There's a very simple objection to the simulation hypothesis that I came up with, and I've read others come up with this same idea. So I assume it's something philosophers should already be aware of.

There are 2 possibilities:

  1. We're in a simulation. In which case we cannot infer anything about the world or worlds above our own (be it also a simulated world or non-simulated real world). We can only make inferences about simulations that will be constructed within our own world.

  2. We're not a in a simulation. In which case there is no higher level world to make inferences about. We can only make inferences about simulations we may construct within our world.

Either way we can only make inferences about simulations constructed 'within' our world. We can't make any inferences about whether there's a world above our own. So all of Bostrom's arguments tell us nothing about the probability that our world is itself a simulation or not. All of Bostrom's references to worlds/civilizations have to refer to world/civilizations within our own. So when he calculates probabilities he's restricted to worlds under our own.

So Bostrom's paper is a bunch of elaborate reasoning about how worlds constructed within our world are probably simulations. It's really a bunch of useless reasoning, because worlds under our own are by definition simulations.

What is the response from philosophers to this objection?

Here's a link to his paper:


  • Are you familiar with Bostrom's "self-sampling assumption", a form of anthropic reasoning? The simulation argument is mainly just saying that given that assumption, we should take it as very unlikely that both of the following claims are true: 1) many civilizations in our universe reach some kind of transhuman breakout point that results in them performing huge numbers of ancestor simulations, so that simulated beings in such ancestor sims vastly outnumber real beings at a comparable level of technological development, and 2) we are not simulated beings of this type. – Hypnosifl Apr 18 at 14:28
  • @Hypnosifl, no I'm not familiar with that. Does he explicitly mention it? If he's making that assumption, isn't it just begging the question? – Ameet Sharma Apr 18 at 14:33
  • The self-sampling assumption doesn't refer to the statements 1) and 2) I mentioned, rather it's an idea used in deriving the conclusion that 1) and 2) are unlikely to both be true. I tried to summarize the self-sampling assumption and some of its implications in this answer, and if you want to get into more details you can look at Bostrom's book Anthropic Bias which is posted on his site. – Hypnosifl Apr 18 at 14:37
  • @Hypnosifl, ok. Thanks. I see what you mean. So the question is the validity of the assumption. – Ameet Sharma Apr 18 at 14:44
  • 1
    Bostrom says at 1:15:21 of this podcast interview that the bland indifference principle is basically a weaker version of the self-sampling assumption, they are both part of his project of "trying to figure out what are the legitimate ways of reasoning about these indexical facts when observer selection effects are in play, in other words, developing a theory of anthropics." – Hypnosifl Apr 18 at 16:20

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.