I read this article about how this guy in Switzerland did an experiment that he thought proved the Simulation Hypothesis of reality (link: http://arxiv.org/abs/1210.1847).

I have also been reading the original philosophy paper by Nick Bostrom articulating the simulation trilemma.

What are some good arguments that I (or anyone else with an extant mind) could use to try to convince themselves that they are not a brain in a vat? Or, failing that, if they are just a brain in a vat, that they could in theory download their mind out of the brain into some robot body in the “real” world (or somehow have potential access to “real” reality, at some point)?

Or, failing that (these are in decreasing order of preference; good arguments for earlier scenarios are more preferable, arguments further down the list are more like last resorts), that reality isn't just a computer simulation, or if it is just a computer simulation, that it's a computer simulation being run on a substrate of spacetime such that it's really no different from a non-computer universe in principle, or at least that it's not a computer simulation being run on an electronic computer by some dodgy aliens who could shut it off at any moment, and/or is being run for some depressing, banal reason?

  • This is definitely a philosophical question so I wouldn't vote to close it unless it is duplicated. (Now the "meta" part of your question is not required and I saw that you removed it so that's fine) – Quentin Ruyant Dec 28 '15 at 16:21
  • @quen_tin now that you mention it... :) --It does seem at least related to philosophy.stackexchange.com/questions/4486/… – Joseph Weissman Dec 28 '15 at 16:35
  • 1
    personally, my main argument against it rests on the implausibility relative to "the brain-in-a-skull-of-a-human-being-in-reality", given the stimulus we experience. there are two different popular movies to refer to for ideas (not proof): The Matrix and The Truman Show. the problem with any other ontological explanation than the obvious one (that we are sentient and sapient biological beings living on a small rocky planet) is that any constructed simulation of reality that is not real is subject to making errors (even small ones) that will eventually be noticed by the subject. – robert bristow-johnson Dec 28 '15 at 22:50
  • The paper you linked doesn't claim to have proven that the universe is a simulation. It shows that if the universe is being simulated and space-time points are represented on a discrete lattice with certain properties (i.e. a model similar to the ones they use in their own small-scale simulations, only better), then this would imply various things involving symmetry-breaking that I would have presumably been able to follow if I was a physicist. (Also, it's from Seattle and Bonn, not Switzerland. Did you link the right paper?) – Ray Dec 29 '15 at 5:45
  • 1
    @Josh : It doesn't show that the universe has lattice-like properties. It shows that if the universe is being modeled using a lattice structure (for computational tractability), then this would imply various things that could be tested for. But it doesn't look like they actually tested any of that experimentally, and they appear to state in section 4.3 that the measurements we can make now are still an order of magnitude too large to observe these predictions. (And Beane may be Swiss; I was just looking at the university affiliations and was worried I was looking at the wrong paper.) – Ray Dec 30 '15 at 0:23

11 Answers 11


Your question is about metaphysical realism and skepticism. There are indeed radical sceptic arguments against realism such as Descartes's demon, brain in a vat or the idea that one is actually dreaming, but also reasons to resist these arguments.

First note that there can be no empirical evidence for or against such radical scepticism because these arguments purports to undermine empirical evidences themselves (even if we could go "outside the vat" the question would remain at the higher level), so both realism and radical scepticism are dogmatic positions. The question then moves on a priori rational grounds: is it a priori rational to entertain such a radical scepticism?

Some philosophers would argue that it is not. For example Wittgenstein argued that all doubts must hinge on some background knowledge otherwise it becomes meaningless: doubt occurs within the context of things undoubted (you can read "on certainty"). After all, claiming that we are brains in vats requires that "brain" and "vat" are meaningful words. Similarly Putnam argued that the idea that we are brains in vats is inconsistent because then "vat" wouldn't have the meaning we think it has ("vat" has meaning only if it refers to things external to ourselves) and this belief is self refuting.

One could also provide a transcendental argument to the effect that a minimal form of realism is a prerequisite for knowledge, so that realism is a priori justified.

There are also more pragmatic arguments to the effect that systematic doubts about the world would make no practical difference (whether this table is real or is a simulation makes no difference for all practical purpose). Following a pragmatic attitude, one could be a quietist: one chooses to remain silent about the question of realism or to consider it meaningless, because what matters and what is meaningful is what has practical imports.

There are also arguments for realism that say that it would be a miracle that our concepts are so successful if they were not corresponding to real entities. Brain in a vat scenarios are kind of conspirational, and would need extraordinary evidence to be warranted (but they cannot have evidence by their very nature).

EDIT: here is a text that defends this kind of arguments against scepticism: http://www.konversari.com/2015/12/27/scepticism-and-metaphysics.html

All this is related to the debate on the foundations of knowledge. You can have a look at these resources: http://plato.stanford.edu/entries/skepticism/ http://plato.stanford.edu/entries/epistemology/

  • From a personal perspective, I found the actions which one considers "good" as a real person or as a brain in the jar are very effective at working along side phrases like the OP's phrase "...is being run for some depressing, banal reason." The pragmatic approach you mention can be quite useful! – Cort Ammon Dec 28 '15 at 17:10
  • "First note that there can be no empirical evidence for or against such radical scepticism" – not true, see e.g. the Bostrom article the OP linked – Xodarap Dec 28 '15 at 18:11
  • @Xodarap this is a philosophical argument that can be discussed, not empirical evidence. – Quentin Ruyant Dec 29 '15 at 0:04
  • @quen_tin Wow, thank you! This is all really good discussion, this helps me a lot with my thinking on this issue. I do find your final item, about the arguments for realism that say that it would be a miracle that our concepts are so successful if they were not corresponding to real entities, to be the most convincing argument against radical skepticism, to me personally. – Josh Zmijewski Dec 29 '15 at 12:38
  • @Xodarap And this is true, however, as quen_tin says, it's still a philosophical argument that can be discussed, not empirical evidence (which is why I asked this question in the first place!). I actually, maybe a bit surprisingly, don't necessarily agree with Bostrom's first two propositions, and can't understand how the third one could be true unless I accepted the first two as definitely true. – Josh Zmijewski Dec 29 '15 at 12:42

In my opinion, the best response to ontological uncertainty is to strive to live in a way that is meaningful regardless of the true nature of reality. While it may seem implausible, it may be less so than it seems.

Consider the following --we don't know how our universe originated, we don't know what its fate is, we don't know with any certainty our own afterlife, if such a thing exists, and we don't know whether or not our own personal lives might end at any given moment due to circumstances beyond our controls. Those are all things that are true even if we assume material reality exists just as it appears to. Given that, the search for meaning is hardly any more difficult if material reality is an illusion. In fact, it might actually be considered easier, due to the fact that some common options --pursuing material wealth, for example --would be revealed as dead ends.

One of Plato's main themes was how to live a meaningful life in an illusionary, unreal world. I think his answer still stands --try to be as good a person as you can be, because virtue has its own reality that transcends whatever reality it is embedded within.

  • 1
    "So, Malunkyaputta, remember what is undeclared by me as undeclared ... And why are they undeclared by me? Because they are not connected with the goal ... They do not lead to disenchantment, dispassion, cessation, calming, direct knowledge, self-awakening, Unbinding. That's why they are undeclared by me." - The Pali Canon: Cula-Malunkyovada Sutta – Chris Degnen Dec 28 '15 at 20:29
  • @ChrisDegnen I would agree to an extent, at least with the first two paragraphs. I'm not entirely sure that I would necessarily agree with Plato about virtue being transcendental like that; I think you can only do meaningful good if your actions eventually have some kind of a positive impact on other conscious, sapient beings, (like not p-zombies) right? But again, I don't think that people are p-zombies, really, it's just annoying when the skepticism/doubt starts to eat at me. But I do think, you know, Occam's Razor and all that. – Josh Zmijewski Dec 29 '15 at 12:57
  • 1
    @JoshZmijewski - You're assuming that simulated beings must be p-zombies. That's not necessarily true. Even using the dream/coma scenario, we might argue that the other figures in the dream are not so much zombies as they are other aspects of yourself, in the disguise of external figures --like a person with multiple personalities. If you divide into several people in your dream, by what standard to we judge one as more real than the others? – Chris Sunami Dec 29 '15 at 19:22
  • @ChrisSunami That's very true. – Josh Zmijewski Dec 30 '15 at 19:44

There is an important distinction between reality-is-a-simulation and brain-in-a-vat, namely that in the former one's brain is simulated, whereas in the latter one's brain exists outside of the simulation and only one's experience is simulated.

Because of this "stitching together" of two realities, brain-in-a-vat offers far more opportunities to discover the simulation.

But first, we must ask some questions:

  • How different is the "real" reality from the simulated one? (Perhaps entities in the "real" reality are not human. Maybe they are advanced beings seeking to understand simpler life forms or vastly different laws of physics.)
  • Why is my brain in the vat? (Entertainment? Research? Healthcare? Space travel? Imprisonment?)
  • What percentage of other people in this simulation are also brains-in-vats? (If I am the only one, the simulation can do a lot to both hide itself from me, and to only simulate very small parts of the world at a time.)

The easiest way to determine that you ARE a brain in a vat is to be "brought back" into the real world. Short of dying in the simulation, your best bet at doing this is to find a way to force the simulation to do a lot of work, using a lot of processor power, which might annoy your captors sufficiently to get you out. (For example if you build a huge room lined with TVs displaying detailed real-time satellite images of populated parts of the Earth it might force the simulation to make sure that all those areas are simulated in high detail so that you don't notice any discrepancies. Maybe you will be watching pre-recorded footage but if you try enough such ideas you may find something the simulation can't cope with.

There may be some sort of failsafe "get me out of here" thing you can do, but that seems a bit pointless if we aren't given the memory of what it is. On the other hand one might argue that those who have "transcended" or "ascended" through meditation have found the secret to breaking out of this existence. Perhaps when we stop paying attention to our senses we can start to sense the vat we are in---or perhaps it is necessary to disconnect ourselves from our senses in order to be safely disconnected from the simulation!

If you are a brain in a vat, then your simulated brain will appear to contradict the laws of physics at times. If you get a knock on the head in the simulated world, this cannot be reproduced exactly in the real world, and so either your brain and the simulated brain respond differently, or the simulated brain mirrors the real brain and therefore seems to contradict the laws of physics. This might be very hard to detect, but it will be there. Of course, can you really trust the MRI machine or the neurosurgeon not to be manipulated by the simulation so that everything seems fine? Perhaps your best bet is to build your own brain imaging machine. Even then, the simulation could twist reality to lie to you, but the bigger the lies, the greater the chance an inconsistency will arise.

Onto your next question: if I'm a brain in a vat, can I in theory experience the real world?

Sure! If technology exists to feed your senses with an artificial world, it should be even easier to feed your senses with input from a camera, microphone, touch and motion sensors, etc.

Of course, you could be told that you are visiting the real world but are in fact in another simulation. But that's beside the point.

If the real world is utterly different from the simulated one (e.g. it's 6 dimensional, there's no light or sound, and objects can overlap) there might be some serious challenges in interfacing it with a brain that is designed for our world. But consider that we can use infra-red goggles to sense a somewhat unfamiliar world. If we are slowly introduced to a weird universe perhaps we can adapt.

If the universe is just a computer simulation, we cannot tell who is running it or why. Any discrepancies or artifacts of simulation might give us clues, but at the end of the day, we can't be completely sure of anything. The simulation designers have full control of what it's like inside the simulation, so they can hide ALL of their tracks and put in red herrings if they like.

  • Wow… this is insanely clever and original! I never actually expected practical instructions on physical things I might try in the possibly simulated world to try and test it! Though I guess that's what Beane was trying to do, although he had access to highly specialized and most likely highly expensive equipment. This is an astonishingly great answer. I feel like this should be published somewhere for future philosophically-inclined people to read! – Josh Zmijewski Dec 29 '15 at 13:16

I am going to focus solely on Bostrom's argument for the simulation hypothesis, as opposed to metaphysical skepticism in general.

First, note that Bostrom himself doesn't find this overwhelmingly likely:

Personally, I assign less than 50% probability to the simulation hypothesis – rather something like in 20%-region, perhaps, maybe. However, this estimate is a subjective personal opinion and is not part of the simulation argument. My reason is that I believe that we lack strong evidence for or against any of the three disjuncts (1)-(3), so it makes sense to assign each of them a significant probability.

Second, consider that it is a trilemma stating that one of the following must hold:

  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

Basic probability theory tells us that if these three things must add up to 100%, then we can decrease the probability of #3 by increasing the probability of #1 or #2.

Why might we be likely to go extinct? Wikipedia has some answers:

In 2008, a small but illustrious group of experts on different global catastrophic risks at the Global Catastrophic Risk Conference at the University of Oxford suggested a 19% chance of human extinction over the next century

Nanotechnology and problems with super intelligence were the two most likely reasons the experts gave for human extinction in the next century. You might also believe that that humans will go extinct before becoming "post-human" simply because post-humanity isn't possible (e.g. it simply isn't possible to simulate a human mind on a computer chip).

The second fork of the trilemma seems harder to argue for: perhaps people will view this as unethical, or perhaps simulations will be extremely expensive, but it seems harder to argue for this.

  • 1
    Yeah, I actually agree completely with your answer here. I already actually kind of disagree with his first disjunct: I actually believe it's really likely that the human species, specifically our human species, not even just any random iteration of the human species in any random real or virtual possible world, is actually going to survive into a "posthuman" stage. – Josh Zmijewski Dec 29 '15 at 13:05

The Brain-in-a-Vat, using the language of 50s pulp science fiction, vividly demonstrates the dilemma confronted by Descartes, and earlier Al-Ghazali; and in between Avicenna, with his floating man argument:

One must suppose he was created at a stroke, fully developed and fully-formed but with his vision shrouded from perceiving all external objects ... then let the subject consider whether he will affirm the existence of his own self. There is no doubt that he will do so.

Descartes innovation here, his famous cogito, is to fill in a gap in this argument (preceding the last sentence); but this leads later to a distinction, as in Kant, between the phenomenal world of appearances; and the things themselves, though affirmed, are shrouded behind a veil.

Over-coming this dichotomy has been a strong theme in early continental philosophy; it motivated the development of Phenomenology, by Husserl - whose slogan was:

To the things themselves

In a way, they no longer considered the question of a Descartian divide, a fruitful question; thus they bracketed out this question, and their ontological reduction, returned them into the world, so they were being-in-the-world; and actually looking at it, as it stands before them in it's teeming midst.

  • Do you have a ref for Al Ghazali discussing this problem? – Alexander S King Dec 28 '15 at 19:11
  • @alexander king: it's in the deliverance of error. – Mozibur Ullah Dec 28 '15 at 19:13
  • There's a translation of it here, by Richard McCarthy. – Mozibur Ullah Dec 28 '15 at 19:21

One of my favorite answers to the brain-in-a-vat hypothesis is not a direct refutation, but rather the argument that even if we are brains in vats, it doesn't mean that all our beliefs are false, or that the reality we know is "a lie." Specifically, it frames the hypothesis as a metaphysical one, on par with (and, in fact, related to) the questions of mind-body dualism, creationism, and empiricism.

See the full argument here: http://consc.net/papers/matrix.html

  • Hmm, thanks. I have to wonder, should The Matrix at this point be considered a legitimate philosophical text that eventually all philosophy students will watch as a part of their studies lol? – Josh Zmijewski Dec 29 '15 at 13:19

You can apply Occam's Razor to the hypothesis that you are a brain in a vat as opposed to really experiencing the outside world you sense: "Among competing hypotheses, the one with the fewest assumptions should be selected." (Wikipedia)


Consider, if:

  • you are just a brain in a vat
  • whoever owns/maintains the vat doesn't want you to know this
  • the vat lasts for eternity, with your brain in it
  • the vat's simulation is perfect and gives you no clues or escape whatsoever
  • nothing you do or don't do will affect any of the above

Then you are totally, utterly, hopeless but to continue seeking your best interests in the long run, without ever even giving thought to the vat...

However, there will always be the possibility that any or all of the above are not true, in which case you should actively pursue an understanding of what the heck is going on behind the scenes; it will always be in your best interest to do so. (Can you tell I'm Christian? :P)

P.S. Great movie (The Matrix): (from The Matrix)

  • 1
    Welcome to Philosophy.SE! Could you improve this post by adding references? – Keelan Dec 29 '15 at 10:12
  • @Andrew I would concur with that, as well! – Josh Zmijewski Dec 29 '15 at 13:07
  • References? What's your problem? He asked for arguments, and I gave him an argument. Also, don't edit my posts. – Andrew Dec 29 '15 at 16:58
  • 1
    @Andrew PhilosophySE is significantly different from most other philosophy sites. We aren't here to debate philosophy. Our actual goal is to communally build a quality reference library about philosophy in Q&A format. You might find this post useful: meta.philosophy.stackexchange.com/questions/474/… – Chris Sunami Dec 29 '15 at 19:32
  • What an awful purpose and a wasted website. Philosophy StackExchange is here to quote philosophers of old. Great. Well, I'm a well-known philosopher; haven't you heard of me? P.S. Do other philosophers even exist? – Andrew Dec 29 '15 at 23:34

Daniel Dennett gives a strong argument against brain in a vat experiments being possible in the real world at the beginning of "Consciousness Explained". Is is pretty simple: Simulating reality to the point that the subject can tell the difference between the simulation and real life is so computationally expensive that it is not practically feasible.

A simulation that is capable of taking into account every single possible choice I can make in a given situation would have to be to handle and store a super-astromically large number of variables.

The fact that it has to take into account all the possible choices an individual can make in its simulation is the key point. It is totally possible that a computer can simulate relatively stable scenarios with real world accuracy - to an extent that is what many video games do.

But in video games the number of possible courses of action a player can take is always limited. For a brain in a vat scenario to be convincing, the player has to be able to choose whatever course of action is possible, and that's when the combinatorial explosion happens that no computer will be able to handle.


There is at least one ludicrously simple "proof" that you are not a brain in a vat.

First, we must give up the idea that such a scenario can be "true" or "false" in any absolutely inarguable sense. In other words, we address the question as "moderns" free from dogmatic or authoritative answers. I am assuming this is okay with you, as a premise... for now.

Second, as so-called moderns, without appeal to dogmatism, we must ask: what is our best recourse to knowledge? Our current position as "moderns" seems to be that we can appeal neither to strict rationalism nor to strict empiricism as grounds of certainty. (Without going into detail, I refer here to Kant's critique of Leibniz and Hume, but there are many others.)

We appeal to an "in-between" combination of the two positions, theoretical reason and sensory evidence, called very generally the "scientific method." This is a global skepticism limited and constrained by replicable, sensory demonstration--or, put the other way around--falsifiability.

All boring. But let us now ask about brain in vat (BIV) proofs. You asked about this possibility relative to "yourself." This self "you" has a "world" and wonders skeptically if it can be globally falsified.

Let's start from that position, your "world." And let's skip over the infinite regress in which your world in in your brain and yet brains are in your world.

Given your best recourse to knowledge, in your world is there scientific evidence that there could be such a thing as a "brain" alive and thinking in a "vat"? Not only has such a thing never been demonstrated, its "physical" possibility seems exceedingly remote.

As far as science can currently confirm the "vat" capable of recreating a world would have to be as complex as a human body and the planetary ecosystem that sustains it. In other words, big, very complex and currently irreducible "vat."

There are even bigger problems with this ludicrous idea that your total reality can be reduced to the two external objects "brain" and "vat." First, no matter how you define them, they are objective subsets of your "world." The idea of "brain in vat" simulating the world it is "inside" has never been demonstrated and cannot be falsified.

Lastly, the very idea of a (BIV) simulating a world containing you-me as the simulators only occurs to us through the accepted "truth" that this is itself a "false" simulation. The very idea of simulations subsuming any scientific standard of "reality" is always presented to us, quite self-consciously, as a self-evident falsification, a movie, etc. Never, even roughly, as science.

So your choice is to opt either for our best current definition of "justified belief" or for the representational regress of "simulations" captured in the liar's paradox.


What are some good arguments that I (or anyone else with an extant mind) could use to try to convince themselves that they are not a brain in a vat?

The short version of the argument is as follows. The theory of computation as it currently stands claims that any physical system that can perform a certain set of operations can do universal computation, regardless of its physical constituents. As such, if your experiences are all a simulation it is impossible in principle for you to discover anything at all about the computer on which you are being simulated. As such, that hardware can play no role in any explanation of anything that happens to you. So this alleged simulator amounts to slapping a label on stuff saying it's not real while explaining all of its actions with no reference at all to that label. So its takes your current explanations and adds an unexplained complication, so the simulation idea should be rejected.

I have also been reading the original philosophy paper by Nick Bostrom articulating the simulation trilemma.

This paper takes for granted that the right way to count instances of yourself is to count the number of simulations. For a start, it doesn't explain what counts as a simulation. My brain may represent information about my thoughts in a highly redundant way for the purposes of error correction because brain hardware is noisy. A simulation of me on less error prone hardware might represent my thoughts less redundantly. Should I then say I am more likely to be the real me rather than a simulation? My point isn't that one way is right, and the other is wrong, but that we have no argument for deciding between them.

Depending on the relevant way of counting instances we may continue to exist indefinitely far into the future.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.