Involves

Idea

Bostrom's trilemma

Bostrom argued that at least one of three statements must be true. First, civilizations almost always go extinct before they can run realistic ancestor-simulations. Second, advanced civilizations that could run such simulations almost never choose to. Third, we are almost certainly living in a simulation. The logic: if many civilizations run vast numbers of detailed simulated minds, then simulated observers would vastly outnumber real ones, so a randomly chosen observer, like you, would most likely be simulated. To escape that conclusion, you must accept option one or two.

This is the structure of Bostrom's 2003 paper. Its cleverness is that it does not claim we are in a simulation; it shows the three options are jointly exhaustive, forcing you to reject something you might have assumed: that technological civilizations survive, that they would run ancestor-simulations, and that we are not in one.

The argument doesn't say we are simulated. It says you must reject at least one comfortable assumption.

A good argument can trap you between options without telling you which one to pick.

Where it breaks down

The trilemma leans on contested assumptions: that consciousness can be simulated computationally at all, and that we can meaningfully assign probabilities across hypothetical observers. Many philosophers and physicists regard the hypothesis as untestable and therefore outside science; the argument is a piece of careful reasoning, not an empirical finding.

The enduring value is the method, not the verdict: reframing an unanswerable yes-or-no question as a set of exhaustive options is a reasoning trick that clarifies many debates far removed from simulated worlds.

How it connects

Bostrom's move was to stop asking “are we simulated?” and prove that at least one of three uncomfortable options must be true.

This idea appeared in Is this base reality?, the Involves connection for August 29, 2026, which asked: How could you tell whether the world around you is real?

Check yourself

What does Bostrom's simulation argument actually conclude?

At least one of three propositions is true, one of which is that we are almost certainly simulated.. Right. The argument is a trilemma: it forces one of three options, not a flat claim that we are simulated.

What the sources establish

  • Bostrom's argument concludes that at least one of three propositions holds: near-extinction before simulation capability, near-universal disinterest in running ancestor-simulations, or that we are almost certainly simulated.

Sources