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

What is Information?

  1. Dec 29, 2011 #1
    They (physicists) say that information can't be lost and I heard Susskind saying that he thought information of objects going into a black hole was stored in the outside of the black hole while the material was sucked in. I thought information was almost like a fossil, where an imprint of an event is left on lots of other things, however that can be destroyed so I must have that wrong. What is information (in the simplest terms you can think of)?
  2. jcsd
  3. Dec 30, 2011 #2
    Great Question!! I've been puzzling over this myself.
    I think we are talking about physical information, as in one spin can contain 1 bit of information. it is either up or down. (or is it two since it also has direction?). When we measure it, to know it is there, must we be recieving information from it, so that that scrambles the information that is in it? eg when we measure the spin along the X or the Y direction when it starts off in the Z.

    Wiki has a nice article that probably wont answer your question with satisfaction (http://en.wikipedia.org/wiki/Physical_information): [Broken]

    Information itself may be loosely defined as "that which can distinguish one thing from another"
    Last edited by a moderator: May 5, 2017
  4. Jan 1, 2012 #3
    Yeah I tried to understand it but I couldn't completely.

    So would physical information be something that occurs simply at the quantum level? I got the impression from the Hawking Susskind black hole debate that it had to do with reversibility, so since things happening at our level are part of reversibility why wouldn't everything be included?

    If it did include things happening at our level it seems like alot of it would get destroyed but I guess if you're good enough you can still figure it out. It also sounded like physical information has to do with messages being sent off.
  5. Jan 1, 2012 #4


    User Avatar
    Science Advisor

    Last edited: Jan 1, 2012
  6. Jan 1, 2012 #5
    Padraic: there is no simple explanation of information that is very useful. At least I have never found one. Maybe someone will surprise!

    One thing we can say: In general, information as the answer to some question should reduce your uncertainty about which among possible answers is correct.

    A source of a message should have HIGH uncertainty, otherwise you'll know what the message contains before it is sent...we say it has high entropy, yet we want a received message to have LOW uncertainty...so we can understand it!!!

    yet a stream of bits like 1,1,1,1,1,1,1,1,1,1,1,1,1...........doesn't tell you much...you know what the next bit will be....
    The above ideas pertain most directly to communications 'information'. From an engineering [applications] perspective, that much is pretty well understood.

    Let's say I want to gather some "information" about local particles: Accelerated detectors will register different particle counts than an inertial detector! This is called the Unruh
    Effect. So how many local particles and the energies (temperature) I observe is not so simple.

    Let's switch and talk about manipulating 'information' (bits) in a computer:
    only erasure costs energy...that means it increases entropy...so when entropy increases information decreases. This is called Lanudaur's Principle.


    So at the end of the universe as entropy is at a maximum...information will be at a minimum.
    Reversible processes behave one way, irreversible another.

    As another perspective: someone posted this elsewhere and I liked it:

    The above means different observers see different things based on different available information but they should not be inconsistent.This also applies in cosmology.

    Finally, you might want to look at 'entanglement', an idea of quantum mechanics.
  7. Jan 2, 2012 #6
    Wow...here is something crazy:

  8. Jan 2, 2012 #7
    And another perspective:
    from Post #1 here:
  9. Jan 4, 2012 #8
    Sorry it took me so long to get back to you guys.

    How is a mixed state a real thing and not just something we don't know? If something is 50% probable this way and 50% probable that way isn't it really just one or the other but we don't know which one?

    I'm pretty familiar with entanglement so if you can explain it from that it might help. I still don't see why something like what you are discussing can't be destroyed. I still don't get what is being stored in the outer part of a black hole while it's mass falls in.

    That sounds like a whole nother can of worms but it sounds incredible. I've started reading it.
  10. Jan 4, 2012 #9
    I'll tackle one small piece:
    join the crowd....there are some explanations, but are they correct????.....

    you can check in wikipedia I'm sure....

    ok, here is one good place:


    So information appears to be related to enclosed surface area NOT volume!!!! What a crazy
    situation. Information in a black hole is FINITE>>>>

    Leonard Susskind in his book THE BLACK HOLE WAR (his controversy with Stephen Hawking)
    has some really interesting insights on information and horizons.....like the horizon of a black
    hole is "stringy"...it can be described in terms of quantum strings.....and so hiodden information is proportional to the total LENGTH of a string!!!!....and Hawking radiation can be viewed as string bits breaking loose from just outside the horizon...due to quantum fluctuations....a perspective akin to virtual particles causing the Hawking radiation.

    Also: consider acceleration....that also causes a horizon to form...which leads to a temperature (called the Unruh effect)...that means different particles are observed under
    acceleration....this is the same as sitting stationary outside a black hole horizon......
    so just what information is available locally?? not as obvious as once thought. A stationary and accelerating observer passing each other locally read different temperatures!!!!!
    Last edited: Jan 4, 2012
  11. Jan 4, 2012 #10


    User Avatar
    Gold Member

    No, most emphatically not. Quantum indeterminacy is NOT a measurement problem it is reality.

    The most graphic demonstration of this, to me, is the concept of an electron going through a slit (see the "double slit" problem) and hitting a phosphor screen. At any specified tiny amount of time prior to it striking the phosphor, we have NO idea where it is except that there's a low probability that it is anywhere near the place where it strikes the screen. In classical physics, which seems to be how you are looking at things, this is nonsensical. Welcome to the quantum world.
  12. Jan 4, 2012 #11

    Ken G

    User Avatar
    Gold Member

    It was my quote above about information available locally, and I confess I can't give a precise meaning to what that is (maybe that is a crucial question of current research). But I think Unruh radiation doesn't break the importance of thinking in terms of the locality of information, because for an accelerating observer to read a different temperature requires some time for the observation, and so they cannot remain in the same reference frame as the observer reading a different temperature. Thus they cannot have access to the same "local" information, because reference frames are also local, so changing reference frames requires changing the available information. In special relativity, a reference frame is viewed as global, but Unruh radiation requires general relativity, and there reference frames (AFAIK) should be regarded as local, and should also be regarded as having access to local information.

    So there might be a direction connection between information and reference frames-- the information that any observer could call "complete" is connected with the reference frame of that observer, and some "global union" of complete information would then require uniting the information of all the reference frames, like putting together a jigsaw puzzle. There would be significant redundancy there, but it would not be completely redundant, so I think understanding the redundancies and independences of the information available in different reference frames might be a good way to connect information in relativity with information in quantum mechanics. I feel that physics must always be about information, so unifying theories of physics can be framed as ways of unifying the various treatments of information.
  13. Jan 4, 2012 #12

    Ken G

    User Avatar
    Gold Member

    Actually, that question is at the heart of a lot of the distinctions between the various interpretations of quantum mechanics. We should distinguish this from the concept of quantum indeterminacy, which actually applies to pure states (superpositions of states with definite outcomes for some given measurement). Quantum indeterminacy does indeed seem to be part of reality, but mixed-state indeterminacy is harder to say. In the Copenhagen interpretation, for example, a mixed state is interpreted as you say-- it is one or the other we just don't know which. In Everett's many-worlds interpretation, the mixed state is interpreted as a projection onto some measured subsystem of what is in reality a much larger pure state that includes the entire system (including the observer), and the larger state includes all the possible outcomes even if we don't perceive them.

    So in the two-slit experiment, for example, if the apparatus does not determine which slit the particle goes through, then the particle is in a superposition of going through both slits. I wouldn't say it actually does go through both slits, merely that it is in a superposition that does not answer the which-slit question. But if the apparatus does determine which slit, even if we are not privy to that information, then we have a mixed state of going through each slit, and in the Copenhagen approach, we can say the particle actually did go through one or the other. So the information about the particle has changed, and that is irreversible, whether or not we know that information.

    As to whether or not a mixed state "is the real thing", you need to take a stance on some flavor of realism. To Bohr, a state is never a "real thing", it is always a description in the mind of the physicist who is trying to understand the real thing. But the same holds for information and entropy-- they are live in the mind of the physicist doing some analysis, and their connection to reality appears only in how they work for some goal. They are the ways we talk about reality, which is not the same as the reality itself, but nevertheless are effective in theories like thermodynamics so connect perfectly well with the concepts of information and entropy.
  14. Jan 5, 2012 #13
    Ken G:
    that seems to be the situation.

    yes, the only thing I'd mention that frames are also local in SR.

    quite possibly; I don't recall the details but the appearance of a horizon is behind the Unruh effect.... like Hawking radiation just outside a black hole horizon...And as Leonard Susskind points out ,changing the encompassing horizon changes the "location" of the information to a new horizon...say a bigger one....so not only may information be two dimension, it is not so local as one might think. Apparently, evey 'bit' of information in our universe is stored on the enclosing boundary(horizon)....Maybe the cosmic censorship hypothesis is also linked to all this "information"....
  15. Jan 5, 2012 #14
    Wouldn't a simple description of information be like this:

    "Information is anything that can be used to help find the cause to an effect"?

    effect: a neutrino hits a detector
    cause: a star went supernova
    information: any form of data that could possibly be collected from the neutrino that would be needed to help find out the who, what, when, where, or why of the effect (in this case a supernova).

    In the listed example information could be how fast was the neutrino going, what direction was it going, when did it hit the detector, how energetic was the particle, all of which could be used in some way to find the who, what, when, where, and why of the source.

    Why, as far as I am aware, that Hawking radiation loses information is because the particle that is created is not capable of giving all the information of the cause of it's existance (the effect).

    Back to my neutrino example, you can find out exactly when where how and why it was created from it's information theoretically speaking that is. You can know how long it's existed theoretically, how fast it was going which way it was going, etc. etc. True Hizenberg's Uncertainty Principle won't allow you to know all these things at the same time, but the information is there. You can find any of these things individually from a particle, therefor the information exists.

    With the Hawking's radiation since the particle's anti-particle pair was sucked into the black hole if you track the particles information back to the moment of it's original creation you still would not be able to find out the everything about it's creation since it's anti-particle pair technically doesn't exist anymore because it has been condensed down to a singularity where, according to our best available understanding and math, is a place where infinities and all sorts of chaos occurs, the information needed to be able to find out using pure math the cause for the effect of the existence of the particle is lost. Thus information is lost.

    TL:DR version:
    Laymans version of information that's probably wrong.
  16. Jan 6, 2012 #15

    Ken G

    User Avatar
    Gold Member

    There might indeed be some deep duality between how we think about what local means in spacetime, and holographic kinds of thinking about information on surfaces or horizons. Or maybe what we think of as local information is actually some kind of change in information or expectation across some surface. I don't have a clear sense of what information really is, other than how we use it in specific situations, but it does seem to have something to do with changing expectations, with winnowing down the possibilities.
  17. Jan 16, 2012 #16
    Mathematically there are currently two types of information: macroscopic and quantum. My knowledge of quantum Information, accounted (not measured) with units of qbits is currently minimal and I will not address that here. Macroscopic information can be further divided into two sub-classes. First there is an absolute information, defined by Shannon's entropy measure and which involves an implicit comparison of an observed system against a virtual uniformly random background. This is the measure of information used in communication theory, and is of particular importance in showing how to send the most message content in a given bandwidth within a noisy universe. A second macroscopic measure is known variously as relative entropy, relative information, or relative information. An important example of this type of information measure is the Kullback-Liebler divergence, which for some time has been known to provide a basis for unification of diverse statistical methods. The Kullback-Liebler divergence is a measure of the number of bits of information available to make a discrimination in favor of one probabilistic description over another as a correct description of a physical system. This is potentially useful in physics since cause and effect states suggested by an earlier correspondent's note as information are, at their core, different states which can be subject to statistical comparison. The Kullback-Liebler divergence has several interesting properties: it is invariant under a non-singular transformation of random variables, and the lowest order expansion in probability parameter differences is a quadratic form based on the Fisher information matrix; the Fisher information matrix has been used as the basis of some derivations of quantum mechanics.

    There have been some suggestions that a truly fundamental physical theory (potentially more fundamental than current quantum theories) might be based on information. Such an approach has the curious potential to make some thermodynamic concepts, which are now understood to be based on the underlying statistics and information constraints, to be regarded as more fundamental than quantum theory.
  18. Jan 16, 2012 #17

    Ken G

    User Avatar
    Gold Member

    That's an interesting comment. Personally, I feel that the next great theory will need to move closer to an understanding of how we do physics, and farther from the model that physics is about things that have nothing to do with how we think. Recognizing the importance of information might be a step in that direction, if we also recognize our own role in information processing.
  19. Jan 21, 2012 #18
    For those interested in some different perspectives on information, consider the book
    DECODING THE UNIVERSE, by Charles Seife...
    from biological reproduction to Heisenberg uncertainty.....
  20. Jan 21, 2012 #19
    i agree, i think this is the essence of heisengber uncertainty and infromation.

    for something to be ontological (that it can exist), it must have effect, and must be also epistomological (knowable), we must be able to know about it, to learn it is there.

    we can only deal with things we know are there and exist, that is basically science (in some coarse outline :) )

    but to know it is there we have to intearact with it, and the thing with the microscopic world is that any interaction is bidirectional, coupling to the system affects the system and also affects the measurmnet system as well.

    mesurment is no different in essence than any system being measured.

    to be more concrete, consider how qubits are read out dispersively,
    the qubit energy levels get coupled to a SHO resonator (cavity qubit can talk to),
    if we put the cavity near the qubit at a smaller frequency,
    the interaction shifts the energy levels of the cavity slightly,
    the one photon in cavity qubit levels get pushed up and down and by the zero photon exited energy level of the qubit, and so the cavity resonant frequency shifts depending on the state of the qubit.
    i think this picture is at the essence of measurement.
  21. Jan 21, 2012 #20

    Ken G

    User Avatar
    Gold Member

    Yes, I agree that we learn about things by interacting with them, it was always kind of a cheat to pretend we could "observe" rather than "interact."
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook