Are we living in a simulation? The evidence

29

17

I am not questioning whether the simulation topic is outside science. I am asking what evidence there is or could be to resolve whether we are or not.

Living in a simulation has been a topic for philosophy since at least Descartes, with his 'evil demon not less powerful than deceitful'. Zuangzhi's 'Now I do not know whether I was then a man dreaming I was a butterfly, or whether I am now a butterfly, dreaming I am a man' could be argued to be a statement of the same idea. What I am interested in is, is can we get beyond speculation and make headway on this? How does the evidence stand, what tools do we have, and what implications might the different answers have?

Here I quickly summarize some of the main positions:

1. "Are you Living in a Computer Simulation?" [PDF] (2003), Nick Bostrom, makes the case we are in a simulation, on the balance of probabilities

2. "Quantized gravitational responses, the sign problem, and quantum complexity" (2017), Ringel & Kovrizhin, makes the case such a simulation is impossible using classical computers. But that leaves quantum computers.

3. "The Case for Libertarian Compatibilism: A Brief Overview" (2014) gives a more detailed idea of what being in a simulation might mean, as a kind of peer-to-peer reality

4. "Fermi Paradox: a simulation solution" [PDF] demonstrates a serious argument that the simulation hypothesis is the best answer to the Fermi Paradox. Although as recently discussed on here, the existence of this paradox is open to dispute, and relies on a range of assumptions.

Anyone got anything else evidence based? Any models, or discussions of implications?

1

Comments are not for extended discussion; this conversation has been moved to chat.

– None – 2018-02-04T12:38:26.393

FYI - My new series "Saints and Simulators" will explore the answer to your question at length.

– Chris Sunami supports Monica – 2019-01-24T14:15:21.873

Reword "simulation" as "magic show" and then read this so-called eastern story in 2 versions

– Rusi-packing-up – 2019-06-03T04:46:36.967

I personally do not see what is the difference between being in a simulation and non-simulation, because : if we are in a simulation then this Universe is simulated in another computer, machine, mind...etc. And if it is not, then we already know that everything is a simulation in our brains like this cup of coffee I hold in my hands now. My brain simulates some kind of information that is unknown, shapeless, spaceless, timeless to construct what I perceive and taste as a cup of coffee. For suppose if we are in a simulation , then our brains do not exist, but there is computer where it happen – SmootQ – 2019-08-28T16:05:10.757

It's just a matter of flavor ... not a matter of whether it is true or not, whether it is a simulation running in your material brain, the mind of God, a computer... it is all considered a simulation, no matter who or what material runs it .

But in the end, there must be something that is fundamental, that is not simulated, and that is itself shapeless, colorless, spaceless and timeless because it is not simulated by a brain to take these qualities. – SmootQ – 2019-08-28T16:06:42.560

If you really want to know this, study the works of Gurdjieff. – Marino Proton – 2020-12-21T20:39:50.583

@Rusi-packing-up - I love that story. Always been one of my favorites. It hardly provides anything resembling evidence though. – dgo – 2021-02-23T15:36:42.143

2

I am a bit surprised by the hostility to the concept of this question, given that it is a classic problem restated by nearly every serious philosopher. If your stance is, 'I don't want to think about it', fair enough. No one is making you. This thread is clearly stated as being about evidence for or against. You might have saved everyone time if you didnt want to address that.

I found this great article, that I'd say answers my question as thoroughly as it's going to be. Please consider what it has to say if you have been argueing this is a meaningless question.

"David Chalmers has argued that we should consider the 'simulation hypothesis' not as a skeptical hypothesis that threatens our having knowledge of the external world but as a metaphysical hypothesis regarding what our world is actually made of. " http://philosophycommons.typepad.com/flickers_of_freedom/2014/08/the-case-for-libertarian-compatibilism-a-brief-overview.html

The Peer-To-Peer Reality model framed here seems to me directly comparable to the Yogacara Mind-Only school's stance. The objection around nested or recursive realities can be seen as being directly addressed by Nagarjuna, as described here: https://absoluteirony.wordpress.com/2014/09/17/nagarjuna-nietzsche-rorty-and-their-strange-looping-trick/ Buddhist thought acknowledges reality is confusing, it does not throw hands up and refuse to engage. This avenue of thought looks to be a route towards recognising the sophistication of some Buddhist thinkers.

24

I would like to argue that the topic (especially the search for proof for it) is rather useless. With a search for evidence, this topic is very similar to the topic of the search for evidence that God exists - which is, naturally, an assumption that brings the term "God" into a field it simply cannot exist in. You can't search for an ontological evidence for something that by definition isn't in the ontological field (or, at least not in the human sense of ontology).

If you want to talk in the logical field, then sure, you can present many arguments for and against the simulation argument, but from your description I see that you know at least some that's enough.

So, my point is, the question is pretty much "worthless" (not really, or else I wouldn't want to answer it, so it is worth replying to at the very least), as a search for evidence in simply not realizing the difference between the fields in which evidence lies and the simulation argument exists.

And for a final note, I think Kurzgesagt (a Youtube channel) presented some excellent points on the hard time it'll be for "aliens" to create a simulation in which we exists (although it does assume these "aliens" live in the same world we live in, with the same physical laws limitations we have, but it's still a nice video to watch - they have amazing animations skills ;).

3It involves testable predictions, surely? The Bostrom argument argument points out we will run simulations sooner or later, but they must have limits in scale and complexity. So we look for those. It's as ontological as the anthropic principle, in that even if it may not be falsifiable in the end, it is indicative and can guide our thinking. Whereas god is usually just a deus ex machina, a magic black box – CriglCragl – 2018-02-01T20:51:58.510

13@CriglCragl, any simulation can only be detected from the inside by performing a measurement and finding a discrepancy between the results and what would have resulted if the measurement had been performed outside the simulation. Without access to the outside of the simulation, there's nothing to compare against and no way to tell simulation from reality. – Mark – 2018-02-01T21:59:59.540

2@CriglCragl what Mark said should suffice (and I would like to add that in the experimental-view, I think it would be amazing to try and get AI to try and learn about itself from inside, as many scientists [probably falsely] consider AI to be more advanced then human brain, so I think it could be a perfect experiment for such arguments). And I would like to state that what Mark said is exactly what I mean by different fields of knowledge - you can't possibly assume to get evidence (or more generally, any kind of empirical experience) from the non-ontological field. That is the "outside". – Yechiam Weiss – 2018-02-01T22:17:15.383

@Yechiam Weiss That May be true, but surely depends on which of a set of assumptions you make? – CriglCragl – 2018-02-02T00:30:34.990

3@CriglCragl which assumptions would you like to make that'll bring the possibility to claim otherwise? – Yechiam Weiss – 2018-02-02T00:32:01.790

@Yechiam Weiss The typical assumption runs, what will the theoretical limits be on a simulation we could run in this universe? Then you can use those conclusions to examine whether our own experiences could be in a simulation, of a a universe with the same character (physical laws etc). So scale, granularity, complexity, discontinuities/inconsistencies, all could offer clues. Potentially we could crash the program. – CriglCragl – 2018-02-02T12:12:46.843

2@CriglCragl OK so those aren't evidences, those are theoretical ideas, which of course can be made, but on the exact same level I can just tell you "just imagine an outside world in which the physical laws we know simply don't exist, the boundaries we have are way wider in that world", for example. – Yechiam Weiss – 2018-02-02T13:01:26.097

1It isn't worthless to me. For the almost nothing it's worth: I believe in a truly just God. For such justice to exist, all injustices that ever happened have to be righted. I very strongly suspect I am living in a 'simulated' reality, one of many which will be 'superimposed' together to determine the truly just solution (God can't just determine it, we have to have it shown/proven to us). Of course this might just be me slowly going insane and has nothing to do with philosophy. – anon87987 – 2018-02-02T22:34:55.090

1@anon87987 I didn't mean to make it sound like this question is worthless TO ASK, I meant it's an entire topic that's worthless to talk about simply because you can't, by definition, know anything about it. The "beyond the human knowledge limits". But I must say, I'm not sure I understand what you said about God and the "just" solution. – Yechiam Weiss – 2018-02-02T23:02:56.170

"You can search for an ontological..." should rather be "you cannot... " ? – leonbloy – 2018-02-05T01:28:59.963

@leonbloy oh wow can't believe I missed that, thanks! – Yechiam Weiss – 2018-02-05T12:38:15.643

@CriglCragl: I think you are ignoring a fact you cannot get around: There is no evidence whatsoever that reality itself, when in interaction with an observing subject, may very well show a certain scale, granularity, complexity, etc. that is typical for simulations as well. AFAIK, that's been one of the insights of early QM theorists. Does that mean reality is a simulation? No, it doesn't. They may share these properties. That's why most philosophers reject this kind of thought experiment as a serious metaphysical position. It multiplies ontologies without reason, nothing more. – Philip Klöcking – 2018-02-05T18:57:09.540

@CriglCragl Even assuming it is physically the same as this one, we cannot rightly take discrepancies as evidence that there is an external reality since it might be a simulation also. And without a measure to determine a "base" reality we loose the need for distinction, i.e. reality is simulation. Actually I suspect that any internal proof that we are in a simulation will also prove that we cannot know that outside is a base reality. - As the old lady said: "it's turtles all the way down". – christo183 – 2019-06-18T14:48:51.630

interesting... should 'ontological evidence' be 'ontic evidence'? also, do all simulationists think that actual reality is not a part of ours? – None – 2019-08-22T10:55:48.617

14

I propose that we cannot know whether we are living in a simulation.

• One of the ways to detect a simulation supposedly is to detect errors from the inside. But there is no reason whatsoever to believe that our version of the universe has anything to do with the real reality. Every last single piece of our physics etc. could be artificial, hand-picked by our creators (or just by a deep learning machine of theirs, maybe?). We cannot reason about their reality, having only our simulated reality as base. We have no way to see if they made an error in our simulator by perceiving a seeming contradiction in our own universe. There will always be the possibility that our theories are simply wrong or incomplete.
• The complexity of creating such a simulator: We feel that we know a lot about quantum mechanics, black holes and other complicated topics. But compare what we know today (and what kind of technology we have today, both for doing physics, as well as our computing power) with what we knew/had 50 or 100 years ago. Think about what we could have in 50 or 100 years. Absolutely impossible to tell. Now place our simulators, say, 1000 years into the future. There is no hope that we would even recognize what they are doing. We cannot argue in any form or fashion whether building such a simulator would be physically or logistically possible or not (plus, we do not even know the limitations of the "real" universe).

(As an aside, Nick Bostrom does not make the case that we are living in a simulation; instead he - using pure logic, no philosophy, gives 3 possible scenarios; he does not give a probability of which one would be the most likely in our specific case. He says we should look at his three solutions with equal probability, which, to me sounds like he correctly states that we just don't know.)

(Second: that other paper about Fermis Paradoxon contains no hint about us living in a simulation, at all. The authors wrote a simple simulator which makes it possible for them to simulate the long-term life of civilizations - in the scale of 1 million years being one step of the simulation. Hence totally unrelated.)

Logic, is not part of philisophy now..? Extinct, boring, or in a simulation? He says we should give equal weight. We aren't talking about a Cartesian demon. Clearly our reality can be investigated logically, systematically, and responds in consistent ways. We can look for ways to put limits on options, given conscious assumptions. You seem to prefer, only make unconscious assumptions, and don't play with them or everything would be up for grabs. Granularity, scale, highly complex events with unpredictable outputs that would demand great computations, discontinuities. All could give insights – CriglCragl – 2018-02-02T00:12:23.197

2@CriglCragl, I do not really understand what you are trying to say. But let me address the pure logic; no philosophy thing. It means that he is building a simple mathematical model of what he is trying to say and then assigns probabilities. As he has no knowledge about the "input probabilities" he can make no assertions about the "output probabilities", and he doesn't do so. His well-done paper is much closer to physics/maths than philosophy (and sure, you can treat all them as one if you prefer). On the rest of your comment, indeed, I prefer the viewpoint I wrote in my answer. – AnoE – 2018-02-02T00:27:55.757

I'm think the general gist here is good, but I think point three could be elaborated upon. I'm not sure why you think it's likely that the simulator would even know if any minds/life sprung up inside the simulation. If your goal for the simulation was anything other than observing all possible minds, and possibly even then, it would be very possible to create a simulation which may or may not give rise to some form of conscious minds as just a side-effect, without implementing any introspection that could detect those minds. – mtraceur – 2018-02-02T03:57:57.247

@mtraceur, thanks for the point. My last two arguments were somewhat weak, I have removed them and elaborated the other two a bit more. – AnoE – 2018-02-02T07:22:10.323

9

To focus on the title of the question, there is rather little active research being done around this question, but I'll share what I know.

First is a paper from 2012, Constraints on the Universe as a Numerical Simulation (Beane, Davoudi, Savage). This paper talks about running low level physics simulations that scientists currently perform, and draws comparisons to a theoretical computer simulation of our universe. They theorize about measuring how certain high energy particles are distributed, and provide an upper bound to the size of a grid-like simulation of the universe. However, our most accurate measurements are still orders of magnitude away from being able to determine anything. They close with the following:

Nevertheless, assuming that the universe is finite and therefore the resources of potential simulators are finite, then a volume containing a simulation will be finite and a lattice spacing must be non-zero, and therefore in principle there always remains the possibility for the simulated to discover the simulators.

Next is a recent paper from 2017, On testing the simulation theory (Campbell, Owhadi, Sauvageau, Watkinson). This paper is a bit more philosophical than the previous. There's a bit of discussion on the nature of reality, i.e., "the collapse of the wave function" and to paraphrase "what does it mean to be seen?" The focus is on quantum mechanic effects and present experiment(s) to probe how observers interact with and affect (and effect even) particle interactions. They draw a comparison to video games: often, compute resources are saved by not rendering something that can't be seen. Does something like that happen in the real world (going back to the wave function collapse discussion mentioned earlier)? The paper assumes that observing such an instance of reality suddenly being rendered (or suddenly being rendered in higher detail) would imply that we are in a simulation, but I think the conclusion is not so strongly implied by the premise.

I can't speak to how feasible any of the experiements in the second paper are.

I'm sure there are other relevant scientific papers, but these were all I came across when I was researching the topic (briefly) not too long ago.

5Whatever world is simulating us, they picked our physics to suit their goals. For us, simulating QM/QED/QCD would be a computational nightmare, especially because of the Bell inequality (it precisely says our universe computes gigantic amount of unneeded stuff that gets immediately thrown away). Apparently they decided to use this model... maybe such simulation is quite easy in their physics? – kubanczyk – 2018-02-01T22:52:20.450

2@kubanczyk I think part of the motivation for the "we live in a simulation" theory is the idea that quantum effects are defects in the simulation, rather than desired effects in the simulation. Something like "the rules don't know what to do in this circumstance, so just pick something so the simulation doesn't crash". Of course QM is more delicate than that, but that's the idea. – Ian – 2018-02-02T16:06:06.597

I couldn't disagree more @Ian. Theories help us to predict stuff when we look around. Mixing it with emotions - like calling what you genuinely see a "defect" or "glitch" - makes you only more liable to errors. QM is incompatible with human intuition, but it's a fact about humans, not about QM. Consider that thought experiment http://lesswrong.com/lw/ps/where_physics_meets_experience/

– kubanczyk – 2018-02-24T10:16:31.890

7

Of course it is possible that you are "living" in a simulation right now!

I'd say it is about 50% chance. we call it dreaming and it happens every night.

Well, you might say, that is pure nonsense, for I know I'm not dreaming right now. But do you really?

I dreamed last night that I was in Germany, in a house which looked out on a ruined church; in my dream I supposed at first that the church had been bombed during the recent war, but was subsequently informed that its destruction dated from the wars of religion in the sixteenth century. All this, so long as I remained asleep, had all the convincingness of waking life. I did really have the dream, and did really have an experience intrinsically indistinguishable from that of seeing a ruined church when awake. It follows that the experience which I call “seeing a church” is not conclusive evidence that there is a church, since it may occur when there is no such external object as I suppose in my dream. It may be said that, though when dreaming I may think that I am awake, when I wake up I know that I am awake. But I do not see how we are to have any such certainty; I have frequently dreamt that I woke up; in fact once, after ether, I dreamt it about a hundred times in the course of one dream. We condemn dreams, in fact, because they do not fit into a proper context, but this argument can be made inconclusive, as in Calderon’s play, La Vida es Sueno. I do not believe that I am now dreaming, but I cannot prove that I am not. I am, however, quite certain that I am having certain experiences, whether they be those of a dream or those of waking life. — Bertrand Russell, Human Knowledge: Its Scope and Value

As for the possibility that we are living in a simulation run by some other "people" one world up, I'd say that idea, in my opinion, is just as sensible as last night's dream, and attempting to seriously analyze it, just as silly.

Reality is incomprehensible and mysterious as it is, and the idea of an imaginary one world up, doesn't make an iota of a difference.

Unless of course, it is one of the ideas you are currently playing with, on your way to personally waking up.

And now that I have claimed above that a serious analysis of the simulation hypothesis is silly, I will add my personal silly analysis of it. It is written from a Buddhist / Advaita / Sufi / Hasidic standpoint:

1. Reality is mysterious, in the sense that it transcends reason and logic and human capacity to comprehend it. It is unspeakable. Unthinkable. The simulation hypothesis on the other hand is a simple thinkable and intelligible idea, just like life after death, or any other silly metaphysical idea about our reality. As such, it does not add or change, or explain anything really. In particular, it does not make anything less real than it already is.

2. Most proponents of the simulation hypothesis conceive of it as a computer simulation or a computational simulation. However, the day you find that which Tibetan Buddhist call self-knowing awareness, or Advaita people call Atman, or Sufis and Hasidic people call Godliness, you will know with certainty that reality cannot possibly be explained as a computation. In fact, the belief that reality may possibly be a computation, is an indication that one is, like almost everybody, still asleep.

So, you conclude, Bertrand Russell is just silly? Lucid dreaming offers various techniques for determining if you are in a dream. Buddhist theology concludes all beings are in a dream, and offers wide ranging and complex analysis of it, towards waking up. All, just silly? Or, is that description better applied to dismissing the concept without analysing it..? – CriglCragl – 2018-02-03T16:37:02.820

1You misunderstood me. I did not imply Russell is silly. On the contrary, I quote him as a reference to my claim that one cannot be sure one is not dreaming right now. I am aware and experienced in lucid-dreaming. It does not offer techniques that let you determine with certainty that you are dreaming. I recall numerous occasions where I have failed to realize I was dreaming in fantastic circumstances. in one occasion I was examining my hand, which had three fingers, to see if it looks normal, and I concluded I was not dreaming. – nir – 2018-02-03T19:43:39.177

As for Buddhism, I am familiar with it and I practice some of its teachings. It is from a Buddhist standpoint that I say that reality is mysterious enough to render the simulation hypothesis nonsensical. I will add an explanation to the answer. – nir – 2018-02-03T19:57:33.613

Even if you count dreaming as being in a simulation and assume that you always believe yourself to be awake when dreaming, we aren't asleep 50% of the time, so that's only a ~33% probability that we're in a simulation. That may seem pedantic, but if we're to do any sort of analysis of a problem that can't be answered definitively, we need to be careful about how we use probabilities, and shouldn't be so careless as to discard the prior probabilities when they're easily available. – Ray – 2018-02-04T23:32:56.207

1Great answer. The only "loop hole" is your claim in #2, that we will know with certainty that reality is not a simulation. While I do agree on that 100%, for the rational mind it is just an unprovable claim that adds nothing to the understanding of the topic. However, as I see it, that is just the way it is. Our mind simply won't understand reality on that level, no matter how much we try. – jackthehipster – 2018-02-05T07:29:57.947

"You will know with certainty that reality cannot possibly be explained as a computation" sounds unnecessarily controversial given what you actually have in mind (something with capabilities of a Turing machine, I assume). It would also help to explain why, the experiences you refer to do not manifest themselves in such technical terms. – Conifold – 2018-02-06T05:36:55.317

@Conifold, it is impossible to explain why said aspect of awareness cannot arise in computation. I don't understand what you mean by "unnecessarily controversial given what [I] actually have in mind" – nir – 2018-02-06T20:51:53.917

@jackthehipster, you are correct. I would love it if you contact me by email at nir@winpdb.org. I would like to ask you some questions. cheers – nir – 2018-02-06T21:00:40.547

If it is indeed impossible to explain then there is no credible certainty. As for "computation", there are conceptions of it that would turn Hinduist Māyā or God dreaming up the world before creation into "computations". – Conifold – 2018-02-06T21:13:02.337

Further to the comments on Buddhism, - the entire point of the practice is to establish the truth and this would includes the knowledge of whether we are in a simulation. It would be ridiculous to suppose that skilled practitioners do not know whether or not they are in a simulation, There is no empirical way to establish we are not, but sensory information is only a small part of what is available to us. , . – None – 2018-02-07T13:04:40.390

Why 50% ? I have mentioned the defect of the usage--'simulation' (in my answer). – SonOfThought – 2018-02-08T06:27:02.793

@nir: Why not just ask right here? – jackthehipster – 2018-04-11T10:19:52.203

@jackthehipster Comments are not for extended discussion – CriglCragl – 2019-08-27T23:11:30.267

6

No, we're not living in a simulation. See my question https://physics.stackexchange.com/questions/377516/ and especially see John Rennie's answer to it, citing the relevant research.

In addition to that, see Mermin's article http://physicstoday.scitation.org/doi/10.1063/1.880968   A (legal) pdf is downloadable from http://www.physics.smu.edu/scalise/EPR/References/mermin_moon.pdf     In particular, the "gedanken experiment" section, discussed on pages 4-9, involves an EPR apparatus, and demonstrates that the observed results preclude the possibility of instruction sets. That is, the behavior of the separated components of the entangled system cannot be reproduced by (computer-like) instructions.

So, to repeat, no. This entire universe-is-a-simulation stuff is utter nonsense. Utter nonsense. Utter nonsense. Utter nonsense. (I seem to be stuck in a loop)

Liking the Mermin paper (also, it makes me wonder if it's written by a merman-moomin). Rennie's answer and link doesn't seem to settle anything though. It's a prediction based on specific dubious assumptions, tested only to a certain unidentified degree which clearly won't correspond to the Planck scale.. – CriglCragl – 2018-02-02T12:45:08.050

2I think you are misunderstanding hidden variable theory as presented by Einstein, Podolsky, and Rosen. (1) The linked paper uses "instruction," but it doesn't mean "computer instruction as part of code in the simulation hypothesis," it means hidden variables. (2) EPR paradox/"instructions"/hidden variables isn't evidence for or against the simulation hypothesis. – BurnsBA – 2018-02-02T14:34:45.640

2

I am not sure why you summarize https://physics.stackexchange.com/questions/377516/relation-of-cosmic-ray-anisotropy-to-whether-or-not-were-living-in-a-computer-s as "No, we're not living in a simulation". See "Since the conclusions are only as reliable as the assumptions that went into them, and since we have no idea what technology super intelligent aliens would be using to simulate us I think it would be wise to treat this area with care." from https://physics.stackexchange.com/questions/377516/relation-of-cosmic-ray-anisotropy-to-whether-or-not-were-living-in-a-computer-s/377517#377517

– reducing activity – 2018-02-03T11:59:51.537

"That is, the behavior of the separated components of the entangled system cannot be reproduced by (computer-like) instructions." — This is not true. What is impossible is that it can be reproduced by local hidden variables. But what is non-local in our universe need not be non-local in whatever simulates it. Just because a signal cannot reach one event from another inside our universe does not mean that no signal from the simulation of the first event can reach the simulation of the second event in a computer that simulates our universe. – celtschk – 2018-03-17T11:31:12.370

The SE.Physics citation seems mistaken; the sort of simulated-universe assessed in popularized Physics experiments isn't about any sort of simulation, but a very specific sort. A negative result is fairly meaningless. It'd be like citing a study that shows evidence against human-like aliens on Mars as proving that there aren't any sort of aliens anywhere in the universe. – Nat – 2019-08-21T01:04:01.493

5

Statistics

As soon as you establish that creating a simulated universe is possible it becomes more likely that you are within a simulated one.

Imagine someone creates a simulation of a town, an exact replica. If you're a member of this town it is now a 50/50 chance that you're in a simulation vs the 'real' world. Now imagine, to compare, the person loads up a second simulation and runs the two side by side, you only have a 1 in 3 chance of being in the 'real' world.

Run this a few times and you'll find suitably complex simulations will have their own simulations. The number of simulations will vastly outnumber the number of 'real' worlds (if multiple) so your chances of being inside a simulation will always either equal or outweigh those of being in a real world.

6This is Nick Bostrom's argument, which I listed. I find it dubious because it doesn't consider energy or information processing limitations. – CriglCragl – 2018-02-01T16:09:59.757

1@CriglCragl He doesn't seem to argue as much for us living in a simulation but for us creating them, our society evolving into a simulated one and any descendants of ours being unlikely to create a simulation of their evolutionary past (us). Sure there will always be limitations (thats why I used a simulation of a town as an example, rather than a perfect replica of our universe. Each simulation can only create a simulation smaller than itself to conserve information but it isn't an argument against it. – FreeElk – 2018-02-01T16:26:17.767

2The far more dubious presupposition, here, is mentioned in section II of Bostrom's article: substrate independence, that "mental states can supervene on...physical substrates." – brianpck – 2018-02-01T17:54:19.720

@CriglCragl Do those limits apply to those running the simulation? When you run a simulator, the passage of time inside the simulator isn't attached to that outside the simulator. If it takes a year to process a second, to those inside the simulator only a second has passed. – Lawtonfogle – 2018-02-02T16:20:53.133

2@Lawtonfogle Given the assumptions, that's certainly possible, but you would then get an explosively exponential ([time needed to simulate 1s]^[number of nested simulations]) amount of time passing in the "real" world for every second in the "bottom" world. I don't think anyone claims the universe will exist for an infinite amount of time, so that definitely compromises the probabilistic argument based on the possibility that there are an unlimited amount of nested worlds. – brianpck – 2018-02-02T17:23:53.820

1What's the justification for your claim of a 50/50 chance? The fact that there are two possibilities does not make them equally likely. For example, people have programmed computers to play chess. If you play chess against one of them across the internet, that's a pretty good simulation of a person playing chess. But it doesn't mean that there's a 50/50 chance of anyone you play against over the internet being a simulation. – David Richerby – 2018-02-02T19:43:13.893

"Imagine someone creates a simulation of a town, an exact replica. If you're a member of this town it is now a 50/50 chance that you're in a simulation" NONSENSE. Do you really think the people in your simulated town have consciousness and self-awareness? That's a huge hole in Bostrom's argument as well. Unless when you play a video game you actually believe the characters are self-aware beings. On the contrary. If I find myself in a state of self-awareness, I can be sure I'm not in a simulation. Because simulated beings don't have self-awareness. – user4894 – 2018-02-02T20:08:17.507

"Imagine someone creates a simulation of a town, an exact replica. If you're a member of this town it is now a 50/50 chance that you're in a simulation vs the 'real' world.". No, just no. If the town has say 1000 simulated inhabitants, then the odds of living in that simulated town are 1000 to about 7.5 billion. – Polygnome – 2018-02-02T21:18:59.713

@brianpck That's a good line of thinking--inefficient simulations are unlikely if they have much less "volume" than the real world. But here is where it gets interesting. The world as we see it might have very little volume. There are several possible definitions of volume, including physical volume and consciousness volume (eg number of people x time). Physical volume looks big but might be faked, and consciousness volume can be observed to be tiny. If we imagine a universe similar to ours but filled with simulators simulating worlds similar to ours, the sims' total volume will be huge. – Aleksandr Dubinsky – 2018-02-03T13:04:55.190

Hence, it is perhaps more likely we're in a universe that is running 10^22 simulations that contain 10^32 people who live for a long time (age of universe times simulation speed) than in a universe that only contains 10^10 people who live, in total, for very little time. – Aleksandr Dubinsky – 2018-02-03T13:11:04.900

2@user4894 You seem to have arbitrarily settled on an answer to The Hard Problem of consciousness. Did you prove true AI is impossible, and forget to tell tge world how? – CriglCragl – 2018-02-03T16:15:11.980

@CriglCragl Simply pointed out one of the flaws in Bostrom's argument. – user4894 – 2018-02-03T17:32:18.940

2@Polygnome I was specifically talking about the members of the town when I talked about the 50/50. So, yes, if we're talking about the whole world then its 1000 vs the 7 billion but if you're just talking about a "member of this town" then you're back down to our 1000 vs 1000. – FreeElk – 2018-02-04T16:37:12.207

@FreeElk How do you come up with that 50:50 chance? Just because its a binary choice its not 50:50. Say I take a die. Then there are two possible outcomes: It rolls a 6, or it does not. So, by your logic, the chance of rolling a 6 on a die should be 50%. Its not, it is 1/6 or 16.6% (if the die is fair). That argument with the 50:50 chance simply does not work mathematically / statistically without strong supporting evidence why this chance should be 50:50. That argument looks more like pseudo-science, trying to make it convincing by saying its "statistics" the way you provide it. Add support. – Polygnome – 2018-02-04T16:43:22.310

@FreeElk Say there are 1000 real towns with 1000 inhabitants each (there are a lot more in reality, but lets use that), and one town with 1000 inhabitants that is simulated. Then for every being in the simulated town, as well as for every other being that gets randomly picked, the chance is being in the simulation is 1:1000. That is how maths and statistics work. You do not get to decide that the chance is 50:50 by re-phrasing the problem to two choices. Again, binary choices aren't necessarily 50:50, that is the very first thing that is taught in an elementary statistics course. – Polygnome – 2018-02-04T16:48:02.623

1@Polygnome I'm not just splitting it into two choices. In the town there are 1000 inhabitants in the simulated world and there are a 1000 in the real world, regardless of the rest of the population of earth if we just look at the town then we're comparing the 1000 in the real world to the 1000 in the simulated one. Imagine we make a copy of your household, there are now two copies of you. You, as a person, have two existences a simulated and real version. Same for every member of your household, you all now have a 50/50 chance of being simulated. – FreeElk – 2018-02-05T08:22:55.023

No, my own chance of being simulated is still 1 to 7.5 billion, because your chance to pick my household to copy is abysmal small. You could easily have picked my neighbor or some other household from around the globe. – Polygnome – 2018-02-05T09:31:17.097

5

We know a lot about simulations, and how they work. Lets look for some of those artifacts in the real world.

Simulations use code, which includes various types of numbers - integers, floats etc. These all have a level of precision that they operate to, and thus a minimum positive value. If the 'real world' was in fact a simulation, we would expect to see physical parameters such as length/distance being quantised, with some minimum value.

Simulations run by advancing one 'frame' at a time, using a loop that updates all objects once, and then paints them to the screen simultaneously. If we were living in a simulation, we would expect time also to be quantised.

As well as minimum values for many parameters, there are usually maximum values in simulations. This is again related to the type of numbers used - after a certain point they will overflow. Sensible programmers put in limits on, say the size of the simulation and maximum velocity.

So, we're currently sitting on 3-4 out of 4 on the 'are we in a simulation' scale. Not looking good.

Exactly! We need to look not just at how we run simulations, but theoretical limits on them though, surely? If we could run evolutionary-iterative calculations on how to do simulations, we might do them very differently, and far more efficiently – CriglCragl – 2018-02-02T12:18:48.507

5The universe does not "update all objects once, and then paint them to the screen simultaneously" – user22917 – 2018-02-02T15:24:16.923

5You seem to be assuming that it has to be a digital simulation. Why? – Michael Kay – 2018-02-02T22:41:57.340

1You haven't shown that the universe is likely to be a simulation. You've shown that, if the universe is a simulation, then it probably isn't being simulated on an analog computer. You haven't provided any evidence to distinguish a quantized universe from a universe being simulated on a digital computer. – Mark – 2018-02-02T23:01:12.003

2@user22917 Could you clarify what you mean by that? No-one's arguing that we're literally appearing as images on some extrauniversal screen; but very many of the simulations we perform do have quantised time, and a quantisation of time is a pretty good way to give the illusion of continuity. – Patrick Stevens – 2018-02-03T15:07:25.713

@Patrick Stevens the screen analogy is not important. The point is that the universe doesn't have a 'tick' or framerate. There is no evidence to suggest that reality is a bunch of evenly spaced frozen moments. – user22917 – 2018-02-04T16:00:43.190

@user22917 Neither is there much evidence to suggest the contrary. We simply don't know whether spacetime is quantised or not. You are confident in asserting, though, that it is not. – Patrick Stevens – 2018-02-04T16:12:22.077

5

It is impossible for us to find evidence that we are living in a simulation. All computers that can perform more than a very minimal set of operations are universal computers. A universal computer can simulate any physical system and in particular can simulate any other computer, see:

https://pdfs.semanticscholar.org/e61b/fbc6a38778e1b83088b124a97f2e5009b464.pdf

As a result, there is no way for a person in a simulation to know anything about the hardware he is running on, or even whether he is in a simulation. The laws of physics are incompatible with you knowing that you are in a simulation.

Bostrom's probability arguments are pure handwaving. A probability can only be calculated with an explanation specifying how the probabilities are relevant to the laws of physics

Bostrom provides no such explanation, so the simulation argument doesn't work.

Relevant XKCD: https://xkcd.com/505/

– Mark – 2018-02-02T23:09:34.217

"It is impossible for us to find evidence that we are living in a simulation." - are you sure? In case of simulation whatever entity is doing that may revel itself. I agree that it is impossible to disprove simulation. – reducing activity – 2018-02-03T12:02:12.673

Another relevant XKCD: https://xkcd.com/395/

– DarcyThomas – 2018-02-05T01:51:46.907

4

Nick Bostrum writes in “ARE YOU LIVING IN A COMPUTER SIMULATION?

Suppose that these simulated people are conscious (as they would be if the simulations were sufficiently fine‐grained and if a certain quite widely accepted position in the philosophy of mind is correct).

John Searle’s Chinese Room Argument (CRA) (“Minds, Brains and Programs”) provided evidence that computers running programs are not adequate to generate human understanding of language. Since we can understand language, we can conclude that we are not living in a computer simulation based on the CRA.

One can pick with confidence between the three alternatives Bostrum offered in the paper’s abstract:

This paper argues that at least one of the following propositions is true: (1) the human species is very likely to go extinct before reaching a “posthuman” stage; (2) any posthuman civilization is extremely unlikely to run a significant number of simulations of their evolutionary history (or variations thereof); (3) we are almost certainly living in a computer simulation.

The posthuman stage is when we have enough knowledge and technology to simulate humans with computers. Given CRA, there is no posthuman stage since the proposed technology is impossible. So the first alternative is true.

3Lots of people (myself included) don't believe the CRA; I think it is worth pointing out in your answer that the CRA by no means closes the book on whether computers can be conscious. – Patrick Stevens – 2018-02-03T15:09:09.263

1@PatrickStevens I think it does close the book on strong AI producing conscious human understanding through programming on a Turing machine. The problem with strong AI is that it has presented a complete solution in using a Turing machine to simulate mind. It is easy to counter. The next question to ask is how far can physicalists, such as Searle, reduce consciousness to the brain (or body). Their position is too vague to counter as easily. I suspect some panpsychist approach would be necessary to describe mind such as what Chalmers proposes. However, even that might not be enough. – Frank Hubeny – 2018-02-03T21:24:55.237

But you do agree that many people disagree with you, at least? I merely propose the edit "While not everyone agrees with its validity, John Searle's Chinese Room Argument…". – Patrick Stevens – 2018-02-03T21:28:26.090

@PatrickStevens I have already mentioned in my answer that my conclusions are "based on the CRA" and "Given CRA". If you don't accept the CRA you may come to other conclusions. – Frank Hubeny – 2018-02-03T23:42:39.687

3

I'd argue there are, at least, 3 good identifiers that we're in a digital simulation:

1. That there are globally applicable physical laws i.e. the simulation is algorithmic
2. That there's a limit to the granularity of the simulation and, if you probe below that granularity, the simulation would start to get a bit fuzzy
3. There is an enormous, flaming message half way between here and Jupiter that says "Level 4"

As Meatloaf once said. Two out of three ain't bad.

Globally applicable = algorithmic..? – CriglCragl – 2018-02-01T12:31:19.190

@CriglCragl In the sense that the simulation is just following prescribed formulae that are applicable to everything in the simulation. Oh, and the rules appear to be reversible but they only ever move monotonically forward. Which is kind of weird unless you're in a simulation. – Alex – 2018-02-01T13:17:07.450

7For (1), I fail to see the point you are making. Even if you established that an algorithmic simulation would imply universal laws (which is itself highly dubious), that is eons away from the claim that universal laws imply an algorithmic simulation. For (2), you seem to imply that a non-simulated world would have unlimited "granularity" (again, dubious), and you also don't bridge the gap between observable granularity and objective (for lack of a better word) granularity. – brianpck – 2018-02-01T17:47:01.577

@brianpck While I think the answer could express its reasoning far better, I think the logic being implied is that it would be computationally simpler for a simulation to have universally consistent laws of physics, and to have a finite granularity. While I do not consider this sufficient evidence to soundly conclude a higher likelihood of us being in a simulation, I can see the intuition/appeal: we have some reason to think (from a basis of understanding modern computing) that a simulation is likelier to be this way than not this way, while we know nothing about the odds for a non-simulation. – mtraceur – 2018-02-02T03:48:41.883

3

Suppose that what we call this reality is a simulation. Then by analogy, it is plausible that within this reality we can build machines that contain simulations. Also by analogy one can infer that it is plausible that the reality that simulates our current reality can itself be simulated. This implies that for any particular one of these 'realities', it may be simulated by one, and also simulate one (or more).

Then the question is ... where does it stop? One can infer that you can keep making simulations inside of simulations, and likewise infer a simulator for every one. In either direction it can keep going. So whether it stops above or below or keeps going forever in either direction, we are just one along that chain.

It's fun to imagine these situations, as in popular culture: The Matrix movies and derivatives, dreams within dreams, counterfactual time travel, a Minecraft world that simulates a Turing Machine that simulates a Game of Life that simulates a 3D-printer that prints off a VR-headset.

Thought can reflect on itself. That is all this is.

What about energy and entropy limitations? Computations involve energy costs – CriglCragl – 2018-02-01T20:43:51.200

4@CriglCragl: In this universe, computations involve energy costs. :) Outside of this reality, we can't even say what we know as energy exists, much less any limitations our universe has imposed on it -- and within this reality, we could create a simulation where computations create "energy", if we wanted. – cHao – 2018-02-01T21:01:39.490

@CriglCragl What about them? Those 'costs' are taken into account by the simulation. If each level of simulation takes up costs at the higher level, then there might be a limit to the number of sub-sub-sub-simulations can be made. But then it's turtles all the way up. – Mitch – 2018-02-02T00:00:00.567

Could we? Not everything is a free variable. We think there are 20 something fundamental constants, forming a landscape, plus various points non-determinate quantum events interact with the macro scale making a further landscape (& maybe many worlds). If anyone can show a point the clock's mechanism altered, skipped etc, fair enough. But otherwise, it is logical to assume it was set running and then not interfered with. – CriglCragl – 2018-02-02T00:20:54.640

@CriglCragl Maybe your god walks away from his creation, but my gods are all over the place, messing with mortals, abducting them, like in their face. It's madhouse. So since this is all thought funhouse, anything goes, and you can't really assume one way or the other. – Mitch – 2018-02-02T01:00:05.083

@Mike You can't critique a theory or argument, if you just make up what you think it is – CriglCragl – 2018-02-02T12:14:10.870

@CriglCragl You keep mentioning energy costs. Can you elaborate? Is that a new concern or is it mentioned in any of the linked references? Are you hinting that energy costs limit the ability to simulate? – Mitch – 2018-02-02T14:12:37.287

@CriglCragl I only realize now that I am mostly just venting and in a less articulate manner than Yechiam's answer. We have the ability to simulate (very poorly) our current reality. This encourages us to consider that our own reality is itself simulated. My argument is that you can continue this process in both directions, the simulator is simulated itself, and within a simulation we can create sub simulations, all as far as you can. My next point then is that this makes it all meaningless; we are in one level of these simulations, and it's no more interesting than any other. That's it. – Mitch – 2018-02-02T14:18:51.783

@CriglCragl: There's no requirement that the universe simulating this one have the same properties that this one does. The "fundamental constants" only make sense in the context of those properties. Likewise, there's nothing requiring that we simulate a universe similar enough to ours to impose the same limitations on it that exist here -- only that its rules do not prohibit the development of intelligence, if the cycle is to repeat itself. – cHao – 2018-02-02T15:50:06.590

@cHao There's also no requirement or prevention that a simulation is possible in any system. All these arguments can really do is assume something like 'Suppose a simulation is possible,...'. The parameters allowed in the simulation are pretty broad. – Mitch – 2018-02-02T16:36:12.640

@Mitch Your approach is like answering Zuangzhi's or Descartes' questioning of the nature of reality, by saying you refuse to think about it. Energy, entropy, and information are fundamentally linked: https://en.m.wikipedia.org/wiki/Entropy_(information_theory) Have a look into Maxwell's Demon for a thought experiment that makes the link easier to understand. The clock in my analogy was 'wound' with the original amount of entropy. Energy is conserved, entropy is finite

– CriglCragl – 2018-02-03T16:26:43.410

@CriglCragl You haven't linked energy and entropy to simulation though. – Mitch – 2018-02-03T16:55:11.200

@Mitch You think information isn't obviously linked to simulations? Is that your critique? – CriglCragl – 2018-02-04T17:19:56.597

@CriglCragl Argh. No, I'm not critiquing at all. I'm sure somehow energy could be involved. I just have no idea how. I'm hoping you'll give a hint how energy concerns impact on the simulation arguments. I haven't heard that here. Can you give any indication how energy limits one way or the other how us being in a simulation is possible or not? – Mitch – 2018-02-04T17:23:50.260

@Mitch You can't have infinite regress of simulations within simulations, with finite energy. Decisions like whether a simulation relies on a governing set of rules or many interventions/sets of rules will involve energy/information costs – CriglCragl – 2018-02-05T13:59:53.417

2

If we say "yes" to this question, its implications would confuse most people and they might say that it is nonsense. But we cannot blame them. Those who follow logical thought cannot admit that idea...there would be a discontinuity in the logical line of thought. Also, there is a possibility of misinterpretation of the words of great men who transcended the dual-state.

You might have heard that Advaita Vedanta gives very great importance to logic/reasoning.

If you say that we are living in a simulation, it implies that the original/real 'we' are living/existing somewhere else.

Also, (if the answer is "yes"), since 'we' are living in a simulation each 'I' would also be living in a simulation. [Each 'I' must be a part of 'we'.] This means, since all 'I's are always living in a simulation, nobody can come out of it (simulation) and realise the Truth or speak it out. I mean, they can't say about that permanent state without realising it. Also, then Truth realisation/liberation would always be a mirage in all aspects. But the truth is not so.

If one's duality is not ended completely he/she can say this to his followers: "'we' are not awaken".

But when his duality is ended completely he would say this to himself: "Actually not we, I didn't know that I was 'sleeping' (means: what I believed as, I was in ignorance)". Or some great men may keep mum about it. This is because they are in a so higher level that ordinary people can't understand.

A person who is not awaken can wake up. Then he has another state. But a person who is living in a simulation has no other choice--just live in simulation and die there. If all of us were living in a simulation we would have no other choice.

So, instead of confusing by 'splitting' the term jagat into 'we', 'living' etc they (those who could transcend duality ... when they speak about the Absolute Truth) denied all the plurality and said, "Brahma satyam jagan-mithya". [Brahman is real, the universe is mithya (it cannot be categorized as real or unreal)].

Since everybody's abilities are not alike, their (those great men's) advice is for self-realisation, for serving fellow beings etc. They didn't say that we are living in a simulation. By such statements (approval) people may become inactive/disappointed/gloomy because if it is true there would be no scope for realising the Truth). If such a society is created by this confusing statement, I would say that this is the ugliest face of philosophy. But we can say, we (those who didn't realise the Truth) are not awaken in the true sense. Since there is a higher level of awaken state, there is no mistake in it.

You should compare the meanings of mithya and simulation. You can see the variance if you could replace the phrase--'in a simulation' by 'mithya'.

https://www.speakingtree.in/blog/brahman-satyam-jagat-mithya

2

Science requires proofs. Sometimes scientific theories are built on a number of Axioms or Assumptions.

We cannot prove or disprove that we live in simulation. Though there are many so called "hints" proving that we live in simulation. There are very few "hints" that we don't live in simulation. I welcome everyone to collect more hints proving or disproving the simulation theory. When we get enough hints we can judge the probabilities of one or the other side being more likely.

Some of the hints proving that we live in simulation:

1. Duality of light. It is both particle and wave. Scientific community didn't find any good explanations of that. This duality is very nicely fitting into simulation theory because waves are less computationaly expensive to simulate at the large scale, than simulating each individual particle. Read more on Double slit and Delayed choice experiments.

2. Limitation of the speed of light and connection of speed and time at relativistic speeds. This makes no sense unless looked from the point of view of simulation. "Swapping","Refresh rate" and "Lagging" are well known terms in computer games. When games try to go faster than the computer hardware you get lagging, refresh rate and swapping issues. So limiting speed to certain constant makes perfect sense.

3. Quantum Entanglement makes absolutely no sense unless looked from the prism of simulation hypothesis. Quantum entanglement is instantly connecting two quantums regardless of distance. This can be easily done in computer memory but not in the physical space.

There are many more hints listed in the book I just finished called "Answers In Simulation" by Iurii Vovchenko.

Maybe explain what "prism of simulation" is? – christo183 – 2019-08-20T15:43:15.440

oh, sorry it is just an expression. It means "from the simulation theorie's perspective" – SanityMaker – 2019-08-20T15:44:17.137

1

Maybe I am very naive, but I know I am not living in a simulation. My evidence is my free will!
I interact with other people and see that they also have free will. Therefore, I conclude that we are not living in a simulation.
I am aware of the controversy relating to free will, but from my point of view, there is no controversy.

You seem to be assuming consciousnesses with free will can't be simulated. Any reasoning for that? Being in a simulation doesn't mean you can't know anything, any more than if you are in a dream. – CriglCragl – 2018-02-05T23:04:57.860

1You've picked a pretty flimsy concept to base your rejection on. If it is possible for any being that can observe this universe (from within or without) to know its future with certainty, then "free will" is ignorance, simply the rats' view of the maze. – cHao – 2018-03-13T17:35:37.297

1

The idea of the illusory nature of reality has been around since the beginning of humanity. Perhaps the Vedas 1100-1700 BC were the first ones who deeply pondered about it. The simulation hypothesis is just a new technological possibility to create fake realities or distort them which is no different than holding people captive in a dark cave (Plato) or blinding them(Bible).

Ordinary dreams and astral projections produced by sleep, deep meditations or psychedelics remind us of the illusory nature or reality and how persistence and consistency are its defining characteristics.

"Reality is merely an illusion, albeit a very persistent one" Albert Einstein

What evidence is there? It's all about probability. If something may happen then it will happen or it has already happened.

"There’s a ‘One in Billions’ Chance Reality Is Not a Simulation" Elon Musk.

What if we awoke from this reality into another reality? Would this new reality be "real" or just another illusion? How can we know? What if it was a simulation within a simulation? Could we have endless simulations embedded into one another?

I guess the problem with the whole simulation hypothesis thing is within the human psychology and mind. Why do we want reality to be "real"? Perhaps we are afraid of digging any deeper because we are emotionally dependent on our past, ego, physical bodies, skills, culture, social status etc. and many other impermanent things and that illusory/impermanent nature of reality confuses us and makes our ego feel threatened but it's only reminding us of what we are not as said in Buddhism.

"I knew that most people never see this reality because they attach to the material aspect of the world. Illusions of self and other fill their vision. I also realized there are those with little dust limiting their vision." Buddha

I guess the problem with the simulation hypothesis is the same problem with religion and a creator god. Sometimes when there is no proof in favour or against it's hard just to be agnostic.

"It's all about probability." Surely you have to reason to that, induction can't justify induction? What about unlikely or unrepeatable events? – CriglCragl – 2019-08-27T22:35:09.770

0

Categorically this question only has one correct answer:
(a) If we are living in a perfect simulation such that there is no discernible difference between the simulation and reality then by definition there is nothing telling the difference.

(b) If there is a discernible difference then discerning this difference tells this difference. Instead of hypothesizing every possible way that this difference could be discerned let's cut to the chase. The difference that proves that we are living in a simulation is that physically manifest reality can be changed by thoughts kept silently in the mind.