The Wachowskis' The Matrix (1999) drew on Descartes' evil demon, Plato's cave, and Jean Baudrillard's Simulacra and Simulation to ask a deeply unsettling question: how would we know if everything we experience is a computer-generated illusion? It is one of philosophy's oldest puzzles dressed in cyberpunk clothing — and since 2003, it has had a rigorous mathematical formulation thanks to philosopher Nick Bostrom.
The Simulation Argument
Bostrom's simulation argument, published in the journal Philosophical Quarterly, is not a claim that we live in a simulation. It is a logical trilemma: at least one of three propositions must be true.
Bostrom's Trilemma (2003)
- Almost all civilisations go extinct before reaching the technological maturity to run detailed simulations of minds.
- Almost all technologically mature civilisations choose not to run such simulations — perhaps for ethical or resource reasons.
- We are almost certainly living in a computer simulation run by a post-human civilisation.
The argument's logic is probabilistic. If civilisations routinely reach the ability to simulate ancestor minds and choose to do so, then simulated minds will vastly outnumber biological ones — and any given observer is far more likely to be a simulated mind than a biological one. The only escapes are that civilisations regularly fail before reaching this capability (option 1) or that they deliberately refrain (option 2). If neither applies, option 3 follows as a statistical necessity.
Bostrom himself has stated he finds it roughly equally probable that each of the three options is true. Elon Musk famously claimed the probability we are in a base reality is "one in billions." Most physicists are considerably more sceptical.
What Physics Says: The Computational Cost Problem
The deepest objection to the simulation hypothesis comes from physics itself. Simulating a universe requires computing resources — and the universe is extraordinarily complex. Physicist Sabine Hossenfelder has pointed out that simulating quantum mechanics faithfully requires computational resources that scale exponentially with the number of particles. Our observable universe contains roughly 1080 atoms, each interacting quantum mechanically with its neighbours.
The Bekenstein bound — a result from thermodynamics and quantum gravity — places an absolute upper limit on the information content of any physical region, proportional to its surface area. Any computer capable of simulating a region of spacetime would need to contain more information than that region itself. This seems to rule out a simulation inside the same kind of physics as our universe.
A simulation could sidestep this by only rendering what is being "observed" — the quantum measurement problem's wave-function collapse could, on a sufficiently creative reading, look like lazy evaluation in a game engine. This is speculative, but it has been taken seriously enough that physicists at the University of Washington proposed a way to test it: look for signatures of a computational lattice at the smallest scales of spacetime. No such signatures have been found.
Emergence from simple rules
The Cellular Automata simulation shows how extraordinarily complex behaviours emerge from simple computational rules — the same class of phenomena that makes the simulation hypothesis intriguing. Conway's Game of Life produces apparent "life" from just four rules.
The Discreteness Clue
One feature of physics that simulation proponents find suggestive is quantisation. The universe appears to have a smallest length scale — the Planck length, roughly 10-35 metres — below which the concept of distance may not be meaningful. Energy comes in discrete quanta. Spacetime may itself be discrete at the Planck scale. These features look, superficially, like the pixel resolution and minimum time step of a simulation.
Physicists caution against reading too much into this. Quantisation arises naturally from quantum mechanics and has precise mathematical derivations that do not require computation. But the analogy is aesthetically compelling, and it has inspired research programmes like loop quantum gravity that treat spacetime as a network of discrete quanta rather than a continuous manifold.
The speed of light as an absolute maximum velocity could be analogised to a maximum data transfer rate. The fine-tuning of physical constants for a universe hospitable to complex chemistry could look like deliberate parameter-setting. These arguments are not evidence, but they keep the simulation hypothesis alive as a philosophical possibility rather than disposing of it immediately.
The Matrix's Physics Is Wrong — But Its Philosophy Is Real
The film's actual mechanics are scientifically incoherent. Using humans as a power source generates less energy than the food needed to grow them — the machines would be better off burning the food directly. The rules Neo learns to bend (gravity, momentum) are arbitrary and inconsistent. The 1999 computer aesthetic of the Matrix's "code" has nothing to do with how actual simulations work.
But the film's philosophical framing is legitimate. Descartes' original argument — that an evil demon could be feeding you false sensory experiences — is logically coherent. You cannot rule it out from the inside. The philosophical tradition of brain-in-a-vat thought experiments and scepticism about the external world predates computing by centuries. The Matrix gave those ideas mass-market exposure and a visual grammar that has made them permanently accessible.
See a network learn in real time
Our Neural Network simulation shows how simple computational units — like neurons — can produce surprisingly sophisticated behaviour through layered connections. The question of whether sufficiently complex networks could produce consciousness is genuinely open.