The Argument

a four-book life · book no. 1 · the argument

Almost everyone gets it wrong. The "simulation argument" does not claim we live in a simulation. It claims something more careful and more interesting: that at least one of three propositions must be true — and only the third is "we're simulated." This book states the argument exactly.

The trilemma — at least one is true

Extinction first

Almost no civilizations at our stage survive to reach a "posthuman" level capable of vast computing.

we don't make it

No interest

Almost no posthuman civilizations bother to run many "ancestor-simulations."

they don't bother

We're simulated

Then we are almost certainly living in a computer simulation.

the hypothesis

The key distinction

The whole trilemma — "at least one of three." This is what's defensible.

the disjunction

The hypothesis

Only proposition 3 alone. A much stronger, separate claim.

just leg 3

Not a verdict

Bostrom does not claim to know which leg is true.

no answer

The sting

You can't comfortably reject all three at once.

the bite

How the argument works

Substrate-independence

The starting assumption: consciousness depends on how information is processed, not on whether it runs on neurons or silicon.

premise minds could run on computation

so a detailed enough simulation could house real experience.

+1 if you deny this, the whole argument loosens — it's the keystone assumption, not a proven fact.

Ancestor-simulations

A "posthuman" civilization with enormous computing power could simulate its own evolutionary history — minds and all.

idea high-fidelity simulations of beings like us

so simulated people with experiences like ours could exist.

+1 Bostrom carefully says they could — "may run," not "will run." The possibility is the hinge, not a prediction.

The simulated swamp the real

If such simulations run at all, simulated observers would vastly outnumber the original biological ones.

result sims ≫ non-sims

so a random observer is far more likely simulated.

+1 one civilization could run astronomically many simulated histories — the count tilts overwhelmingly.

The indifference principle

Lacking a reason to think otherwise, you should reason as if you're a random sample from all such observers.

step treat yourself as typical

so if most observers are simulated, probably you are.

+1 this anthropic step is doing real work — and is one of the most debated parts (Book 3).

What it actually concludes

A disjunction, not a claim

The conclusion is only that one of the three legs holds — not which one.

form (1) OR (2) OR (3)

so "we're simulated" is just one branch, not the verdict.

+1 as Bostrom puts it, it does not directly argue that we live in a simulation — a point endlessly lost in headlines.

The uncomfortable corner

The bite is that rejecting legs 1 and 2 forces you into leg 3.

logic deny extinction + deny disinterest ⇒ simulated

so optimism about our future has a strange price.

+1 believe we'll become simulators and survive to do it? Then you should think we're probably simulated now.

Base reality & nesting

Simulations could run simulations, stacking levels — but all rest on one physical "base reality."

terms nested sims, "basement level"

so the idea naturally allows worlds within worlds.

+1 this is exactly the ephemeral-dot's "−0.01 layer" — a level index below your own, now given rigor.

Not The Matrix

Bostrom is explicit that the movie's "humans as batteries" is silly; his scenario is software, not captivity.

contrast a serious line, not sci-fi

so the argument stands on logic, not spectacle.

+1 in a simulation there'd be no "tank" to wake from — you'd be the software, not a body plugged in.

The argument, exactly

The simulation argument is a trilemma: at least one of three propositions is almost certainly true — not the bare claim that we're simulated.
(1) nearly all civilizations go extinct before a posthuman stage; (2) nearly no posthuman civilizations run ancestor-simulations; (3) we are almost certainly simulated.
It rests on substrate-independence (minds can run on computation) and an indifference principle (reason as a random observer) — both assumptions, not proofs.
If simulations run at all, the simulated vastly outnumber the original beings — so a typical observer is probably simulated.
The sting: rejecting (1) and (2) forces you into (3). Only proposition 3 alone is the "simulation hypothesis."

What's valid — and what's assumed

Logically valid given its premises: the disjunction follows. The honest action is in the premises, not the deduction.
Substrate-independence is an assumption — if consciousness needs biology (or something non-computational), the argument weakens sharply. It is not established science.
The indifference principle is contested: whether you may treat yourself as a random sample from "all observers" is a real philosophical fight (Book 3).
The argument is often flattened to "we're in a simulation" — which is not what it says. Hold the difference between the argument (all three) and the hypothesis (leg 3).
This is the argument; Book 2 traces the ancient lineage it descends from, and Book 3 the objections and afterlife.

nick bostrom · book no. 1 · a trilemma, not a verdict · the argument, stated exactly