5 ms·
If we're in a simulation, you have no basis for drawing conclusions about any of our observations, whether those observations have taken place, whether anything
by vidarh 5y ago
If we're in a simulation, you have no basis for drawing conclusions about any of our observations, whether those observations have taken place, whether anything outside of this moment exists, or anything outside your mind, or both.
The notion that non-determinism can be proven, exists, and/or has a relevance to a simulation all make assumptions about reality that either won't or don't have to hold if you're in a simulation.
- Ieghaehia9 5y ago> If we're in a simulation, you have no basis for drawing conclusions about any of our observations, whether those observations have taken place, whether anything outside of this moment exists, or anything outside your mind, or both. But then simulation is unfalsifiable. Whatever you observe, the simulation could have paused an arbitrary number of times, or modified you or the environment to an arbitrary degree, hence it's compatible with simulation. Which makes simulation "a difference that makes no difference".
- macguillicuddy 5y agoWithout wanting to speak on their behalf, I think that's exactly vidarh's point. That simulation is indeed unfalsifiable and therefore mjburgess' conjecture that it's false doesn't hold.
- WJW 5y agoSimulation being unfalsifiable is indeed the main problem with the theory.
- klyrs 5y agoA theory known to be unfalsifiable cannot be conclusively ruled "false." It can only be deemed outside of the realm of science. To make such a strong claim puts one back in the realm of faith in pseudoscience. I like the word "bullshit." Bullshit isn't necessarily true, isn't necessarily false, it's simply not worth my time to wade into it.
- GoblinSlayer 5y agoThe simulation theory rests on a hypothesis that there are certain conditions outside of simulation, but if external conditions are certain, then the simulation has limited flexibility and is thus falsifiable. For it to be unfalsifiable external conditions must be at least unfathomable, which takes away the premise of the simulation theory that it correctly guessed external conditions.
- vidarh 5y agoIs sounds like you're discussing a specific simulation hypothesis. I'm assuming you're referring to Bostrom's Simulation Argument, but that is only one possible variant of simulation. Bostrom's Simulation Argument posits that one of three propositions are likely to be true, of which only one posits that we're likely in a simulation. The "simulation proposition" of the argument could be falsified by e.g. humanity going extinct (proving the first proposition of the argument true, but the 3rd - the simulation one - to be false). But falsifying that proposition would not falsify the notion of simulation.
- vidarh 5y agoAs macguillicuddy pointed out, the point was that the general concept of simulation is indeed unfalsifiable, and we can only ever hope to falsify very limited subsets. E.g. we can certainly say certain things about what features a simulation must have. It does however not necessarily make a simulation a difference that makes no difference, because if we're in a simulation it is possible that we are in one where it is possible to falsify the theory that we're not in a simulation. E.g. we could potentially find bugs that reveal telltale signs we're in a simulation, or outright vulnerabilities. It's even possible we could finds ways of "escaping" the simulation. The problem of course being that you could well devote a lifetime to it and find nothing and it would tell you nothing about whether we're in a simulation or not, so unless you run into some anomaly that hints at it, it'd seem a rather wasteful pursuit. In reality, of course, chances are none of us will ever see anything to give us reason to pursue that idea, but it's fun to think about. I write short stories about this subject, and I have a long document with headache-inducing scenarios to write up.
- GoblinSlayer 5y agoIf simulation is unfalsifiable, then it's also unjustifiable.
- vidarh 5y agoFalsifiability does not work that way. To avoid hypotheses that are unfalsifiable is good practice absent evidence because you can construct an infinite number of such hypotheses. That does not mean that a given unfalsifiable hypothesis can not be true, however, just that pursuing it is generally likely to be a waste of effort. Asserting that simulation is definitely true absent evidence would be unjustified, but very few people would make that claim, so it's not very relevant. However the inverse - the hypothesis that we're not in a simulation - is falsifiable by proving that we are in a simulation. And any number of variations over the idea that we might be in a simulation can also be testable and falsifiable in various ways. The original point of contention was that QM and non-determinism proves we're not in a simulation, however, and that is not a viable way to falsify the hypothesis that we're not in a simulation. The hypothesis that non-determinism proves we're not in a simulation is falsifiable by coming up with a way wherein non-determinism can be accommodated. If the world the simulator runs in has access to non-determinism this is trivially done by simply forwarding non-determinism into the simulation (many other approaches to attack that hypothesis are possible, e.g. using a additional simulations as oracles to determine if supposedly non-deterministic events would be revealed as such, and use that to pick values that ensures a failure to simulate non-determinism never gets detected, so access to a source of non-determinism is not necessary to falsify that hypothesis either, though it would complicate matters). As such the hypothesis that non-determinism alone proves we're not in a simulation is false.
- mjburgess 5y agoI'm taking "being a simulation" to mean something like "implementable by a universal Turing machine". If "simulation" is just an alleged property everything is meant to have (ie., sim = "seeming-as-if real but actually being a... dream, ..."). Then the argument against it is simply pragmatic: rather than add "seeming-as-if" to everything we say, drop it. Adding it can make no difference, which is stipulated by this definition of "simulation". You could also make a kind of rationality argument: consider the family of all such seeming-as-if properties, say S; and the family of all real properties R. Given conditioning on S essentially makes all claims false, eg., P(I am a human awake on earth | S) = 0 Ditching the S properties is equivalent, essentially, to reasoning being possible. So if we are in a simulation essentially every judgement we make is false. So to beleieve we're in a simulation, ie., to condition on being in one, prevents any thought/reasoning/etc. To actually believe we're in a simulation, in this sense, you'd be completely incapable of essentially any thought wahtsoever.
- vidarh 5y agoWell this is basically the philosophical argument between George Berkeley's subjective idealism[1] vs. philosophical materialism (though Berkeley as far as I remember - it's been a couple of decades - at least thought minds were real, and if in a simulation that is also not a given), so it's a well trodden area of philosophy going back to the 18th century, with parts of the ideas dating back much further. You're right that since we can't falsify the general idea of simulation, it makes sense to simply discount it in most discourse, because we just have to act as if the world is real. (EDIT: I'll add that "implementable by a universal Turing machine" or not does not change any of this in either direction) This is the most straightforward rejection of Berkeley's idealism as well. But that is different from discounting the possibility. Because while we can't falsify the hypothesis that we're in a simulation, as I've expanded on in another comment too, if we are in one there's at least a possibility that we are in one where the hypothesis that we are not in one can be falsified (e.g. a doorway could open in front of me to let me "exit" the simulation; as wildly unlikely as it is; or we could travel into space and hit a literal wall, or an infinite number of other things) Going further, it's in particular different from discounting the possibility because of details of the world, because all of what we think we know is a product of processes that if we are in a simulation could have been manipulated or outright manufactured. So you're right that in our day to day lives it makes sense to just not worry about whether we are in a simulation or not and act as if we're not. I don't agree that a belief in one would affect our ability to reason, though, as we are good at compartmentalising and saying exactly what we're saying here: As long as / until we can tell we're in a simulation, you just go on acting as if you're not. Whether you believe you're in one or not doesn't really make a difference as long as you have no evidence either way. [1] https://philosophybreak.com/articles/george-berkeley-subjective-idealism-the-world-is-in-our-minds/ https://philosophybreak.com/articles/george-berkeley-subject...