• flannel jesus
    1.8k
    he says it right after quoting me. If it's not for me, who is he directing the question "what is your evidence" towards?
  • Patterner
    996
    IF determinism rules all things..." but does it? What's your evidence?Ludwig V
    It's Laplace's premise. It's not mine. I don't believe it to be the case.


    Interesting that he doesn't mention that God would be such an intellect.Ludwig V
    My guess is that he didn't want to get into God, because that would be a discussion about why God set things in motion in exactly that way. What's the plan, what's the purpose. That kind of thing. He only wanted to discuss the positions, properties, and forces.
  • Ludwig V
    1.7k

    I'm sorry if I confused things. I'm happy to take answers from anyone who is moved to provide one.

    One need not think of specifically the physical realm as a closed system, one can instead imagine (physical realm plus mind realm) as a combined closed system. And an LD that's fully aware of what's going on in all the relevant realms of the combined closed systems is still conceivable.flannel jesus
    All I was saying is that it makes no difference whether one thinks of the universe as a closed system or as a combined closed system. It is just one way of thinking about the universe. It may be useful, but is it true? What is the evidence one way or the other?

    It's Laplace's premise. It's not mine. I don't believe it to be the case.Patterner
    That's fair enough. I'm just trying to say that it isn't an empirical idea - no amount of empirical evidence will confirm it, or refute it.
  • flannel jesus
    1.8k
    It is just one way of thinking about the universe. It may be useful, but is it true? What is the evidence one way or the other?Ludwig V

    "Closed" just means "everything that matters for calculating the future of this system is here." So what's the evidence that, regardless of whether the physical universe is closed itself, there's SOME closed system that contains the uinverse? Well, I don't have scientific evidence, but consider this intuition: there is a set of things that are the answer to the question, "what are all the things, physical or otherwise, that go into deciding future states?"

    Future states are, in fact, realized, so something must realize them, so there must be a set of things relevant to the process of realizing the future. That set of things is "the closed system", whether that's exclusively physical or also contains other "realms".

    The alternative is the claim "there is no set of things that go into deciding the futre states".
  • Patterner
    996
    That's fair enough. I'm just trying to say that it isn't an empirical idea - no amount of empirical evidence will confirm it, or refute it.Ludwig V
    I agree. I think Laplace was just saying something with all knowledge of where everything is and of all the forces would be able to calculate everything for the future. Sure worried be nice, since it could tell us about any asteroids that are going to impact the Earth. It could probably solve cold fusion pretty easily, also.

    And it would be able to tell us if quantum events are truly uncaused, and if there is free will.

    but, we don't have this marvelous demon. It's just a fun thought. Although the demon could help us quite a bit, the thought cannot.
  • flannel jesus
    1.8k
    It should be noted, and maybe already has, that even in principle the demon cannot exist inside of the same universe it's capable of predicting, even if that universe is 100% physicalist and 100% deterministic.
  • Patterner
    996
    ↪Patterner It should be noted, and maybe already has, that even in principle the demon cannot exist inside of the same universe it's capable of predicting, even if that universe is 100% physicalist and 100% deterministic.flannel jesus
    I don't know what you have in mind. But if it's the first thing I can think of, I disagree. First, it could, itself, be ruled entirely by determinism.

    Second, if it has free will, it could fully calculate the consequences that all of its own actions would have on the rest of the universe.
  • flannel jesus
    1.8k
    I have no idea what the first thing is that you're disagreeing with
  • Patterner
    996
    ↪Patterner I have no idea what the first thing is that you're disagreeing withflannel jesus
    I don't know why you think that, even in principle, the demon cannot exist inside of the same universe it's capable of predicting, even if that universe is 100% physicalist and 100% deterministic. But actually, two thoughts came to mind. My first thought was that you thought something that is not 100% physicalist and 100% deterministic would be, as it were, breaking the rules.

    My second thought was that you thought it's own non-determined actions would make calculations impossible.

    If neither of my guesses was correct, what is the reason you think that, even in principle, the demon cannot exist inside of the same universe it's capable of predicting, even if that universe is 100% physicalist and 100% deterministic?
  • flannel jesus
    1.8k
    If neither of my guesses was correct, what is the reason you think that, even in principle, the demon cannot exist inside of the same universe it's capable of predicting, even if that universe is 100% physicalist and 100% deterministic?Patterner

    Let's imagine a super simplified case. Forget quantum mechanics, imagine the world is classical, space and time are Cartesian, the world is composed of atoms which are more or less like tiny little billiard balls bouncing around.

    Why, in such a simplified world, could an LD not possibly be able to predict the future with perfect accuracy?

    Well, our LD is made of atoms, is he not? Some fraction of his atoms are for his brain, the rest are for his body. Even just to calculate his own future, only 1 second into the future, he would have to know the precise location and velocity of every atom inside his own body and brain, and know the location and velocity of every atom that's going to interact with his body in the next second.

    He doesn't have enough atoms in his brain to store all that information, never mind do calculations on it.

    And then you've got the computing speed problem - you can't compute the universe faster than the universe can compute itself, from within the universe . I mean there are some scenarios maybe where you could jump ahead because you know this particular thing is flying in a straight line and won't interact with anything, but mostly you don't have a bunch of simplified things like that, you have thousands of things bouncing into thousands of other things all the time. Not a lot of space for computational shortcuts available.

    So a leplace demon is impossible to exist inside the universe - you could have one outside the universe looking in, but not inside
  • Patterner
    996

    Demons are made of atoms? I had no idea.
  • flannel jesus
    1.8k
    did we not just say in a physicalist determinist universe? If this demon is in this universe, then yeah.
  • Patterner
    996

    It's a supernatural being in a thought experiment. I don't know how to put limits on that. It's all silly.
  • flannel jesus
    1.8k
    It's a supernatural being in a thought experimentPatterner

    So it doesn't sound like you're disagreeing with me then, when I say "it can't exist in the universe it's predicting". If it's predicting a universe of atoms, it can't just exist in that universe as a thing made of atoms and also be able to perfectly predict the future faster than it happens - it has to be "super natural" - super meaning ABOVE, meaning above the nature of the universe it's supposed to predict. I agree, it has to be SUPER to the nature of the universe it can predict.
  • flannel jesus
    1.8k
    So, classical in-universe Leplace Demons are strictly impossible, I don't see any way around that - so if one wants to allow for a supernatural or extra-universal Leplace Demon, that shakes things up a bit.

    If we had a demon outside of the universe, it could predict the future even given Quantum Mechanics - the catch is, if we live in a quantum universe (and I think we do), its predictions must be probabilistic (probably). It could in principle perfectly predict the probability distributions of various future states.

    Although some interpretations of quantum mechanics go a step further, like Bohmian Mechanics, and say that actually underneath it all there's a true single deterministic path to the future, so if that were the case, that type of QM would still allow for a normal Leplace Demon, who can still perfectly predict a single future.
  • Patterner
    996

    Sure. Demons are supernatural. I've never heard of a story with one that wasn't.

    Hey, how about this idea... To try to put a natural entity in this role, it would, obviously, need to be made of particles. A quantum computer, or whatever the next step would be, made up of enough particles could calculate the rest of the particles. If there are finite particles in the universe. It wouldn't work for infinite particles.
  • flannel jesus
    1.8k
    A quantum computer, or whatever the next step would be, made up of enough particles could calculate the rest of the particles. If there are finite particles in the universe.Patterner

    No, it still couldn't. It couldn't do it faster than the universe. To represent the location and velocity of a single particle, you need MANY particles. Hundreds, probably thousands, maybe millions at minimum. So just to calculate what 100 particles are going to do, you'd need to have hundreds, at minimum, particles per particle you want to predict - and that's just for storing information about them, not even running computations on that stored information.

    By the time you start your simulation, the particles you gathered the information for have already been evolving into their future, leaving the computer simulation in the dust, and the simulation will never catch up. It will necessarily be many many many times slower.

    In fact there's a real example of this, an example of computing a universe within a universe - Minecraft. People built Minecraft in Minecraft, and that's an amazing accomplishment, but there's that catch - Minecraft in Minecraft always necessarily runs many many many times slower than the first layer of Minecraft.

    You can't simulate reality perfectly, faster than reality can do it itself.

    You can simulate it imperfectly faster - we do that all the time, it's easy. But not perfectly.
  • flannel jesus
    1.8k
    Here's a conversation on the topic elsewhere:

    https://www.reddit.com/r/compsci/s/nsfBtiBGE5

    It's almost too intuitive for me to explain..

    But think of the counter-consequences.

    If a computer could simulate itself faster than it could run, then you could run a faster simulator inside the simulator, and in turn have an ever increasing speedup.

    Anyway, the simulator has to do things like fetch memory, but fetching the memory in simulator always takes as long as doing all the prep work in the simulator, then doing a memory fetch in hardware equivalent to what is being simulated.. So basically every thing you do has to be done in hardware anyway, but with more overhead on everything

    And here's another one that's probably even more relevant than the above: https://news.ycombinator.com/item?id=40071773

    If you consider the physically optimal implementation of any function (e.g. the optimal NAND gate), that system cannot be simulated in real time: the simulation will always be slower, pretty much by definition. Insofar that physics optimally implements itself, you cannot simulate reality in general without a massive performance hit (think about the recursive absurdity of the simulator simulating itself).
  • Patterner
    996

    Why are we simulating? Where will Voyager 1 be in fifty years? We don't simulate it's existence for every moment of the next fifty years. We just calculate.

    Also, I don't think LD supposedly knows where every particle will be at every moment for the rest of time. I think it only has to be able to calculate the answer to specific questions. After it tells me where Voyager 1 will be in fifty years, I ask it where a particular particle in Jupiter's red spot will be next week. Then I ask it something else.

    Of course, the farther into the future I'm asking about, the more other things will have an impact on it. But LD's intellect is vast enough.

    And I really don't think Laplace was trying to convince us that such a demon is likely, or possible. He was just saying, in a universe where everything is deterministic, anything at any point in the future would be, in theory, calculable.
  • flannel jesus
    1.8k
    Why are we simulating? Where will Voyager 1 be in fifty years? We don't simulate it's existence for every moment of the next fifty years. We just calculate.Patterner

    Because Leplace Demon is supposed to be able to predict EVERYTHING perfectly, not just simple toy examples. Chaos, right? When a system is chaotic, you can't just do a simple calculation, you can really only find the answer with a simulation.
  • flannel jesus
    1.8k
    And I really don't think Laplace was trying to convince us that such a demon is likely, or possible. He was just saying, in a universe where everything is deterministic, anything at any point in the future would be, in theory, calculable.Patterner

    That's right
  • Harry Hindu
    5.1k
    Again your not getting the point. That turn hasn't happened yet, it's in the future. The pilot is flying the aircraft ordinarily, because the aircraft hasn't been attacked. He's looking at the potential AA site, but as the pilot observes he's not fired upon, no reason for evasive manuevers. Maybe the site is simply a fake or the gunners simply haven't observed him. The LD giving the firing solution and the firing of the gun only alerts the pilot to make evasive maneuvers. The LD solution is defined from the LD solution itself, you cannot get around it, sorry.ssu
    As I pointed out before, you are speaking from a position of ignorance. You simply don't know what LD knows. As I said, LD has a "Law of Everything". You do not, yet here you are arguing what would be impossible for LD.

    Let's just remember how the LD makes the forecast in general. It knows everything at the present, and it can then extrapolate perhaps one nanosecond at a time to the future to millions of years from now. But this isn't anymore a simple extrapolation: here the correct model of future has to take into consideration the model itself. The LD solution happens partly because of the LD solution. That's circular reasoning. And here we come to the interesting philosophical issue at hand: here the LD has to make a subjective decision. It cannot be just an objective observer here. If it would be, then it wouldn't give any LD solution, the anti-aircraft gun wouldn't be fired and the pilot could perhaps fly aircraft in a straight line through the airspace where the AA gun could reach the aircraft. The gunners would angry at such fire control.ssu
    Which would certainly be a possible (non)action by LD as it knows more than the gunner. Let the gunner learn his lesson by firing at the pilot and never hitting them. If LD's goal was to bring down an incoming bomber then his knowledge would have given him some other options that you and gunner could not comprehend, much less think of yourself.

    Maybe you should make a flowchart of how LD would make the decision. I think that the only thing that matters is what LD knows the pilot will do. LD's solution is based mostly on that.

    Again, the pilot alters his flight if the aircraft is attacked (sees the muzzle flashes), that happens only after the LD's firing solution, so LD cannot just extrapolate from the present something that isn't yet done.ssu
    Yet here you are without a Law of Everything predicting that the pilot would perform evasive maneuvers in the future. Now, expand that to an infinite level of precision as LD would have and would you be able to say the same thing about what LD can forecast?

    Here's the most important issue: LD just cannot extrapolate from the past, it has to make a choice when to give the firing solution and what firing solution. That's different what Laplace had in mind. There's many ways to do this, but it isn't simple extrapolation.ssu
    How so? Isn't that how we make any decision in that we must choose what and when to do it? We have to choose how and when to launch a space probe to Mars and we've done it multiple times successfully, more than just random chance would allow.

    My question is, why did the NASA scientists not need to account for the solution to get to Pluto in the solution to get to Pluto and New Horizons still arrive at Pluto? Sure, it seems that if they tried to include the solution in the solution the New Horizons project would never moved past the planning stage, but it did by not accounting for it and the solution was a success. As I said before, some information is irrelevant to the forecast being made. NASA scientists also did not account for the speed at which the weeds in my yard grow to get to Pluto either.

    This actually is very crucial to our usual way of looking at this: if there's determinism, can there be free will? That's the typical way to look at it. The LD example gives another way to look at this: here the LD has to make a subjective decision because it cannot be just an objective onlooker. And once it does, so, then not all computations can be done as earlier. A lot sure, but not all.

    Perhaps in a way our free will simply limits our ability to calculate/prove/extrapolate everything about the future, if it is deterministic.
    ssu
    I've never denied that determinism does not allow for free will. LD has no free will because it knows everything about everything in the present and can then extrapolate what it will do based on this understanding. Sure, it must account for it's own actions and decisions which makes it exponentially more complicated, but that just means it is impossible for you, not LD.
  • ssu
    8.6k
    As I pointed out before, you are speaking from a position of ignorance. You simply don't know what LD knows. As I said, LD has a "Law of Everything". You do not, yet here you are arguing what would be impossible for LD.Harry Hindu
    And as I pointed, Laplace never talked about and LD or a "Law of Everything" that we don't know, but assumed if some extremely well informed entity could make the extapolation from the present (or past), into the future. Laplace wasn't speaking of any divine power. As I said, what he was talking about is simple "Newtonian" physics extrapolation. That should be clear.

    However coming back to your idea of LD having the "Law of Everything":

    Let's first discuss this as this is one crucial factor here and should be discussed. Actually you aren't the first to make this argument.

    Your argument (and please, do correct me if I'm wrong) is basically the "Black box" argument with LD: we don't know what logic, information and laws which LD is using (that we don't know, which is the Law of Everything. LoE) and hence for LD solving the problem is easy, even if it's not for us.

    Ok,

    The first question is then: If LD solves this problem using LoE, is then LoE equivalent to our logic that we use? Well, when one situation is that the correct forecast is a forecast that the LD doesn't give, obviously it isn't so, or then we really have understood very wrong basic logic.

    I've never denied that determinism does not allow for free will. LD has no free will because it knows everything about everything in the present and can then extrapolate what it will do based on this understanding.Harry Hindu
    Well, now you went ahead of me. Assuming that LD has no free will because it knows everything about everything and can extrapolate the future from the past with (LoE) is definately not something the Laplace had in mind. The point that LD would have no free will is quite a statement.

    In fact, this is my point: One can say it that our free will limits this kind of simple extrapolation. Yet is this the correct way to state that theorem? Would it be perhaps better to say that simply there are limitations to what we can compute (or give a direct proof or), because we have free will?
  • dimosthenis9
    846
    One can say it that our free will limits this kind of simple extrapolation. Yet is this the correct way to state that theorem? Would it be perhaps better to say that simply there are limitations to what we can compute (or give a direct proof or), because we have free will?ssu

    Well neither.Even proving Laplace's Demon wrong in some forecasts cause of the veto ability that doesn't consist as a proof for free will existence.

    Supposing you are right,even that veto could occure randomly in human brains.And remember neither randomness is on favour of free will.

    But you are right about what Laplace had on mind about his Demon.He never mentioned LoE and stuff like that.
    That's Hindu's Demon or maybe God :)
  • ssu
    8.6k
    But you are right about what Laplace had on mind about his Demon.He never mentioned LoE and stuff like that.
    That's Hindu's Demon or maybe God :)
    dimosthenis9
    Exactly, "Hindu's Demon" or God is beyond logic.

    Yet when we keep to the logic, it is a limitation because of logic, not just the assumption "that we don't have enough information" that hints that we may have, at least theoretically.

    Supposing you are right,even that veto could occure randomly in human brains.And remember neither randomness is on favour of free will.dimosthenis9
    Where it occurs isn't the question, that it occurs is the important point. Remember that with Turing Machines nobody is suggesting they would have free will, but they fall to the same problem. Turing machines cannot compute functions that are not computable by any algorithm.

    The important issue here is that we are talking about logical, mathematical limitation here.
  • dimosthenis9
    846
    that it occurs is the important point.ssu

    Well i mean that it occurs might also be a random brain function.Just another human thought that pops up from randomly neural activity.So i guess we agree that this isn't free will proof right?

    Logic limitation was something that Gödel proved also with his theorem so i don't doubt about that.
  • ssu
    8.6k
    Logic limitation was something that Gödel proved also with his theorem so i don't doubt about that.dimosthenis9
    Did he at the time? The undecidability results (Gödel, Turing, Tarski etc.) aren't so directly understood as you say.

    For example, just read this article in Stanford Encyclopedia of Philosophy done by Panu Raatikainen in 2013 (revised 2020) about the Gödel’s Incompleteness Theorems. Not only going through the two theorems of Gödel, Raatikainen also in the writes about of the "Philosophical Implications—Real and Alleged". Raatikainen writes nothing about this.

    In fact, the only paper I find talking about this is from David Wolpert (from 2008) in his Physical limits of inference, where the abstract states the following:

    I show that physical devices that perform observation, prediction, or recollection share an underlying mathematical structure. I call devices with that structure "inference devices". I present a set of existence and impossibility results concerning inference devices. These results hold independent of the precise physical laws governing our universe. In a limited sense, the impossibility results establish that Laplace was wrong to claim that even in a classical, non-chaotic universe the future can be unerringly predicted, given sufficient knowledge of the present. Alternatively, these impossibility results can be viewed as a non-quantum mechanical "uncertainty principle". Next I explore the close connections between the mathematics of inference devices and of Turing Machines. In particular, the impossibility results for inference devices are similar to the Halting theorem for TM's. Furthermore, one can define an analog of Universal TM's (UTM's) for inference devices. I call those analogs "strong inference devices". I use strong inference devices to define the "inference complexity" of an inference task, which is the analog of the Kolmogorov complexity of computing a string. However no universe can contain more than one strong inference device. So whereas the Kolmogorov complexity of a string is arbitrary up to specification of the UTM, there is no such arbitrariness in the inference complexity of an inference task. I end by discussing the philosophical implications of these results, e.g., for whether the universe "is" a computer.

    As Raatikainen doesn't write at all about anything above what Wolpert is saying or perhaps isn't aware of this, this isn't a clear cut deal. Wolpert maybe just one of those "anti-mechanists" Raatikainen is talking about. So I think is really something still debated...and hence something really worthy of a discussion here.

    Notice that the similarities aren't so easy to pick, even for Gödel it took a while to understand that his and Turings findings are equivalent. And quite often many people still attack Gödel's results accusing him of just finding the paradox of "this statement is false" and simply purpose banning all self referential statements. And not just on the PF site. These opinions just show how people haven't got around to understand the undecidability results.
  • dimosthenis9
    846
    These opinions just show how people haven't got around to understand the undecidability results.ssu

    Well that's normal I guess, since the undecidability results are connected to the root of the deepest philosophical questions.Such as free will, the observer problem,etc and downline even consciousness itself.

    Of course interpretations of what Gödel's theorem actually shows vary.I read an article long time ago about how Gödel's theorem proves God's existence!(wtf?!!!?).People still debate for less complex things than that.So I guess this isn't a surprise.

    For me though (from the things i have read at least so far) his theorem actually shows logic limitation when it comes to self-refference systems.If we actually accept that mathematics are the highest form of Logic of course.
    Shows that Logic simply isn't enough when it comes to statements about self refference systems.Maybe not about all statements but at least to some of them it fails.

    Well which exactly are these statements and what these Logic limitations actually mean though on issues like the free will problem for instance, is a different thing and still open of course.
  • Ludwig V
    1.7k
    And I really don't think Laplace was trying to convince us that such a demon is likely, or possible. He was just saying, in a universe where everything is deterministic, anything at any point in the future would be, in theory, calculable.Patterner
    Yes, that's right. But that form of determinism does not amount to anything that could threaten freedom. There's a difference between being able to determine which horse will win the race, in the sense of being able to predict the result of the race and being able to determine which horse will win the race by fixing the race. Laplace's demon can do the first, but not the second.
  • Patterner
    996
    Yes, that's right. But that form of determinism does not amount to anything that could threaten freedom. There's a difference between being able to determine which horse will win the race, in the sense of being able to predict the result of the race and being able to determine which horse will win the race by fixing the race. Laplace's demon can do the first, but not the second.Ludwig V
    Agreed. The question of freedom arises when asking whether or not the decision to fix the race is anything other than physical interactions. Are we anything other than extraordinarily complex wind up toys?
bold
italic
underline
strike
code
quote
ulist
image
url
mention
reveal
youtube
tweet
Add a Comment