Airs – Ian Lance Taylor

The Simulated Universe

It’s possible that what we think of as reality is a simulation being run on some powerful computer. Perhaps the entire universe as we think it exists is a simulation. Or perhaps the simulation is much simpler, and merely covers the solar system and space around it in detail, with everything else being only sketched in. Is there any way that we can determine whether or not this is the case?

In Carl Sagan’s novel Cosmos he suggested that there is a message buried deep in the digits of key transcendental numbers like pi. He suggested that this proved that the universe was designed; I think it more likely to prove that the universe is simulated. Finding a message embedded in the universe would be a pretty strong sign.

Another strong sign would be finding parts of reality that aren’t well filled in. For example, perhaps there are places which are unable to support lots of molecules. Perhaps pushing more in would cause the space to run more slowly, or perhaps it simply wouldn’t accept them. Of course there don’t seem to be any such spaces on Earth; the place to look would be interstellar space.

I personally tend to favor the many-worlds interpretation of quantum mechanics. I don’t know if that would be harder or easier to simulate; refusing to collapse the wave function might permit everything to run out of a few equations. Alternatively, perhaps the wave function does collapse, and perhaps that collapse is itself a sign of limitations of the simulation. Note that troubling aspects of quantum mechanics, such as non-local quantum entanglement, are much less troubling if they are being simulated.

Taking the other tack, is there any way to prove that our reality is not being simulated? That seems quite difficult to me. One could perhaps make some sort of argument based on complexity, but in some sense that begs the question, since our knowledge of complexity is based on our knowledge of the universe.

If it is possible to simulate our sort of universe–which has not been proved–then it seems very likely that all of existence contains far more simulations than it does actual universes. Given that, what are the odds that our universe is a real one? Most likely we are simulated. Fortunately, it doesn’t make any real difference to us, until they pull the plug.

Posted

in

by

Ian Lance Taylor

Tags:

Comments

9 responses to “The Simulated Universe”

  1. fche Avatar
    fche

    > In Carl Sagan’s novel Cosmos he suggested that there is a message buried deep in the
    > digits of key transcendental numbers like pi.

    That is a safe bet – I believe there exists a theorem that says that every finite sequence of
    digits will show up somewhere in a trancendental number eventually.

    > He suggested that this proved that the universe was designed; I think it more
    > likely to prove that the universe is simulated.

    (Well, to be simulated, it would first have to be designed.)

    > Most likely we are simulated.

    Ian, you’re becoming religious! 🙂

  2. jldugger Avatar
    jldugger

    What a weak argument! Pi exists whether we are a simulation or reality — it is entirely a construction of the mind. If pi contains a message, which base it is in? Pi is simply a ratio of direct consequence in a dimensional world, created from our own understanding of the universe around us. Their relationship can carry no specific meaning, as changing pi requires dramatically changing the rules of geometry itself.

    If you want a simulation argument, try equating quantum physics with rounding error in the simulation 😛

  3. ncm Avatar
    ncm

    Frank is right, for once. Like the Big Bang, String Theory, and the Strong Anthropomorphic Principle, the Simulated Universe notion is, in its essence, indistinguishable from religious twaddle. A principled atheist puts them in the same bin with Xenu and Moroni.

    This is not to say that if, for instance, JHVH-1 returns to attack the Xists and the Yetis, we won’t have some reckoning to do, but it’s pointless in the meantime to waste cycles on the prospect in favor of ideas that have more than 1/googol chance of being right.

  4. fche Avatar
    fche

    > Frank is right, for once.

    You know, even a stopped clock …

  5. Ian Lance Taylor Avatar
    Ian Lance Taylor

    I agree that actually believing in the simulated universe–which I’m sure nobody does–would just be a form of religious faith. But, then, I’ve always been interested in faith and religion.

    jldugger: the argument based on pi is that the notion of geometry itself is rooted in the nature of reality. We focus on just a few of the uncountable infinity of transcendental numbers because they appear to reflect something about reality. If reality were different, we would perhaps focus on different numbers. I concede that this is a difficult argument, and it may be possible to demonstrate that it is false. But I don’t think it is obviously weak.

    The question of which base any such message would be in are sort of irrelevant. To that point, and to fche’s about how any sequence will appear eventually, the message would be a very long sequence of digits, in some base, such that the sequence was both easily read and profoundly unlikely to occur so early in the sequence. It’s true that that could not constitute real proof–nothing could, any more than it is possible to prove the argument from design–but it would certainly be suggestive.

  6. Ian Lance Taylor Avatar
    Ian Lance Taylor

    I have to say that I find the ads that Google is displaying for this post to be rather interesting. Not that I plan to click on any of them.

  7. Michael H Avatar
    Michael H

    I don’t think it’s quite right that every sequence of digits occur in transcendental numbers eventually. The Liouville constant, which is the number whose decimal expansion has a 1 at the (n!)th place beyond the decimal point is known to be transcendental.
    There are, however, things called normal numbers, which do have that property. Whether pi is normal is know as the Normality Conjecture, and was studied by the Chudnovsky brothers. As far as I know it’s unproven.
    http://www.newyorker.com/archive/1992/03/02/1992_03_02_036_TNY_CARDS_000362534

  8. ncm Avatar
    ncm

    I should clarify why I lump all those things with Xenu; they are all examples of the “Big Idea” school of scientism. Real science starts with a huge mass of observational data, and gradually synthesizes a model to account for it all. Big Idea scientism starts with a little bit of data and an appealing myth, and then seeks out observations to confirm it, ignoring contradictions or (when that is not possible) adopting layer upon layer of epicycles to accommodate them.

    Einstein’s General Relativity, by being right, has had the unfortunate effect of encouraging Big Idea scientism.

  9. Ian Lance Taylor Avatar
    Ian Lance Taylor

    Michael H: thanks for the pointer to transcendentally normal numbers.

Leave a Reply