Why Do Some People Believe We Are Living in a Simulation?

The idea that our reality might be a sophisticated computer simulation is not new. Plato’s cave allegory gestured at it 2,400 years ago. Descartes asked how we could rule out an evil demon feeding us false perceptions in the 17th century. But in 2003, philosopher Nick Bostrom gave the idea a form precise enough to take seriously in a technical sense, and it has not quite left serious intellectual conversation since.

The question worth asking is not “is the simulation hypothesis true?” (we do not know, and there is currently no way to find out). The more interesting question is why smart, rigorous people take it seriously at all, and what that tells us about reality, consciousness, and the limits of what we can know.

Bostrom’s Trilemma: The Precise Version

The usual pop-science version of simulation theory is: we live in a Matrix-style digital reality and someone is running it on a computer. This is close but loses something important in the translation.

Bostrom’s actual 2003 argument, published in the journal Philosophical Quarterly, is structured as a trilemma. He argues that at least one of the following three statements must be true:

1. Almost all civilizations at our current stage of development go extinct before reaching the technological capacity to run realistic simulations of minds.
2. Almost all civilizations that reach that capacity choose not to run such simulations, for reasons we can only speculate about.
3. We are almost certainly living in a computer simulation right now.

The logic runs as follows. If even a small fraction of civilizations reach technological maturity and run simulations of their evolutionary history, the number of simulated minds would dwarf the number of biological ones. A randomly selected conscious entity would then be statistically far more likely to be simulated than biological. Unless something consistently prevents civilizations from either reaching or exercising this capability, we should assign significant probability to being simulated ourselves.

Notice that Bostrom himself has not argued that option three is true. He has argued that the structure of the reasoning makes one of these three options very likely. Which one is correct requires evidence he acknowledges we do not have.

Why Serious People Take It Seriously

The simulation hypothesis attracts genuine philosophical and scientific interest for several reasons that go beyond its entertainment value.

The computational universe parallel. Theoretical physicist John Wheeler proposed in the late 20th century that information might be more fundamental than matter: that “it from bit,” the physical world emerges from information-theoretic processes rather than being made of stuff in any traditional sense. This framing is not the same as the simulation hypothesis, but it rhymes with it. If reality is fundamentally informational, the question of whether it runs on a substrate of silicon or quantum fields becomes less metaphysically sharp than it first appears.

Quantum mechanics’s strange behavior. The universe behaves, in some respects, as if it is only rendered when observed. Wave functions collapse under measurement. Particles have no definite position until detected. Various physicists have noted, usually with considerable caution, that this resembles how a computational system might work if optimizing resources by only rendering what is being directly examined. This is a weak analogy, not evidence, but it is the kind of observation that keeps the conversation alive among people who know the physics.

The mathematical fine-tuning problem. The fundamental constants of the universe (the strength of gravity, the charge of the electron, the cosmological constant) are calibrated with extraordinary precision. Small changes in several of them would produce a universe with no stars, no chemistry, no life. The simulation hypothesis offers one explanation for this: the constants were set by a designer. This is not a better explanation than others on offer (the anthropic principle, the multiverse), but it is a coherent one.

The Main Objections

Unfalsifiability. The most serious scientific objection is that the simulation hypothesis makes no testable predictions. If we are in a simulation, what observable difference would that make? Without a prediction that could be falsified, it fails the basic criterion of scientific theory. It remains in the domain of philosophy rather than physics, which is not a dismissal but a categorization.

Computational impossibility arguments. Some physicists argue that certain properties of the universe cannot in principle be simulated algorithmically. Quantum entanglement, the behavior of quantum fields, and Gödel’s incompleteness theorems have all been invoked as arguments that a simulation, in any conventional computational sense, cannot reproduce what we observe. These arguments are disputed, but they represent serious technical pushback.

The self-defeating problem. If we are simulated, our knowledge of physics and computer science, which forms the basis of the simulation hypothesis itself, could be an artifact of the simulation rather than a reliable map of any deeper reality. The argument undermines the very tools it uses to construct itself. This is a version of the Cartesian problem with no obvious resolution.

The infinite regress problem. If we are in a simulation, who simulated the simulators? And who simulated them? The hypothesis potentially generates an infinite chain of realities rather than an explanation of why there is something rather than nothing.

What the Hypothesis Is Actually Doing

The simulation hypothesis is, in a useful sense, a modern version of a very old philosophical puzzle: how do we know that our perceptions correspond to an external reality? Descartes asked this with his evil demon. Berkeley asked it with his idealism (the idea that matter does not exist independently of perception). The simulation hypothesis asks it with 21st-century technology as the framing.

The reason these questions recur across centuries is that they do not have clean answers. We operate on the working assumption that our perceptions are roughly tracking reality because that assumption is useful and because we have no compelling evidence against it. But “useful assumption without compelling counterevidence” is not the same as “proven.”

Whether the simulation hypothesis is true may be permanently undecidable. What it does is push us to be precise about what we mean by real, what assumptions we are making when we do science, and what the limits of what we can know actually are. For a hypothesis with no confirmed evidence, that is a considerable philosophical contribution.

Elon Musk’s often-quoted claim that the odds we are not in a simulation are “one in billions” represents one end of the credence spectrum. Oxford philosopher David Pearce places the probability at under 1%. Most people who have thought carefully about it land somewhere between these poles: uncertain enough to find the question worth engaging with, appropriately skeptical enough not to treat it as established fact.

That epistemic position, holding the question open without resolving it prematurely in either direction, is probably the most honest place to stand.