What Is the Simulation Hypothesis?

Direct Answer

The simulation hypothesis is the claim that the universe we inhabit — including our bodies, our minds, and our apparent memories — is a computer simulation created by a more advanced civilization. On this view, what we call "reality" is a computational artifact: the laws of physics are the rules of the simulation, the stars are rendered scenery, and our experiences are the outputs of a vastly powerful machine. The hypothesis is not an empirical theory of physics but a philosophical possibility — a modern descendant of ancient and early modern skepticism.

The hypothesis was given its most influential modern form by Nick Bostrom in his 2003 paper "Are You Living in a Computer Simulation?" Bostrom's simulation argument runs as follows. Suppose that at least one civilization reaches a "posthuman" stage at which it commands enormous computing power. Such a civilization could run "ancestor simulations" — detailed simulations of the minds and environments of its predecessors. If posthuman civilizations are likely to reach this stage, and if they are likely to run many such simulations, then the number of simulated beings with minds like ours would vastly exceed the number of nonsimulated beings. Unless you have reason to think you are special, you should conclude that you are almost certainly in a simulation.

The argument is usually presented as a trilemma: either (1) almost all civilizations at our level go extinct before reaching posthuman computing power, or (2) almost all posthuman civilizations decline to run ancestor simulations, or (3) we are almost certainly living in a simulation. Bostrom argues that we do not know which disjunct is true, but that if the first two are false, the third follows. The argument does not prove we are simulated; it shifts the burden of proof and forces us to take the hypothesis seriously.

Philosophers have responded in various ways. Hilary Putnam argued, in the related brain in a vat case, that the hypothesis is self-refuting: if the meaning of our words depends on causal contact with the things they refer to, then a being entirely inside a simulation could not use "simulation" to refer to simulations. David Chalmers responded in "The Matrix as Metaphysics" that even if we are simulated, the simulation is real for us: simulated minds can be genuinely conscious, simulated worlds can be genuinely real, and a scientific inquiry carried out inside a simulation can yield genuine knowledge of the simulated world. On this view, the simulation hypothesis changes less than it seems.

Historical Context

The simulation hypothesis belongs to a long skeptical tradition. Descartes opened the Meditations on First Philosophy with the hypothesis of an evil demon who deceives him about everything — the seventeenth-century ancestor of the simulation. Blaise Pascal speculated about the infinite regress of worlds within worlds. In the twentieth century, the hypothesis was renewed by the brain in a vat scenario made famous by Putnam, and by Wittgenstein's reflections on whether our forms of life could be systematically different. The popular imagination was captured by The Matrix (1999), and Bostrom's 2003 paper turned the cinematic idea into a rigorous philosophical argument. Chalmers' "The Matrix as Metaphysics" then showed how the hypothesis could be made compatible with realism about the world we experience.

Philosophical Significance

The simulation hypothesis matters because it is the most powerful contemporary form of radical skepticism about the external world. If we cannot rule it out, then a great deal of our ordinary confidence about reality is undermined — unless, with Chalmers, we hold that the simulation hypothesis is compatible with realism about the world as we find it. Either way, the hypothesis forces us to say what it is about our world that makes it real, and what would have to be true for our beliefs and values to remain legitimate.

The hypothesis also intersects with the philosophy of mind and artificial intelligence. If the universe is a simulation, then minds are computational processes, which lends support to functionalism about consciousness — the view that mental states are defined by their causal roles, not their physical substrate. And if simulated beings can be conscious, then the question of whether machines can be conscious gains a striking new instance: we ourselves might be the machines. The hypothesis thereby connects to the Turing test, the Chinese Room argument, and the ethics of artificial minds (AI ethics, machine ethics).

Examples & Thought Experiments

The ancestor simulation is the canonical example: a posthuman civilization runs billions of simulations of the lives of its ancestors, complete with simulated physics, biology, and psychology. Each simulated being has an inner life indistinguishable from ours. If you are such a being, nothing in your experience can reveal it — except, perhaps, if the simulation has "glitches" or computational artifacts that physicists could detect in principle.

The evil demon of Descartes is the same structure without the computer: a deceiver ensures that all your beliefs about the world are false while your experience remains coherent. The brain in a vat updates it with technology: your brain is kept alive in a vat and fed experiences by electrodes, and your beliefs about having a body are false. The simulation hypothesis is the maximal version — not just a brain, but the entire environment, is simulated.

The nested simulation is a further twist: if posthuman civilizations run simulations, those simulated civilizations might themselves reach posthuman power and run their own simulations, so simulations could contain simulations. The regress is dizzying, but Bostrom notes it only strengthens the argument: with many levels, the ratio of simulated to nonsimulated minds grows further. Chalmers' response, meanwhile, offers a deflationary thought: even if the regress is real, each level is a genuine world for its inhabitants, and the questions of ethics and meaning at each level are undiminished.

Related Concepts

• The brain in a vat
• Skepticism and knowledge and truth
• AI and consciousness and functionalism
• Consciousness and the hard problem
• The Turing test and the Chinese Room
• Metaphysics

Further Learning

• Nick Bostrom, Are You Living in a Computer Simulation? (2003)
• Stanford Encyclopedia of Philosophy, Skepticism
• Stanford Encyclopedia of Philosophy, Artificial Intelligence
• David Chalmers, "The Matrix as Metaphysics" (2005)