Suppose that there was a headline in every newspaper that said that science has now proven that we do indeed live in a simulation, without any doubt. Or, some godlike aliens that created our simulation appeared to us and told us. Suppose that fact was settled.

Starting from that point, and working our way backwards, what are some of the pieces of evidence we would use to say, "Well, that makes sense, because of this fact."

I'm looking at it mostly as a computer programmer, thinking of how the simulation would work from a programming standpoint. Here are a few examples I'm looking for, see if you can think of some other similar arguments...

Individual and indivisible computations: A simulation, by definition, is an approximation. Having indivisible pieces of matter, or subatomic matter, or even strings, would be good evidence. They would be the individual pixels of our universe as it were. Or, perhaps the Planck length itself would be the invisible pixels, and the maximum resolution of our universe, and particles (or strings) would be variables.

No rendering until necessary: In a video game, there's no need to render objects that aren't in frame. You can save processor power that way. In the same way, it seems that our universe doesn't decide on a definite position, momentum, spin, etc, for a particle until it is pinned down and has to make a decision. That is the simulation saving expensive computation power.

Approximating complex systems: A black hole could be seen as merely a local system that is too complicated for a small area, considering the large number of particles are in such a small area. Especially if the processing is done as a distributed system, meaning one processor for one area of the universe, and perhaps a black hole taxes one processor too heavily. Instead of trying to simulate every particle and every particle interaction, at a certain density, just save the mass of the total amount of particles and delete all the extraneous detail about each individual particle. Approximate it with one mass parameter, one amount of angular momentum, and one amount of charge, the only 3 parameters you need for a black hole. For interactions with particles far away, the simulation would be unchanged, yet it would be much less computationally expensive locally.

Speed of light is speed of framerate: it would be prohibitively expensive to simulate every particle's interaction with every other particle in the entire universe simultaneously. It would be much easier to limit the amount of things that can influence other things, at least on a frame by frame basis. You don't have to simulate how a supernova will effect every star system right away, just simulate how it effects local particles, and then simulate the particles after that, then the ones after that, in a ring. Perhaps the speed of light is simply the set speed for computations of interactions between particles, in order to lighten the load of how fast things can influence other things.

Time dilation: It would likely be more computationally expensive to update all the info for an object such as a spaceship moving closer and closer to the speed of light, as it gets closer and closer to the edge of the framerate of the simulation. To help counter that, the framerate of the local simulation (inside the fast moving spaceship) would go down, much in the same way that video game framerates go down during fast moving parts of the game. If you run into the limit of processor power, you either have to drop frames, or slow down the speed of the frames.

Quantum entanglement: Perhaps entangled particles are created as a single class, together, and when they need to be updated, they are ran as one function, one subroutine, meaning they are both updated simultaneously, short-circuting the usual speed limit of c found in interactions between separate unrelated particles.

Nuclear fusion and the uncertainty principal: Our sun should not be hot enough or have high enough pressure for particles to overcome the Columb barrier in order to fuse, but they do anyway, and it's because of the uncertainty principle. The particles can't get close enough or move fast enough to fuse, but occasionally, the uncertainty principle makes it such that the particles were close/fast enough after all, because their position/momentum weren't really nailed down, only approximated. As a result of the imprecision of their true position/momentum, it turns out that they were close enough, and they fuse. In a way, laziness of computation of the particle interactions in the core of the sun is what makes the sun shine.

Of course there are statistical arguments about how simulations can make their own simulations and so forth, and we are unlikely to be at the top level. But that is speculation and probabilities, I'm looking more for evidence we can see in our own world, such as evidence of approximations, computational shortcuts, imprecise calculations, etc. Tricks that a programmer or a designer would make that would make a simulation easier to run.

Can you think of other clues we would use to explain why we live in a simulation?


marked as duplicate by David Z Apr 24 '18 at 0:26

This question has been asked before and already has an answer. If those answers do not fully address your question, please ask a new question.


There's basically nothing about physics as we understand it, that is especially suggestive of a simulation. The arguments you make are based on misconceptions regarding what our physical theories actually say.

The concept of computation we are used to, involves discrete variables and discrete timesteps.

Physics as we know it involves continuous variables (real or complex). If they became discrete at a certain scale, that would affect rotational symmetry, propagation of light, and many other things, and there would be effects analogous to aliasing. No such effects have been detected, putting strong upper bounds on the scale at which such discreteness could exist.

(It's true that certain classically continuous quantities can become discrete in quantum mechanics, but it's still based on an underlying continuum mathematics.)

Physics also has relativity of time. This doesn't just mean that fundamental physical processes are slowed down at high speeds (which was the conception of length contraction and time dilation before Einstein and Minkowski). It involves an altered conception of the four-dimensional union of space and time in "space-time", which makes the simultaneity of events (whether or not they happen at the same time) dependent on how the four-dimensional coordinate system defines the three-dimensional slices of equal time. The length and time effects of relativity actually arise from the change of space-time coordinates. Every special relativity tutorial explains the details, but the point is, relativity does not involve the universal time-ordering implied by standard models of computation.

Then there's everything to do with quantum mechanics. You speculate that quantum properties are undefined until they are measured, in order to conserve computing resources of the simulation. The problem is that measurement results take specific values according to probabilities that we can calculate, and the calculations are computationally intensive! So it appears that nature is very busy between measurements, as if sampling from a probability distribution that does require work to simulate.

All this can be simulated in discrete deterministic computers, but in some ways it's not a natural fit. In fact you could argue that the universe is a relativistic quantum continuum, which at macroscopic scales is "simulating" a discrete deterministic system with universal time.

  • $\begingroup$ Calculating a random quantum event or property from a statistical equation at a time when it becomes necessary could still be less resource intensive than storing all of the values of all of the parts of the universe, in the same way that a computer program randomly selecting a number every so often uses less resources than storing a pre-determined seed with a quintillion digits for every possible random action that could happen in the future. And with Bell inequalities implying that events and properties truly are random, it would seem that the universe is 'rolling the dice' when required. $\endgroup$ – Tazz250 Apr 24 '18 at 2:32
  • $\begingroup$ Of course, using these types of arguments as 'clues' that we are in a simulation requires some assumptions, such as us living in a universe with discrete units, necessitating such things as hypothesized planck units being a real limit to our world, instead of just a mathematical concept that could be real. But we DO know of some real physical limits that could be interpreted as limits of computation, such as the speed of light, the schwarzschild radius, the Pauli exclusion principle, etc, and those 'clues' are what I am looking for. $\endgroup$ – Tazz250 Apr 24 '18 at 2:56
  • $\begingroup$ Except that when you get into the details, none of those clues actually favors simulation. Let's sum up what the world appears to be, as a 'quantum relativistic continuum' (QRC). We are asking whether there is any evidence in physics, that reality is actually a digital computer. One possibility is that our QRC is rough around the edges - with the limits of its digital underpinnings showing - but there is no sign of that... $\endgroup$ – Mitchell Porter Apr 24 '18 at 4:16
  • $\begingroup$ Another possibility would be that the QRC paradigm itself is favored by simulation - that if you set out to simulate a world, you would have reason to make it that way. You seem to be making this argument - when you say lightspeed limit or various aspects of QM, could be design decisions motivated by a desire to economize on computation... $\endgroup$ – Mitchell Porter Apr 24 '18 at 4:19
  • $\begingroup$ But I'm saying that the details of how those things work in physics, go against the argument in favor of simulation. A QRC is not a natural thing for a digital computer to simulate. $\endgroup$ – Mitchell Porter Apr 24 '18 at 4:22

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