In the inverted world, the burden of proof flips. The smug position is not 'we're obviously in base reality.' It's the opposite. In 2003, the Oxford philosopher Nick Bostrom published a peer-reviewed paper in The Philosophical Quarterly arguing that, under a few plausible assumptions, the rational bet is that you are a simulation, and the person insisting they're a flesh-and-blood original is the one who owes you an argument.

Here is the actual structure, because it is more disciplined than the meme version. Bostrom does not claim we are in a simulation. He proves a trilemma: at least one of three propositions must be true. One, almost all civilizations at our level go extinct before reaching technological maturity (the 'posthuman' stage). Two, almost all mature civilizations that could run vast numbers of ancestor-simulations choose not to. Three, we are almost certainly living in a computer simulation. The logic is a disjunction, not a prophecy. You are free to pick door one or door two to escape door three, but you cannot reject all three at once.

The engine driving it is a counting argument and a 'bland' principle of indifference. Suppose a mature civilization can run ancestor-simulations, conscious historical re-creations of beings like us. The computing power required, Bostrom estimates from physical limits, is trivial for such a civilization; a single planetary-scale computer could host astronomically many simulated lives. So if even a tiny fraction of mature civilizations run even a few such simulations, the number of simulated experiences like yours swamps the number of original biological ones by orders of magnitude. If almost all minds with your exact experiences are simulated, then absent a special reason to think you're in the rare biological minority, you should assign that same low probability to being biological. That is the whole move, and it is valid.

The proof, importantly, is in the published mathematics, not in any spooky 'glitch in the Matrix' anecdote. The paper is in the Oxford University Research Archive and on Bostrom's own simulation-argument.com, with the formal expression of the fraction of all observers who are simulated. This is not a TED-talk hunch; it cleared peer review and has been picked apart by other philosophers for two decades. The standard response is not 'that's silly,' but a serious debate over the indifference principle and over Bostrom's later patch addressing it.

The argument escaped the journals and went mainstream when figures like Elon Musk began citing it, declaring at conferences that the odds we are in 'base reality' are 'one in billions.' That is a sloppier, more confident claim than Bostrom's, and worth separating from it. Bostrom himself has consistently refused to say which of the three doors is shut. His position is genuinely agnostic: maybe we go extinct, maybe posthumans find ancestor-simulations boring or unethical, maybe we're code. Pick one.

The honest skeptical reading is that the trilemma is only as strong as its premises, and the load-bearing one is that consciousness is substrate-independent, that a sufficiently detailed simulation of a mind would actually have experiences rather than merely model them. If minds cannot be computed, the whole counting argument never starts. There are also live objections about whether a simulation could ever be detailed enough to fool its own physicists without simply being a universe. None of these have been settled.

So the unresolved question is not really 'are we simulated.' It's which exit you're willing to take. To insist you are flesh in a base reality, you must commit to believing either that civilizations like ours almost never survive, or that the technically capable ones almost never bother to run histories like the one you're reading. Each of those is a strong empirical claim about the future of intelligence. Bostrom's quiet, unsettling point is that 'I'm obviously real' was never the safe default. It was always a bet, and one you now have to defend.