations and explorations of consciousness suggest it is the latter.
    – user20253
    Commented Aug 1, 2019 at 12:00
  • 1
    @PeterJ Indeed simulation isn't strictly necessary to explore these questions. Though I find it handy a paradigm to classify some of the concepts, i.e. upper-, lower, or derived ontologies. More pointedly, using simulation as a sort of modern(ist) language, for contemplating the "conceptual imputation", forces the relegation of scientific fundamentalist notions. A sort of circumvention allowing materialists to think outside the box, so to speak.
    – christo183
    Commented Aug 1, 2019 at 12:42
  • 1
    @christo183 - I get that it's a handy paradigm. The trouble is with simulation theory is that it makes us think outside the box only to find ourselves in another box and an endless regress of simulations. This is the problem The Matrix scriptwriters did not solve and which leads onwards to the principle of non-duality and Buddhism proper. I've met many young people whose interest in mysticism and metaphysics was sparked by this film so as you say, at least simulation theory gets us thinking out of the box. .
    – user20253
    Commented Aug 1, 2019 at 15:22
  • This infinite layers of simulations problem precisely illustrates one of the weaknesses with simulation theory in my opinion. Simulation theory assumes reality to be either the "true reality" or a simulation just like in the Matrix. But what does it mean to be in the matrix while playing a virtual reality game in which you are playing another game? Which level is your subjective experience really engrossed in and how many upper layers are you conscious of at every moment?
    – virtore
    Commented Aug 5, 2019 at 18:17
1

Yes we can know 'something' about 'outside' world, if we live in a simulation.

If Super Mario somehow figured out that his world is a simulation, what could he have judged about the world of its creators?

He might have thought:

  • Well, I can jump, probably my creators can jump, too.
  • I can walk upstairs and downstairs, probably there are stairs in 'real' world, too.
  • They, probably, also have underground and overground levels.
  • Likely, there are creatures that crawl, and creatures that fly in the 'outer' world, etc.

He wouldn't have ever had any actual clue of our world, nor any ability to peek into our world, but, he could have made some rough approximations.

Actually, I do believe that we live in a simulation, but I don't believe that they watch us real time. They just run millions of simulations and then look at the end results. Just like researchers at OpenAI do: https://www.youtube.com/watch?v=Lu56xVlZ40M

3
  • 1
    Simulations do not necessarily emulate 'reality'. "Well, I can teleport to other places by jumping into pictures, so …" Commented Mar 12, 2020 at 0:32
  • Yes, that's why I said 'rough approximations'. Not all properties of a simulation would differ from the 'reality' and, of course, not all properties of a simulation would resemble 'reality'. Commented Mar 12, 2020 at 8:20
  • Do you seriously claim that Ms. PacMan has a rich inner life?
    – user4894
    Commented May 31, 2022 at 18:05
1

Yes, looking for the "hand of a designer" is a legitimate activity, and could be applied by the inhabitants of a simulation. It is a key feature in anthropology, in the SETI program, and in the tests of a Creator God claim (tests for the Problem of Evil, the Problem of non-optimization, etc). The tests for a Creator god are the only ones that answer your particular question, but they are an application of the same search for design traces that informs the other two fields.

I have seen one author who applied one aspect of this method to our physics, and that was Paul Davies in The Hand of God. In one discussion, he evaluated whether our universe was "computable", and concluded NO, because it includes so many irrational numbers such as Pi, and asymptotes, which cannot behave appropriately in a digital calculation. His presumption is that any calculation would need to be digital not analog, and that one cannot have infinitely fine digital gridding to always get "close enough" for all purposes to an analog answer. If our universe is not computable, and he thought it was not, then presumably it cannot be a simulation.

Alternatively, one might possibly infer that if we are a simulation, then our designers must be able to compute in analog.

9
  • SETI (and, in general, signals intelligence?) is an interesting avenue.
    – christo183
    Commented Mar 21, 2020 at 14:15
  • 1
    Note as an aside, the "simulation" assumption presupposes that consciousness == algorithms, IE the functional version of Identity theory. I think we have more than enough examples of non-conscious functional implementations in our world to know that functional Identity Theory is utterly false. I did not address this problem with the question in my reply, as you pretty clearly are just using "simulation" to try to think your way to bigger questions.
    – Dcleve
    Commented Mar 21, 2020 at 20:10
  • Argument about pi is not a good argument because pi is computable. There's a finite-length algorithm that, given n, will halt after a finite number of steps, giving the n-th decimal digit of pi. That is, pi encodes only a finite amount of information. After all nobody complains that 1/3 = .333... requires an infinite amount of information to express. All you need is the high school long division algorithm. A better argument is the noncomputable numbers; but there's a good (constructive) argument that they don't exist; namely, that they're not computable!
    – user4894
    Commented May 31, 2022 at 6:29
  • @user4894 An nth decimal digit approximation of of pi is just an approximation to pi, and in formal logic terms is therefore NOT pi.
    – Dcleve
    Commented May 31, 2022 at 9:01
  • @user4894 -- if two terms which are not equal (B=/=A) are set equal in a computation (B=A) then that computation is now at risk of a logic explosion. This can lead to computational errors, or a computation crashing. Approximations work well enough that this characteristic of computational approximating rarely leads to computational catastrophe, but the potential is always there.
    – Dcleve
    Commented May 31, 2022 at 14:42
1

We can.

There are two different reasons given in physics that say this universe cannot be a simulation.

Reason 1: The universe cannot be a simulation because the computational cost of the hall effect grows exponentially with the number of particles with no limit.

Reason 2: While the uncertainty principle provides a limit to the linear precision of the universe, photon angle and angular precision measurements reveal the lack of any limit in the angular simulation of the universe.

Both of these lead to the same inexorable conclusion; no simulation of the universe is possible in first order logic. The host system must run on second order logic and be able to solve the halting problem for the Turing machine, and upturns the foundations of our philosophy.

4
  • Could you explain the reasons a bit more or provide some further reading that explain their relevance?
    – christo183
    Commented Jun 2, 2022 at 9:15
  • @christo183: To what level? These are all hard concepts; and the science of computer science has gotten to the point where it can say things about computing problems in arbitrary physics. But someone who doesn't know in their bones "there are more atoms in a thimble of water than there are thimbles of water in the oceans" will have a hard time of it.
    – Joshua
    Commented Jun 2, 2022 at 14:03
  • Doesn't this assume that the simulator doesn't use heuristics
    – Darrell
    Commented Jun 26 at 22:31
  • @Darrell: No, no it does not. It would seem that has to be the case but it really doesn't. The amazing answer is quantum mechanics doesn't change when you're not studying the system and this has been put to the test by light beam tricks that would require heuristics to solve the harder problem of rolling back the simulation after discovering after the fact that oops somebody was looking at the exact answer after all.
    – Joshua
    Commented Jun 27 at 0:29
1

If we look at simulators today, we dont actually compute the laws of physics precisely to simulate an environment. We use heuristics to approximate laws of physics, saving an awful lot of compute in the process but still achieving a result that seems realistic. It would make sense that this trend continues.

One heuristics that a simulator could use would be to:

  1. Synchronise the speed of the simulation with the compute. In other words if there is a lot of load placed on the simulator CPU, the simulated world "clock speed' "slows" down - its inhabitants completely oblivious as their brains slow down too along with their laws of phsyics. The simulator can catch up on load. Therefoee the speed of the simulation can be variable and "wait" on simulator resources such as CPU if necessary.
  2. The simulator can rewrite memory / history - if the contents of any observers memory can rewritten then they only need to think they witnessed a complex observation or situation play out - the end result needs to be consistent with the laws of physics but how that end result was derived has some room for heuristics.

In any case, it could be for example that a simulators need to use heuristics leads to an imperfect rendering with inaccuracies that we might detect as inhabitants of a simulated world.. For what if "dark matter" is an artifact of our reality being an imperfect rendering of certain laws of physics due to heuristic used by the simulator to render the simulation.

I'm not actually suggesting this is the case of that this is what I beleive- just trying to make a point.

1
  • P.S I don't like this line of thinking very much for a few reasons however. It implies we exist in a simulation but at some level there must be a "base" reality of "non simulated" people. How this base reality came into existence goes unexplained. It's ultimately this explanation we are seeking and "location" of where we are doing our reasoning (inside or outside of a simulation) doesn't help you with this explanation apart from adding an extra level of abstraction / indirection.
    – Darrell
    Commented Jul 11 at 10:15
0

The simulation hypothesis only appears plausible because of universal computation: the ability of universal computers to simulate any physical system including any other computer. There is a very wide range of ways of implementing universal computation. For example, almost any two qubit quantum gate can be composed into networks that can perform any physically possible computation.

As a result, the universality of computation implies that it is impossible for anyone in an alleged simulation to know what hardware it is implemented on. For more details see:

https://www.daviddeutsch.org.uk/wp-content/ItFromQubit.pdf

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .