The Simulation Hypothesis: Are We Living in a Simulation?
You've probably met this idea in its tossed-off form: a billionaire says the odds we're in "base reality" are billions to one, someone brings up The Matrix, and everyone nods. Stated that way it sounds like a stoner thought or a tech-guy affectation, easy to wave off.
The actual argument is sharper and stranger than the pop version, and most retellings get it wrong in a specific way. It doesn't claim to prove we're in a simulation. It proves something more slippery. Let me lay out what it really says, the one assumption everything hangs on, whether it can ever be tested, and why the question turns out to be a very old one wearing new clothes.
What is the simulation hypothesis?
The simulation hypothesis is the proposal that the reality we experience is a computer simulation, run by some technologically advanced civilization, and that everything we take to be physical, including our own minds, is information being processed on a computer we can't see from the inside.
It's worth separating two things right away, because they get merged constantly. The simulation hypothesis is the bare idea that this could be the case. The simulation argument, made by the philosopher Nick Bostrom, is a specific piece of reasoning about how likely it is. The argument is the interesting part, and it does not conclude what people think it concludes.
Bostrom's simulation argument: the real version
In a 2003 paper in The Philosophical Quarterly called "Are You Living in a Computer Simulation?", Bostrom showed that at least one of three propositions has to be true. Not "we're probably simulated." One of these three:
One: almost every civilization at our stage goes extinct before it becomes "posthuman," meaning technologically mature enough to build vast computing power.
Two: such advanced civilizations exist, but almost none of them bother to run "ancestor-simulations," detailed simulations of beings like their evolutionary predecessors.
Three: we are almost certainly living in a simulation.
The logic tying them together is a counting argument. If civilizations usually survive to maturity (so one is false), and if even a few of them run ancestor-simulations (so two is false), then those civilizations would run enormous numbers of simulated worlds full of simulated minds. Simulated people would outnumber real ones by a staggering margin. And if almost all minds in existence are simulated, then a randomly chosen mind, yours, is almost certainly one of them.
So the argument is a fork, not a verdict. It forces you to pick one of three branches, and only the third says we're in a simulation. You could just as coherently conclude that civilizations tend to wipe themselves out, or that mature civilizations lose interest in running ancestor-simulations. What you can't easily do, Bostrom argues, is believe that we'll become simulation-running posthumans ourselves and that we're not in a simulation, because that combination is exactly the one the counting argument rules out.
The assumption everything hangs on
The whole argument rests on one premise that's easy to miss: substrate-independence, the idea that a mind is a pattern of information processing and doesn't care what it runs on. If consciousness can arise just as well from silicon as from neurons, then simulated beings would be genuinely conscious, and the counting argument works.
But that premise is doing enormous work, and it's exactly the thing we don't know. If consciousness turns out to require biological wetware, or something else that a computer can't reproduce, then simulated "people" would be empty animations with no one home, and there'd be no vast population of simulated minds for you to probably be one of. The entire argument would deflate. So the simulation question is downstream of the hardest question in philosophy of mind, which is how physical processes produce experience at all. That's the same knot I pull at in does consciousness create reality?
It's not quite The Matrix
The movie version is a real brain in a vat, fed fake signals while your body floats in a pod somewhere in the real world. Bostrom's version is more thorough and more unsettling. In an ancestor-simulation there is no pod and no biological you. Your brain, your body and your memories are all part of the simulation, computed rather than housed. There's no "real you" elsewhere waiting to wake up. If it's true, you are the software, not a prisoner watching it.
Can the simulation hypothesis be tested?
Mostly, no, and this is the strongest objection to treating it as science rather than philosophy. A good simulation would, by definition, render everything inside it consistent with its own rules, so any evidence you found would just be part of the program. That makes the hypothesis very hard to falsify, and an idea that no possible observation could ever count against sits outside normal science.
People have looked anyway. A 2012 proposal by the physicists Beane, Davoudi and Savage suggested that if our universe were computed on a discrete lattice, like the ones used to simulate particle physics, that grid might leave a faint signature in the spectrum of the highest-energy cosmic rays. Others point to apparent "resolution limits" in nature, like the finite speed of light or the Planck scale, as suspiciously computer-like. None of this has produced anything conclusive, and there are ordinary physical reasons for those limits that have nothing to do with simulation. Treat the "physics proves it" headlines with caution. The honest position is that the hypothesis is, for now, essentially untestable.
Would a simulated reality even be fake?
Here's the turn that makes this more than a party trick. Suppose we are in a simulation. Does that mean nothing is real? The philosopher David Chalmers argues no, in his 2022 book Reality+. A simulated tree is still a real pattern you can bump into, made of information instead of atoms. On this view the simulation hypothesis isn't the terrifying claim that nothing exists. It's a claim about what reality is ultimately made of, that at bottom it's computation and information rather than fundamental matter.
Put that way, it stops being a trapdoor under reality and becomes a metaphysical position with close relatives. "Reality is fundamentally information or mind, and the physical is downstream of it" is not far from philosophical idealism, and it rhymes with a lot of what the older traditions were reaching for.
A very old question in new clothes
The thing worth noticing is that this is not a new idea at all. It's the latest costume for one of the oldest questions humans ask: is the world as it appears the ultimate reality, or a layer over something else? Every era has posed it with the most powerful metaphor it had on hand. Ours happens to be the computer.
Plato had prisoners mistaking shadows for reality. Zhuangzi woke from dreaming he was a butterfly and couldn't be sure which one was dreaming the other. Descartes imagined an all-powerful demon feeding him a false world, and used it to ask what he could still be certain of. The Hindu concept of Maya calls the everyday world a kind of veil over a deeper reality. The simulation hypothesis is that same suspicion, formalized with computation and probability. That lineage doesn't make it true, but it does mean it's tapping something perennial rather than something Silicon Valley invented. I put several of these worldviews side by side in the God models map.
The bottom line
The simulation argument is a genuinely clever piece of reasoning, and it earns less than it's usually credited with. It doesn't show we're in a simulation. It shows that three propositions can't all be false at once, and leaves you to decide which to accept. It stands or falls on whether consciousness can run on a computer, a question nobody has answered. It's close to untestable. And even if it's true, a careful reading says reality would still be real, just made of information. Which lands you back at the oldest question there is, now asked in the language of our own machines: what is the world actually made of, underneath the appearances?
Related reading: Does consciousness create reality? · The God models map · Panentheism vs pantheism · Spirit & philosophy