The simulation argument and the Boltzmann brain paradox
The simulation argument:
The Boltzmann brain paradox:
Two different arguments that try to show that a "brain in a vat"-type situation is more likely than a common sense realism situation. Unlike traditional skeptical arguments, these don't simply try to argue that a situation like this is possible, but that we could have more reason to believe that a situation like this is in fact the case.
Any thoughts?
1. The fraction of human-level civilizations that reach a posthuman stage (that is, one capable of running high-fidelity ancestor simulations) is very close to zero, or
2. The fraction of posthuman civilizations that are interested in running ancestor-simulations is very close to zero, or
3. The fraction of all people with our kind of experiences that are living in a simulation is very close to one
...
Bostrom goes on to use a type of anthropic reasoning to claim that, if the third proposition is the one of those three that is true, and almost all people with our kind of experiences live in simulations, then we are almost certainly living in a simulation.
The Boltzmann brain paradox:
Boltzmann proposed that the state of our observed low-entropy universe (which includes our existence) is a random fluctuation in a higher-entropy universe. Even in a near-equilibrium state, there will be stochastic fluctuations in the state of the system. The most common fluctuations will be relatively small, resulting in only small amounts of organization, while larger fluctuations and their resulting greater levels of organization will be comparatively more rare. Large fluctuations would be almost inconceivably rare, but inevitably occur if a universe lasts infinitely long. Even if the universe does not have an infinitely long past, modern cosmological theories of the Big Bang do suppose that the latter occurred via stochastic fluctuations in a larger meta-universe; the paradox is retained by incorporating our brief-but-finite past into the random fluctuation.
Furthermore, there is a "selection bias": we observe our very unlikely universe because those unlikely conditions are necessary for us to be here. This is an expression of the anthropic principle.
If our current level of organization, having many self-aware entities, is a result of a random fluctuation, it is much less likely than a level of organization that only creates stand-alone self-aware entities. The number of self-aware brains that spontaneously and randomly form out of the chaos, complete with memories of a life like ours, should vastly outnumber the brains evolved from an inconceivably rare local fluctuation the size of the observable universe.
Two different arguments that try to show that a "brain in a vat"-type situation is more likely than a common sense realism situation. Unlike traditional skeptical arguments, these don't simply try to argue that a situation like this is possible, but that we could have more reason to believe that a situation like this is in fact the case.
Any thoughts?
Comments (66)
The problem is in a sense that our reasoning and beliefs are the same for a common sense universe and the two other possibilities. In other words we have no ability to decide what is the case, because our reasoning and beliefs are the same in all variants.
So it leads to the 3rd answer as being the most reasonable:
1: We live in a common sense universe.
2. We live in a non- common sense universe.
3. We don't know.
I can't remember the details but I think I came to the conclusion that the 'two-thirds' position in that problem uses the same sort of argument as Bostrom, and that the argument makes unjustified assumptions. The problem is in how one determines what are 'equally likely events'. If one just assumes that whatever events one is talking about are equally likely then one can change the probabilities just by subdividing events and declaring the subdivided events equally likely to the un-subdivided alternative events. eg subdivide the event 'I am in a simulation' into events E(1), E(2), E(3) etc where E(k) is the event 'I am in a simulation and the program has between k thousand and (k+1) thousand lines of code'.
If I can recall my thoughts a bit more clearly I'll try to post them here because it was a bit of an epiphany for me. Up till then I had considered Bostrom's argument strong and could not point to a reason to reject it, even though I felt intuitively it was flawed. But I think I managed to convince myself that there was an explicit flaw one could point to, along the lines of my rather vague preceding paragraph.
I'm a 'halfer', by the way. Or at least I recall that I was, the last time that I understood the question.
Given 3., "... if even a tiny percentage of them were to run 'ancestor simulations' (that is, 'high-fidelity' simulations of ancestral life that would be indistinguishable from reality to the simulated ancestor), the total number of simulated ancestors, or 'Sims', in the universe (or multiverse, if it exists) would greatly exceed the total number of actual ancestors."
If there are more Sims than there are non-Sims then we're more likely to be Sims.
I suppose the only "out" here is to claim that there wouldn't be enough computing power to simulate that many ancestors, and so that the number of actual ancestors is still greater.
Or you could reject a physicalist account of consciousness and argue that simulated consciousness is impossible.
There are vastly more Boltzmann brain scenarios that are inconsistent than there are that are consistent. Tegmark suggests for example, if our memories recall a famous piece of music as white noise then we might well have reason to believe we are a BB. Otherwise, not.
Sure, but then there are vastly more Boltzmann brain scenarios that are consistent than there are non-Boltzmann brain scenarios (or at least that the former are more likely).
Given the above, I don't think so. We only have reason to believe that we are not in an inconsistent Boltzmann brain scenario.
Given you agree that an inconsistent BB is far more likely than a consistent one, we should expect to live in an inconsistent one rather than a consistent one, regardless of whether BBs outnumber non BBs.
Sure, but we can see that we don't. So we can rule out that possibility. That then leaves us with a consistent BB or no BB, with the former more likely.
And there's the rub ....
I think you would struggle to show that there are more consistent BB scenarios than there are suitable variant universes of all the physical constants, say....
That's what Boltzmann does: "The number of self-aware brains that spontaneously and randomly form out of the chaos, complete with memories of a life like ours, should vastly outnumber the brains evolved from an inconceivably rare local fluctuation the size of the observable universe."
His argument (from physics) isn't just that spontaneous brains are more likely, but that spontaneous brains with memories of a life like ours are more likely.
yeah, but that's titchy!
Is this spontaneity a form of last Thursdayism? Also, the spontaneous brains would involve the formation of an environment I can survive in long enough to remember, and will likely involve interaction with other people and technology, such as what I'm using right now.
You could reject that consciousness could be emulated (not just simulated where the Sims act as if they were conscious) and still be a physicalist about it. Physicalism doesn't commit one to functionalism about subjectivity. Searle would probably reject the simulation argument, for example.
They say nothing about doubt or knowledge. They only say that it's more likely that we're simulations or Boltzmann brains. Doubt may follow from recognising this, but that's just a side effect of the arguments.
Regarding the Boltzmann brain paradox, that it depends on a posteriori premises is probably why it's considered a paradox. If our knowledge of the external world is genuine then it's more likely that we're Boltzmann brains.
And what a posteriori premise(s) does the simulation argument use?
But what does a 'simulation' mean? Doesn't it entail that all your experiences are illusory?
Yes, that is the point for all aposteriori premises including your version, that we can trust our experiences. They are all begging the question and so are you. All claims to what the world really is as far as claims of knowledge beg the question. They all show that the trust in reason and the experience are limited.
It is funny though that you can't spot the problem in your own claim to knowledge.
If all your experiences are illusory, are they then real and how the world works?
Where did I make such claims?
So what the OP is telling us is that these events are boring to us but meaningful and enjoyable to a post-human civilization? Why would the post-human civilization have simulations inside simulations to pass the time in the boring parts of the first simulation? That doesn't sound very efficient to me.
But whether a Jupiter-sized computer could actually be built and perform such incredible feats is pure speculation. And whether an advanced civilization would be motivated to build it, who knows.
You claim by extension of your demands for logic and e.g. coherence in regards to knowledge, that knowledge is possible with logic and e.g. coherence, unless your point is that there is no knowledge, which abides to logic.
So I suspect that you believe that there is such a sensible concept as knowledge and that it must abide to reason, logic and so on.
So if knowledge must abide to reason, logic and so on, then that is also the case for your claims to knowledge. So what do you claim about knowledge?
That there is a world, which is non-solipsistic and for which our experiences (in general) match the world. The problem is that is begging the question and thus not logical.
So for once, try to give reason, logic and evidence for the fact, that there is a screen in front of you, which you read this on and don't follow into the following trap:
It is absurd, if we can't trust our senses, therefore we can trust our senses.
P1: We can't trust our senses.
P2. That is absurd.
Conclusion: Therefore we can trust our senses.
As a deduction it is invalid.
PS: "How often have I said to you that when you have eliminated the impossible, whatever remains, however improbable, must be the truth?" Sir Arthur Conan Doyle
You in this thread dismiss an argument, because it is incoherent, That is fair and well, but it misses the following possibility: That all claims with knowledge in mind about what the world is, is not possible with reason and/or logic, because all such claims run into Agrippa's Trilemma.
In other words, it only matters that a specific argument is incoherent if you can make an argument for a given area which is within reason and/or logic. So I am trying to get you to understand that all strong claims to metaphysics and knowledge run into Agrippa's Trilemma.
In general words as in the regards to regards to knowledge within methodological naturalism, then we, humans, live inside a cognitive bubble. We can't know what the world is, we just hold differently beliefs about what the world is.
What's "Agrippa's Trilemma"? Would you mind explaining?
Agrippa's Trilemma is a variant of skepticism about knowledge.
It can be explained in the following manner:
When you claim something, you ask yourself how you know that. Then you make a reasoned argument about that and repeat - "How do I know that?". When you continue to do that, you realize that you run into these 3 problems. You will do one of the following:
#1: Run into an infinite regress turning to ground your knowledge in something, for which you can stop asking how do I know that.
#2: You dogmatically declare that it is so.
#3: You beg the question.
So as a Skeptic, I have stopped claiming I have knowledge and just explain what I believe in.
Relevant to this thread, I believe that there is a world like you believe, but I don't know that nor do I claim that I know it. I believe it.
The obvious answer to this is to say that not all knowledge is based on arguments with premises as you claim. You can know many things non-inferentially, say by basing your beliefs on a perceptual experience which you take to reveal to you directly how things are in the world, and not by virtue of being part of some argument.
That is a variant of a dogmatic claim. There is a screen in front of me, because that is how it is and it is not worth questioning and anyway it is absurd, illogical, meaningless, wrong and what not to question if there is a world as I know there to be. THERE IS A SCREEN IN FRONT OF ME - PERIOD!!! Stop asking silly questions. :)
For which you run into Agrippa's Trilemma, because now you have to make a reasoned argument as how you know that. Or you simply declare dogmatically that it is so. hit infinite regress or beg the question.
What you said was this: "I know, that you can know many things non-inferentially."
For which I answer: "How do you know that?"
And off we go into Agrippa's Trilemma. :)
The hidden assumption you hold can be viewed like this:
# We can explain knowledge with reason, logic and experience.
Versus these 2:
# Knowledge is meaningless, unless we can explain knowledge with reason, logic and experience.
# The idea, that we can explain knowledge with reason, logic and experience, is just that - an idea, which doesn't hold up to reason and logic. In other words reason and logic are limited and can't explain knowledge including if we can fundamentally trust our experiences.
My point is that your argument doesn't prove that we don't know things non-inferentially, and since most people believe that they do know things on the basis of their sense experience (and not arguments), your argument simply doesn't engage the most plausible view out there regarding knowledge. I'm not trying to assert that I'm right that we do know things non-inferentially, I'm just saying that your argument doesn't show that we don't (and so I don't have to prove that we do, in order to show that your argument doesn't succeed).
I don't know ;) how to get this across. It is pointless to point out that my argument about knowledge fails, if all arguments about knowledge fail.
In short:
Me: Argument X.
You: It fails.
Me: Argument Y.
You: It fails.
Me: Argument Z.
You: It fails.
...
If all arguments about knowledge fails, then why single out the skeptical ones?
What is the point, Fafner?
That you believe you have knowledge, means you have knowledge? Is that your point?
It doesn't hold up! I believe that you don't exist, means that you don't exist? Or rather do you believe that all beliefs work?
Further you claimed that there are things, which are know non-inferentially, so would you please explain, how you know that?
I don't understand what you mean. If your argument fails, then it fails, which means that you haven't proved your conclusion (skepticism or whatever it was). So it's not pointless.
Quoting Mikkel
No, my point is that your reasons for denying that we have knowledge aren't very good (and I'm not trying to prove to you that I do have knowledge - I've got better things to do).
Quoting Mikkel
I already told you - by perceiving them. This is what 'non-inferential' means - the reason for your belief is not in the form of an argument which you can give to someone, it just suffices to have the right sort of experience without needing any additional reasons.
How do you know that?
Well, because you told me so, which makes it a dogmatic claim. You don't get Agrippa's Trilemma, because you keep claiming that it is so, because you say so. That is what makes it dogmatic.
Know what?
Quoting Fafner
That you are perceiving them!
How do you know that?
What do you mean? You know that you perceive something by perceiving it, how else?
How do you know that it is something and not an illusion?
You take for granted that there is a screen in for of you. How do you know that there is? How do you know, that it is not a simulation or that you are not a Boltzmann Brain? How do you know, that this something is, what you claim, it is? How do you, that it is not something else?
How do you know???
Now you are just using the Cartesian argument which I discussed in my other post. I thought that your 'trilemma argument' was a different argument from the classical argument from illusion (and by the way, you said previously you agreed with me that the argument is incoherent).
It is incorrect, because I believe despite evidence that you are something even if I wasn't around. Of course, I believe I will die one day.
Now you take for granted that there is something. E.g. I exist :) But you have given no evidence. Further there is a problem with your notion of taking for granted. Let us say, I take for granted that there is a God. Now that is not evidence for a God, so why should the fact, that I take for granted that you exist, be evidence for your existence.
You still haven't given any evidence. It boils down to that I have no evidence, therefore you have evidence. Namely that when between P and non-P, non-P is wrong, then P is true. That is not how logic works.
You have made a naive realistic claim - "You know that you perceive something by perceiving it, how else?", but you have given no evidence.
#I know, that I perceive something by perceiving it.
#I know; that I don't perceive something by perceiving it, because I know, that I am a Boltzmann Brain.
#Either case is not knowledge.
You don't get the last one, do you?
To be an old school Skeptic means in some sense not to believe in knowledge, just like some people don't believe in the concept of a god/gods.
I use the word knowledge as an idea that some people believe in, but I don't believe in knowledge.
You don't seem to understand doubt.
I doubt that I am in the world, you know you are in.
I doubt that I am a Boltzmann Brain and I doubt that we are in a simulation.
But none of that is evidence for the fact that we have knowledge.
The evidence is simply my perceptual state (of seeming to see that something is the case in the world).
Again, you are misunderstanding what 'non-inferential' means. It means that your evidence is simply the state on which you base your belief. And you don't need further evidence (to prove that your original evidence is good, or is not the result of an illusion etc.) precisely because your experiential state has intrinsic justificatory power to support what is believed, which is independent of any further premises (hence it is called 'non-inferential').
How do you explain a hallucination?
I once lived at the coast and could look out on a bay. Once day I saw an island, which I hadn't seen before and never saw again; i.e. a hallucination. How is that possible?
How can an experience be wrong?
Now if you claim that hallucinations are not real, then I have to ask - How do you know that?
They are sometimes real, but not always (at least they need not be, or you can't prove that they have to). I tried to show in that other post on skepticism that hallucinations (for their intelligibility as hallucinations) presuppose that there's also a possibility to perceive things just as they are in reality. Think about it this way: what makes something into a hallucination? The fact that when you have it, it looks to you as if things around you are in a certain way (when in fact they aren't). But this means that there's also the possibility for things to be in the world just the way they appear, which will not be a state of hallucination, since your experience will portray accurately the way things are in reality (so if it is possible for your experience and the world no to match, then it is also possible - at least logically - for them to match).
This is the first step. The second step will be to show that we are justified to rule out or ignore the sorts of hallucinations or illusions of the kind to which the skeptic appeals in his argument. I also discussed this to some extent in the other post, though a bit more needs to be said about that, since people didn't get the idea (I can elaborate on this if you wish).
How can something be not real?
If you have a hallucination, then is that a not real or unreal hallucination or is it real?
What is real and how do you know that?
What is not real and how do you know that?
In other words within methodological naturalism and science a hallucination is a process in a brain, so how can it be unreal? How can something, a brain, produce something unreal?
How can something, which is real, produce something, which is unreal?
Take time, space and processes in the world. Now it is a brain(real) then time and space passes and it becomes unreal. At what time, in what space and what process make something real not real/unreal?
Do you understand that you have made an incoherent argument, because you claim that in one moment some is real and then it turns unreal, but how can something become unreal, which is real? It is no different than ontological dualism, we have two ontological categories; real(matter) and unreal(mind), and something real can turn unreal.
You are taking for granted that something can turn unreal, but that this is real. I.e. unreal is real, because it is really unreal and it is something, which take place in the real; i.e. a brain.
Fafner, you have simply hidden the problem with your claim of real and not. But it doesn't stand up and your claim that all experiences including hallucinations can be explained non-inferential, doesn't hold, because you can't explain the non-real with your model of the world
When you have a hallucination of course your experience is as real as any (and that it is a state of your brain etc). The point is rather that not all experiences are necessarily hallucinations, since the very concept of hallucination presupposes that it's at least also possible to perceive directly how things are in the objective reality (and by "unreal hallucination" I just meant a genuine experience - e.g. seeing a real tree as opposed to just hallucinating a tree etc. - I didn't mean that the experience of hallucinating itself is not real).
Now I see a tree. That is non-inferential, I will accept that for now. But what that have to do with the objective reality? What do you mean by objective reality and how do you know that there is such a thing as an objective reality?
The fact that we're able to do science, and objectivity has a meaning and use.
"Objective reality" simply means that things are exactly as they appear to you perceptually (when you see a tree, there's really a tree in front of you etc.). Really, there's nothing to explain here.
And about the "how do you know?" part, it's a bit complicated, but I'll try to keep it simple. The basic idea behind the skeptical argument (that is, the argument from illusion) is that even what we take to be the best cases of knowledge aren't really knowledge because we cannot rule out every conceivable possibility of us being wrong. So we have two options according to the skeptic for ways for the world to turn out to be: for every imaginable experience (say of seeing a tree) either a. things in the world match exactly the way they appear to you or b. you are completely deceived about everything in your experience since you are perpetually hallucinating or a brain in a vat etc. And since no piece of perceptual evidence could support favoring possibility (a) over (b), you cannot know that you are living in (a) and not (b) (and hence you can't know anything).
But notice the difference between scenarios (a) and (b). Suppose that scenario (a) is the actual one, what this world would be like? Well, that would be a world in which things are almost exactly correlated with the way they appear to you perceptually, at least in the typical cases (since the skeptic allows us to choose any case we wish as the best candidate for knowledge; and as I've argued before, for every experience of things appearing to you to be such and such, there's a possibility that things in the wold are indeed such and such, even if they are not so in the actuality). And now let's consider world (b). The way the example set up, world (b) is designed in such a way that things look to you exactly as if you are in world (a) (but you aren't), and you have no way to tell (based on your experience). So in this world, reality has absolutely nothing to do with the way things appear to you, but somehow, by virtue of some miracle or extremely clever design, your appearances are correlated with the way the world would've been if world (a) were the actual one (because otherwise it would not be a case of an illusion).
And now, keeping all this in mind, let us assess the relative probabilities of worlds (a) and (b). Given that you are having right now an experience as if you are living in an (a) type world, what are the odds that you really are? I think they are pretty high (or even extremely high - and conversely, they are very low for (b) being the actual world), and this is shown by the mere fact that in order to persuade you that you might be mistaken, the skeptic had to concoct this very bizarre and unlikely sci-fi story about brains in a vat or whatever. But the world as you perceive it, is such (if we assume for a second that it is the actual world) that in it scenarios of the kind that the skeptic envisions are extremely unlikely; that is, cases in which reality matches appearances are way more common than when it doesn't (because by hypothesis, it is a world where whenever you see a tree there's really a tree, since normally no one tries to deceive you etc.). And that means that it is much more likely for your experiences (taken by themselves) to turn out to be right than wrong, since it would not be a coincidence if your appearances corresponded with reality, but it would be an extreme coincidence if (despite their internal coherence etc.) they actually didn't.
So I believe that we should conclude that we have a very good reasons to favor the belief that we actually live in world (a) rather than (b), since given the extreme unlikeliness that all our experiences are systematically deceptive (as the skeptic imagines), we are perfectly entitled to ignore such possibilities and trust our senses at face value. And now if you supplement all this with my previous account of non-inferential justification, you can actually get from all of this an account of a genuine perceptual knowledge of the world.
You don't get that you are doing a case of begging the question- ( -when you assign the odds.
Now I don't do metaphysics and epistemology as knowledge. I do it as different beliefs and the telling sigh is here that so do you:
Now Fafner, the objective reality does not conform to what you believe, not matter what you believe. That you can't understand, that you have just admitted that it is a belief and not knowledge, tells me, that you don't understand the skeptic position - You don't know what reality is and you only believe.
Further you are biased because of theses words:
You turn it being about God, if reality is not as it appears. You don't understand the examples in this thread, do you? None of them has to do with God. Both examples are naturalistic.
So you have shown your hand - You are begging the question if you assume you can determine the odds. You admit that you believe and you argue as if this thread could be about God. I do get that you properly don't believe in God, but neither do I. I believe reality is as it appears, but I don't know that, nor do I know the odds.
In the end no matter what you believe won't determine what reality you are in. No matter how you believe, it won't affect what reality you are in. You don't control reality based on how you think, reality controls you. That is part of what objective reality entails.
You are doing magical thinking, because you believe that your thinking determines, what reality is. It is similar with some religious believers, namely that some people don't like the unknown and will try to offer beliefs as evidence. You have given no evidence, because you are begging the question, when you believe you can assign actual odds. Further it appears that you don't understand that you believe and don't know, despite you use words like belief and trust. You believe and trust in a naturalistoc reality, you just have to give up the notion that you know
Quoting Michael
This whole set of logic presupposes that what we are is the instantiation, or execution, of whatever physics makes us up. But a simulation of thing X is not an instance of X. X already exists (is defined, and the definition is enough), and the simulation simply imparts some truth about X to whatever is executing the simulation.
So the list above leaves off the 4th possibility: that what we are is not the execution of the steps. Is 3+3 not equal to 6 until something performs that arithmetic? There are arguments for both sides of that debate, so I will not present it as truth, just a 4th option.
This whole set of logic presupposes that what things are is the instantiation, or execution, of whatever physics makes defines. But a simulation of thing X is not an instance of X. X already exists (is defined, and the definition is enough), and the simulation simply imparts some truth about X to whatever is executing the simulation.
That said, there is another presupposition that IS about us and not just things: That what we are is just a physical thing/process and not in need of some immaterial experiencer of the thing X, simulated or otherwise. I agree with that assumption, but it is still an assumption. This sort of groups under item 1: An ancestor simulation is not possible because a simulation of an ancestor is not a full simulation of that ancestor unless there is a way to grant the immaterial experiencer to the simulation. Something like that.
Possibility 5 then: The real world (the one running the simulation) is not a human civilization at all, but a much higher tech thing which can execute simulations of trivial things like a human civilization at will. The simulation need not be a simulation of its native underlying reality. I don't think that scenario is covered in any of your options.
My possibility 4 stands in such a scenario: Such a simulation is not the creation of this reality, but merely an analysis tool to learn about the already-defined reality.
Here what I can do:
Suppose that we start with the assumption that there is no knowledge out there in the objective reality.
Then I construct evidence:
Knowledge is a cognitive construct, an idea just like the idea about God. Some people treat God as real, some people treat knowledge as real.
Knowledge is not real like a tree, you can't see, touch, hear, smell and so on knowledge. The idea that there is an objective reality independently of the mind, which you none the less, have access to is a contradiction, because reality can't be independent of you and yet you have access to it.
Reality is your experiences and your beliefs about how to make sense of it. There is no "The objective reality", because then subjectivity is not a part of reality.
That facts, evidence, truth, knowledge and so on matter to you is because, that it matters to you, is what makes it subjective.
Why not? That's a really extravagant claim to make for someone who calls himself a "skeptic".
Quoting Mikkel
This simply begs the question. If this is meant as an argument to prove what you said in the first quote ("mind independent reality is a contradiction"), then you cannot start from a premise which equates reality with your experiences. You indeed experience reality through you subjective states, but it doesn't follow that reality is identical with them.
Quoting Mikkel
What you say here doesn't follow. If dogs matter to me subjectively, that doesn't prove that dogs themselves are 'subjective'.
I don't believe that you don't believe you know anything. Do you know how to find your way home? Do you know the difference between a tree and a pork sausage? Do you know how to swim or ride a bike?
It's true that we don't know what are the metaphysical conditions that give rise to our experience of a world of familiar, more or less invariant objects, landscapes, day and night sky, animals and people. About that we may only speculate. And we have good reasons to think some speculations are more plausible than others, even if we cannot prove that they are.
On the other hand we may indeed have knowledge about those metaphysical conditions, but we can never achieve discursive certainty. In other words we may know, but we cannot know that we know, because that would involve (per impossible) the absurd discursive regress of the ancient skeptics; the so-called 'criterion' problem.
The kind of doubt you are valorizing seems trivial and empty.
So given my assumption, it is easily proven that a perfect simulation of some actual state of the universe (say a chosen person) is impossible. But an imperfect one is not, say simulated as a neural-chemical body with a finite world to inhabit. This simulation, being finite and imperfect, would have a different history than the real person who is being modeled, but the simulated person would be conscious (again, given the assumption of consciousness is a physical process, and for the most part, not a quantum process). If the simulation took a scan of my body as part of the initial state, then the simulated person would not know it was not me, but might quickly figure it out if the finite nature of the world is at all obvious.
Now the main part of my point: The simulation need not run at all for the simulated person to be conscious. All that is needed is the initial definition of state and the definition of how the simulation is to proceed from that state. This is what I meant by the fact that 3+3 is equal to 6 even if the arithmetic is not actually performed by some mechanism.
Never said you assumed BIVs. I merely used the case of BIVs to demonstrate what I think is a wild assumption with your approach. As far as I see it, the only way for your argument to undermine the simulation argument to work is for you to somehow point out some feature or element about the nature of our understanding of the world (or our experience of it) that cannot in principle be replicated by a computer program. If it can be, then we cannot ascertain whether or not we are looking at a simulation when talking about the world we live in.
Of course, this sort of discovery seems as likely as the discovery that there is a feature of our experience that is impossible to replicate, whether by a BIV, or demon, or a vivid dream scenario. Now you may not be alone in thinking this. I believe this is the sort of suggestion made in the "Answering the Skeptic" thread, but as for my own take on it I find it to be a bit too extraordinary for my own liking.
Your example deals with a conscious person, but again, I must point out that the simulation argument (as well as the BIV argument) is more about the world we find ourselves in rather than who we are.
I'm not undermining the argument. I'm listing additional possibilities than the three listed. I have no problem with a computer program simulating consciousness.
This statement presumes that the simulation is the simulated. Not sure if I've driven home the difference. The simulation is a tool for yielding information to the simulator (the creator of the simulation). So if say a human set up and ran a simulation of a bat, the running of the simulation would behave like a bat, but would not impart the human with knowledge of what it is like to be a bat. Meanwhile, the bat in the simulation would know what it is like to be a bat, whether or not the simulation is actually run or not.
Hope this helps.
Hence my disclaimer/presumption about experience being essentially a macroscopic physical process. If it isn't, not sure if such simulation is possible, so I'm considering only the case where it is.
The simulation runs on macroscopic rules, and suddenly the simulated guy starts doing non-macroscopic experiments in his simulated lab and the simulation cannot handle that. He'd be able to tell. So the simulation has to be good enough to mimic even that, and at that point I have a hard time agreeing that it is possible even in principle.
Likewise, the simulation needs to be confined somehow, limiting resources. An infinite universe cannot be fully simulated even macroscopically. The simulated guy would possibly notice that he is in the center of a finite place, just like we were centuries ago. So I have a hard time with the assertion that the simulation could be so good that we can't know we're in one.
I believe there are a couple of assumptions that you've made that I've pointed out below. I am not sure what you mean by dualism though, if you consider BIV to require it. As far as I can tell BIV works perfectly fine with most positions about the mind.
Quoting noAxioms
Your possibility tries to undermine the simulation scenario just as much as the first two possibilities do. That's just how it is.
Quoting noAxioms
Okay, then we are not sure if we are talking about something simulated when we are talking about the world we live in. Whichever way you word it, the fact is that we still have no clue.
Quoting noAxioms
The simulation's macroscopic rules don't necessarily have to be the same as the rules in the world of the simulation. And even if the simulation world's rule do mimic the macroscopic rules, there is no requirement that it has to be an exact representation. Though either way, if we are living in the simulation we will not be able to determine what those macroscopic rules are so from our POV we can't compare them to our own world's rules.
Quoting noAxioms
The true universe (the world that isn't simulated) does not necessarily have to be infinite either. If we find out that our universe is finite that does not mean we live in a simulation, that could just mean that the true universe is finite. But of course we aren't able to know either way, so that doesn't help in deciding the issue.
And even if that isn't the case (and that the true universe must be infinite), there is nothing in principle that says that we should be able to determine ourselves if the universe is finite and thus simulated. Assuming we are in a finite simulation, if the creators of that simulation really wanted to play the role of Descartes's Demon, then they can probably do something to prevent us from being able to tell.
As for determining if the universe is infinite and therefore not a simulation, I am not sure what kind of experiment could be done to even determine such a fact anyways (though I am open to hearing proposals), so it seems like we're lost there too.
Quoting AlecI don't consider BIV to require it. There is a mind, and it has zero access to the nature of itself or reality, so it can trust no knowledge. I am discounting that scenario from my discussion.
A more dualistic BIV scenario is a very immersive video game. You are Lara Croft in Tomb Raider, and so convinced of it that you forget you're something else. But you can tell: Lara's head is empty and contains no mechanism for thought and experience, which is served by the 'immaterial' player. The physics of Tomb Raider is obviously not deterministic, so no simulation of Lara physics will result in her acting like she does with the player component.
Don't know how to explain it better. The simulation is of a real thing. The simulation is not the thing, and thus at no point are we in a simulation, be we simulated or not. So my stance is that we cannot be a simulation. If a simulation is run and it is not perfect (does not simulate what was intended), then the simulation is just of something else with different physics, but the simulated thing is still not a simulation.
It perhaps comes down to a language quibble, but one I feel is important. I can be simulated in principle, but I cannot be a simulation. Running a simulation is not an ontological act any more than the running of a 'real' universe. All it does is execute a simulation process.
Yea, I guess it does. It seems that my position is not just another possibility, but it debunks the whole argument. The argument works from the presumption that the states universe are things (objects??) that happen or are executed, which I find absurd.
Agree, but is it an 'ancestor simulation' as described by the OP if the simulation is of a different world that the one running the simulation? Conway's Game of Life is a simulation, but not one of our physics. No structure in that game can detect that it is in a simulation because no structure is actually in one. They are being simulated, but are not simulations. See what I'm saying?
It indeed cannot be exact, and yes, it works fine that way. I suppose our universe could be simulated down to the quantum level, but not on any simulating equipment that I can envision. It seems unnecessary unless there is a quantum amplifier (like the one that kills Schrodinger's cat) that needs to be simulated. Biology seems not to have them, nor does electronics for the most part. History cannot be reproduced, but a functional world of sorts can be simulated. It would work. A transistor for instance has very simulatable macroscopic behavior, and I've actually worked with transistor simulations which are cheaper to test than real chips. But transistors rely on quantum effects to function. They wouldn't work without that. Doesn't matter: The simulation does not simulate at the quantum level. It simply simulates it as a macroscopic switch with this and that delay, threshold, and EM noise. The neuro-chemical nature of brain cells seem to have similar macroscopic function. The simulation need not be more detailed than that.
I have to correct my own comment there. If the simulation is so 'imperfect' or size-limited, then it is simply a different thing being simulated. The thing can know it is not in a simulation. It simply exists in a small limited simpler universe.
A true universe I would say, or 'ours'. I speak repeatedly of other universes (say the Conway GoL one) in this post. There is our universe, but I don't think ours is any more or less true than another. Perhaps we need a more concrete definition of 'universe'.
My solution solves this problem. Can't be a simulation because things are not simulations (by definition), even if simulated.
And my point is that it doesn't matter. Sure, a simulation of Paris is not the same as Paris. That is why it's called a simulation. But that doesn't affect the fact that we do not know if the Paris we know is one or the other. Apart from that, I must admit I don't see your point.
And again, I must stress that you move away from talk about the self. The simulation argument deals more with the world we are in rather than ourselves and I feel like ignoring that would lead to more confusion than not.
Quoting noAxioms
Didn't know you were implying quantum phenomena here by the use of "macroscopic". I must first start off by saying that my knowledge of physics is only basic (some Pop Sci. Books and a rough knowledge of the history), but I do not see how quantum phenomena cannot in principle be simulated. Indeed, aren't some physicists already simulating quantum phenomenon in their research? I can imagine it is impractical sure but not impossible. In addition, we should not imagine that the world running a simulation of our physics needs to run on the same rules. That was the point of what I said earlier. The ancestor world can run on an entirely different set of laws, one that makes the simulation of the quantum more practically feasible. However, the simulation would still run on the same fundamental principles of computation that we have for our own computers.
And that is the key point in all this. In order to demonstrate that something cannot in principle be simulated, it must not be able to run on a computer. Computers, as far as I can tell, are digital, they run on binary, they use an algorithm and are finite.
You mentioned infinity, which is something that our ideal computer cannot simulate,due to its limitations. Other examples of phenomena which cannot be simulated are continuity and true randomness. Unless the intelligent race is somehow able to tap into the infinity and create the ultimate computer, then we can safely assume that their simulations are limited (thus excluding the infinitely small and the infinitely big). And the same goes for true randomness, as computers are necessarily deterministic. However, the problem with these possibilities, when I was thinking about them, was that it was impossible for us to know whether or not they were true. Unless we are able to go to the ends of the universe, then we cannot determine if there really is an end to space or not. Same for continuity and randomness. Unfortunately, I don't see how quantum mechanics fits the bill in all this. All in all, I remain unconvinced that you have shown or solved anything.
Evidence of this lack of distinction is that it is impossible to do a "I'm in a simulation" test armed with only subjective knowledge, no matter how inaccurate the simulation.
Note I did give a distinction there. Given non-subjective knowledge, there is a distinction since there is suddenly existence of information from outside the universe. It isn't really a simulation then anymore, but just a sub-process within a larger universe. So for instance, if God created this universe and communicates in any way (imparts state inconsistent with the physics), then the universe is not really a universe, but just a sub-process/object (not a simulation), part of a larger universe containing the full explanation of all empirical data including the said communication.
No self needed here. Paris has a library with books describing different physics than the physics of the Paris being inaccurately simulated. That's inconsistent. Such inconsistencies are detectable.
It indeed cannot be exact, and yes, it works fine that way. I suppose our universe could be simulated down to the quantum level, but not on any simulating equipment that I can envision. It seems unnecessary unless there is a quantum amplifier (like the one that kills Schrodinger's cat) that needs to be simulated. Biology seems not to have them, nor does electronics for the most part. History cannot be reproduced, but a functional world of sorts can be simulated. It would work. A transistor for instance has very simulatable macroscopic behavior, and I've actually worked with transistor simulations which are cheaper to test than real chips. But transistors rely on quantum effects to function. They wouldn't work without that. Doesn't matter: The simulation does not simulate at the quantum level. It simply simulates it as a macroscopic switch with this and that delay, threshold, and EM noise. The neuro-chemical nature of brain cells seem to have similar macroscopic function. The simulation need not be more detailed than that.
— noAxioms
Never said that. I said macroscopic rules will usually do unless the simulation needs a quantum amplifier, which cannot be simulated properly with a macroscopic simulation.
It can be, but not with simulator running macroscopic rules, and not even in principle if predictable results are expected, since quantum events are not predicable.
Sure they do, but those simulations do not predict unpredictable events like when the atom will decay. The simulations necessarily have randomness built into them, something not particularly needed at the macroscopic level.
That's right, but it wouldn't be an ancestor simulation then, but merely a dumbed down fictional virtual reality.
No, that cannot be. The nature of our universe is not one that can be accurately simulated on the principles on which our computers operate. I cannot even figure out how to express the position and state of the most trivial particle given unlimited computing resources. An accurate simulation of us would require something fundamentally different I would think.
OK, so we can simulate Paris, at least on a macroscopic level. Anything that can be put in a box. Hard to put Paris in a box since it has continuous interaction with its neighbors, but perhaps the whole Earth with pared down simulation of celestial activity which is pretty easy to predict most times.
A simpler case is a locked room. The whole simulation can be confined to that. I assert that barring information inconsistent with the simulation, it is impossible for the simulated thing to distinguish reality from the simulation. If it were possible, it would be a simulation of a different reality, and thus the distinction would be in error.
Hence the need for a box. The box confines some of the infinities.
That is an interesting point. Does randomness need to be true? I think so, since if it was not true randomness, it would be predictable, and that conflicts with QM. It would not be a simulation of our reality. If it is a macroscopic simulation, I don't think randomness is needed at all, true or otherwise.
There are computers with true randomness. It just requires a small randomness device, and they probably fit on a chip. Not so hard.
I've seen 3-body simulations that had poor continuity implementation. The objects seemed to always gain energy. So point taken again.
That the running-on-the-same-rules point. Said super-race might have physics that effortlessly let them do infinities in their computations, but again, they would not be running an ancestor simulation by simulating us here.
Not too hard to know that. People didn't know where the Earth ended either, nobody having visited the edge. Then they found out it had a geometery with no edge and the problem went away. The geometry of the universe has no edge, and no center (not in space at least). There is no vantage from which there are stars only on one side and not the other.
A computer-simulation can't create a world. Infinitely-many possibility-worlds are already "there", as possibility-worlds, complex systems of inter-referring inevitable if-then facts. Each one existent and real only in its own local inter-referring context, and quite independent of eachother, or anything objective or global.
A computer-simulation can't "create" what's already there.
The only thing that a computer-simulation could create would be a demonstration, a portrayal, of something that already is. ...for the benefit of the simulation's viewing-audience.
Are we and our world being simulated? If it's possible to simulate a universe with a computer-program, and if there'd be any conceivable reason for doing so, then in the infinity of possibility-worlds, no doubt someone is simulating our universe with a computer-simulation. But that didn't create our world. ...merely duplicated, portrayed it.
You simulation program can amount to a duplication of our possibility-world, but some transistor-switchings in a computer has no effect on anything.
Michael Ossipoff