What if gravity, a sunset, and the equations that describe them are lines of code? The simulation hypothesis has a rigorous 2003 philosophical formulation behind it, an actual Bayesian analysis putting the odds at close to a coin flip, and a striking, well-documented resemblance to religious ideas thousands of years older than computing itself. It also has What if gravity, a sunset, and the equations that describe them are lines of code?
The simulation hypothesis has a rigorous 2003 philosophical formulation behind it, an actual Bayesian analysis putting the odds at close to a coin flip, and a striking, well-documented resemblance to religious ideas thousands of years older than computing itself. It also has at least one widely repeated quote that turns out to be praise from a close colleague, not a neutral verdict, worth separating from the parts of the argument that actually hold up.
Bostrom’s Actual Argument
The Matrix gave the idea its cultural moment in 1999, but the rigorous version belongs to Oxford philosopher Nick Bostrom, who laid it out in a 2003 paper. Bostrom’s argument isn’t a claim that we’re definitely simulated. It’s a trilemma: either civilizations almost always go extinct before they can build convincing reality-simulations, or advanced civilizations that could build them choose not to, or we are almost certainly living in one right now, since a civilization capable of running such simulations would likely run vast numbers of them, making simulated observers vastly more numerous than the original, “base reality” ones. At least one of those three has to be true, Bostrom argued, though the paper itself doesn’t say which.
In 2020, Columbia University astronomer David Kipping put actual numbers to the question, applying Bayesian statistics to Bostrom’s trilemma in a paper published in the journal Universe. Kipping’s analysis, assigning a neutral 50/50 prior to whether simulation technology is even achievable, landed on odds of roughly 50.2 percent for base reality versus 49.8 percent for simulation, close enough to call a coin flip. Kipping added an important caveat: if humans ourselves ever succeed in building a convincing ancestor-simulation, that achievement would immediately collapse the odds in the other direction, since it would prove the technology is achievable, leaving almost no weight on the “it can’t be done” branch of the trilemma.
The Famous Quote Needs Context
Transhumanist philosopher David Pearce, not “Peirce,” did call Bostrom’s simulation argument “perhaps the first interesting argument for the existence of a Creator in 2000 years.” That’s accurate, and worth quoting precisely rather than paraphrasing “argument” into “evidence,” a small but meaningful shift toward overstating what Pearce actually claimed. It’s also worth knowing who’s speaking: Pearce co-founded the World Transhumanist Association with Bostrom himself, making the line a close colleague’s compliment rather than an outside philosopher’s independent assessment. Bostrom’s own official website hosting the paper lists the quote with a parenthetical of its own: “(exaggerated compliment).” Even the argument’s author treats the line as generous praise, not a serious theological verdict, and it’s worth reading it the same way.
The Rendering Problem
A genuine objection to the hypothesis, raised in papers as recently as 2025, is computational: rendering every particle and quantum fluctuation in an entire universe with perfect fidelity would require processing power that seems implausible even for a civilization vastly more advanced than ours. The standard rebuttal within the simulation-hypothesis literature is that a sufficiently efficient simulation wouldn’t need to render everything at once, only the parts currently being observed, an economization strategy familiar to anyone who’s played a video game with a limited draw distance. Whether that’s a satisfying answer or simply pushes the objection back a step remains genuinely disputed among people who take the hypothesis seriously.
A Genuinely Old Idea in New Language
Where this gets interesting has less to do with computer science than with how closely the simulation hypothesis echoes ideas that predate computing by millennia. Hindu and Buddhist traditions describe the physical world as maya, illusion, a dream from which awakening is possible, language the Samadhiraja Sutra renders directly: “all things are alike, a mirage, a castle on a cloud, dream, vision, devoid of essence.” Paramahansa Yogananda, who introduced meditation and yoga to Western audiences in the 1920s, reached for the newest technology of his own era, cinema, to explain the same concept, comparing ordinary perception to actors absorbed in a film. The Persian poet Rumi described the body as clothing the spirit puts on and later removes, and the Bhagavad Gita uses close to identical language: the soul discards worn-out bodies “as a man puts on new clothes when he takes off his old clothes.” None of these traditions anticipated computing. What they share with the simulation hypothesis is a specific philosophical structure, ordinary perceived reality as a layer sitting on top of a more fundamental, less accessible truth, arrived at independently and centuries apart.
Google DeepMind’s Genie 3, an actual AI system released August 5, 2025, adds a strange, literal footnote to this parallel. The model generates navigable, physics-consistent virtual environments directly from a text prompt, rendered live and explorable for several minutes at a stretch. It’s a working system today, not a hypothetical, and it makes the “world spoken into being” language of Genesis feel, for the first time, like a description of an actual (if far more limited) technical process rather than pure metaphor.
Where the Odds Actually Sit
None of this settles the question, and it isn’t supposed to. Bostrom’s trilemma is a serious piece of philosophy, Kipping’s Bayesian treatment is a serious piece of applied statistics, and neither one proves anything about a creator, a programmer, or an underlying “base reality.” What the exercise does show clearly is how naturally this specific modern argument maps onto a philosophical shape that Hindu, Buddhist, and Sufi thinkers worked out independently, centuries before anyone had a word for computation. Whether that convergence means something deep about the structure of reality, or simply that humans have always reached for the best available metaphor to describe the sense that appearances aren’t the whole story, is very much still an open question, one Kipping’s coin-flip odds aren’t going to resolve on their own.
The actual CIA Gateway Process document sits in the same evidentiary category, real, declassified, and inconclusive.
