Some days, the idea that we live in a simulation seems completely absurd. On other days, you try to print a document, the printer denies that it exists, you restart the computer, and suddenly everything works. On those days, the hypothesis gains a few points.
Behind the joke, however, lies a question that is difficult to forget: what if the universe, our history, and even this conversation were part of a simulation created by a far more advanced civilization?
What if we were not the users of the system, but characters inside it?
The idea sounds like The Matrix or a conversation that started far too late at night. Yet there is a serious philosophical argument behind it. It does not prove that the world is a simulation, but it forces us to consider several uncomfortable possibilities.
That is what I want to explore: not to convince you that there is a programmer outside our world, nor to dismiss the whole subject with a joke, but to understand what the argument actually says, what it needs to assume, and how far science can take us.
The question is older than computers
The suspicion that reality may not be what it appears did not originate in Silicon Valley.
Plato imagined people confined to a cave, watching shadows and believing those shadows were the whole of reality. Descartes asked whether something could deceive him about everything he perceived. Later came the thought experiment of a brain connected to a machine, receiving enough signals to believe it has a body and an ordinary life.
The question has always been similar: how can I know that the reality I experience is fundamental reality?
Computers did not invent that doubt. They made it easier to imagine.
Today we create video games, virtual worlds, weather simulations, and artificial agents capable of operating within rules we define. So instead of imagining only a demon deceiving us, we now imagine a programmer running us.
In making that leap, we often blend three separate claims:
- The universe can be described mathematically.
- Some physical processes can be understood as information processing.
- Our universe was intentionally constructed and is running inside an external reality.
One claim does not prove the next. The fact that we can simulate a storm does not mean the storm exists inside a computer. The fact that natural laws can be expressed through equations does not prove that somebody typed them into a cosmic keyboard.
The computational metaphor is powerful. It remains a metaphor until additional evidence appears.
What Bostrom actually argued
In 2003, philosopher Nick Bostrom published Are You Living in a Computer Simulation?, the paper that turned this intuition into a probabilistic argument.
Its conclusion was not “we live in a simulation.” Bostrom argued that at least one of the following three propositions must be true:
- Almost no civilization like ours reaches a posthuman technological stage.
- Civilizations that do reach it almost never run large numbers of simulations of their ancestors.
- If they reach it and produce vast numbers of simulations inhabited by conscious beings, almost all observers like us would be simulated.
The intuition behind the third option is a matter of numbers.
Imagine one original civilization with billions of inhabitants. If that civilization creates a million historical worlds, each containing billions of conscious people, simulated inhabitants would vastly outnumber the originals.
If we possess no special information that distinguishes us, it would appear more likely that we belong to the enormous group rather than the small one.
This is a conditional argument. It reaches that conclusion only if we accept that civilizations survive, possess the required resources, want to create enormous numbers of worlds, and produce genuine conscious experience inside those worlds.
Bostrom himself makes this clear in the updated version of his FAQ: a person can accept the trilemma while rejecting the hypothesis that we are simulated. The argument does not tell us which of the three doors is correct.
The simulation would need to feel
The assumption that interests me most is one that often remains hidden.
For the numbers to work, simulated people cannot be empty characters who merely appear human from the outside. They must feel. There must be someone or something experiencing joy, fear, pain, love, and boredom within that world.
Bostrom relies on what is called substrate independence: the idea that consciousness might arise not only from biological neurons, but from another medium capable of reproducing the relevant organization and processes of a mind.
We do not know whether this is true.
A computer can simulate a fire without burning. It can model a hurricane without getting wet. So does a perfect simulation of a brain produce consciousness, or does it merely produce a perfect description of conscious behavior?
We still lack an accepted theory that can answer this. If simulated minds have no subjective experience, they stop counting as observers of the same kind as us, and the statistical advantage driving the argument becomes weaker.
The hypothesis does not depend only on building enormous computers. It also depends on solving what consciousness is.
A famous number that is not a measurement
Elon Musk helped popularize the idea by noting that video games progressed from Pong to three-dimensional worlds shared by millions of people in only a few decades. If that evolution continued for centuries, he reasoned, simulations might become indistinguishable from reality. In that scenario, the odds that we occupy base reality would be one in billions.
The phrase is memorable. The number is not a scientific measurement.
It does not come from observing the universe. It comes from projecting a technological trend and accepting all the previous assumptions: continued progress, survival of civilizations, artificial consciousness, abundant resources, and an interest in recreating worlds like ours.
There is also something cultural at work. Every era tries to explain the universe through its most powerful technologies. At one point, the cosmos looked like a great clockwork mechanism. Today, surrounded by software, abstraction layers, and digital worlds, we naturally imagine it as a program.
Perhaps that tells us something profound about reality. Or perhaps it tells us a great deal about ourselves.
A less spectacular probability
In 2020, astronomer David Kipping published a Bayesian analysis of the argument. His starting point was that the technical possibility of creating conscious simulations remains unproven, and that uncertainty cannot simply disappear from the calculation.
In his model, the probability that we are simulated remains below 50 percent, although it approaches that value as the hypothetical number of simulations increases. This is not a universal percentage or a definitive answer: it depends on how the models and prior probabilities are defined.
What matters is the change in perspective. We cannot jump from “perhaps one day we can create simulated worlds” to “therefore we almost certainly live in one.”
Kipping adds a provocative observation. If humanity ever creates a simulation containing genuinely conscious beings, we will have shown that one of the argument's major conditions can be satisfied. The evidence that changes the debate may not appear while looking at the sky, but when we switch on a world capable of asking whether it is real.
At that point, the question would stop being merely philosophical.
Does physics reveal the universe's pixels?
The internet is full of supposed clues. The Planck length is said to be a cosmic pixel. The speed of light becomes the processor's limit. Quantum mechanics supposedly proves that the universe renders things only when somebody observes them. Déjà vu becomes a system error.
These are attractive comparisons, not evidence.
The Planck length marks a scale at which our current theories no longer work independently; we do not know that it is an indivisible unit of space. The speed of light is a limit within relativity, not evidence of an external computer. Quantum measurement is a genuine and profound problem, but physics does not require a graphics engine to describe it.
There have been more concrete scientific attempts. In 2012, Silas Beane, Zohreh Davoudi, and Martin Savage studied what traces a simulation built on a space-time lattice might leave. Such a structure might slightly break symmetry between directions and leave a signature in extremely high-energy cosmic rays.
The scope matters. Finding that signal would be interesting, though alternative physical explanations would still need to be excluded. Failing to find it would constrain that specific class of simulation, not every imaginable simulation.
We cannot conclude that every anomaly is a bug. Nor can the absence of bugs prove there is no system.
The problem with an explanation that accepts every result
Here we reach the strongest scientific difficulty.
A useful hypothesis has to risk being wrong. We need to be able to imagine an observation that counts against it.
Yet a broad version of the simulation hypothesis can accommodate almost any outcome. If we find a grid, it is the code. If we do not, the creators hid it. If the laws are elegant, the system was well designed. If they are chaotic, the designers wanted them that way.
At that point, the hypothesis moves closer to metaphysics than to a testable physical theory.
That does not make the question foolish. It only means there may be no experiment capable of settling it definitively from the inside. We can update probabilities as new information arrives, but we do not have a button that returns “base reality” or “simulation.”
Computational does not mean programmed by someone
Even if we someday discover that the universe can be described entirely through computational rules, one more leap remains.
A computational universe does not need to be a copy of another world or to be running on a cosmic laptop. Computation could simply be the physical activity of the universe itself.
A whirlpool exists in the dynamics of water. We do not need a more authentic whirlpool in another dimension to run it. In the same way, fundamental rules could produce the cosmos without an external programmer.
Adding a creator also fails to resolve the final question. If someone simulated our universe, who created theirs? The simulators might also be simulated. We can add layers, but we still have not explained why there is something rather than nothing.
If it were a simulation, would this life be less real?
Suppose undeniable proof appears tomorrow. A window opens across the sky and says, “Trial version. Seven days remaining to renew your license.”
After the panic and the search for whoever is responsible for the credit card, we would need to ask whether our lives had lost their value.
I do not think they would.
If we are conscious within this world, our experiences still occur. Pain does not stop hurting because information sustains it. Love does not lose its importance because another layer exists underneath. A promise still creates responsibility, and an injustice remains an injustice.
Discovering a simulation would change what we believe about the deepest level of reality. It would not automatically erase the meaning we create within it.
Nor would it prove that the creators hear us, can intervene, or have a moral purpose. They might be researchers, historians, an artificial intelligence, or students finishing a project the night before it is due.
The existence of a programmer does not guarantee the existence of a plan.
The question that turns back toward us
Today, we do not have enough evidence to say that we live in a simulation. We have an ingenious argument supported by contestable assumptions about future technology, consciousness, the choices of advanced civilizations, and the correct way to count ourselves among possible observers.
What we do have is a question arriving at a very particular moment.
We are building artificial intelligences, agents, and increasingly complex digital worlds. We remain very far from creating a conscious civilization inside a machine, but for the first time we can imagine a technological path toward something resembling it.
So the question changes direction: if we could one day create conscious beings within a system, would we do it?
Would we create millions of lives for research, entertainment, or rehearsing possible futures? Would they have rights? Could we switch their world off? What responsibility would fall on someone capable of creating a consciousness that can suffer?
The simulation hypothesis may never tell us who built the universe. But it already acts as a mirror reflecting the kind of creators we might become.
We may live in fundamental reality. We may exist inside an extraordinary simulation. Or we may be using the most powerful metaphor of our era to understand a universe stranger than any technology we know.
Until we have an answer, one thing remains unchanged: simulated or not, this life is the only interface available to us. And if our decisions affect other conscious beings, they matter.
You can continue this reflection and tell me which of Bostrom's three possibilities seems most likely to you in the original YouTube video.