14 ms·

Constraints on the Universe as a Numerical Simulation (2012)

77 points by aq3cn 10y ago

aq3cn 10y ago
For those, who want to read it in lay man's language, they can go through following articles: https://www.technologyreview.com/s/429561/the-measurement-that-would-reveal-the-universe-as-a-computer-simulation/ https://www.technologyreview.com/s/429561/the-measurement-th... http://www.huffingtonpost.in/entry/physicists-may-have-evide_n_1957777 http://www.huffingtonpost.in/entry/physicists-may-have-evide...
- sakopov 10y ago
  Interesting. So we're a simulation that can simulate 15 (femto)meters of cosmos that is indistinguishable from reality. This isn't much but it's still absolutely mind-bending. Reminds me of the movie The 13th Floor [1] [1] http://www.imdb.com/title/tt0139809/ http://www.imdb.com/title/tt0139809/
  - codesnik 10y ago
    femtometers?
    - sakopov 10y ago
      Yep. Typoed.
- yodon 10y ago
  The tech review summary seems to be saying this super intelligent universe-simulating computer builder doesn't know about Bresenham's line drawing algorithm [0] and as a result it can only walk high energy particles along a limited number of paths. [0] https://en.m.wikipedia.org/wiki/Bresenham%27s_line_algorithm https://en.m.wikipedia.org/wiki/Bresenham%27s_line_algorithm
  - komali2 10y ago
    Possibly? I can see why one would want to assume that an entity that can simulate a universe would be able to know more than the collective ability of that simulation, but on the other hand, we simulate things and learn unexpected results from our simulations. I think it's entirely feasible that a simulation can generate more knowledge, or different knowledge, than exists outside the simulation.
    - yodon 10y ago
      Certainly, but the idea that some hyper intelligent entity would develop the tech to build a universe-scale finite element simulator without groking anti-aliasing along the way is simply absurd. Bresenham's algorithm is a big deal because it's essentially the first thing everyone needs to invent when they start trying to render continuous processes on a discrete grid.
      - to3m 10y ago
        Bresenham's algorithm draws aliased lines.
      - raattgift 10y ago
        Where would you introduce the trajectory-correcting algorithm? In General Relativity, fundamental particles are simply taking shortest paths through spacetime, and where there are multiple short paths, the one that extremizes the timelike distance is favoured. In a perturbative theory of gravitation, you introduce corrections on the particles by way of graviton interactions. Worse, whether a local patch of energy-density amounts to a particle is observer-dependent; in both approaches above an accelerated observer sees more particles than an unaccelerated one does. That includes more gravitons in the latter approach. It would be a lot of work to write down a theory in which particles (whose count isn't global) decide on corrections to their own trajectories that return themselves to a path similar to one along a smooth manifold even though they are "really" travelling along a non-smooth one, so long as you are trying to match well-tested results of General Relativity. (A theory which doesn't reproduce those results is maybe an interesting toy; "cheating" by using a nearly-but-not-quite minimal length that has no direct observables is an uninteresting toy, and is not really a good argument for the existence of Simulators (or Creator Gods or whatever).) In particular, the non-linear contributions of gravity will likely lead you into an explosion in the amount of knowledge that a particle would have to acquire in order to assess its trajectory; you'd probably also want to propose an explanation of how it goes about adjusting its trajectory when it discovers it's not on the "right" one. Don't forget to make this correct for all possible observers, including the ones that will see no particle and those that will see two or more particles. In practice, attempts to do away with infinitesimals of length that do not clearly fail to reproduce known physics also tend to produce infinities of state in matter and/or additional gauge fields, and the infinities don't succumb to information-reduction techniques like renormalization by power-set counting. If you're looking to encode that knowledge somewhere, like in a spacetime-permeating field, the gravitational field of General Relativity is right there and already serving that purpose in a non-toy theory that (in explorable parts of the known universe) behaves completely correctly with quantum field theories of matter, including The Standard Model. :-) Best of all, in a block universe model, the whole thing is "pre-rendered", and there is no difference between the block universe model and an initial-values-surface formalism in the presence of determinism. So you only have to fully render one frame, and you get all the others, past and future. (Quantum mechanics as we know it is deterministic -- that's unitarity.)
CuriouslyC 10y ago
It kind of irks me that they basically assume a cubic lattice. Nature does not form cubic lattices, when nature forms a lattice (for example, in a foam) it tends to form roughly dodecahedral cells (with coordination number 12). Given this property, a space-time lattice would be very nearly isotropic.
- SCAQTony 10y ago
  Would a dodecahedral lattice explain why the universe stays symmetrical while expanding?
  - pklausler 10y ago
    Does it?
    - SCAQTony 10y ago
      Yep, it's expanding symmetrically in all directions in that the distances may increase between bodies but the masses such as suns and planets do not grow. http://nautil.us/blog/the-universe-expanding-symmetrically-and-eternally http://nautil.us/blog/the-universe-expanding-symmetrically-a...
      - raattgift 10y ago
        Introducing more edges with your dodecahedral (or an icosahedral or some large-n polytope) cell worsens the edge-vs-face problem below, because a "cell shape" discretizes rotation and we have excellent evidence that rotation is smooth in our spacetime. Small regions of our spacetime are locally Lorentz-invariant, meaning that a number of observable quantities (magnitude of angular momentum and mass are two) do not change for a particle held at the origin of a system of coordinates as the particle is arbitrarily rotated or boosted. Arbitrariness is important. If we are observing a distant and predictable multifrequency radiator and define spherical coordinates with the radiator at the origin, then any movement our observation gear makes that isn't exclusively radial is equivalent to a rotation of the distant radiator. You can do this by holding the radiator at the (spacelike) origin, and the observer at its fixed (spatial) coordinates; in order to keep these coordinates constant, you have to rotate the system of coordinates to counter wholly non-radial relative movement. At large distances, the rotation at the origin becomes extremely small. Small or large, Lorentz invariance means the radiator has the same mass (it's by definition rest mass since it's always at the coordinates [0,0,0,t]) and momentum. So arbitrarily small rotation goes hand-in-hand with arbitrarily distant observers, and also with nearer observers who can displace themselves tiny amounts. So is there a minimum rotation? The arbitrariness of rotation is a in conflict with "cell shape", as when your "cell" distinguishes between edges and faces, the discrete nature of rotated mass-energy-momentum exposed through the (corner-filled, discrete) cell structure leads to different observables under rotation compared to that of the (smooth, continuous) spacetime of (either theory of) relativity. This deviation is larger for objects of higher momentum; and remembering Einstein's relation for rest-massless particles, E = pc = \hbar\omega = h / \lambda, that means that for different-wavelength photons emitted from the same source, the higher-frequency photons will arrive later than the lower-frequency ones. We have good observational evidence against frequency-dependent arrival times from bright distant objects (supernovas, gamma-ray bursts, even millisecond pulsars). We can conceive of a "cell shape" which is uniform under infinitesimal rotations, but at that point you have shifted one set of infinitesimals to another, and in the context of this topic (a simulation that among other things saves on state by abolishing real numbers in dimensions of length, rotation, boost and/or translation) is pretty much a non-winner. (Additionaly, ignoring the simulation context, you would also almost always run into difficulties if your "cell size" -- a minimum length in space or a minimal interval in spacetime -- is large enough to produce observables.)
kornork 10y ago
If we're living in a simulation, the question of whether AIs can have consciousness is solved.
- skc 10y ago
  But what if consciousness itself is simulated. The problem here is that a simulation would have to adhere to a set of programmatic rules meaning that the only way consciousness could arise as we like to think we know it, is via some weird unintended side effect.
- mslev 10y ago
  Is there a difference between a conscious being and a being who has been programmed to believe it has a conscience?
  - kornork 10y ago
    I'm no expert on philosophy, but I feel like consciousness is a subjective experience. So externally there would be no difference, but I'm pretty sure I'm conscious. Of course, I've also heard of the illusion of consciousness, but that's where my brain explodes.
- maxerickson 10y ago
  What if there is a rule in the simulation preventing intelligences created by simulants from becoming sentient? I suppose you can point at all the artificial simulants and their sentience, but as pointed out in a sibling comment, so what.
- pklausler 10y ago
  The question would be mooted.
- astrodust 10y ago
  What if the simulation is actually being "run" inside an entity or organism that was the natural product of its environment?
smrtinsert 10y ago
What sort of moral issues result from creating a simulation universe that gives rise to sentient beings? I feel creeped out and reminded of Source Code with Jake Gyllenhaal for some reason.
- hosh 10y ago
  That's assuming sentience is an emergent property of the simulation rather than something else, such as being the ground from which the simulation arises from.
  - Ygg2 10y ago
    If it's any consolation. It's probably calculating wind patterns or something.
- knome 10y ago
  https://qntm.org/responsibility https://qntm.org/responsibility
- maxerickson 10y ago
  As ever, it depends on your morality. I mean, plenty of people don't take issue with a lot of things I find reprehensible.
- kornork 10y ago
  As creator, are you responsible for the well being of the sentient beings? Are you morally prohibited from turning them off? Can you reset to a save point, possibly undoing the creation of some of the sentient beings?
- teraflop 10y ago
  Greg Egan has written a (slightly ham-fisted, but still interesting) short story about that question: http://ttapress.com/553/crystal-nights-by-greg-egan/ http://ttapress.com/553/crystal-nights-by-greg-egan/ The author has commented that he regrets having overlooked the moral implications of simulated life in his previous works (see e.g. http://gregegan.customer.netspace.net.au/PERMUTATION/FAQ/FAQ.html http://gregegan.customer.netspace.net.au/PERMUTATION/FAQ/FAQ..., which contains spoilers for Permutation City) so I wouldn't be surprised if that was the inspiration for this story.
retox 10y ago
(2012)
astrodust 10y ago
Keep in mind that if you wanted to create a simulated universe you can ensure nobody ever finds out. If someone does find out, you can simply fix the problem and revert to a point in time just before the data leading to the discovery was captured. Speaking academically, you might be playing with fire if you probe the outer limits of the universe and we are in a simulation run by some entity concerned with keeping that a secret.
- 4ad 10y ago
  > you can simply fix the problem and revert to a point in time just before the data leading to the discovery was captured. Only with classical computers, not quantum computers.
  - aswanson 10y ago
    Who's to say "real" computing, or "reality" at all, has to adhere to the laws we observe within the simulation? That's like Luigi in Super Mario Brothers saying you can't kick a turtle into a brick in the our world and not expect a gold coin to pop up in front of our faces.
    - ajuc 10y ago
      The difference is mathematical, not physical, so it should still hold.
      - Nadya 10y ago
        The mathematics depend heavily on physics. If physics outside of our universe differ - the math wouldn't add up. I can make 2 + 2 = 5 in a simulation. That doesn't mean 2 + 2 = 5 outside of the simulation.
        ajuc 10y ago
        > I can make 2 + 2 = 5 in a simulation I'm not sure you can. Yes you can make a chest with 2 apples contain 5 apples if you put 2 more, but it's impossible (IMHO) to make it consistent. How many apples do you have if you place 2 chests besides 2 other chests and put 1 apple in each? What if you repeat the process 2 times and then another 2 times? Also - when some things in physics aren't intuitive (QM for example, but even adding velocities in relativity) we don't think math is wrong - we think math doesn't apply in that way to this particular phenomena.
mentos 10y ago
What is interesting is if we are the result of an unfathomable number of transistor gates, the consciousness that arises from it would never be able to understand the 'platform' (hardware) that it sits on.. would it? The idea being that you are creating a new dimension that cannot see the dimension that gave rise to it. I feel like no matter what we are in a 'simulation' where another dimension gave rise to ours whether it was intentional or not. What dimension is thought in?
- tonmoy 10y ago
  Nobody said the simulation is being done with transistors. Even if it was why wouldn't it be possible? I don't see any constraints. One day humans will fully understand neurons. One day we will make an AI using transistors who will be able to "understand" transistors
  - mentos 10y ago
    I meant it more in an abstract sense. They don't have to be silicon they could be enormous metal planets and moons..
- Razengan 10y ago
  Indeed, just like how a computer program will never detect the exact positions of all the electrons in the computer it's running on. Not even if it tried to get that information from an external observer.
- CoryG89 10y ago
  Is this really true? What about a self-aware AI? I can imagine an AI that resides on a chip and developed consciousness could conceivably control a camera/robot in the real world and literally inspect and even reverse engineer or repair/improve itself or the hardware it runs on.
  - komali2 10y ago
    True, but in the end their ability to perceive will be limited by the wires to/from those cameras to their hardware, and thus it is still possible that even the external world they thing they're observing, including themselves and their place in it, is also simulated. As in, someone could plug a "fake" camera into the AI and fool it.
rihegher 10y ago
13th floor style? http://www.imdb.com/title/tt0139809/ http://www.imdb.com/title/tt0139809/
lossolo 10y ago
I would be surprised if we are not living in simulation. Look at superposition and function collapse in quantum physics, it's equivalent to what we are doing in games today = do not render if player is not watching, save resources. Think about it, look for example at GTA V, what we can achieve today and think about how world looked like 100 years ago. I imagine that in 1000 years from now, simulating world similar to ours will not be a problem at all. Maybe not as complex because every level you go deeper you have less resources at your disposal. Who knows which level deep we are? Maybe we are simulation in simulation in simulation...
- comicjk 10y ago
  This is a common misconception, but superposition has nothing to do with "do not render if the player is not watching." As far as we know, every particle in the universe is interacting with every other (with a light speed delay) and thus "measuring" it (modifying and responding to its wavefunction) at all times. In the two-slit experiment, for example, the wavefunction of a particle moving through a slit interacts with the wavefunction of the particle used to detect it, causing both wavefunctions to be more narrowly distributed in the space around the slit. But before or after this interaction, the whole wavefunction still needs to be simulated to produce the effects we see experimentally. The wavefunction itself is the real thing being simulated, and it's always "running." You may be right that it will become easier to simulate worlds like ours, but it will require innovations that make a quantum computer look like an etch-a-sketch. The universe is really big and really detailed.
  - Vraxx 10y ago
    I see the distinction you're making (I think), but I still don't see how that invalidates GP's analogy. Surely "running" a wavefunction is computationally cheaper than re-evaluating the particle's characteristics at every universal frame. Perhaps it's not so much about our own "watching" but rather "do not calculate when irrelevant". My own analogy might be an online map frame like google maps, where tiles outside the viewing box are not loaded. This is because we can tell what needs to be calculated by the bounding box of the viewer without needing all the other finer details. I see it more as a lazy eval.
45h34jh53k4j 10y ago
Sigh. Why would you dedicate your life to proving this? You will either find you cannot prove it, in which case you have wasted your entire life on falsehood, or you prove it true in which you undo your dilute your existence and everyone you have ever known and loved. Kinda like religion I guess...
- Justsignedup 10y ago
  though, if we are a simulation... maybe we can exploit bugs in it and give ourselves advantages! After all, the only bug-free program is one that was never started.
  - slackware_linux 10y ago
    What are you trying to tell me? That I can dodge bullets?
    - Justsignedup 10y ago
      I'm trying to say that when the time comes, you won't have to!
  - allemagne 10y ago
    let's just hope we don't find the bug that crashes our server
- zzleeper 10y ago
  You don't want to show that it's true. You want to find out whether it's true or not.
  - Pinckney 10y ago
    It's not entirely clear what, exactly, it would mean for this to be true. It's also not clear that defining what we mean by "Is the universe a simulation?" has any value if it is not testable.
- 10y ago
codezero 10y ago
The title of the arxiv article is: Constraints on the Universe as a Numerical Simulation The only time they use the word evidence is in this quote: __ Therefore, there is a sense in which lattice QCD may be viewed as the nascent science of universe simulation, and, as will be argued in the next paragraph, very basic extrapolation of current lattice QCD resource trends into the future suggest that experimental searches for evidence that our universe is, in fact, a simulation are both interesting and logical.
- sctb 10y ago
  We've updated the submission title from “Physicists May Have Evidence Universe Is a Computer Simulation”, which breaks the guidelines by being editorialized.
  - codezero 10y ago
    Awesome, thanks!
AnimalMuppet 10y ago
Per the arXiv abstract, physicists may be able to find evidence. That's very different from may have found.
3chelon 10y ago
Let's all just think, very hard, of the number 42 and see if the simulation ends.
AnimalMuppet 10y ago
So we stopped believing that God created the universe, and us as part of it. We believed only in science and natural processes. Following that through, we started believing that we live in a computer simulation with, presumably, some kind of programmer. That's pretty ironic.
- code_sardaukar 10y ago
  The root distinction seems to be whether you prioritize reality or abstractions. If reality comes first, you might not even believe that there is a set of universal mathematical laws governing nature (Feynman expresses this viewpoint, and raises the possibility that no such laws exist here[0]). Edward Nelson came up with similar ideas on the limits of abstractions in mathematics, based on Christian philosophy [1]. [0] https://www.youtube.com/watch?v=QkhBcLk_8f0 https://www.youtube.com/watch?v=QkhBcLk_8f0 [1] https://web.math.princeton.edu/~nelson/papers/warn.pdf https://web.math.princeton.edu/~nelson/papers/warn.pdf
codesnik 10y ago
I'm really surprised (every time!) when someone talks about conscience or discoveries or whatever so human-centric and local and unimportant in the cosmic scale while talking about simulations. If it's simulation, it certainly isn't made because of us.
- inimino 10y ago
  We are on the only planet we know about with life on it, so it is fairly easy to assume this is the most interesting place in the universe. Whether that's rational or not is another question, but it doesn't seem surprising.
3chelon 10y ago
On a semi-serious note, I've always had this nagging feeling ever since I first heard of Planck time and Planck length. Why would there be a lower limit to the resolution of the universe, unless there was something spooky going on? (I know they're just theoretical, but...)
- niftich 10y ago
  I'm not a physicist, so I'm speaking as 100% layman, but I can better accept a fixed-resolution universe, where forces have a certain magnitude and beyond them (in either direction, smaller or larger) matter doesn't behave in productive ways. For a long time, we thought atoms were these building blocks; now we think it's quarks; the Standard Model does a pretty good job at relating all observed phenomena to each other, except for, you know, gravitation, and that pesky observed-but-unaccounted-for thing we deemed 'dark matter'. A few things about the Standard Model make my head hurt, but the idea of a fractal universe that is infinitely scalable up and down and assumptions about spacetime and forces still hold the same way is truly mindbending that I don't know how to reason around it. So perhaps conversely, can we flip the question and ask, why wouldn't the universe have a finite resolution in terms of matter actually existing and behaving like we know and love?
  - raattgift 10y ago
    That's a good question you asked. The Standard Model has as a subgroup the Poincaré group, which is the isometry group of flat spacetime, which is both the (global) spacetime of Special Relativity and is the local spacetime of General Relativity for a pretty narrow definition of local, although we regularly construct Local Inertial Frames which are regions of spacetime that are so approximately flat that the difference is negligible. The Poincaré group has as a subgroup the Lorentz group, and its generators include rotations about the three spacelike axes of Minkowski (i.e., flat) spacetime. The other symmetries of the Standard Model are invariant under these rotations. A distant observer moving transversely relative to a particle observes a tiny rotation of the particle. The particle's fundamental properties do not change under that rotation, which can be arbitarily small. Rather than descending into group theory to reason about a minimum translation or boost, we can look to the energy-mass-momentum equivalence E_r^2 = (m_0 c^2)^2 + (pc)^2 and focus in on photons so we can ignore the m_0 term. Here we have E = E_r = (pc) = \hbar\omega = hc/\lambda = hf. Photons come at arbitrary frequencies, and we have good blackbody radiators all across our sky. If there were a minimal length scale, we would expect that derivatives of position would incorporate that minimal length, and so we would impose observables on things like the doppler shift, for example if we accelerate in a straight line towards a radiator emitting extremely high energy photons. So it's not so much "why wouldn't it have" but rather, "Q: does observation or experiment support a minimum scale that is large compared to the Planck scale for spacetime intervals in our universe? A: no" The answers for minimum length et al. scales that are small compared to the Planck scale are subtler (essentially by definition of the Planck scale, at those scales quantum corrections to account for gravitation become significant, spoiling the observability of short spacetime intervals) but so far still 'no'. For every particle in the universe there can be an (ultrarelativistic) observer who sees the particle wavelength shrink below the Planck length; there is nothing special about this observer -- there is likewise a possible observer who sees the particle at some much longer wavelength -- and the point of relativity is that neither observer is more correct than the other. Furthermore, if the former type of observer manages to see a star ultra-blueshifted and extremely Lorentz-FitzGerald-contracted, that observer does not create a black hole; no event horizon forms (event horizons are a global feature of the causal structure of spacetime[0]), and so we can turn an argument about Planck length into an argument about localized Planck energy: there is nothing obviously special in spacetime about the Planck scale. However, our ultrarelativistic observer will certainly see some very strange stuff courtesy of the Unruh effect: different observers observe different particle counts, and one would expect our observer would see an explosion in the number of particles compared to a more typical observer, and those particles -- like all others -- will also interact gravitationally, and we don't know yet how to decomplicate the picture enough to do useful calculations on them. [0] different observers may disagree on precisely where the event horizon is, what its shape is, and even the count and energy of the particles just outside the event horizon. but they will all agree that there is an event horizon.
zizzles 10y ago
Should we perhaps hypothesize the motive of this simulation, if any? Death seems to be a very important focal point: All living organisms are condemned to their aging bodies, eventually begin decaying, then die. Death of the universe is also inevitable, "Heat Death". Then there are other existential issues our feeble brains have to deal with. Then of course, our Earth is replete with suffering around every corner. Perhaps other terrestrial planets operate differently, for better or worse. You would think our maker(s) should have known the possible outcomes and scenarios before hitting the "start existence now" button, or perhaps it is just part of the grand plan. You have to wonder: Are these entity/entities sadistic pranksters of some sort?
- nzp 10y ago
  > Death of the universe is also inevitable, "Heat Death". This is far from settled. Heat death applies only to closed systems. Although it is not unreasonable to assume our universe is a closed system, we simply do not know this to be a fact.
- 3chelon 10y ago
  A cursory reading of most religious texts would certainly indicate a degree of sadism and lack of empathy with our human feelings. Human sacrifice? Plagues? Genocide? As an experiment, they would make sense. Virtual vivisection. The old God-as-a-Programmer argument certainly has its appeal :)
- nefitty 10y ago
  Entropy could just be an inevitable factor in this simulation. As the simulation steps forward in time, with the rules set forth at the beginning, these different emergent properties of the universe begin sprouting. There may not be a motive at all for this or an other simulation, besides gathering data on the effects of different rulesets.
mkstowegnv 10y ago
IMHO as a biologist infused with awe of the natural world from observation and reading, it is laughable that gods or computers could design or simulate the infinite levels of mind numbing complexity and unimaginable patterns that evolution has produced. But most people who are more focused on, and hungry for, the order created by technology and society in general, find it easy - with occasional cognitive dissonance, my favorite expression of which is the poem "Design" by Robert Frost: I found a dimpled spider, fat and white, On a white heal-all, holding up a moth Like a white piece of rigid satin cloth- Assorted characters of death and blight Mixed ready to begin the morning right, Like the ingredients of a witches’ broth-- A snow-drop spider, a flower like a froth, And dead wings carried like a paper kite. What had that flower to do with being white, The wayside blue and innocent heal-all? What brought the kindred spider to that height, Then steered the white moth thither in the night? What but design of darkness to appall?-- If design govern in a thing so small.
- pmoriarty 10y ago
  A beautiful poem. But who says that the "creators" couldn't be evil, or amoral, or just have different ethics/morals than humans, or who don't think in terms of ethics/morals? Also, while a human mind might marvel at some level of complexity, there's no reason why a superhuman mind would have to. I mean, imagine a "brain the size of a planet" (like Marvin's in The Hitchhiker's Guide), or one the size of a star, or larger. What could their limits be? Now imagine a mind not bound by the laws of physics as we know them -- for if they're "outside" the simulation of the universe as we know it, they'd not necessarily be limited by its physical laws. Really, at that point it's anything goes. Finally, consider that you, or any given human, or all of them, could simply be mistaken about what they think or observe. Just given that potential for error, anything could be possible.
- 3chelon 10y ago
  But evolution would be part of the process. The programmer (or god) just sets the thing in motion. From an evolutionary biology perspective, evolution might be the main deal, the whole point of the simulation. It would make some sense, to create processes, systems and structures that no being could ever come up with. As far as I know, there are some creationists that believe something along these lines already. Dawkins used to call it smuggling God in the back door.
alphanumeric0 10y ago
I'm not sure why people insist on assigning agency to this phenomenon. If you wanted, couldn't you invoke the 'many worlds' theory and suggest that our particular universe exhibits the same properties as what you might find in a simulation? This would do away with the need for an agent who is simulating.
jackcosgrove 10y ago
Even if we live inside a simulation inside a simulation inside a simulation, ultimately there is a base reality that influences the design of the simulations contained within it, and can impinge upon the simulations at any time. Infinite regression doesn't work here - using the term "simulation" implies some special pleading when moving from the outermost simulation to the base reality. If you want to regress forever, then there is no qualitative difference between layers of reality and the word "simulation" is meaningless. We live in the base reality. All a simulation could do is obscure complexity. Interestingly this isn't a new idea. The gnostics of the first two centuries BCE had a theology of this, in which a flawed world, in which humans lived, was created by an imperfect creator called the demiurge. There was a perfect realm beyond this world with an intercessor named Sophia (wisdom) who would rescue us from our flawed world, the simulation created by the demiurge.
- Smaug123 10y ago
  Sorry, I think I might be misunderstanding you. By "We live in the base reality", do you mean "Even if we live in a simulation, the simulation itself is hosted somewhere; following back the chain, eventually you reach the base reality. Therefore in the same sense as one should say that Microsoft Word runs/exists in our real world, one should say that (even if we're simulated) we exist in the base reality"?
  - jackcosgrove 10y ago
    Correct.
- akvadrako 10y ago
  You imply twice in your post that you assume there is a base reality, but why? I could argue it makes just as much sense that every layer is simulated by the layer "above". That could continue ad infinitum or perhaps create a cycle, so eventually it's simulating itself. These may be what we refer to as simulations, like what's running on my PC, or some kind of emergent behavior. Either case seems to work. It's natural to define the universe / all of reality / all layers as that which is self-defining, which is hard to describe without circular reasoning.
botw 10y ago
The other day I saw a post on social media that how he explained spaces from 0-dimension to 10-dimention in simple drawings. At the end, he claims that 10-dimentional space is just a dot, and there is no higher dimensional space more than 10-dimention any more. And he justifies his claims based on the string theory. At that time immediately I thought about the billions of parameter space in deep neural network and fractal spaces such as Mandelbrot Set. How could one simulate the universe based on only 3D space-time lattice?
danbruc 10y ago
In my opinion this seems all pretty pointless. To be able to differentiate between a simulation and reality you have to know at least how one of those works in quite some detail. What if we discover floating point rounding errors in the laws of physics we observe? Who is to say that this is not a feature of reality? What if it is the other way round, what if we live in a simulation with laws of physics made out of real real or complex numbers but the reality outside of our simulation has laws of physics behaving like floating point numbers? It seems pretty impossible to ever figure out what reality or a simulation must look like so that we can determine what surrounds us. Maybe there is some incredible self-consistency law at the bottom of reality, forcing reality to be what it is because there is one and only one way reality can be. But besides something like that?
- pmontra 10y ago
  Not so pointless. Think that one of the NPC in a computer game gets intelligent. It could learn how to trigger bugs or cheats in the game or features thought only for the player. Examples: crash the game, which hopefully restarts from the last save. Get where no NPC was supposed to be, only players, and get teleported somewhere else. Trigger a stack overflow and perform actions in the host, maybe changing the programming of the game or the data in the save files. In our case it could be the way to FTL. Did any Sci-fi author considered that yet?
  - Smaug123 10y ago
    "hopefully restarts from last save" - that's an awfully big gamble to take when the stakes are all of human existence.
  - danbruc 10y ago
    But how would that NPC know that it is in a game? What if we figured out how to do teleportation or time travel? How would you determine whether that is a feature of reality or whether we discovered a bug in a simulation?
    - pmontra 10y ago
rwallace 10y ago
> At fixed quark masses, the CRR of a lattice ensemble generation (in units of petaFLOP-years) scales roughly as the dimensionless number... The formula they give, indicates that the largest QCD calculations to date have taken 7.9e11 petaflop-years; this is consistent with the graph they give, but not with actual feasibility. What am I missing?
xenny 10y ago
Can someone summarise the paper? Seems to complex for a lay-man. For example, what is Lattice QCD?

77 points by aq3cn 10y ago

aq3cn 10y ago

For those, who want to read it in lay man's language, they can go through following articles: https://www.technologyreview.com/s/429561/the-measurement-that-would-reveal-the-universe-as-a-computer-simulation/ https://www.technologyreview.com/s/429561/the-measurement-th... http://www.huffingtonpost.in/entry/physicists-may-have-evide_n_1957777 http://www.huffingtonpost.in/entry/physicists-may-have-evide...

sakopov 10y ago
Interesting. So we're a simulation that can simulate 15 (femto)meters of cosmos that is indistinguishable from reality. This isn't much but it's still absolutely mind-bending. Reminds me of the movie The 13th Floor [1] [1] http://www.imdb.com/title/tt0139809/ http://www.imdb.com/title/tt0139809/
- codesnik 10y ago
  femtometers?
  - sakopov 10y ago
    Yep. Typoed.
yodon 10y ago
The tech review summary seems to be saying this super intelligent universe-simulating computer builder doesn't know about Bresenham's line drawing algorithm [0] and as a result it can only walk high energy particles along a limited number of paths. [0] https://en.m.wikipedia.org/wiki/Bresenham%27s_line_algorithm https://en.m.wikipedia.org/wiki/Bresenham%27s_line_algorithm
- komali2 10y ago
  Possibly? I can see why one would want to assume that an entity that can simulate a universe would be able to know more than the collective ability of that simulation, but on the other hand, we simulate things and learn unexpected results from our simulations. I think it's entirely feasible that a simulation can generate more knowledge, or different knowledge, than exists outside the simulation.
  - yodon 10y ago
    Certainly, but the idea that some hyper intelligent entity would develop the tech to build a universe-scale finite element simulator without groking anti-aliasing along the way is simply absurd. Bresenham's algorithm is a big deal because it's essentially the first thing everyone needs to invent when they start trying to render continuous processes on a discrete grid.
    - to3m 10y ago
      Bresenham's algorithm draws aliased lines.
    - raattgift 10y ago
      Where would you introduce the trajectory-correcting algorithm? In General Relativity, fundamental particles are simply taking shortest paths through spacetime, and where there are multiple short paths, the one that extremizes the timelike distance is favoured. In a perturbative theory of gravitation, you introduce corrections on the particles by way of graviton interactions. Worse, whether a local patch of energy-density amounts to a particle is observer-dependent; in both approaches above an accelerated observer sees more particles than an unaccelerated one does. That includes more gravitons in the latter approach. It would be a lot of work to write down a theory in which particles (whose count isn't global) decide on corrections to their own trajectories that return themselves to a path similar to one along a smooth manifold even though they are "really" travelling along a non-smooth one, so long as you are trying to match well-tested results of General Relativity. (A theory which doesn't reproduce those results is maybe an interesting toy; "cheating" by using a nearly-but-not-quite minimal length that has no direct observables is an uninteresting toy, and is not really a good argument for the existence of Simulators (or Creator Gods or whatever).) In particular, the non-linear contributions of gravity will likely lead you into an explosion in the amount of knowledge that a particle would have to acquire in order to assess its trajectory; you'd probably also want to propose an explanation of how it goes about adjusting its trajectory when it discovers it's not on the "right" one. Don't forget to make this correct for all possible observers, including the ones that will see no particle and those that will see two or more particles. In practice, attempts to do away with infinitesimals of length that do not clearly fail to reproduce known physics also tend to produce infinities of state in matter and/or additional gauge fields, and the infinities don't succumb to information-reduction techniques like renormalization by power-set counting. If you're looking to encode that knowledge somewhere, like in a spacetime-permeating field, the gravitational field of General Relativity is right there and already serving that purpose in a non-toy theory that (in explorable parts of the known universe) behaves completely correctly with quantum field theories of matter, including The Standard Model. :-) Best of all, in a block universe model, the whole thing is "pre-rendered", and there is no difference between the block universe model and an initial-values-surface formalism in the presence of determinism. So you only have to fully render one frame, and you get all the others, past and future. (Quantum mechanics as we know it is deterministic -- that's unitarity.)

CuriouslyC 10y ago

It kind of irks me that they basically assume a cubic lattice. Nature does not form cubic lattices, when nature forms a lattice (for example, in a foam) it tends to form roughly dodecahedral cells (with coordination number 12). Given this property, a space-time lattice would be very nearly isotropic.

SCAQTony 10y ago
Would a dodecahedral lattice explain why the universe stays symmetrical while expanding?
- pklausler 10y ago
  Does it?
  - SCAQTony 10y ago
    Yep, it's expanding symmetrically in all directions in that the distances may increase between bodies but the masses such as suns and planets do not grow. http://nautil.us/blog/the-universe-expanding-symmetrically-and-eternally http://nautil.us/blog/the-universe-expanding-symmetrically-a...
    - raattgift 10y ago
      Introducing more edges with your dodecahedral (or an icosahedral or some large-n polytope) cell worsens the edge-vs-face problem below, because a "cell shape" discretizes rotation and we have excellent evidence that rotation is smooth in our spacetime. Small regions of our spacetime are locally Lorentz-invariant, meaning that a number of observable quantities (magnitude of angular momentum and mass are two) do not change for a particle held at the origin of a system of coordinates as the particle is arbitrarily rotated or boosted. Arbitrariness is important. If we are observing a distant and predictable multifrequency radiator and define spherical coordinates with the radiator at the origin, then any movement our observation gear makes that isn't exclusively radial is equivalent to a rotation of the distant radiator. You can do this by holding the radiator at the (spacelike) origin, and the observer at its fixed (spatial) coordinates; in order to keep these coordinates constant, you have to rotate the system of coordinates to counter wholly non-radial relative movement. At large distances, the rotation at the origin becomes extremely small. Small or large, Lorentz invariance means the radiator has the same mass (it's by definition rest mass since it's always at the coordinates [0,0,0,t]) and momentum. So arbitrarily small rotation goes hand-in-hand with arbitrarily distant observers, and also with nearer observers who can displace themselves tiny amounts. So is there a minimum rotation? The arbitrariness of rotation is a in conflict with "cell shape", as when your "cell" distinguishes between edges and faces, the discrete nature of rotated mass-energy-momentum exposed through the (corner-filled, discrete) cell structure leads to different observables under rotation compared to that of the (smooth, continuous) spacetime of (either theory of) relativity. This deviation is larger for objects of higher momentum; and remembering Einstein's relation for rest-massless particles, E = pc = \hbar\omega = h / \lambda, that means that for different-wavelength photons emitted from the same source, the higher-frequency photons will arrive later than the lower-frequency ones. We have good observational evidence against frequency-dependent arrival times from bright distant objects (supernovas, gamma-ray bursts, even millisecond pulsars). We can conceive of a "cell shape" which is uniform under infinitesimal rotations, but at that point you have shifted one set of infinitesimals to another, and in the context of this topic (a simulation that among other things saves on state by abolishing real numbers in dimensions of length, rotation, boost and/or translation) is pretty much a non-winner. (Additionaly, ignoring the simulation context, you would also almost always run into difficulties if your "cell size" -- a minimum length in space or a minimal interval in spacetime -- is large enough to produce observables.)

kornork 10y ago

If we're living in a simulation, the question of whether AIs can have consciousness is solved.

skc 10y ago
But what if consciousness itself is simulated. The problem here is that a simulation would have to adhere to a set of programmatic rules meaning that the only way consciousness could arise as we like to think we know it, is via some weird unintended side effect.
mslev 10y ago
Is there a difference between a conscious being and a being who has been programmed to believe it has a conscience?
- kornork 10y ago
  I'm no expert on philosophy, but I feel like consciousness is a subjective experience. So externally there would be no difference, but I'm pretty sure I'm conscious. Of course, I've also heard of the illusion of consciousness, but that's where my brain explodes.
maxerickson 10y ago
What if there is a rule in the simulation preventing intelligences created by simulants from becoming sentient? I suppose you can point at all the artificial simulants and their sentience, but as pointed out in a sibling comment, so what.
pklausler 10y ago
The question would be mooted.
astrodust 10y ago
What if the simulation is actually being "run" inside an entity or organism that was the natural product of its environment?

smrtinsert 10y ago

What sort of moral issues result from creating a simulation universe that gives rise to sentient beings? I feel creeped out and reminded of Source Code with Jake Gyllenhaal for some reason.

hosh 10y ago
That's assuming sentience is an emergent property of the simulation rather than something else, such as being the ground from which the simulation arises from.
- Ygg2 10y ago
  If it's any consolation. It's probably calculating wind patterns or something.
knome 10y ago
https://qntm.org/responsibility https://qntm.org/responsibility
maxerickson 10y ago
As ever, it depends on your morality. I mean, plenty of people don't take issue with a lot of things I find reprehensible.
kornork 10y ago
As creator, are you responsible for the well being of the sentient beings? Are you morally prohibited from turning them off? Can you reset to a save point, possibly undoing the creation of some of the sentient beings?
teraflop 10y ago
Greg Egan has written a (slightly ham-fisted, but still interesting) short story about that question: http://ttapress.com/553/crystal-nights-by-greg-egan/ http://ttapress.com/553/crystal-nights-by-greg-egan/ The author has commented that he regrets having overlooked the moral implications of simulated life in his previous works (see e.g. http://gregegan.customer.netspace.net.au/PERMUTATION/FAQ/FAQ.html http://gregegan.customer.netspace.net.au/PERMUTATION/FAQ/FAQ..., which contains spoilers for Permutation City) so I wouldn't be surprised if that was the inspiration for this story.

retox 10y ago

(2012)

astrodust 10y ago

Keep in mind that if you wanted to create a simulated universe you can ensure nobody ever finds out. If someone does find out, you can simply fix the problem and revert to a point in time just before the data leading to the discovery was captured. Speaking academically, you might be playing with fire if you probe the outer limits of the universe and we are in a simulation run by some entity concerned with keeping that a secret.

4ad 10y ago
> you can simply fix the problem and revert to a point in time just before the data leading to the discovery was captured. Only with classical computers, not quantum computers.
- aswanson 10y ago
  Who's to say "real" computing, or "reality" at all, has to adhere to the laws we observe within the simulation? That's like Luigi in Super Mario Brothers saying you can't kick a turtle into a brick in the our world and not expect a gold coin to pop up in front of our faces.
  - ajuc 10y ago
    The difference is mathematical, not physical, so it should still hold.
    - Nadya 10y ago
      The mathematics depend heavily on physics. If physics outside of our universe differ - the math wouldn't add up. I can make 2 + 2 = 5 in a simulation. That doesn't mean 2 + 2 = 5 outside of the simulation.
      - ajuc 10y ago
        > I can make 2 + 2 = 5 in a simulation I'm not sure you can. Yes you can make a chest with 2 apples contain 5 apples if you put 2 more, but it's impossible (IMHO) to make it consistent. How many apples do you have if you place 2 chests besides 2 other chests and put 1 apple in each? What if you repeat the process 2 times and then another 2 times? Also - when some things in physics aren't intuitive (QM for example, but even adding velocities in relativity) we don't think math is wrong - we think math doesn't apply in that way to this particular phenomena.

mentos 10y ago

What is interesting is if we are the result of an unfathomable number of transistor gates, the consciousness that arises from it would never be able to understand the 'platform' (hardware) that it sits on.. would it? The idea being that you are creating a new dimension that cannot see the dimension that gave rise to it. I feel like no matter what we are in a 'simulation' where another dimension gave rise to ours whether it was intentional or not. What dimension is thought in?

tonmoy 10y ago
Nobody said the simulation is being done with transistors. Even if it was why wouldn't it be possible? I don't see any constraints. One day humans will fully understand neurons. One day we will make an AI using transistors who will be able to "understand" transistors
- mentos 10y ago
  I meant it more in an abstract sense. They don't have to be silicon they could be enormous metal planets and moons..
Razengan 10y ago
Indeed, just like how a computer program will never detect the exact positions of all the electrons in the computer it's running on. Not even if it tried to get that information from an external observer.
CoryG89 10y ago
Is this really true? What about a self-aware AI? I can imagine an AI that resides on a chip and developed consciousness could conceivably control a camera/robot in the real world and literally inspect and even reverse engineer or repair/improve itself or the hardware it runs on.
- komali2 10y ago
  True, but in the end their ability to perceive will be limited by the wires to/from those cameras to their hardware, and thus it is still possible that even the external world they thing they're observing, including themselves and their place in it, is also simulated. As in, someone could plug a "fake" camera into the AI and fool it.

rihegher 10y ago

13th floor style? http://www.imdb.com/title/tt0139809/ http://www.imdb.com/title/tt0139809/

lossolo 10y ago

I would be surprised if we are not living in simulation. Look at superposition and function collapse in quantum physics, it's equivalent to what we are doing in games today = do not render if player is not watching, save resources. Think about it, look for example at GTA V, what we can achieve today and think about how world looked like 100 years ago. I imagine that in 1000 years from now, simulating world similar to ours will not be a problem at all. Maybe not as complex because every level you go deeper you have less resources at your disposal. Who knows which level deep we are? Maybe we are simulation in simulation in simulation...

comicjk 10y ago
This is a common misconception, but superposition has nothing to do with "do not render if the player is not watching." As far as we know, every particle in the universe is interacting with every other (with a light speed delay) and thus "measuring" it (modifying and responding to its wavefunction) at all times. In the two-slit experiment, for example, the wavefunction of a particle moving through a slit interacts with the wavefunction of the particle used to detect it, causing both wavefunctions to be more narrowly distributed in the space around the slit. But before or after this interaction, the whole wavefunction still needs to be simulated to produce the effects we see experimentally. The wavefunction itself is the real thing being simulated, and it's always "running." You may be right that it will become easier to simulate worlds like ours, but it will require innovations that make a quantum computer look like an etch-a-sketch. The universe is really big and really detailed.
- Vraxx 10y ago
  I see the distinction you're making (I think), but I still don't see how that invalidates GP's analogy. Surely "running" a wavefunction is computationally cheaper than re-evaluating the particle's characteristics at every universal frame. Perhaps it's not so much about our own "watching" but rather "do not calculate when irrelevant". My own analogy might be an online map frame like google maps, where tiles outside the viewing box are not loaded. This is because we can tell what needs to be calculated by the bounding box of the viewer without needing all the other finer details. I see it more as a lazy eval.

45h34jh53k4j 10y ago

Sigh. Why would you dedicate your life to proving this? You will either find you cannot prove it, in which case you have wasted your entire life on falsehood, or you prove it true in which you undo your dilute your existence and everyone you have ever known and loved. Kinda like religion I guess...

Justsignedup 10y ago
though, if we are a simulation... maybe we can exploit bugs in it and give ourselves advantages! After all, the only bug-free program is one that was never started.
- slackware_linux 10y ago
  What are you trying to tell me? That I can dodge bullets?
  - Justsignedup 10y ago
    I'm trying to say that when the time comes, you won't have to!
- allemagne 10y ago
  let's just hope we don't find the bug that crashes our server
zzleeper 10y ago
You don't want to show that it's true. You want to find out whether it's true or not.
- Pinckney 10y ago
  It's not entirely clear what, exactly, it would mean for this to be true. It's also not clear that defining what we mean by "Is the universe a simulation?" has any value if it is not testable.
10y ago

codezero 10y ago

The title of the arxiv article is: Constraints on the Universe as a Numerical Simulation The only time they use the word evidence is in this quote: __ Therefore, there is a sense in which lattice QCD may be viewed as the nascent science of universe simulation, and, as will be argued in the next paragraph, very basic extrapolation of current lattice QCD resource trends into the future suggest that experimental searches for evidence that our universe is, in fact, a simulation are both interesting and logical.

sctb 10y ago
We've updated the submission title from “Physicists May Have Evidence Universe Is a Computer Simulation”, which breaks the guidelines by being editorialized.
- codezero 10y ago
  Awesome, thanks!

AnimalMuppet 10y ago

Per the arXiv abstract, physicists may be able to find evidence. That's very different from may have found.

3chelon 10y ago

Let's all just think, very hard, of the number 42 and see if the simulation ends.

AnimalMuppet 10y ago

So we stopped believing that God created the universe, and us as part of it. We believed only in science and natural processes. Following that through, we started believing that we live in a computer simulation with, presumably, some kind of programmer. That's pretty ironic.

code_sardaukar 10y ago
The root distinction seems to be whether you prioritize reality or abstractions. If reality comes first, you might not even believe that there is a set of universal mathematical laws governing nature (Feynman expresses this viewpoint, and raises the possibility that no such laws exist here[0]). Edward Nelson came up with similar ideas on the limits of abstractions in mathematics, based on Christian philosophy [1]. [0] https://www.youtube.com/watch?v=QkhBcLk_8f0 https://www.youtube.com/watch?v=QkhBcLk_8f0 [1] https://web.math.princeton.edu/~nelson/papers/warn.pdf https://web.math.princeton.edu/~nelson/papers/warn.pdf

codesnik 10y ago

I'm really surprised (every time!) when someone talks about conscience or discoveries or whatever so human-centric and local and unimportant in the cosmic scale while talking about simulations. If it's simulation, it certainly isn't made because of us.

inimino 10y ago
We are on the only planet we know about with life on it, so it is fairly easy to assume this is the most interesting place in the universe. Whether that's rational or not is another question, but it doesn't seem surprising.

3chelon 10y ago

On a semi-serious note, I've always had this nagging feeling ever since I first heard of Planck time and Planck length. Why would there be a lower limit to the resolution of the universe, unless there was something spooky going on? (I know they're just theoretical, but...)

niftich 10y ago
I'm not a physicist, so I'm speaking as 100% layman, but I can better accept a fixed-resolution universe, where forces have a certain magnitude and beyond them (in either direction, smaller or larger) matter doesn't behave in productive ways. For a long time, we thought atoms were these building blocks; now we think it's quarks; the Standard Model does a pretty good job at relating all observed phenomena to each other, except for, you know, gravitation, and that pesky observed-but-unaccounted-for thing we deemed 'dark matter'. A few things about the Standard Model make my head hurt, but the idea of a fractal universe that is infinitely scalable up and down and assumptions about spacetime and forces still hold the same way is truly mindbending that I don't know how to reason around it. So perhaps conversely, can we flip the question and ask, why wouldn't the universe have a finite resolution in terms of matter actually existing and behaving like we know and love?
- raattgift 10y ago
  That's a good question you asked. The Standard Model has as a subgroup the Poincaré group, which is the isometry group of flat spacetime, which is both the (global) spacetime of Special Relativity and is the local spacetime of General Relativity for a pretty narrow definition of local, although we regularly construct Local Inertial Frames which are regions of spacetime that are so approximately flat that the difference is negligible. The Poincaré group has as a subgroup the Lorentz group, and its generators include rotations about the three spacelike axes of Minkowski (i.e., flat) spacetime. The other symmetries of the Standard Model are invariant under these rotations. A distant observer moving transversely relative to a particle observes a tiny rotation of the particle. The particle's fundamental properties do not change under that rotation, which can be arbitarily small. Rather than descending into group theory to reason about a minimum translation or boost, we can look to the energy-mass-momentum equivalence E_r^2 = (m_0 c^2)^2 + (pc)^2 and focus in on photons so we can ignore the m_0 term. Here we have E = E_r = (pc) = \hbar\omega = hc/\lambda = hf. Photons come at arbitrary frequencies, and we have good blackbody radiators all across our sky. If there were a minimal length scale, we would expect that derivatives of position would incorporate that minimal length, and so we would impose observables on things like the doppler shift, for example if we accelerate in a straight line towards a radiator emitting extremely high energy photons. So it's not so much "why wouldn't it have" but rather, "Q: does observation or experiment support a minimum scale that is large compared to the Planck scale for spacetime intervals in our universe? A: no" The answers for minimum length et al. scales that are small compared to the Planck scale are subtler (essentially by definition of the Planck scale, at those scales quantum corrections to account for gravitation become significant, spoiling the observability of short spacetime intervals) but so far still 'no'. For every particle in the universe there can be an (ultrarelativistic) observer who sees the particle wavelength shrink below the Planck length; there is nothing special about this observer -- there is likewise a possible observer who sees the particle at some much longer wavelength -- and the point of relativity is that neither observer is more correct than the other. Furthermore, if the former type of observer manages to see a star ultra-blueshifted and extremely Lorentz-FitzGerald-contracted, that observer does not create a black hole; no event horizon forms (event horizons are a global feature of the causal structure of spacetime[0]), and so we can turn an argument about Planck length into an argument about localized Planck energy: there is nothing obviously special in spacetime about the Planck scale. However, our ultrarelativistic observer will certainly see some very strange stuff courtesy of the Unruh effect: different observers observe different particle counts, and one would expect our observer would see an explosion in the number of particles compared to a more typical observer, and those particles -- like all others -- will also interact gravitationally, and we don't know yet how to decomplicate the picture enough to do useful calculations on them. [0] different observers may disagree on precisely where the event horizon is, what its shape is, and even the count and energy of the particles just outside the event horizon. but they will all agree that there is an event horizon.

zizzles 10y ago

Should we perhaps hypothesize the motive of this simulation, if any? Death seems to be a very important focal point: All living organisms are condemned to their aging bodies, eventually begin decaying, then die. Death of the universe is also inevitable, "Heat Death". Then there are other existential issues our feeble brains have to deal with. Then of course, our Earth is replete with suffering around every corner. Perhaps other terrestrial planets operate differently, for better or worse. You would think our maker(s) should have known the possible outcomes and scenarios before hitting the "start existence now" button, or perhaps it is just part of the grand plan. You have to wonder: Are these entity/entities sadistic pranksters of some sort?

nzp 10y ago
> Death of the universe is also inevitable, "Heat Death". This is far from settled. Heat death applies only to closed systems. Although it is not unreasonable to assume our universe is a closed system, we simply do not know this to be a fact.
3chelon 10y ago
A cursory reading of most religious texts would certainly indicate a degree of sadism and lack of empathy with our human feelings. Human sacrifice? Plagues? Genocide? As an experiment, they would make sense. Virtual vivisection. The old God-as-a-Programmer argument certainly has its appeal :)
nefitty 10y ago
Entropy could just be an inevitable factor in this simulation. As the simulation steps forward in time, with the rules set forth at the beginning, these different emergent properties of the universe begin sprouting. There may not be a motive at all for this or an other simulation, besides gathering data on the effects of different rulesets.

mkstowegnv 10y ago

IMHO as a biologist infused with awe of the natural world from observation and reading, it is laughable that gods or computers could design or simulate the infinite levels of mind numbing complexity and unimaginable patterns that evolution has produced. But most people who are more focused on, and hungry for, the order created by technology and society in general, find it easy - with occasional cognitive dissonance, my favorite expression of which is the poem "Design" by Robert Frost: I found a dimpled spider, fat and white, On a white heal-all, holding up a moth Like a white piece of rigid satin cloth- Assorted characters of death and blight Mixed ready to begin the morning right, Like the ingredients of a witches’ broth-- A snow-drop spider, a flower like a froth, And dead wings carried like a paper kite. What had that flower to do with being white, The wayside blue and innocent heal-all? What brought the kindred spider to that height, Then steered the white moth thither in the night? What but design of darkness to appall?-- If design govern in a thing so small.

pmoriarty 10y ago
A beautiful poem. But who says that the "creators" couldn't be evil, or amoral, or just have different ethics/morals than humans, or who don't think in terms of ethics/morals? Also, while a human mind might marvel at some level of complexity, there's no reason why a superhuman mind would have to. I mean, imagine a "brain the size of a planet" (like Marvin's in The Hitchhiker's Guide), or one the size of a star, or larger. What could their limits be? Now imagine a mind not bound by the laws of physics as we know them -- for if they're "outside" the simulation of the universe as we know it, they'd not necessarily be limited by its physical laws. Really, at that point it's anything goes. Finally, consider that you, or any given human, or all of them, could simply be mistaken about what they think or observe. Just given that potential for error, anything could be possible.
3chelon 10y ago
But evolution would be part of the process. The programmer (or god) just sets the thing in motion. From an evolutionary biology perspective, evolution might be the main deal, the whole point of the simulation. It would make some sense, to create processes, systems and structures that no being could ever come up with. As far as I know, there are some creationists that believe something along these lines already. Dawkins used to call it smuggling God in the back door.

alphanumeric0 10y ago

I'm not sure why people insist on assigning agency to this phenomenon. If you wanted, couldn't you invoke the 'many worlds' theory and suggest that our particular universe exhibits the same properties as what you might find in a simulation? This would do away with the need for an agent who is simulating.

jackcosgrove 10y ago

Even if we live inside a simulation inside a simulation inside a simulation, ultimately there is a base reality that influences the design of the simulations contained within it, and can impinge upon the simulations at any time. Infinite regression doesn't work here - using the term "simulation" implies some special pleading when moving from the outermost simulation to the base reality. If you want to regress forever, then there is no qualitative difference between layers of reality and the word "simulation" is meaningless. We live in the base reality. All a simulation could do is obscure complexity. Interestingly this isn't a new idea. The gnostics of the first two centuries BCE had a theology of this, in which a flawed world, in which humans lived, was created by an imperfect creator called the demiurge. There was a perfect realm beyond this world with an intercessor named Sophia (wisdom) who would rescue us from our flawed world, the simulation created by the demiurge.

Smaug123 10y ago
Sorry, I think I might be misunderstanding you. By "We live in the base reality", do you mean "Even if we live in a simulation, the simulation itself is hosted somewhere; following back the chain, eventually you reach the base reality. Therefore in the same sense as one should say that Microsoft Word runs/exists in our real world, one should say that (even if we're simulated) we exist in the base reality"?
- jackcosgrove 10y ago
  Correct.
akvadrako 10y ago
You imply twice in your post that you assume there is a base reality, but why? I could argue it makes just as much sense that every layer is simulated by the layer "above". That could continue ad infinitum or perhaps create a cycle, so eventually it's simulating itself. These may be what we refer to as simulations, like what's running on my PC, or some kind of emergent behavior. Either case seems to work. It's natural to define the universe / all of reality / all layers as that which is self-defining, which is hard to describe without circular reasoning.

botw 10y ago

The other day I saw a post on social media that how he explained spaces from 0-dimension to 10-dimention in simple drawings. At the end, he claims that 10-dimentional space is just a dot, and there is no higher dimensional space more than 10-dimention any more. And he justifies his claims based on the string theory. At that time immediately I thought about the billions of parameter space in deep neural network and fractal spaces such as Mandelbrot Set. How could one simulate the universe based on only 3D space-time lattice?

danbruc 10y ago

In my opinion this seems all pretty pointless. To be able to differentiate between a simulation and reality you have to know at least how one of those works in quite some detail. What if we discover floating point rounding errors in the laws of physics we observe? Who is to say that this is not a feature of reality? What if it is the other way round, what if we live in a simulation with laws of physics made out of real real or complex numbers but the reality outside of our simulation has laws of physics behaving like floating point numbers? It seems pretty impossible to ever figure out what reality or a simulation must look like so that we can determine what surrounds us. Maybe there is some incredible self-consistency law at the bottom of reality, forcing reality to be what it is because there is one and only one way reality can be. But besides something like that?

pmontra 10y ago
Not so pointless. Think that one of the NPC in a computer game gets intelligent. It could learn how to trigger bugs or cheats in the game or features thought only for the player. Examples: crash the game, which hopefully restarts from the last save. Get where no NPC was supposed to be, only players, and get teleported somewhere else. Trigger a stack overflow and perform actions in the host, maybe changing the programming of the game or the data in the save files. In our case it could be the way to FTL. Did any Sci-fi author considered that yet?
- Smaug123 10y ago
  "hopefully restarts from last save" - that's an awfully big gamble to take when the stakes are all of human existence.
- danbruc 10y ago
  But how would that NPC know that it is in a game? What if we figured out how to do teleportation or time travel? How would you determine whether that is a feature of reality or whether we discovered a bug in a simulation?
  - pmontra 10y ago

rwallace 10y ago

> At fixed quark masses, the CRR of a lattice ensemble generation (in units of petaFLOP-years) scales roughly as the dimensionless number... The formula they give, indicates that the largest QCD calculations to date have taken 7.9e11 petaflop-years; this is consistent with the graph they give, but not with actual feasibility. What am I missing?

xenny 10y ago

Can someone summarise the paper? Seems to complex for a lay-man. For example, what is Lattice QCD?