Elon Musk has a talent for boiling down huge questions into a single story. On 21 March 2026, he did it again. He replied to a post on X that linked quantum physics to the idea that reality only “renders” itself when someone is watching.
Musk shared a story from a dinner he once attended with two leading experts in very different fields. “I had dinner once with a top physicist and a top computer scientist and asked what they thought the probability was that we were in a simulation,” Musk wrote. “They answered simultaneously at 0% and 100% respectively. It was like a double-slit experiment, but with humans.”
Yesterday, Musk had already sparked debate with a viral cartoon on corporate planning.
To understand why it spread so widely, and why this debate keeps coming back, you need to understand two things. The first is a 200 year old physics experiment that still puzzles the smartest people alive. The second is a philosophical argument from 2003 that will not go away.
What the double slit experiment actually shows
The double slit experiment is, on the surface, very simple. The British scientist Thomas Young first carried it out in 1801. He shone a beam of light through two narrow slits cut side by side into a barrier.
On a screen behind the barrier, the light did not form two bright lines, as you would expect if light were made of tiny particles flying in straight lines. Instead, it created a pattern of alternating light and dark bands. This pattern is the signature of overlapping waves.
This was a major finding at the time because it showed that light acts like a wave. But the truly strange part came much later. In the twentieth century, physicists repeated the experiment with single particles, including electrons, photons, and eventually molecules made up of thousands of atoms.
When these particles are fired one at a time through the two slits, they hit the screen as single dots, like tiny bullets. But after hundreds and thousands of repetitions, those dots slowly form the same wave pattern.
Each particle, travelling alone, somehow seems to pass through both slits at once and interfere with itself.
In the language of quantum mechanics, the particle exists in a “superposition” of states. It takes every possible path at the same time until it hits the screen and is detected at a single point.
That alone is strange enough. But the truly puzzling result comes when scientists try to watch which slit the particle actually goes through. The moment a detector is placed at one of the slits to record the particle’s path, the wave pattern disappears.
The particles start acting like ordinary bullets. They pass through one slit or the other and form two simple clusters on the screen. Take the detector away, and the wave pattern comes back.
This is the observer effect. It sits at the very heart of quantum mechanics. The physicist Richard Feynman once said the double slit experiment contains “the only mystery” of the quantum world.
As recently as mid 2025, researchers at the Massachusetts Institute of Technology carried out a new version of the experiment using atoms trapped in a grid of light.
They confirmed once again that you cannot observe a photon’s wave nature and its particle nature at the same time. The more exactly you work out which path the photon takes, the more the wave pattern fades. The act of gaining information about the system changes how it behaves.
It is this property, the way physical reality seems to depend on whether or not it is being watched, that has made the double slit experiment a starting point for one of the most daring ideas in modern philosophy.
The simulation argument: not science fiction, but not science either
In 2003, the Oxford philosopher Nick Bostrom published a paper in the Philosophical Quarterly called “Are You Living in a Computer Simulation?” He did not argue, as people sometimes claim, that we are definitely living inside a computer program. Instead, he laid out three options. At least one of them, he argued, must be true.
First, that nearly all civilizations at our level of technology die out before they ever build computers powerful enough to simulate conscious minds.
Second, that civilizations which do reach that level are, for whatever reason, almost entirely uninterested in running such simulations. Or third, that we are almost certainly living in a computer simulation right now.
The logic works like this. If even one advanced civilization ever builds the ability and has the desire to create detailed simulations of its own history, complete with conscious beings who believe their world is real, it would almost certainly create not just one such simulation but billions.
And if billions of simulated worlds exist alongside one original “base reality,” then any randomly chosen conscious being is far more likely to be inside a simulation than outside one.
Bostrom himself has been careful about which of the three options he thinks is most likely. In an FAQ on his website, he once said he gave about 20 percent probability to the simulation idea.
He later raised this to a “substantial probability.” The argument does not prove we are in a simulation. What it does is show that dismissing the idea outright means you have to accept one of the other two options.
Each of those carries its own uncomfortable meaning about the future of our species or the behavior of advanced civilizations.
Why the double slit experiment keeps getting pulled into this
Most working physicists do not accept the link between quantum mechanics and simulation theory. But it is easy to see why the idea grabs people’s attention.
Think about how a modern video game works. To save processing power, a game engine does not build the entire world at once. It only creates the visual detail that the player is currently looking at.
Turn your character to face a wall, and the landscape behind you is not being drawn. The system uses resources only when needed, based on what is being watched.
Now think about the double slit experiment. Particles seem to exist in a fuzzy, wave like state of possibility until the moment they are watched. Then they “collapse” into a definite result.
To simulation believers, this looks a lot like a system that only fills in the details when someone is looking. Why bother calculating the exact path of every particle in the universe if nobody is watching? Just produce the result when a measurement is made.
The astrophysicist Neil deGrasse Tyson has been one of the most well known scientists to discuss this comparison in public. At the 2016 Isaac Asimov Memorial Debate at the American Museum of Natural History, Tyson said the chances of our universe being a simulation “may be very high.”
He pointed to the rapid progress of computing technology. He has since pulled back somewhat, settling on roughly 50/50 odds after hearing opposing arguments. One such argument is that if we were in a simulation able to create smaller simulations within it, we would probably have built that ability ourselves by now.
Tyson has also pointed to the Planck scale, a basic limit at roughly 10 to the power of negative 35 metres. Below this scale, the idea of distance stops having its usual meaning.
To simulation fans, this looks a lot like a pixel, a smallest possible unit built into the fabric of reality. The holographic principle adds another layer. It says that the information in a three dimensional space can be stored on its two dimensional surface.
To some, this suggests our three dimensional experience might be projected from a simpler layer underneath, much as a video game creates depth from flat code.
It must be said clearly: mainstream physics does not support the conclusion that the observer effect proves we are in a simulation. The “observer” in quantum mechanics is not a conscious person looking at a particle.
It is any physical interaction that takes information from the system, including an electronic detector with no awareness at all.
The wave pattern disappears not because the universe “knows” someone is watching. It disappears because the act of detection physically disturbs the quantum state through a process called decoherence.
The Wikipedia entry on the observer effect in physics says clearly that the popular reading linking conscious observation to the collapse of the wave function is “not supported by scientific research” and reflects a misunderstanding of quantum measurement.
Yet the comparison will not go away, and not without reason. The deeper mystery remains: why should gaining information about a system change how it physically behaves? Why should the universe care, at any level, whether a measurement has been made?
Quantum mechanics can describe this mathematically, but it cannot fully explain it in a way that satisfies human thinking. And it is exactly in that gap between description and explanation that simulation theory finds its place.
Musk’s dinner table thought experiment
Which brings us back to that dinner. Musk’s story about the physicist and the computer scientist captures something truly helpful about the simulation debate.
The two fields approach the question from completely different starting points and reach completely opposite answers.
A physicist, trained to build models of reality based on evidence and mathematical precision, tends to push back against the simulation idea because it cannot, right now, be tested.
There is no experiment you can run that would definitely prove we are in a simulation. Any result could, in theory, be explained as a feature of the simulation itself. For a physicist, an idea that cannot be tested is not very useful. So, zero percent.
A computer scientist, on the other hand, thinks about what is possible to compute. They know that virtual worlds are getting better at an extraordinary rate.
They understand that if it is theoretically possible to simulate a conscious mind, then the probability argument becomes very hard to escape.
If billions of simulated worlds can exist and only one “real” world does, the math points strongly toward us being inside one of the simulations. So, one hundred percent.
Musk’s comparison to the double slit experiment is not just a joke. It is a way of saying that the answer you get depends entirely on how you look at the question.
Look at reality through the lens of physics, and you get one result. Look at it through the lens of computing, and you get another. The two ways of thinking produce clashing pictures of the same underlying reality.
This is much like how the double slit experiment produces clashing pictures of the same particle depending on whether you measure its path.
The idea that will not die
It is worth asking why this debate keeps drawing attention from people like Musk, Tyson, and the late Stephen Hawking, rather than staying locked inside philosophy departments.
Part of the answer is cultural. We live in an age shaped by computers, and the idea of reality as software speaks to something deep in the modern mind. Part of it is about technology.
The rise of artificial intelligence, virtual reality, and brain computer interfaces like Neuralink makes the line between “real” and “simulated” experience feel less firm than it once did.
But part of it is truly philosophical. The simulation argument, whatever its flaws, forces us to face a question that most people prefer not to think about: what is the nature of the reality we live in, and how much of what we take for granted rests on ideas we have never tested?
Physicist Tom Campbell, a former NASA researcher, has gone further than most. He has designed tabletop optics experiments meant to test whether reality behaves like a simulation.
His proposed setups involve firing laser beams through detailed sequences of slits, mirrors, and detectors. He is looking for results that would be impossible if reality were fully set in advance but perfectly consistent with a system that produces results on demand. He has written that a positive result would be “an unambiguous indicator that our reality must be simulated.”
Other scientists are more careful. Marcus Noack, a computational physicist at Lawrence Berkeley National Laboratory, has argued that it is impossible to test the simulation idea as a whole.
The best we can do, he says, is explore a “limited neighbourhood” of ideas about how a simulation might work. We can only hope that the designers, if they exist, were not clever enough to stop us from finding their work.
The honest answer, for now, is that we do not know. We do not know whether the universe is a simulation. We do not know why watching a quantum system changes how it behaves.
We do not know whether consciousness is something that can be copied in silicon, which is needed for Bostrom’s argument to hold. And we do not know whether the apparent fine tuning of physical constants reflects design, luck, or something else entirely.
What we do know is that Elon Musk sat at a dinner table, asked two brilliant people the same question, and got two answers that could not both be right. And that, as he noted with his usual dry humor, is itself a kind of quantum experiment.
The particle went through both slits. The answer was both zero and one hundred. And the interference pattern, for those willing to look at it, is still there on the screen.
