I was thinking a while back about the Simulation Hypothesis.  For those who haven’t come across it before, the Wikipedia article does a pretty good job of laying out the basics.  The really short version, though, is that it postulates that we live in the Matrix and tries to offer some anthropic reasoning in favor of this argument.  In general, I find anthropic arguments (arguments that try to use the fact that you are here and thinking about the argument as key evidence in favor of said argument) highly specious.  They may sound attractive, on the surface, but they never cash out into anything useful outside of very artificial scenarios.

Personally, I think the best evidence in favor of the fact that we live in a simulation is the “weirdness” of quantum mechanics and relativity compared to classical mechanics.  Most people think that the reason why classical mechanics aligns so well to human intuition is that it deals with scales that are within human experience.  When things start moving really fast, act at giant scale, or get really tiny, this alignment between intuition and truth breaks down.  But I find it a little odd that classical mechanics should be so broadly applicable, so simple, and yet be not quite good enough.

When I let myself get fanciful, I postulate that it’s because we live in a buggy (or incomplete) simulation of a classical universe.  For instance, the fuzziness of quantum behavior that seems dependent on information storage (e.g. the famous double-slit experiment) can be seen as an artifact of optimizations that the simulation makers put in.  It would be too expensive to actually tag and track every sub-atomic particle.  So they don’t.  The physics engine just fuzzes particles into probability distributions that all work out right most of the time.  But if you track which slit a particle went through, then the optimizer has to track the trajectories of the particles, and in doing so changes the pattern that gets generated.

It’s like if you were living in a video game and you noticed that people in the distance just disappear once they get a certain magic distance away.  That would seem really weird if you were a physicist in that world doing experiments.  They don’t go away, really, and when you get closer (or look at them through a sniper rifle scope), they appear again.  The reason why this is, to the designers, is because it’s way cheaper to ignore objects that are too far away when rendering a scene.  It won’t change much of the final answer and it saves you a lot of computational effort.

The fixed speed of light in all inertial reference frames could be another one of those optimizations.  Instead of coming up with a really complicated system to track photons classically, the simulation makers just set the speed of light to a high constant, which I bet is set tightly to the clock rate of the physics system (since updates in forces seem to move at the speed of light).  Of course, having a fixed speed causes a whole bunch of distortions in space and time, but you only ever notice them when you’re going really fast or measuring time very precisely.  If you’re hacking out a universe simulation, this would probably seem like a good trade-off.

Or the fact that there seem to be so many fundamental physical constants.  The Standard Model has at least 25 of these.  And there are even more physical constants with dimension, like the gravitational constant or the speed of light.  This is strange from a physics perspective.  But it doesn’t seem so weird from a computer science perspective.  After all, people love making configuration files, in case they want to tweak the behavior of a running system without recompiling.  Why not have 25 different switches you can throw to tweak the behavior of the physics engine?

This also serves as a neat solution to the Fermi Paradox.  There’s nobody else out there because alien life is computationally expensive.  And if this simulation isn’t about alien Von Neumann probes colonizing everything, why render them in the first place?  Or why render anything interesting about distant planets until someone actually points a telescope at them?

OK, so let’s just say for the sake of argument that we do live in the Matrix.  What does that imply?  Well, it could mean all sorts of things.  It certainly tends to get people thinking in a religious vein.  Is the Simulator God?  Should we care what he thinks?  Does it matter to us what the outside people think is the purpose of the simulation?  And so on and so forth.  For all of their other faults, the people who made the Matrix sequels were certainly thinking along these lines.

But what occurred to me recently was that, if you take the Simulation Hypothesis seriously, it provides a rigorous framework in which you can define the orthodox versions of every religion I can think of.  By that, I mean that I can build a model of the universe using the Simulation Hypothesis that’s isomorphic to the output of any religious system of thought.  But phrasing it this way, it dissolves a lot of what people have claimed are irreducible mysteries.  Let’s take it for a spin on Biblically-focused Trinitarian Christianity.  This is a neat test case because it’s historically important; is considered to be deeply unfashionable, if not laughable, by modern scientific types; and has a bunch of mysteries that stop being terribly mysterious with the Simulation Hypothesis.

So God starts out by making a universe.  He says “let there be light” to his computer system, which starts the simulator up, and it takes six days to crunch out the starting position for the universe.  The simulator is running on extreme fast-forward for most of this time, slowing down to render more detail when God decides that something is happening that’s interesting.  Since God is trying to make people in his image (meaning he wants a planet full of humans, not intelligent lizards or whatever), he’s pruning the start and nudging things in an attempt to make this happen.  Plenty of crazy things happen during this phase, like the impact that split off the moon (with just the right size and distance to eclipse the Sun as seen from Earth, an artistic touch), the mass extinction that happened when the atmosphere became substantially toxic with all the free oxygen, the asteroid impact to kill off all the dinosaurs, and so on and so forth.

It’s worth noting here that God is omniscient and omnipotent in the sense that he has full access to the state of the physics engine in memory.  He can pause the simulation, run it forward or restore it to a previous state, zoom in to see anything that’s going on in detail, and edit the state of the simulation however he wants.  So he can’t make a rock that’s so heavy he can’t lift it.  God doesn’t lift things by reference to the object’s weight.  When God lifts a rock supernaturally, he changes the z-position (or its velocity, perhaps) of the rock object in the simulator so that it’s above where it used to be.  He’s also outside of time in the sense that worldly time is a property of the physics engine.  But, presumably, since he’s simulating something in his own image, he’s subject to a similar constraint in his world.

It’s also worth noting that God hacks these variables very rarely.  The system is huge and complicated.  It would be really easy to change the wrong memory location and crash the whole thing.  So he does most of his work through subroutines written elsewhere.  These entities are called angels and daemons.  God has that Unix guru style, it would seem.  If they need to be rendered in the physics engine to perform their action, they take physical form and then sometimes stories are written about them.  But most of the time they’re just used to help edit the world state directly without causing widespread chaos.  The effects of these edits are called miracles.

OK, so after the initialization work (including a little interlude in a separate physics instantiation called the Garden of Eden), God starts messing with people.  And he’s kind of a jerk about it.  It’s like when you’re playing Sim City and you summon Godzilla to come knock down houses and power lines.  Just because.  Some of his greatest hits are when he notices a bunch of people building a tower up to the heavens, so he confuses their language so that they collapse into small tribes and start fighting each other for his amusement.  Or when he decides to flood out the whole world.  Probably because he had a flood button and he wanted to see what it would do.

After a while, God gets bored just trolling his simulated peons.  So he leaves them alone for a while.  And they start doing all sorts of cool stuff, like building civilizations and pyramids.  He wasn’t expecting that.  It wasn’t against the rules, exactly, but he’d kept the people from eating from the tree of knowledge in the Eden interlude because he didn’t want them to get too advanced too quickly.  But he finds himself sucked into the story of all these little guys building and fighting.  And it’s the most dramatic and compelling in the Middle East.  Societies are spreading and flourishing all over the place.  Then, all of the sudden, these barbarians come out of nowhere and wipe everybody out.  All the big-time civilizations collapse, with the exception of the Egyptians, who were the coolest and just fall on really hard times.

So now God’s got the civilization bug.  He figures that he’s going to start playing this game like Populous.  Since he’s feeling pretty invincible (after all, he’s the guy running this simulation), he picks the Jews to shepherd to glory.  They don’t have a lot going for them, but he figures it’d be awesome if he could pull it off.  In order to make it fun, he sets one of his daemons to be the “enemy”.  After all, it would be too easy if he could just miracle his way to victory.  But instead of giving the daemon a nation of his own, he just sets the daemon to influence the situation to make things hard for his people.

Pretty quickly, God’s able to set up the Jews in Israel and get them running something resembling a decent country.  He does the occasional miracle to help them out.  But mostly, he decides to do things by talking to particular chosen people.  He gives them very specific instructions that he wants them to follow any time it looks like they’re running off the rails.  It sort of sucks compared to the civilizations that were there before, but they do OK.  They win some military victories, wipe out a few rival tribes, but it’s mostly small-time stuff.  Strikingly, the pattern is mostly God failing at the game.  He gives them a command, like worship no God before me, and the next thing he knows they’re worshiping some golden calf from the next town over.  Or he establishes a monarchy and gets a civil war instead.  Eventually, Israel gets wiped out by the Assyrians.

So God figures that he’ll turn it around by switching up the type of game.  Using his prophets, he lays the framework for the arrival of the Messiah, a great war leader that will unite the Jews and lead them to world conquest.  Then he creates an avatar and plugs himself into the game.  He’s going to do this RPG-style.  Again, because he’s the kind of guy who likes the idea of being the underdog, God incarnates as a poor carpenter – Jesus – instead of a prince or king somewhere.  The trinity is made up of God the Father (in his role as Systems Administrator), God the Son (in his role as RPG avatar), and God the Spirit (in his role as the nonphysical inspiration of people).  They’re all the same guy underneath, but they’re different from the perspective of the Matrix.

He’s on his best behavior as Jesus, making sure to live a sinless life (made much easier by fast-forwarding the boring parts and save-reloading any time temptation overcomes him), and he makes sure to check all of the Messiah boxes that he’d been so careful to set up ahead of time.  He does the requisite number of miracles to establish his divine nature and starts preaching.

But a funny thing happens on the way to world conquest.  God spends enough time in RPG mode that he starts to care about these simulated people.  Before, they were just numbers in a computer to him.  But living amongst them, seeing their aches and pains and sorrows, he can’t help but feel for them.  And he feels just awful about all the things he did to their predecessors.  He has so much blood on his hands.  And so do they.  The world he built isn’t anything resembling a utopia.  But, then again, it wasn’t ever meant to be.

So God decides to spend the rest of his time on Earth as Jesus trying to convince people to be nice to each other.  Trying to fix it for everybody, not just the Jews, though of course the Jews are still special to him.  He could just reboot the simulation, or shut it down and start another one that was better from the start, but he feels responsible for these people.  He wants to make it right.  But it turns out he’s as poor of an RPG player as he is a strategy gamer.  He’s doing a great job of rallying the Jews to his cause, but he gets outmaneuvered by the local authorities and sentenced to death before he makes the big time.

God knows it’s not real in here.  If he dies, he won’t actually die in the real world.  But if he doesn’t log out of the game or perform some miracle, it’ll feel like he’s dying.  It’ll hurt.  It’ll be scary.  After a lot of thought and contemplation, he decides that he has to do this.  He has to go through with it, for both himself and the world he created in a different guise.  So he heads to the Cross and dies for the sins of the world.  Then, three days later in the system, he restores his avatar to life and logs back in.  Then after spending some time telling his guys to go out and spread the word to the whole world, he exits RPG mode for the last time.

The rest of history is mostly God doing his best to steer history in accordance with his newfound appreciation for his simulated people.  At first, he takes a very active role in talking to the people he knew during his life.  He helps them out of jams, sends them inspiration, and every so often he loads the dice on a crucial encounter.  But as time goes on, God progressively loses his stomach for drastic intervention.  It feels like cheating.  After all, the people on the other side are people, too.  So he mostly just watches.  But in order to satisfy his sense of justice, he instantiates two separate physics engines for the afterlife.  Based on the contents of a person’s soul (which is a metadata collection attached to a person at birth – real, but wholly non-physical), he routes people after death to one or the other (by saving them in the one system and instantiating them in the other).

And that’s that.  I find the resulting story surprisingly compelling.  But the most fascinating thing to me about all of this is that if you take the Simulation Hypothesis seriously, the more “supernatural” cosmologies are the more plausible.

For instance, if the point of the simulation is people, then it’s way more likely that God ran the billions of years between the beginning of the universe and the advent of people on hyper fast-forward.  Maybe they weren’t run at all in the normal engine, and the simulation was just started in media res 6,000 years ago.  Or, if we’re in a simulation, why would the physics engine necessarily be all there is?  Non-physical properties (like the soul) could very well exist and have meaning in a broader context, even if there is no possible way to gather material evidence they exist.

Or maybe we’re in a buggy simulation.  If so, then perhaps mystics and sorcerers throughout history have done “magic” by finding exploits in the system from within.  Some crazy stuff along these lines is possible in real life video games.  There is a branch of tool-assisted game speedruns called “execute arbitrary code” where, through just using the right inputs into the controller, the speedrunners are able to reprogram the RAM to do anything they want.  I highly encourage you to check it out in the context of the Simulation Hypothesis.  If this is the case, then perhaps God is even pushing bug fixes into the live running simulation, so that exploits that were possible in the past are no longer possible.

Anyhow, I figure that the upshot of all this meandering is that, to the degree one finds the Simulation Hypothesis compelling, one should find any religion (especially the ones that are popularly professed) almost equivalently compelling, since their claims are isomorphic.  The case for atheism thus rests on the question: “How are you so sure we don’t live in the Matrix?”