Is the Universe computing something?

Discussion in 'Astronomy, Exobiology, & Cosmology' started by arfa brane, Jan 26, 2016.

  1. hansda Valued Senior Member

    Messages:
    2,219
    At any instant of time we can consider "past" as the input and "present" as the output. Then we can consider "past" and "present" as the inputs and "future" as the output. In this fashion, at every instant of time, dynamics of the Universe can be analysed.
     
  2. Google AdSense Guest Advertisement



    to hide all adverts.
  3. PhysBang Valued Senior Member

    Messages:
    2,422
    It can be analyzed, but not as a computation. In order for it to be a computation, it has to take identifiable discrete states into identifiable discrete states, and physical systems do not do this. We can mimic this with some physical systems.
     
  4. Google AdSense Guest Advertisement



    to hide all adverts.
  5. hansda Valued Senior Member

    Messages:
    2,219
    Who knows... Somebody can develope a software analysing the universe. Nothing is impossible...Mind is working. Mind can compute.
     
  6. Google AdSense Guest Advertisement



    to hide all adverts.
  7. Edont Knoff Registered Senior Member

    Messages:
    512
    No, really no. Computers do not approximate. They have limited data types, but within the limits of the data type, the calculations are, well, calculations and no aprroximations. (Some computers (CPUs) were built badly and calculated wrong - this is not a general problem of computers though, but a faul in these models and series).

    There are programming languages with "infinite precision" calculations. These are as precise as the memory of the computer allows - each digit of a number must be remembered, and this is not different between a human brain and a computer. Also there are symbolic calculators which go at full precisions even with numbers like PI or the root of 2, which have an infinite amount of digits and which cannot be written down on a limited storage of any sort - the symbolic calculators still give you the correct and infinitely precise result, by handling such numbers as symbols and not sequences of digits.

    Google "Maple" or "Mathematica" (computer programs for math). Also google "Algorithm" and "Algebra". Please.

    There are many things computers can't do - but calculating, they can. It's what the were made for in the first place, and it's what they do best.

    And furthermore - calculation is a pure mechanical process. It needs no mind or consciousness. It's just combining symbols via rules and tables. There have been mechanical calculators, and that we use electronic ones these days, it's just that they are smaller, faster, and lighter than the mechanical ones. But they are basically not more precise (also not less) - because the scheme of calculation is the same.
     
  8. hansda Valued Senior Member

    Messages:
    2,219
    Computers can do many things but they take some time for their computations. There is some time delay between the input and output of a computer. A computer can not do any computation instantaneously without any time delay. Whereas response of the universe is instantaneous. There is no time delay.
     
  9. PhysBang Valued Senior Member

    Messages:
    2,422
    Yes, really, yes.

    The best that we can do in setting up a physical system is approximation. If one has ever had a laptop crash, then one knows that physical system is not doing perfect computations.

    This is not something about language, it's a restriction on physical possibility.
    Calculation is a mental process, it is just one that does not require creativity. This means that we can approximate it and simulate it with mechanical devices. In order to do so, we have to set these devices up very specifically with our decisions.
     
  10. Edont Knoff Registered Senior Member

    Messages:
    512
    I thought the argument was that it requires a mind to do calculations. I tried to show that it doesn't.

    Yes, you are right, calculations take time. But this also applies to the universe - speed of light limits the transmission of information, and thus even the universe cannot calculate instantly, unless this calculation happens in an infinite small space - which si also the reason why we try to make CPUs smaller and smaller, to reduce the distance a signal must travel, thus speeding up the calculations.

    There is no difference in this between the universe and a computer.

    And there are computers coming up which can use quantum states, e.g. do several related yet distinct calculations in one step, and decide for the "right" result in the last step, while haveing computed thousands of possible results in the first step. This is something, that the universe does, too.
     
  11. Edont Knoff Registered Senior Member

    Messages:
    512
    A crashed laptop happens if the software drove the hardware into an erroneous condition. This is only very indirectly linked to the question if calculation can be done correct and at full precision with a mechanical device.

    Just because the construction somethimes has mistakes, it doesn't mean, that the basic principle is wrong, or as you do, impossible.

    Laptops are just very complex - we should talk a simpler model like a touring machine. The turing machine has be shown to be able to do all calculations at perfect precision.
     
  12. Edont Knoff Registered Senior Member

    Messages:
    512
    Good. I have tried to show you the links, and tried to give you insight. I cannot compress the knowledge of a four year study into a few forum posts. I suggest you study the terms algebra, numeric and symbolic calculations, storage, algorithms and computational complexity.

    If you do, you will see that calculations are a mechanical process, which needs no mind. And you'll see that approximations happen only if the calculation needs more memory than a given calculator has. While the memory is sufficient, the calculator can calculate the full and correct result.

    If you are interested in the question how much time a calculation needs, there is a field of computational complexity, which researches just this. You'll see that there are distinct groups of calculations, some which take a fixed time, some need time protortional to the input data, more than proportioanl up to infinite time. You'll learn why this is so, and also, that the universe or any sort of mind is bound to the same rules - they are universal rules and not limited to computing devices.

    If you want to keep your opinion, I'm fine with that. I just want to let you know, that if a computer scientist reads your messages, and your arguments, you have a good chance that they will either shake their heads in disbelief, or just outright laugh at so wrong conceptions as you hold them, and don't even recognize them to be wrong when one tries to help you.

    I've studied computer science and I mastered the tests in calculation theory ... I know this does not mean I actually know something about it, but believe me, you are well adviced to review your knowledge. At the moment your arguments are not backed with good reason.

    I'm out of this discussion. I gave you the pointers, it's on you to read and learn, or let it be.

    Edit: Just realized I was talking to two different persons. Which makes some of my writing look weird. But I keep the core points, and also the links for further research.
     
  13. hansda Valued Senior Member

    Messages:
    2,219
    See my post #77. I explained why computer is not exactly computing.

    Light travels at its own speed. How it is a calculation problem for the universe? Speed of Light is the phenomenon of the universe. You can not change it.

    Computer at best can simulate the universe but a computer can not make another computer. It lacks the creativity which a mind has.
     
    Last edited: Feb 1, 2016
  14. PhysBang Valued Senior Member

    Messages:
    2,422
    It can also result from overheating, or electrical failure...

    There are many physical things that can happen to an actual computer that can keep that physical system from carrying out a calculation as we wish it to carry that calculation out.

    But the physical system did not fail to evolve properly as a physical system, it merely failed to act as we would wish it to in order to properly mimic a calculation.
    The Turing machine is an abstract machine. We can construct physical systems that mimic Turing machines, but they too can rust, or disintegrate, or otherwise more from one physical state to the next in a way that is not how we conceive of a computation.
     
  15. Edont Knoff Registered Senior Member

    Messages:
    512
    Calculations usually do not happen in an infinite small space. Once, information needs space to be stored, secone, many computations require several compontens to be involved.

    Since these componts are in differen locations, and since the storage of information requires space, the time to transmit a signal from one component to the other, or the space that an information takes (and therefore an examination must travel), space and speed of light put a limit on the speed of computations.

    But as I said, I'm out of this. The concepts that I show here are very basic to people who are into the construction of calculators. You are arguing in a field that is foreign to you, and you lack the concepts to develop the right ideas.
     
  16. Edont Knoff Registered Senior Member

    Messages:
    512
    It's been a discussion about the theory of calculation - if machines must approximate, and if a calculation needs a mind. Even if things like rust might influence any real calculator, it doesn't contradict the fact that a machine can calculate, and that it doesn't have to approximate, given that it has enough storage and time for the calculation.

    Sorry, that I sounded so harsh. I meant the other poster - since you know the turing machine, I assume you also know all the limitations and implications of the concept, and that's quite sufficient for the whole idea of calculations in theory and also the practical limitations - like only finite storage, and the mentioned defects any real machine will have.
     
  17. PhysBang Valued Senior Member

    Messages:
    2,422
    It's fortunate that I studied the foundations of computation then!
    You are confusing the mental constraints on calculation that make it something that we can more-or-less easily mimic with the machinery.
    Some might.
    I am relying mostly on the original work by Turing and others. The machines that Turing describes are meant to capture what people do in computing. They rely on identifiable symbols and to changes of states based on symbols.

    Physical systems can be made to react to symbols, but there is no sign that physical systems on their own react to some kind of ur-symbol that exists without human (or other sentient) identification.
     
  18. hansda Valued Senior Member

    Messages:
    2,219
    We are not exactly discussing a computer in this thread. Focus of discussion in this thread is the "universe".
     
  19. Edont Knoff Registered Senior Member

    Messages:
    512
    For sure. This gives us a common foundation to talk the concepts and the problems. My retort was aimed at the other person which claimed that calculators always must approximate, and that calculations need a mind.

    Sorry that I mixed that with your posts. The time when I studied is a while back and at the moment I think your knowledge is more up to date.

    I like the concept of the abtract machines, since they show which sorts of calculations can be done with a certain type of calculator, and eventually show the limits of calculations - e.g. semi-decidable problems and the ones beyond that. And also, that they show, in pronciple, that calculation is a mechanical process, which needs no mind. Only time and space.

    The argument of yours that real calculators suffer from further limitations is fully valid, and I don't want to oppose it. I just was treating this discussion on a more theoretical level, and therefore used the turing machine as the example of a calculator which can calculate everyting calculatable, and it can do that without approximating.
     
    PhysBang likes this.
  20. Edont Knoff Registered Senior Member

    Messages:
    512
    I thought it was you who said that calculations need a mind - this was the point that I wanted to disprove. Calculations are a mechanical process which does not need a mind. Furthermore, calculators of any make do not need to approximate, if they have enough storage and time. This also applies to the universe, which existed for a limited time and apparently is also limited in dimension, thus it has the same restrictions as any other calculator.

    Furthermore, the problem that information access and transport needs time, and that the speed of light puts a limit there, also applies to the universe. Even teh universe cannot calculate instantly, as you claimed in one of your postings.
     
  21. PhysBang Valued Senior Member

    Messages:
    2,422
    One thing to be careful about is that even the abstract machine has to represent a real number in some way. In many cases (uncountably many?), this will mean that the only representation the machine can produce in a finite number of steps is an approximation. We can set up machines to produce a representation of certain real numbers up to any desired limit of accuracy (can't remember the better term right now), especially if we use a nice system for encoding them.

    This is another reason to deny that the universe is doing a computation: there is no representation that it is doing and no accuracy bound.
     
  22. hansda Valued Senior Member

    Messages:
    2,219
    Mind calculates. Do you want to deny this?

    Dont mix up mind with a calculator or a computer. They are different entities.

    Have you found some fault with the universe?

    Universe is as it is. You have to accept the universe as it is.

    Can you link my post, where i said this.
     
  23. Edont Knoff Registered Senior Member

    Messages:
    512
    I think this is a really tricky point. At times we can use symbols like PI or sqrt(2) to denote numbers shortly which have an infinite amount of digits in decimal or binary representation, and the result of calculations will then be something like 15 * PI or sqrt(2) * 0.5, short and precise.

    In these cases we can avoid the problem of reading/writing/storing an infinite amount of digits. But given the sheer amount of numbers, I feel uncertain how many of them can be handled this way - all roots can, and some of the "magical" constants like PI, E and some others also can. If my brain isn't tricking me, there should be an uncounatble amount of numbers "in between" the roots and other "short" denotable numbers, and therefore an infinite number of cases, when real life calculations will be approximations, because the storage hits a limit. This is very reality vs. theory bites - in theory the process of a calculation is mechanical and doesn't need approximations, in real, it has limits. We need a bigger universe

    Please Register or Log in to view the hidden image!



    To answer that, we need someone with a better brain. For the sceptics - most computatiosn done in computers are doine with 64 bit precision, meaning about 15 decimal digits. I guess this is why many think, computers cannot do better, although this is way below the actual limit of a computer. It's just a compromise that has shown to be working well - precise enough for most applications and fast enough for most as well. Who needs more precision must use something else, and that's why I refered to tools like Maple and Mathematica, because they cannot only do more precise calculations, but also symbolic ones.
     

Share This Page