Popular Boards

Abstract

The Simulation Argument posed by Bostrom suggests that we may be living inside a sophisticated computer simulation. If posthuman civilizations eventually have both the capability and desire to generate such Bostrom-like simulations, then the number of simulated realities would greatly exceed the one base reality, ostensibly indicating a high probability that we do not live in said base reality. In this work, it is argued that since the hypothesis that such simulations are technically possible remains unproven, statistical calculations need to consider not just the number of state spaces, but the intrinsic model uncertainty. This is achievable through a Bayesian treatment of the problem, which is presented here. Using Bayesian model averaging, it is shown that the probability that we are sims is in fact less than 50%, tending towards that value in the limit of an infinite number of simulations. This result is broadly indifferent as to whether one conditions upon the fact that humanity has not yet birthed such simulations, or ignore it. As argued elsewhere, it is found that if humanity does start producing such simulations, then this would radically shift the odds and make it very probably we are in fact simulated.

Tags

Plain English Takeaway

This paper shows that, based on careful math, the chance we live in a computer simulation is at most about 50-50, not nearly certain as some people claim.

Study Aim

The main goal of this paper is to re-examine the Simulation Argument, which suggests we might be living in a computer simulation, using a more careful statistical approach. The author aims to show that previous arguments overstate the likelihood of us being simulated because they ignore uncertainty about whether such simulations are even possible. The paper uses Bayesian statistics (a method that updates beliefs based on evidence and uncertainty) to provide a more balanced estimate of the probability that we live in a simulation. Simply put: The paper wants to find out, using better math, how likely it really is that we live in a computer simulation.

Study Design

The author uses a Bayesian statistical framework (a way to update probabilities based on new information and uncertainty) to analyze the Simulation Argument. Two main hypotheses are compared: one where advanced civilizations create many simulated realities (the simulation hypothesis), and one where they do not (the physical hypothesis). The analysis models a hierarchy of simulations, where each simulated world could create its own simulations, and calculates the odds of being in a simulation versus base reality. The study also considers what happens if we start creating our own simulations. Simply put: The study uses a careful math method to compare the chances of living in a real world or a simulated one.

Findings

The research demonstrates that, when accounting for uncertainty about whether simulations are possible, the probability that we live in a simulation is less than 50%. This probability approaches 50% only if there are an extremely large number of simulations. The findings also show that if humanity ever starts creating realistic simulations of conscious beings, the odds would shift dramatically, making it very likely that we are simulated. Until then, the best estimate is that the chances are about even, not overwhelmingly in favor of being simulated as some popular claims suggest. The author recommends using Bayesian methods to properly account for uncertainty in such philosophical questions. Simply put: The study finds we are probably not living in a simulation, but if we ever make our own, that could change the odds a lot.

Referenced In

Mar 3, 2026 7:34 AM

Hm I don't 100% get the Kipping paper. The starting assumption of 50-50 seems too... arbitrary? Then again I'm definitely not up to speed on advanced probability theory/math! But I get the overall point that, "given we have not made any conscious simulations ourselves..." (yet!) lowers the chances that we are in a simulation.

Season 17, Episode 12: Are We Living in a Simulation?

Hey StarTalkers! Season 17, episode 12 sees Neil and Paul go through some cosmic queries with legendary mathematician Terence Tao. A reader question near the end of the episode opens up a potentially huge discussion: Do we live in a simulation?

Is the Universe a Math Problem? With Terence Tao - StarTalk Radio

(starts at 49:30)

They discuss the idea for a few minutes – and it’s been covered on StarTalk before – but as always with these “cosmic queries” discussions, there isn’t much time to go into detail.

So, what exactly is simulation theory and why do so many people seem to believe it? Let’s dig in.

Bostrom’s Paper

The modern form of the simulation hypothesis comes from a 2003 paper by Nick Bostrom .

First, he argues that it should be possible to simulate consciousness. Bostrom is talking sci-fi more than “real life,” rooted in known science but pushed to the extreme. If we had a computer the size of a planet, he concludes that we could simulate “the entire mental history of mankind.”

In other words, we could simulate every human brain to ever exist.

The Three Possibilities

This leads to three possibilities:

  1. No human-level civilizations reach this technological level. Either large-scale disaster (man-made or not) or technological stagnation could prevent this from happening.

  2. No human-level civilization at this tech level is interested in making such simulations. After developing to this level, human-level civilizations could simply lose interest in making these simulations. Maybe they’re too advanced to benefit, or think it’s cruel to simulate human-like minds.

  3. We are almost certainly living in a simulation. If 1) and 2) aren’t true, then it’s likely there are many simulations. If there are simulations within simulations, the number becomes truly huge (see image below). At this point, a simple count of “real” vs. simulated realities makes it almost certain we’re in a simulation.

Improving the Estimate

Bostrom basically compares numbers (real vs. simulated) to make his famous argument. But there’s a slightly improved, Bayesian calculation that just puts us at just over 50% odds of living in a true reality. One key difference is assuming some limit to simulations within simulations – eventually, computing power runs out.

So toss a coin. If you call it right, maybe we’re real.

4