Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Big numbers don't exist

  1. Sep 17, 2004 #1
    I'm not convinced that big numbers really exist. To convey or represent information requires energy, so if we assume a finite amount of energy in the universe, then that means that the universe can only contain a finite amount of information.

    There are other arguments based on the hubble volume of the universe and how many states it can contain (by calculating the number of particles that can be crammed into that space, that calculation, discussed in a scientific american article was 10^118).

    In any case, that means if you pick a sufficiently large number, then you can say that all the number from 1 to that number can not be represented or expressed in this universe. And if there is no way to even theoretically express a range of numbers, how can you say they exist?
  2. jcsd
  3. Sep 17, 2004 #2
    do you mean by your post that infinity doesn't exist??
  4. Sep 17, 2004 #3
    No, not at all. That has nothing to do with it.

    What I mean is, pick a big integer, we can say 10^10000 for example. I don't think all integers from 1 to this number exist because they can not even theoretically be expressed.
  5. Sep 17, 2004 #4
    what you're saying, does it change in the rules that we consider nowadays?
  6. Sep 17, 2004 #5
    Sorry, your question doesn't make any sense to me. Can you rephrase it?
  7. Sep 17, 2004 #6
    i mean that if your statement is verified, will it change something in the theories we know.. in maths concepts..??
    (i don't know if i made myself clearer :$)
  8. Sep 17, 2004 #7
    Probably not, at least not directly. Which is why I posted it in the philosophy section. However, it could lead to interesting implications ... the idea that there is a finite amount of information that can be expressed in our universe.

    There are more pratical spin offs of the idea that I find strange to think about. For example, assuming your head doesn't keep getting bigger, and assuming that everything you "see" can be expressed as a "brain state" (mostly, but not necessarily limited) to your visual cortex, you get a weird idea. Your head is finite in size, and so there is should be a limited number of states possible inside your head. This means that there is a limit to what you can "see". Which means that if you lived for an incredibly long time and traveled all over the universe, unless you increased your brain compacity somehow, you would eventually see everything you could see ... in other words, new images would look exactly like an old image to you, you would not be able to see the difference, even if they were different.

    One can assume given the size of the universe relative to your brain, that there should be far more visual patterns out there than you have possible brain states.

    To me it's a weird thought, not necessarily practical in any way, which again is why I posted it in the philsophy section.
  9. Sep 17, 2004 #8
    while i started to read ur post.. an interesting idea got up to my mind..
    well u said that integers from 1 to 10^10000 do not exist..
    i guess they do exist but human beings brain is not yet so developed to understand such hard complications..
    and then, who told you..
    one can spend his lifetime counting from 1 to 100^10000 on sheets of papers..
    it can be.. but no one will do it though
    my idead opposes yours
  10. Sep 17, 2004 #9
    But that's the point, it can't be done, not with sheets of paper, not with computers, not with anything even theoretically possible as far as we understand it. The universe can not contain that much information. That's my whole point.
  11. Sep 17, 2004 #10
    i just said.. that it contains it..
    but "we" "human beings" we still don't have enough powers or enough intelligence
    to comprehend everything..
    don't forget..
    we are and we will always be a stain in a huge and wide ocean..
  12. Sep 17, 2004 #11


    User Avatar
    Staff Emeritus
    Science Advisor

    The problem is that gonzo never told us what he meant by "exist". Unless he is using "exist" is some very unusual limited way, like "someone has actually written it down on paper", all numbers exist in the same sense as all ideas exist. Since this is about, presumably, the philosophy of mathematics would say that some number that no has ever even referred to (and you don't have to go to large numbers- there are plenty of numbers no one has ever specifically referred to between 0 and 1!) exists in exactly the same sense that "one" or "two" exist.
  13. Sep 17, 2004 #12


    User Avatar

    For a number this size, 10^10000 can be expressed in distances for instance 1 meter is 10^9 nanometers c is 3E8 meters, so then there are (10^9 nanometers per meter)x(3x10^8 meters per second)=30x10^18 nanometers per second is the speed of light in a vacuum and the units can go smaller than this. 3.1536x10^7 seconds in a year. There are stars millions of light years away. Even if they are not 10^10000 nanometers away, they are a small part and light can travel back and forth, sooner or later it will reach this distance and every measurable distance in between. :yuck:
  14. Sep 17, 2004 #13
    u r talking about the distance..
    it is actually a point of view.
    can anyone talk about time? ;)
  15. Sep 17, 2004 #14


    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    I think your use of the word 'exist' is a little fuzzy. What does it mean for a number to exist? Certainly, assuming that the universe is finite, there is some limit to the amount of numbers that can be implemented / represented physically. (Actually, we would probably also have to stipulate that all physical quantities, including time and space, are quantized. Otherwise, for example, for any number n, we could consider any region of space to be composed of n units of volume.)

    But it seems that the notion of number is sufficiently abstract that to speak of a number's existence, one needn't refer to physically possible implementations of that number. To take your example, it may be that 10^10000 can't be physically expressed by a counting procedure, but it certainly can be theoretically expressed-- after all, I have just theoretically expressed it using only 8 characters. In the sense that I can speak of it at all, it seems to enjoy some kind of existence, even if it's an abstract one.

    This is an interesting idea. More relevant than the size of the universe, though, would be the number of retinal images that could impinge upon the retinas during the minimum salient duration of time needed to form a visual image. This gets a bit fuzzy, though.

    On the one hand, you probably have significant information loss since the brain probably treats large classes of retinal images as identical. (For example, if you modify your current retinal image by adding a few photons of any visible frequencies, anywhere on the retina, you won't notice a difference. Thus for any given perceived visual images, we can already conclude that there are a large number of slight variations of the retinal image that are perceived identically.) On the other hand, you also have higher order gestalt visual processing which could interpret identical retinal images in a number of different ways. (For example, an ink blot literally looks different once you have decided that it resembles, say, a cow, even though the stimulus has not changed.)

    Just for fun, let's estimate a rough upper bound for the number of possible visual conscious states possible in a given time period. One way to tackle this would be to roughly divide the visual field into pixels of minimum visual acuity and multiply the number of pixels by the number of noticeably different colors. But this method seems overly simple, as it neglects important higher order effects like perceptual grouping, foveal vs. peripheral vision, etc.

    Another way to do it would be to count brain states, as you suggest. Assuming:
    1) Conscious processes are supervenient upon brain processes;
    2) A sufficient level of granularity to describe this supervenience is determined by the firing speed of individual neurons;
    3) We can't a priori count out any part of the brain as having a potential impact on vision (although the visual cortex is most associated with vision, it seems to be influenced significantly by interaction with at least motor capabilities and higher order thought, and probably other many other capabilities supported in various brain regions);
    4) The average human brain contains roughly 10^11 neurons (http://faculty.washington.edu/chudler/facts.html);
    5) The maximum firing rate of neurons is 250 Hz ("maximum+neuron+firing+rate"&hl=en)
    6) Neuron firing rates only make a difference to conscious experience in discrete steps of 1 Hz. (There probably is some minimal difference beyond which neural firing rates make identical contributions to consciousness, given that at least some aspects of consciousness, like visual consciousness, apparently comes in discrete steps (eg http://www.nybooks.com/articles/17030), and it's probably lower than 1 Hz-- but our overestimate is going to be large enough as is, and assuming 1 Hz will simplify calculations.)

    So if we have 10^11 neurons, each of which can assume 250 causally significant 'states' (firing rates) with respect to consciousness, then there are 250^(10^11) different ways to construct brain processes that are causally distinct with respect to consciousness in the span of one second. Of course, the vast majority of these permutations probably don't result in consciousness at all (ie when all neurons are cycling at 1 Hz), and in principle most of them are also not physically realistic since the cycle of any given neuron will be a function of the cycles of its dendritic neighbors, whereas here I assumed that they are completely independent. But there you have it: we can be reasonably certain that the number of ways to experience consciousness in the span of one second is less than 250^100,000,000,000. :biggrin:
  16. Sep 17, 2004 #15


    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    To show how your (gonzo) definition of the existence of a number is very selective, and that there do exist "numbers" bigger than any given (real) number, here's a counterexample.

    Let the total number of states be N. Then the cardinality of the powerset of this set of states is 2^N, which is bigger than N. This can be extended indefinitely.
  17. Oct 2, 2004 #16
    In fact big numbers such as 10^10000 place limits on our normal reality all the time. For instance, cryptography is largely based on the fact that the space of possible keys is so large that no one can search it. One Time Pad keys can easy have 10^10000 posssible forms. If only small numbers existed you could easily search all the small ones and find the key. And it wouldn't stop at cryptography. Related to the brainstate/retina comment, imagine searching through every possible 100x100 24 bit image. Most of it's just noisy snow, but it also contains a low res version of every image a human could ever see. But then the large number 16777216^10000 steps in and prevents us looking. Even the space of a stark black and white image 10 pixels by 10 pixels (2^100) is too large to search. The same thing happens with potential songs, books and computer programs. If the number of potential patterns wasn't so huge (i.e. if large numbers didn't exist) we could just find all the good ones by scanning every possibility.

    On the other hand numbers like 10^10000 don't present any problem when in comes to manipulating them. For instance, I can say that if I divide that number by 10 I get 10^9999. Log math is pretty easy. Does counting up to 10^10000 by lots of 10^9999 sound ok? It's not necessary to count by single units, it can make more sense to talk about bags of rice instead of individual grains, without losing accuracy.

    (I made a correction here. I originally said counting by lots of 1^10000. Unfortunately 1^10000=1. Oops. And after me saying log math was easy...)

    By the way, in a more general sense of existance, whether any mathematical concept (even just 1 or 3 or 1/2) exists is a hot philosophical topic. Mathematical Platonism says that they do, abstractly and independent of physical existence. Mathematical Formalism says math is squiggley lines on paper (or computer) that we manipulate according to rules. Cultural Relevatism says math is just another cultural concept like Satan or Shiva or The Simpsons. I think there is some value to all these points of view.
    Last edited: Oct 2, 2004
  18. Oct 17, 2004 #17
    that number will eventually exists, in terms of age of the universe
  19. Oct 24, 2004 #18
    or enough paper
    ... but of course the number theoretically exists:
    10^10000 <--- there you have it. theoretically... i can even devide it by two:
    the issue is, of course, capacity. what if it was written in atoms? would it fit into space then? if not, then the universe cannot contain this number, if yes, then do'h...
    it doesn't seem plausible to me that there arent
    10^10000 / 6*10^23 moles of atoms in this univers...

    another proof:
    kryptographers have discovered a prime number with over 7 million digits... that far surpasses 10^10000...
    Last edited: Oct 24, 2004
  20. Oct 24, 2004 #19


    User Avatar
    Staff Emeritus
    Gold Member
    Dearly Missed

    The representation of a number (say in decimal digits) is not the same as the number.
  21. Nov 8, 2004 #20


    User Avatar
    Gold Member

    All we need for a number to exist is a representation of it. In theory, every single number imaginable can be expressed, because whenever a system starts to fail in representing numbers, we define a new system.

    For example, 1.0 x 10^3 works nicely for representing fairly small large numbers....but
    1.0 x 10^ 100000000000000000000000000000000000000000000000000000000.....continuing out to 1 million zeroes would be annoying to write down. So what do we do? we define a variable, z, to =1.0 x 10^10000...out to 1 million zeroes. Suddenly, this number is z, not a big long expression. And once z x 10^1000000....out to a million zeroes becomes too hard to handle, we define z' to equal this number....in this manner any number, no matter how large, can be represented.
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?