Nov 29 2012

Are We Living in a Simulated Universe?

In short, we don’t know. But let’s explore the problem a bit deeper.

What is meant by a “simulated universe” is that our universe, the one that we experience as our reality, is actually a simulation being run on a fantastically powerful computer, presumably designed and built by a race profoundly more technologically advanced than ourselves. In the movie, The Matrix, most humans were plugged into and experiencing a simulation of the entire Earth. This movie is often brought up as an example of a simulated universe, but it’s not quite what we mean. In the Matrix, humans still had a physical existence. Their brains were just plugged into a simulation. In a simulated universe, however, we would be entirely simulated ourselves, not really floating in a bubble of liquid while information is being fed into our non-simulated brains.

There are several basic approaches to this question – philosophical, theoretical, and empirical. The Rationally Speaking blog recently tackled the philosophical approach to this question. In that series of posts David Kyle Johnson writes:

The simulation hypothesis suggests that the world we inhabit is a simulated world and that we are simulated persons. And Nick Bostrom has argued this hypothesis is much more likely than you might think. His argument goes like this: If we do create a simulated world, we won’t stop there. We won’t just create one. We didn’t stop with one iPhone; we wouldn’t stop with one simulated world. Once Pandora’s box is open, you can’t close it. If there is one, there will be thousands. And beings in simulated worlds could even advance enough to create their own simulated worlds within their world. So a physical universe with a single simulated world isn’t a very likely scenario. Thus, when we consider the possible physical universes that could exist, we realize the following: either there is one real physical universe and in that universe no simulated world is ever created, or, in the one physical universe that exists, thousands upon millions upon billions of simulated worlds are created.

Therefore, some philosophers conclude, it is highly likely that we are living in a simulated universe, since most universes would be simulated. However, there is one massive “if” in that chain of logic – if we create a simulated world.

Is the current universe we are experiencing simulated? Could we empirically demonstrate its simulated nature if it were? It seems that the answer to both questions may lie in an aspect of physics known as lattice-gauge theory. At its most fundamental level the universe is made of fields of energy. The question that is relevant to the simulated universe is this – are those fields ultimately continuous or discrete? Another way to state this is this – is the universe ultimately analog or digital? If digital, then that opens the possibility of simulating the universe in a computer, and therefore that the universe itself is a simulation.

Think of the lattice as a grid, with spacing representing the smallest interval that energy fields can possess, and the increments as advancing time. Physicists have been able to describe the strong nuclear force in terms of such a lattice. If the lattice theory is correct, then nothing can exist that is smaller than the spacing between two points on the lattice. This provides an opportunity to test the lattice theory – as explained in a recent MIT technology review article:

The question that Beane and co ask is whether the lattice spacing imposes any kind of limitation on the physical processes we see in the universe. They examine, in particular, high energy processes, which probe smaller regions of space as they get more energetic

What they find is interesting. They say that the lattice spacing imposes a fundamental limit on the energy that particles can have. That’s because nothing can exist that is smaller than the lattice itself.

So if our cosmos is merely a simulation, there ought to be a cut off in the spectrum of high energy particles.

It turns out there is exactly this kind of cut off in the energy of cosmic ray particles,  a limit known as the Greisen–Zatsepin–Kuzmin or GZK cut off.

They further propose that there will be an asymmetry in the direction of observed cosmic rays, and that they can look for this asymmetry to confirm the existence of the lattice (and therefore a digital universe, and possibly a simulated one).

The same line of reasoning, however, leads to the opposite conclusion – that the universe is analog – when the weak nuclear force is considered. So far (as explained in a recent Scientific American podcast) physicists have been unable to place the weak nuclear force on the lattice. George Musser explains that our current inability to place the weak nuclear force on the lattice may simply be due to the fact that it is a very complex problem that no one has yet solved. The argument that we are living in an analog universe because physicists have not explained the weak nuclear force in terms of the lattice, therefore, is an argument from ignorance. According to Musser, physicists have not proved that the weak nuclear force is analog, they have only been unable to prove that it is digital.

But, he argues, it is possible that this problem has not been solved (despite extensive attempts to do so) because it is insolvable – because at its most fundamental level the universe is analog and not discrete.

Even if some aspects of quantum electrodynamics, chromodynamics, and the standard model of particle physics are ultimately discrete (can be placed on the lattice), if any aspect is not then the universe cannot be simulated.


At present the current answer to the question of whether or not our universe is simulated has no definitive answer. The physics questions above address the question of whether or not it is theoretically possible for our universe to be simulated. If it turns out that the answer is that it’s impossible, then of course our universe is not simulated. If it turns out that the lattice-gauge theory is correct and applies to all of physics, then it will be theoretically possible to simulate our universe, but we still won’t know if our universe is in fact simulated.

That leaves us with the empirical approach – is there any observation that we can make that would demonstrate our own universe is simulated? I think the answer to this is no. In a simulated universe we would be part of that simulation. How could we make an observation from within such a simulation that would demonstrate it’s a simulation? Further, at that level, what’s the difference between a “real” universe and a simulated one (from the perspective of someone inside the universe)? Even if simulated, what we experience is just the universe.

34 responses so far

34 thoughts on “Are We Living in a Simulated Universe?”

  1. Plittle says:

    Thus, when we consider the possible physical universes that could exist, we realize the following: either there is one real physical universe and in that universe no simulated world is ever created, or, in the one physical universe that exists, thousands upon millions upon billions of simulated worlds are created.

    It seems to me this is a false dichotomy.

  2. ccbowers says:

    I think it is a false dichotomy because it assumes that if a universe can be created, it can be created fairly easily such that “millions upon billions” would be created. It may be that creating such a universe is possible, but very hard so that only 1 or a few are possible. The Universe may not be an Iphone, but more like the Large Hadron Collider (as a limited analogy)

  3. xkyzero says:

    Wouldn’t this create a situation of decreasing information? A computer created in a given universe would not be able to store the amount of information needed to completely reproduce the information describing the simulating universe – so the simulated universe would contain less information.

    The universe simulating us is more complex than ours. And the universe we eventually simulate will be less complex than ours.

  4. I agree – I think there are many unknowns. Another one is this – if we can create a simulated universe, that does not mean that you can create a nestled simulated universe within a simulated universe. Power consumption alone might limit the number of simulated universes.

  5. vmes says:

    I have a real problem with degrading granularity when it comes to nested universes or even just a single simulated universe. Seems to me that the information granularity would go down by an order of magnitude each time when you consider that the simulating universe must provide both storage and energy for the simulated universe, when talking about a perfect simulation. However, I believe the original paper on this topic discussed the possibility of granularity being a clever illusion only simulated on an ad hoc basis. Similarly, distant objects would also be simulated with decreased granularity and either brought into working memory as needed or the simulators constrain our ability to visit such places. This significantly decreases storage and energy needs. However, when considering that the simulators are giving us the illusion of granularity and scale, whether we are actually in a simulation or not becomes quickly unknowable.

  6. tmac57 says:

    Well,if we are going down the rabbit hole of speculation,why would we assume that whom or whatever has created this universe is bound by the same constraints in terms of physics,that this universe has? What if in their universe, computing power and resources are limitless,and their laws of physics are nothing like what we understand?There’s no reason to assume that our simulated universe would be directly analogous to their’s,except that we are trying to extrapolate from our own experience.That seems like a huge assumption,and failure of imagination.

  7. Marshall says:

    I don’t understand the “analogy” vs “digital” argument. So what if the universe ends up being one or the other? There’s an underlying assumption that an “analogy” universe could not be simulated, since we (humans) currently can’t conceive of a way in which to do that. But we’re talking about technology so advanced that they can simulate a UNIVERSE. It’s pure hubris to believe that they can’t simulate continuously. Hell, for all we know, infinity is no problem them them.

  8. I have a couple of different takes on this idea.

    First, the concept seems to be that it could be possible to simulate a universe at the size and level of detail that exists in our universe. I suspect this is a fundamental problem, at least if we are the real universe. Moore’s Law will not continue to apply forever. At some point in time, however soon or distant that may be, computing power will run up against limitations imposed by the laws of physics, speed of light for signal transmission, physical limitations on the smallest possible storage & switching nodes, power reqs, undesired quantum effects, etc.

    On the other hand, as TMAC sais, there’s no reason to assume the laws of physics in the simulation would exactly mirror those of the real universe. Indeed, the laws of physics in our universe may well be there to address the limitations of computing power in the real universe. Consider a video game where distant objects are not visible until you get close enough and they pop into view. Maybe the fundamental limit of the speed of light is there to prevent us from getting to a destination before the computer has drawn that section of the universe.

  9. Jim Shaver says:

    If we ever do discover proof that our universe is a simulation running within some enormous computer, I suggest we then attempt to locate the source code for the TSuperstition class, and delete it.

  10. Jacob V says:

    It seems to me that the assumption of a potential simulated universe, or that our universe is simulated, must also contain the assumption of immensely wasteful and impractical beings that clearly have too much time on their hands. V’ger perhaps.

  11. SARA says:

    “Even if some aspects of quantum electrodynamics, chromodynamics, and the standard model of particle physics are ultimately discrete (can be placed on the lattice), if any aspect is not then the universe cannot be simulated.”

    Isn’t this based on the assumption that the larger creator universe is working like ours? Or that they are unable to find a way to overcome this issue?

    In any case, this entire problem seems a great deal like trying to prove there is no God. Although a great deal more entertaining.

  12. Sara – yes, I think that is an assumption, hence “simulation.” The idea is whether or not we can simulate our actual universe or if our universe is a simulation of another actual universe. Otherwise I think the term “virtual” universe would be better.

    Whether or not our universe may be a virtual universe embedded in a hierarchically higher universe that had a different set of physical laws seems an unanswerable question to me.

  13. KDoug says:

    It’s an interesting question but I think we could also consider it from another angle: Cost vs benefit.

    If a civilization had the ability to simulate an entire universe, complete with sentient life, would they really need or even want to do it? Would anyone in the real world be able to examine the simulation of a complete universe in any satisfying detail? Couldn’t they better answer their questions through more simple and more focused means?

    I think it would make more sense to create smaller, more specialized simulations for whatever questions that they wanted to answer. Why bother with simulating an entire universe when it wouldn’t be necessary?

  14. SARA says:

    Steven – I hadn’t really considered the distinction. It brings up all sorts of shades of grey, doesn’t it? I mean, is it simulated if it is largely experienced like the creator universe, or if it is actually a duplicate at such a minute level.

    Where is the line to virtual? I imagine that is one of those semantics arguments that flare up all over the future internet when we create our simulated universe.

    Still, if I was plunking down money, I’d want to create a universe that was not a simulation of an existing one, but a brand new and different thing. Purple People Eaters and Tardises and Flying people would be far more fun.

  15. “Still, if I was plunking down money, I’d want to create a universe that was not a simulation of an existing one, but a brand new and different thing.”

    And yet we don’t tend to create virtual environments that are radically different form our own. We tend to create virtual environments that are mostly analogous to our own reality with relatively few fundamental differences. Sure, we may make virtual worlds (video games, etc) with faster than light travel and portal guns, but 3 lefts usually still make a right, the sum of than angles of a triangle is still 180 degrees and 2+2 still equals 4. The laws of physics, mathematics, and logic in any such “over world” might constrain what can be effectively simulated. On the other hand, maybe 2+2=5 in the real universe, and this is a cunningly effective simulation.

    Maybe our universe is a simulation run with only 1 or two parameters changed in order to see what differences result. Maybe they added dark energy to see how that affects the evolution of the universe over the long run. Scientists in this universe have run (obviously much smaller and dimplier) similar simulations to see what minor differences in the strength of the strong nuclear force would have on the evolution of the universe from the time of the big bang.

  16. tmac57 says:

    Here’s proof* that our hypothetical ‘creators/programers’ (whatever) are not creating us in their own image:
    Are we even remotely,as a species,able to do such a thing ourselves?
    Don’t get me wrong,I think that what we have achieved so far collectively,is quite remarkable,but then I read the comments on Yahoo News items…(sigh!)

    *This statement should not be taken as factual.It is not intended to treat or cure anything,and intended as entertainment only.

  17. Bronze Dog says:

    I think our universe is simulated, and it involved lots and lots of redstone and pistons.

  18. Jared Olsen says:

    Awesome thought experiment! My take is this: If quantum processes are fundamentally unknowable, even in principle (assuming no hidden variables), then how could a computer (no matter how advanced), which must rely on logic, produce those unknowable processes? They would be unknown even to the simulator. Is that possible?

  19. BillyJoe7 says:

    I’d rather believe in god.
    Oh wait…

  20. SimonW says:

    I suspect it is unknowable.

    Although one can have some fun making silly deductions.

    Even if our Universe has properties we couldn’t simulate, it doesn’t mean it isn’t simulated, merely the host Universe (if there is one) must have properties such that our Universe can be simulated. Didn’t these guys hear of analogue computing?

    If the whole of biological history on earth is simulated, we can conclude via the Darwinian Problem of Evil that there is no ethics committee overseeing this research. Of course they could just be simulating the last 5000 years and making it look older, but still probably no ethics committee.

    Even if we live in a Universe with lots of simulated Universes what is the probability we are in one of those simulations? I think that probably permits of an answer like the Drake Equation where the answer may be quite low depending on the assumptions you make. People assume because there are say 1 million simulated Universes and 1 real one, that the odds are a million to 1, but that only follows if the odds are the same for all results. What if most of those simulated Universes never bring forth intelligent life (for example). We know the odds of intelligent life in the host Universe are 1 (someone to make the simulator), the odds could be lower in the others. Similarly how long do these simulations run for?

    Fortunately William of Ockham (or possibly someone else) has an approach that can be used to resolve this train for thought until evidence is forthcoming.

  21. Christopher Newell says:

    I’m intrigued that there’s scientific research being done on this at all. When I first read Bostrom’s paper I thought to myself that it was fun speculation, an interesting hypothetical argument, but ultimately I was doubtful that it could be tested. Bostrom writes in his FAQ’s “We could also obtain strong indirect evidence, such as one day observing that we ourselves have created the appropriate kind of computer simulations.” In other words, if we simulate a universe we can reasonably suspect that our universe is simulated. It may not be possible to conclusively prove any one of the prongs in Bostrom’s tri-partite disjunction, but varying degrees of probability could be assigned to them in light of new evidence. While not nearly as compelling as making our own simulation, I think that the above cited research would be an incremental step toward increasing the odds that one possibility is more likely than another. I’m especially interested to know if there’s an experiment planned to look for an asymmetry in the direction of observed cosmic rays. If an asymmetry is found, that would mean that two out of three postulated characteristics of a digital universe would be manifest in our own, with the weak nuclear force being the final stumbling block to lattice-gauge theory.

  22. shaderology says:

    Jaan Tallinn made an anrgument for simulated universe at Singularity Summit this year.

  23. locutusbrg says:

    I am no physicist obviously so I will ask the dumb question. Why couldn’t the universe simulation be made specifically so that the intelligence in the universe cannot find out. If flaws were purposefully hidden from us how would we know? If we figured it out couldn’t they just wipe the knowledge from our memory, or have a program in place to do it. I mean nice thought experiment but what does it really tell us.

  24. Jared Olsen says:

    I’m an Athiest, but I reckon if there was a God that’s exactly what It would do-at least, hide It’s presence from us, just to see how smart we really are and also not interfere with our ‘free will’.

  25. steven johnson says:

    Analog simulations are possible after all. In a trivial way Edison’s gramophone was an analog simulation! The speculation assumes that the human mind, which is not fundamentally digital, can nevertheless be digitally simulated in this virtual universe. All this about a digital universe is a red herring at best. Or a pseudoscientific gloss at worst.

    The real point is whether the simulation of a universe is computable. Regardless of what I think on this issue, there is no consideration or thought given to this. I’m open to being corrected (not much of a personal virtue, because in science what I feel in the end doesn’t matter,) but aren’t there some computational processes which no algorithm can judge to be terminated? Wouldn’t any genuinely scientific speculation as to the possibility of a virtual universe have an argument that the simulating program can specify when all calculations must terminate? In other words, isn’t it true that this hypothetical simulating program by everything we know about computability impossible unless we can make a case that there is no such calculation needed to simulate our universe? I don’t think these questions give answers that allow us to deem the simulation hypothesis anything but wild conjecture untrammelled by any rational thought.

    There is also another aspect which seems to go unremarked. What is the point of a simulation which does not provide information to its creators? In order for any simulation to be of interest it seems to me, there must be some sort of output to its creators. There certainly is no physical mechanism for information to be obtained from the virtual universe and used to project an image onto a viewscreen in front of God’s Throne. There is also the question of economy, which is why anyone would need to simulate a whole universe in such detail? A cosmologist would not always need to simulate life forms, and a biologist would not always need to simulate dark energy. The hypothesis covertly assumes omniscient observer(s) who would be capable of assimilating unbelievable amounts of data.

  26. BillyJoe7 says:

    “in front of God’s Throne”

    God‘s throne?

  27. RBH says:

    But, he argues, it is possible that this problem has not been solved (despite extensive attempts to do so) because it is insolvable – because at its most fundamental level the universe is analog and not discrete.

    I fondly remember working at Honeywell’s Systems & Research Center in the 1960s, where we built aircraft cockpit simulators (UH-1, AH56-A, F-4. etc) driven by hybrid computer systems using a couple of PACE analog computers and a Sigma Data Systems digital machine.

  28. Christopher Newell says:

    I’m not an expert on computers, but a couple of people have brought up analog computing, apparently to counter the implication that non-digital means non-simulated. Correct me if I’m wrong but aren’t analog computers inferior to digital e.g. slide rules? If so, would a post human society (or an advanced human society) revert to analog computing to produce the kind of simulation we’re talking about? After all, the basis of the argument is that our current reality may be derived from and based on an archetypal reality. So the argument goes, we can anticipate hyper-omi-mega sophisticated simulations in our own future which would make possible the simulation that we’re currently (maybe) living in. Unless I’m mistaken, analog computing is our past and not our future, so we wouldn’t forecast a future where analog computing is prevalent.

  29. tmac57 says:

    This is like a group of amoeba’s trying to understand the purpose and function of the Large Hadron Collider, extrapolating from their experience in a petri dish.

  30. raylider says:

    I think information is the limiting factor here. In order to store all of the information contained within one atom – subatomic particles, spins, locations, momentums, you would need storage at least the size of one atom, but probably more matter than that. And so in order to simulate a whole planet, you would need a computer at least the size of the planet, and in order to simulate the whole universe, you would need a computer the size of a whole universe. Of course you could simulate just a portion of the universe, but then that universe can only simulate a portion of itself and so on, and this would arrive at some finite limit of number of simulated universes. If you were to limit the resolution of information contained in each dimension, you could theoretically simulate the whole universe, but then you would limit the resolution, and thus arrive at the same conclusion.

  31. Murmur says:

    I found this article fascinating, but it made me chuckle a bit because ultimately if we are in a simulation, then we have a Creator. And we all know the many names The Creator is given by our religious friends.

    Is Science confirming, in a roundabout way, the existence of a God?

  32. BillyJoe7 says:


    If you think belief in a simulated universe is reasonable, then I guess you would also think that belief in gods is reasonable.
    Besides, it’s the same old reasonable retort: from whence gods.
    In other words, you haven’t solved anything.

    If this is a simulated universe, presumably it is being simulated by someone in a real universe, so let’s just stick with this universe being real, otherwise we are into that infinite regress of simulated universes with no solution in sight.

  33. dave martyn says:

    BillyJoe7, are you suggesting that we pick a solution (unsimulated), because otherwise we might have no solution? That seems just as limiting as God.

  34. kcwong says:

    The simulated world is a cheap invention machine and not a game. In a game, the Creator inputs HIS knowledge and experience. There is no free will. However, for an invention machine, free evolution with free will is necessary for creation and invention.

    An intelligent being created in HIS world may endanger HIM. It is much safer and cheaper to create intelligent simulated beings to think, invent and work out new inventions to be extracted as required. Judging from the simulated universe that we now know, I think that we are still far away from the problems that HE would like us to involve.

    I also try to explain “Relativity” in the simulated world. I am not studying physics. Assuming there is a speed for our brain to interpretate the data from the Creator to form the successive holograms. Assuming more data in a denser environment (more material) and/or higher speed (larger area), a brain (A) in a lighter and/or slower place produces more holograms than a brain(B) in a denser and/or faster place in a given time measured from A or B. If each sucessive holograms means the passage of a time unit, the time experienced by brain (A) will be faster. Is this the reason of ‘Relativity’?

    There is also limits to the amount of data that our brain can intrepretate. The limits are the speed of light and the black hole. Given that the brain cannot process the amount of data of a black hole, the concept of falling into a black hole or passing a warmhole may be impossible in our holographic world.

    Given that we are simulated beings, what should be the primary purpose for our physics (or science)? My friends, I suggest ‘to find ways to exit the hologram so that we can meet the Creator’.

    Is it possible? Despite our effort, there is still no sign of any other civilisation. Perhaps, the advanced civilsations had already found ways to exit and found the Creator or be invited to serve the Creator.

Leave a Reply