Banner
    Quantum Randi Challenge: Help Perimeter Physicist Joy Christian To Collect The Nobel Prize
    By Sascha Vongehr | June 1st 2011 10:26 PM | 99 comments | Print | E-mail | Track Comments
    About Sascha

    Dr. Sascha Vongehr [风洒沙] studied phil/math/chem/phys in Germany, obtained a BSc in theoretical physics (electro-mag) & MSc (stringtheory)...

    View Sascha's Profile

    You know the James Randi Challenge right? James offers $1000000 to anyone that can demonstrate paranormal abilities under laboratory conditions. Well, today we have something big to announce: A very similar challenge (now officially on the archive here), but much easier to accomplish.

    The rewards are immediate fame (in a few weeks from now if you like) and the reception of the Nobel prize in physics later on – no doubt about it.

    In fact, we have our first contestant, but we need you to help him: Joy Christian at the prestigious physics department of Oxford University, also at the Perimeter Institute of Theoretical Physics, and funded by a grant from the so called “Foundational Questions Institute (FQXi)”.


    Joy Christian is waiting impatiently for you to drive to Sweden with him.

    Local realism has been strictly disproved; here is a lay-person-accessible version of that proof. But this is just me talking.

    The great news for you: Joy Christian claims that he has published a model that gives rise to the predictions of quantum physics via classical, local physics. Joy even managed to get this into a new book published by FQXi. The title of his chapter seems to be “What Really Sets the Upper Bound on Quantum Correlations?

    This great man’s great insight is that the so called “co-domain” is not the usual three dimensional XYZ space “R-cubed”, but a three dimensional sphere that happens to be the equator of a four dimensional sphere. Pure geometry and the formulas are subsequently simple. There are not even any imaginary numbers – just real valued tables that are called real quaternions. Basically high school math.

    Why is this great news? Well, see, if there would be even a single case of somebody coming up with quantum behavior where it is proven impossible to do so, even just in the simplest case, this would go on the front page of every newspaper. Now we have a way to do just that. You can now do it with help of Joy, so hurry up before somebody else does it. He wrote to me in an email that he cannot write a single line of computer code, which may explain a lot, but apart from that: THIS IS YOUR CHANCE TO DRIVE TO SWEDEN WITH HIM!

    Here is the great news that I have for all who believe in local realism; I also already told Joy and he was excited about it – no joke(!):

    Local realism of a model, any model, implies that a classical computer model must (must!) be possible!

    You guys claim physics is local and real, so a setup with usual computers replacing all physical objects and measurement apparatus is possible – that is what local realism basically means: All information is locally available as if particles carry classical computers around, telling them what to do.

    Now you may think that programs could still be very complicated. The good news is: it is simple. Programming the game described in Hidden Variable Madness according to the instructions given in the email from 16 May is in fact a piece of cake for anybody with just a little programming experience (here is the core algorithm already), and this is me saying so; I am one of those late adopters – not a computer geek at all.

    It is a very simple setup that just needs considering few angles. The program does not need any graphics or knowing anything complicated about angles at all. All it does is sending 800 pairs of different message strings to Alice and Bob, who are maybe two players far apart or just two simulated players in the same computer for starters. Their computers select “angles” at random, read out the resulting “measurements” from the received strings, and then the 800 results are analyzed with a very simple formula kind of adding them up. The whole game can be played in a few seconds. Programming time modifying my program with using randomly generated “simple minded hidden variables”: Maybe two weeks - if you are slow.

    The news become better and better: Joy’s claim is that if you just take care of that the computer simulation does not simulate Bell’s “simple minded hidden variables”, but instead use his ‘advanced minded’ four dimensional sphere, then you will be able to simulate quantum behavior with the help of his formulas. That modification of the program should take a day tops! Why?

    Virtual reality in the computer is anyway anything you want it to be: actually the computer does not even know what space it is simulating. Joy insists on that our “simple minded hidden variables” somehow reflect R-cubed space. Well, there is absolutely no problem at all to simulate a four or five dimensional sphere in a computer instead. Gosh, there are simulations of crazy black hole/ ripped worm hole symmetries nowadays on computers. So, just go ahead and change the hidden variables so that their correlations mirror the spherical space instead of the simple minded R-cubed - a sphere is no more difficult than a cube (and I'd write the same if it would have to be a 5 dimensional torus, too). That is all – that is the claim – you are now a famous person!

    If you are interested in becoming famous in about a couple of months at the latest, read the emails where I explain what to do. It is real simple: Bell’s “simple minded hidden variables” are programmed already. The simulation does of course not show quantum behavior. Just modify anything that looks like a simulated R-cubed space into a certain sphere that Joy describes in his archive papers – he will eagerly help you as he does with Albert Jan, who is also interested in simulating his model.

    This is for real! Hurry before somebody else does it. From our email exchange, which by the way is quite entertaining with highlights such as "What are you trying to say? That the recorded data change inside Bob’s lab-log while he drives to Alice’s place?" answered by “Yes! In a sense, that is exactly what is happening.”:

    “The programs are open source; everybody can check that the programs do not secretly establish an internet connection between Alice’s and Bob’s computers after the angles are chosen (in other words: there is no cheating via secret non-locality). If the Bell inequality is violated in this way, it will be a huge confirmation that is going to spread over the internet like a firestorm.”

    Meaning: You do not need to publish anything in the biased peer review journals. The program is simply downloaded as a game and played by interested people who take on the role of Alice and Bob. If the Bell inequality is violated because of how you modified the “simple minded hidden variables”, then the game with your modification will be an immediate super star, just like you.

    And to all those who think they have an even better theory than Joy:

    The Quantum Randi Challenge:

    Anybody with some crackpot “local QM” theory is cordially invited to either write the program (as described here) so that Bell’s inequality is violated or to shut up (precisely like the traditional Randi Challenge)! And we want to see no other angles but those where quantum physics maximally violates Bell’s inequality. We have no interest whatsoever in all the other amazing stuff that you claim you can do with it. Recreate quantum behavior in that simple experiment at the few angles including Pi/8 and become famous; don’t waste our time with emailing! If you are successful, fame is yours.

    --------------------------

    Update: Didactic part of the QRC is now also introduced and defended in:

    S. Vongehr: Exploring Inequality Violations By Classical Hidden Variables Numerically.” Annals of Physics 339: 81-88 (2013), www.sciencedirect.com/science/article/pii/S0003491613001863 Preprint version adds section on realisms and shows programs with output: arxiv.org/abs/1308.6752

    Comments

    lumidek
    This Joy guy is an annoying crook. It's been always difficult to figure out whether all those folks believe their crackpottery or they're just saying it because they may benefit of it. I think that the conclusion is that they want to benefit and to do so optimally, they're working hard all the time to make themselves believe that their crackpottery is actually right.
    vongehr
    that they want to benefit and to do so optimally, they're working hard all the time to make themselves believe
    You scratch on something here that I think is a very important psycho-mechanism that humans employ naturally during rationalization, which is of course one of the main tasks of mind, and this mechanism is going to be the central issue in what is called "new enlightenment". We do precisely this, making ourselves believe what we want to believe, steered from some sort of semi conscious level/module that is aware of us being actually wrong or at least aware of pushing aside what we on that level quite well know to be arguments we should consider.

    Humans do this all the time, all of us. It is a for good reasons evolved ability of intelligent social systems.

    These crackpots and young earth creationists and what not, they are not bad, but with them, this mechanism is "over developed" so that it becomes clearly non adaptive (except in today's publish or perish/publicity/fame environment). But in the end, this mechanism is why humankind is in the quagmire it is in and only who is aware of their own doing this, too, can possibly have the rationality to grasp "the next level".

    Intelligence
    there is plenty! These years I discover that advanced rationality (e.g. in the sense of being aware of such mental rationalizations in yourself) is what is totally missing also and maybe especially among us intellectuals (I write "especially" considering environmental selection processes that made us intellectuals, say academics for example, in the first place).
    lumidek
    Sounds to me as a full-fledged sensible paper in evolutionary psychology. ;-)
    Hfarmer
    I am not going to argue with the rest of what you have written here.  However I will point out one minor flaw. You say
     There are not even any imaginary numbers – just real valued tables that are called real quaternions. Basically high school math.
    In fact one can formulate quantum theory quite nicely using quaternions instead of complex numbers.

    J. Math. Phys. 3, 207 (1962); doi:10.1063/1.1703794

    A new kind of quantum mechanics using inner products, matrix elements, and coefficients assuming values that are quaternionic (and thus noncommutative) instead of complex is developed. This is the most general kind of quantum mechanics possessing the same kind of calculus of assertions as conventional quantum mechanics. The role played by the new imaginaries is studied. The principal conceptual difficulty concerns the theory of composite systems where the ordinary tensor product fails due to noncommutativity.
    In fact the use of quaternions in formulations of quantum theory is so common and accepted that it's done quite often.  Even the great Edward Witten has done so

    http://arxiv.org/PS_cache/arxiv/pdf/1009/1009.6032v1.pdf_

    . The original complex structure J and the holomorphic two-form Ω are part .
    of the hyper-Kahler structure of cM. Indeed, a hyper-Kahler manifold has a triple of complex structure I, J, K obeying the quaternion relations I2 = J2 = K2 = IJK = −1.
    So either your comment on the use of Qaternions is wrong or Edward Witten is wrong.  I choose to believe Edward Wittten.
    Science advances as much by mistakes as by plans.
    lumidek
    Dear Hontas Farmer, the quaternions in the context are not used for anything inherently quantum. They're just used to describe a geometry of a manifold - hyper-Kahler structures are where quaternions appear - but this geometry governed by quaternions exists even at the classical level.
    One may (sometimes) take the resulting classical geometry and quantize it - but the quaternions don't enter into the "engine" of quantum mechanics. In particular, wave functions in the quantized theory Witten talks about are complex numbers, just like in all consistent quantum theories. None of them is a quaternion.
    Hfarmer
    Dear Lubos Motl.  
    Quaternions have all the mathematical richness needed to build consistent quantum theories and can be used in inherently quantum ways.  

    I did not learn that just now by googleing for some papers.  I learned it from the person who taught me quantum mechanics at UIC, Dr. Tomas Imbo.  He teaches it from memory and without a textbook he's just that freaking smart.  s.

    A simple and short way to see this.  

    Quaternions are SUPER SET of the complex numbers.  The complex numbers are contained within the set of quaternions.   Such is evident from the way the multiplication of their basis elements is defined. 

      

    Plenty of rich mathematical structure for building any quantum theory one wishes.
    Understand? :)

    Edited to add:  As a matter of fact one can even represent quaternions as a pair of complex numbers.  Thusly

    (a+bi)1+(c+di)j

    Now if the author that Sacha is talking about is saying that quantum mechanics can be represented just with real numbers (scalars.... like a in the above)  then yes he is a crackpot. 
    That said DO NOT claim that quantum mechanics could not be constructed consistently with quaternions because you are easily proven wrong.  What's more as I cited above not just witten but the earlier paper it was done way back in the 60's. 
    Science advances as much by mistakes as by plans.
    lumidek
    You're still missing the point. Quantum mechanics can only work on complex numbers and the rest of the quaternions, even if you added them, decouples. That's easy to see.
    There's one time coordinate, and the time-derivative of a wave function - and similarly for operators (or in path integral) - is given by -i times Hamiltonian acting on psi.

    Here, "-i" has to be pure imaginary, and you get an equivalent maths if it were j, k, or any combinations of i,j,k. Wave function that is quaternionic could be written as a+bj where a,b are complex numbers. The evolution would act on a,b separately, so the wave function would be equivalent to a mixed state with wave functions a and b, respectively. The whole quaternionic multiplication table would never be used.

    What you write is the same kind of irrational worshiping of the names of these algebraic structures, without understand what the structures can and can't do, that John Baez does especially with the octonions. Quaternions and octonions are OK and they appear in various structures but they're always exceptional structures of a sort and none of these algebras enters as fundamental frameworks as the general postulates of QM.
    The i in the Schrödinger equation can be replaced by any imaginary unit quaternion t, that is, t=ai+bj+ck with a^2+b^2+c^2=1.

    Indeed, one may replace "i" by "j" or "0.6 j - 0.8 k" or any unit quaternion but one gets an *exactly isomorphic* system of equations and the remaining two orthogonal quaternionic imaginary units still decouple.

    The equivalence of all these unit pure imaginary quaternions is guaranteed by the SO(3) automorphism group of the quaternions. Without a loss of generality, one may call the unit used in Schrodinger's or Heisenberg's equation "i", the conclusion is unchanged. The remaining imaginary units are unphysical and decoupled. There's no role for quaternions in the fundamental rules of quantum mechanics.

    The i in the Schrödinger equation can be replaced by any imaginary unit quaternion t, that is, t=ai+bj+ck with a^2+b^2+c^2=1. The decoupling only occurs when only one of a,b,c is 1, with the others zero.

    vongehr
    Well, you fully agree then. All I said is that he does not use any complex numbers, and you fully agree. Where is my "minor flaw"?
    Absolutely brilliant, thanks! But you are wasting your time asking Oxford or PI for help. I have some experience with both, and you might as well be talking to monkeys. If they will let me starve to death in the gutter, I guess they'll do the same for you. After all, professional theory is not actually about Science these days. How could they uphold their fine, outdated, patriarchal pole dance if people were actually allowed to speak the truth!! What were you thinking?

    When the quantum chit chat gets this dense I really struggle to sense the fallacies and understand the true logic, but I agree with Ms. Barratt, it is entertaining.

    Sorry for my questions in this nice discussion.
    But

    - Is an entangled state depending on a special temperature?

    - Is it possible that there is entanglement between photons/particles with different temperatures?

    The aim here is to prove or disprove local realism of course.

    vongehr
    Temperature emerges from quantum physics at the point where the system becomes too complicated to describe and we start allowing uncertainty (about what quantum states are involved, not the fundamental quantum uncertainty). At the level that we are discussing here (e.g. the photons between Alice and Bob, the quantum state of these photons is known exactly), there plainly is no temperature (temperature makes no sense on this level, as it has not emerged yet - it would not even be correct to say that it is zero). Entanglement between systems with different temperatures is possible, but their temperature in that case plays no role in the entanglement.
    Thank you very much Sascha!

    I adore your knowledge. But please show me an example where entanglement e.g. quantum states are possible with different temperatures.

    vongehr
    Generally speaking, we are all entangled. Once the entangled photons are observed, Alice and Bob are entangled just via that fact, and if Bob happens to have a fever ... .
    Hi Sascha,

    RE: Your diagram (above) and this quote: "However, the more fundamental description rests on the fact that the discussed crystals do something very similar: If a linearly polarized photon is going into the entrance channel and the relative angle between a certain crystal axis and the photon’s polarization is delta again, the probability to exit as a horizontally polarized photon through the H-channel is cos2(δ), and the probability of going through the V-channel instead is sin2(δ)."

    Quick question: Are you saying that the Alice-Bob (Bell) experiments are conducted with linearly polarized photons?

    Bell would say that they are unpolarized before they interact with a crystal; right?

    Thanks.

    vongehr
    The part you quote is about the more fundamental description of why there is the cosine square factor (because of QuantMech rather than ElectroMag), not about that the photon is linear polarized in the Alice-Bob experiment. However, each photon (if considered by itself, say when Alice considers the state of the arriving photon) is in a superposition of two polarization states, and these states (regardless the basis, as long as the basis uses two linear polarization states rather than say circular ones) each do indeed behave as linear polarized photons.
    Thank you; but, for me, this reply raises more questions: A superposition of two linear polarization states? Doesn't this apply to a linearly polarized photon? Bell says the particles are unpolarized; right? And they are; right? For the entangled photons do not behave as linearly polarized photons (LPPs) because their joint (entangled) correlation is twice that of such LPPs? And re the cosine-squared factor: Are you invoking Malus' Law? Or a new one?

    Am I therefore right in thinking that you have mistakenly invoked polarized photons in your model; which is therefore wrong as a model of entanglement or Bell's ideas?

    vongehr
    Sorry to be blunt, but at this point there is no shortcut to you sitting down and seriously study what a quantum superpositions is. It is not that difficult - after all we are talking linear mathematics.
    Your bluntness is fine with me. So can you confirm or comment on Bell's position, the essence of my questions: The particles are unpolarized, right?

    vongehr
    Yes, the particles are unpolarized. Nevertheless, for example the photon that goes to Alice can be described by a superposition of two basis states, and these basis states can be linearly or elliptically polarized, however you like. Each basis state will behave according to its polarization, however, the outcome of the experiment is due to the superposition! Because the maths is linear, it does not matter which polarization basis states you use to describe. One can Fourier transform music, but that does not mean that Lady Gaga only ever sings a pure cosine tone.
    Hank
    Amateurs love to invoke Einstein, and so I will: "Make everything as simple as possible, but not simpler".

    The notion that everything can be simplified indefinitely is a myth.  Sometimes physics is hard and it takes math and that is why not everyone can do it so if they aren't willing to do it, there is only so much you can do in the name of outreach.    I am over my head plenty of times but I always feel like I am at least converging on getting a little smarter.    You're pretty good about that.
    vongehr
    is hard and it takes math and that is why not everyone can do it
    Especially in this very context however we should not say this. The mathematics involved here is quite simple, as you can see from the posts on hidden variable madness and so on, which I have all written in order to prepare this very article here. Quantum mechanics is one of the few theories that is linear! It is the crackpots who use this line to support the megalomaniac notion that all others just do not understand their genius quaternion/GA/whatever mathematics. What I meant needs to be studied by Anon is not some higher maths but the understanding what a superposition state is and how quantum evolution works on it. I think the best way to learn such is books that use very little maths like QED by R. Feynman.
    I think that Joy Christian's work is good and he does indeed disprove Bell's Theorem. Others have too, like Han Geurdes, (I will post the reference later but it is on the archives.

    The reason I believe in the Clifford algebra of Joy is because I have come to the same conclusions, but rather than treating mathematically all the spins, I treat individual spins, and resolve his vector I = -isig(x)sig(y)sig(z) into non-hermitian states,
    sig(z)sig(x) = isig(y). It is not that simple, but I have a video blog that relates my work to Joy's and makes this clear.

    http://quantummechanics.mchmultimedia.com/2011/quantum-mechanics/009-dis...

    However on talking to Gregor Weihs he said that Joy's ideas cannot predict the statistical data of his experiments. Gregor says, rightly so,why should people look at a theory unless it generates his data?

    This is what I am working on now. First a product states from my theory works in a very simple way: Assume that Alice's spin gives a click, (the wave function must have collapsed and so the probability is 100% (pa=1) Then the question arises, what will Bob's get? and his is pb=-cos(Thetaa-Thetab)--the experimental result form a product P=pa x cos(Thetaa-Thetab). In order for this to happen, the LHV at Alice and Bob must be the same. Now careful thought leads to the conclusion that equal LHV mean only the two spins come from the same ERP pair (a coincidence click). However all the nay-sayers say it is signalling. That was the criticism of Joy's work too.

    Ok, so there will always be nay-sayers, and to answer them, we must generate the data separately at A and B (reproduce Gregor's data) and that is what I am working on. I am optimistic because my earlier product state (above) works (with the signalling criticism).

    In order to do this, I have to assume a spin is not a point particle but has a 2D structure, maybe something like a Cooper pair, but this is like a qft condensate between two usual spins. I discovered this in 2006 and submitted the first naive ideas on the archives up until 2009

    http://arxiv.org/find/all/1/all:+AND+b+sanctuary/0/1/0/all/0/1

    My work has been mathematically checked but has raised the hackles of the physics community (I'm a chemist) and those in QI theory, because it is clear from my work that the locality assumption is ok, but not his spin assumption. That is, QI rests on Bell's theorem, which is wrong.

    As a result, when I submitted papers, the responses are "do not like" "ludicrous" etc, but nothing objective.

    So now I am working on the statistical test. If it fails, back to the drawing board, and I will soon know.

    I would like to put my latest articles on the Archives but the moderated Black Listed me. They said publish first, but with referees who do not read and say "don't like" my feeling is I have to pass the statistical test first. Although I enjoy the blog, that has become the place where I explain the ideas, at least until I have some conclusions about my theory from the statistical test.

    lumidek
    I send my best regards to the Sanctuary, Bryan.
    Well, I surely hope that there will be "naysayers", as you call the people who understand basics of quantum mechanics. What I am amazed by is the fact that there are still so many "yes-sayers" who are able to defend the indefensible - something that can be seen to be wrong in a few minutes.

    When you say sensational things such as the "disproof" of Bell's theorem, have you actually ever seen Bell's theorem? Take e.g. the first 8 pages of

    http://www.karlin.mff.cuni.cz/~motl/entan-interpret.pdf

    where various versions of Bell's theorem are proved. College physics is enough to follow the text. I recommend you to read all 16 pages with the interpretation of QM and decoherence, too.

    Do you actually claim that the proof I wrote is invalid? Where is exactly the mistake? Could you please focus on the text I wrote which makes sense, instead of escaping to meaningless formalism mutating some formulae that you haven't really understood which is why you keep on writing parodies of it?

    Thank you very much.
    I have seen your pdf which is, in my opinion, the orthodox view with nothing new, but first, why do you say obnoxious things like:; Joy Christian is a crook,or doubling that I have ever read Bell. And you say to me: "instead of escaping to meaningless formalism mutating some formulae that you haven't really understood "

    Please tell me which formulae I mutated.

    All your ranting does not do much for collegiality and is not helpful. So without saying too much about your paper, which is basically the orthodox view, you say things like

    "Entanglement at a distance is probably the most shocking feature of quantum mechanics. This bizarre
    feature of quantum mechanics was first pointed out by Einstein, and later in 1935 by Einstein and his
    collaborators Boris Podolsky and Nathan Rosen (EPR). They hypothesized that quantum mechanics
    had to be wrong because in various contexts, it was predicting correlations between measurements
    that looked like voodoo to EPR. Meanwhile, the proponents of quantum mechanics insisted that the
    predictions of quantum mechanics were right."

    This was NOT pointed out by EPR. All EPR did was show that QM is incomplete, and said NOTING about non-locality.

    EPR also did not think their results were like voodoo; that came in 1964 with Bell.

    EPR did not say that QM is wrong, just incomplete.

    So if you think we are all crooks and crackpots, tell me this: what is the mechanism for long distance entanglement? You cannot tell me (I can: it is on my blog and the link is in my last comment). The best the orthodox people can do is say this is "quantum weirdness" and using your word, that is "shocking". This mean to me you do not understand what is going on in non-locality. (this is not a criticism: Gregor Weihs and Nicolas Gisin told me the same thing so you are in good company).

    Non-locality makes no physical sense, but to cover this up has led to placating comments like "QM and relativity live in peaceful coexistence" complete bunk .The Quantum Info people have circled their quibits to protect their grants because their whole existence depends upon the validity of Bell.

    Two things are needed:

    First the orthodox proponents should keep an open mind and not stick to the idea that Nature is statistical.

    Second, those of us who do not believe in it need to generate the experimental EPR data.

    Over to you Lubus.

    lumidek
    There is nothing new because almost all these things were figured out and settled in the mid 1920s. Well, decoherence in the 2nd part of my PDF is from the 1980s, but it is a relatively minor update....

    The whole idea of yours that one should be searching for "something new" when it comes to the foundations of quantum mechanics is a misguided source of your crackpottery. People like you just refuse to learn basics of physics - they refuse to read any paper, book, or PDF file that makes any sense, just like what you did now once again - and they try to sell this clear defect as an advantage - oh, you're so intellectual and original. Well, you're not. You are just ignorant about basic physics and under proper circumstances, you should have never received a PhD, not even in chemistry.

    You're completely wrong about the history of EPR and nonlocality, too. The idea that there was something nonlocal about quantum mechanics was due to Einstein, in the 1930s, that was the very point of the EPR paper (Einstein was claiming that entanglement couldn't exist because faster-than-light signals would be needed to report the type of measurement done on one place to the other place), and since the very beginning, Bohr et al. were explaining to Einstein (and later others) that this idea was wrong. Quantum mechanics doesn't bring any new nonlocality and quantum field theories are quantum theories that tare perfectly and exactly local as long as we use a proper physical definition of locality that is based on actually measurable phenomena and correlations.

    Bell's theorem didn't say anything new about nonlocality of quantum mechanics whatsoever: Bell's theorem is purely a theorem about theories that are *not* quantum mechanical, namely about local realist theories. Such theories inevitably disagree with the experiments because they predict inequality for correlations - and yes, these inequalities are called Bell's inequalities - that are routinely violated in experiments - and in quantum mechanics that agrees with all the experiments. It follows that a proper theory describing the observation - and quantum mechanics is an example - has to be either non-local or non-realist. Quantum mechanics is local and non-realist. Realist theories that would match the data would have to be non-local, and therefore fundamentally incompatible with special relativity. That's why they can be discarded - Lorentz invariance is known to hold - even at the Planck scale, it works at the accuracy of 14 decimal points.

    So all your basic statements about all these things are just wrong. You should be ashamed of your failure to understand anything about the essence of the most important scientific breakthrough of the 20th century, instead of being as strikingly arrogant as you are. Pompous fools like you is something I simply cannot stand.
    Hank
    I wish you'd write here once in a while.  You and Sascha on the same site carpet-bombing people would be pretty terrific.  Then we could get Ethan Siegel from Scienceblogs to guest post once in a while and he and Tommaso could be our 'nice' physicists.
    lumidek
    Dear Hank, it's clear that you come from the Wild West if Tommaso Dorigo, the obscene backwoodsman who is constantly kicking nice Silvio Berlusconi into his balls, among many other things, is your symbol of a nice physicist.
    At any rate, in your Wild West, you're getting profit out of the writing, right? Why don't you get someone who also writes something for me? The only problem is that your server has to build on quantity rather than quality and TRF can't afford to be overwhelmed by alien quantity. ;-)
    Hank
    The site doesn't make a profit, the money from ads goes to pay for the servers and then the rest to the writers.  People can obviously guest post over there, they are not shackled.  If you have a bigger audience than Science 2.0 we will all quit and join you.
    vongehr
    Lubos, I am honored, seriously, to have you read my column and even spend your time defending my points. However, the mystery maths of crackpots is there to confuse people - it is a tool. By entering into endless arguments about such maths, you let yourself be exploited, since their only aim is to make it appear as if a controversy between experts exists where there actually is none - this is no different than the strategy adopted by "intelligent design" or any other of the many sorts of irrational zealots.

    This article proposes the Quantum Crackpot Randi Challenge (QCRC). Anybody who has a local model would be able to tackle the QCRC in a matter of a few weeks. Of course they cannot, but the main point is: Anybody with even the slightest understanding of the scientific method, even without knowing much QM or much maths, can understand that this is indeed the case! THIS is the main message of my article and I will better write another one not mentioning this Joy crook in order to get the message across. Anyway, I would like to invite you and other science bloggers and skeptics to reconsider the strategy and adopt the same one that is available with the original Randi Challenge.

    1) Do NOT discuss their silly mystery "theories", crank maths etc.
    2) Point out that experimental observation is the judge of scientific models.
    3) Point towards Randi type challenges and stress that these types of challenges (that is why I use the name "Randi") do not suffer from any suppression by establishment science/ peer review as they bypass that hurdle.
    4) Refuse to discuss further until that challenge is met.

    I would really like you to tell us your opinion on this and whether you could help make the QCRC widely known.
    lumidek
    Dear Sascha, your recommendations are probably wise, and I often have the same desire to recommend myself the same thing, and sometimes it's necessary.
    However, one can't ever be sure that every new guy from this camp is deeply misled. So I am making sure that the answer is what it is, and along the way, I am also gaining the privilege to have interacted with pretty much every such crackpot in the world, so that I may say that all of them have been told where their mistakes have been and if they hadn't been completely blinded, they would know.
    vongehr
    Bryan, welcome to my column. I am sure you are a nice guy and teach your students good chemistry and all that - no joking. I am also aware of that you have some sort of deep sitting obsession that makes you, although you apply the moves of the scientific method in other fields, unable to apply it to quantum mechanics. That is also all fine - it does not make you a bad person. In order to be able to communicate with people like you while this communication still is also useful to others, I have written this article. The main message of the article is:
    1) I will not give your obsession another platform by going down to the level of endlessly throwing around trivial mystery maths in order to confuse lay persons about that there is a controversy where there is none.
    2) There is now the Quantum Crackpot Randi Challenge (QCRC)! Read it, understand it, solve it, finished! If you have a local model and even the slightest understanding about real physics even without knowing QM or maths, you will appreciate that a local model must be able to tackle the QCRC in a matter of a few weeks. Fame will be yours, so get going, do not waste your time discussing on the internet or somebody will beat you to it!
    Re: your inference of my obsession.  I am a local realist by logic, not by emotion.  
    I have a question:

    Take an EPR pair: one spin moves to Alice and gives a click +.  That means the probability is 1, (wave function collapse) for the + and 0 for the -.

    Now the spin going to Bob must also give a click.  The two spins must have the same LHV.  

    Here is the question:

    Is it acceptable to assume that coincident spins have the same LHV?

    I would like a Yes or No with a concise reason.  I also do not want to hear what should NOT be.  I think the question is straight forward and clear, and I think it is important for us to have a consensus on this. 


    I request that you do not edit my comments.

    Thanks.

    http://quantummechanics.mchmultimedia.com/category/quantum-mechanics/
    vongehr
    Take an EPR pair: one spin moves to Alice
    Already wrong. No "spin moves to Alice". You need to first understand what the EPR problem actually is about.
    Also: None of your comments have been edited. I can delete, but not edit.

    So, let me repeat, as you seem to have difficulties to grasp basics: I will not discuss with you endlessly nonsense, thereby helping you create the false impression of a controversy between experts. You are plainly not an expert on fundamental quantum physics. If you want to understand the EPR problematic, go to my articles about them (linked above), where it is all clearly explained.

    I respect the quantum chemistry done by chemists like you. I for example attended a class by Curt Wittig at USC, and it was brilliant; his quantum chemistry is top notch. However, quantum chemistry can be done perfectly well without understanding the very core of quantum physics. In fact, I am not able to tell whether Curt for example does or does not understand the EPR problem. I would bet that he does, but I cannot tell from the quantum chemistry that he knows flawlessly.

    UPDATE: Yes, I just deleted your "comment" below. Stop posting insults, thinly veiled threads, and links to pseudo science. For the last time: I will not give you yet another stage to spread nonsense. If you have well thought through arguments concerning the content of my article, lets hear them. Tip: you should first understand my articles! Do you know what understanding means? Understanding means that even if your opponent is wrong, you know why he/she thinks it correct and could successfully feign arguing just like them (much like I did with Joy's mystery math). You are obviously nowhere near that level of understanding.
    Here is the paper by Han Geurdes that disproves Bell

    Sure, it is: Adv. Std. Theor. Phys. 4(20) 495-49, (2010).

    ArXiv: http://arxiv.org/abs/1012.4990.

    lumidek
    This preprint is complete nonsense. In equations 6,7, it defines capital Omega with indices plus, minus, or zero. He or she deliberately omits the fact that these Omega's depend on all variables a,b,x,y - i.e. on the axes with respect to which we may make the measurements. Properly, all of them should be written as Omega_plus_minus_zero (a,b,x,y)....

    So the sets Omega will depend on the decisions of both observers A,B about the axis, so if he redefines the expected correlations of the spins as integrals over a fixed Omega, then the result will be wrong, or if the Omega is allowed to be re-adjusted according to the decisions about the axes a,b,x,y,  it will mean that the theory defining the correlations depends on the decisions about the axes a,b,x,y, so it needs to spread superluminal signals. At any rate, he can't get a local theory that violates Bell's inequality. 

    The simplest way to see that one can't get such a result - and that hundreds of similar crackpot papers that will undoubtedly be written in the future again are wrong - is to go through the 5-line proof of Bell's inequality. Because it can be proved and the proof is easy to be checked, it can't be "unproved" by making a transformation. Any transformation leading to a result that contradicts the proof is obviously wrong. However, the community of the people who are trying to fight against basic results of physics is so intellectually feeble that you would need thousands of years to find this simple bug in the paper above. So you have no chance to see the light in your lifetime.
    It's just like if you're trying to prove that 2+2=5. To do so, you write 2 as 1+1+1-1+1-+1...., and then you add the terms +-1 one by one, and assuming that you have written a sufficient number of terms that is well above your abilities to calculate - i.e. at least 6 times - you will ultimately reorder the terms and get that 4=5. Meanwhile, the other people who try to make things and proofs as simple as possible know how to prove that 2+2=4 (and Bell's theorem) which also allows them to see that any way to rewrite the integrals in the theorem will have absolutely no impact on the result. And the result is, once again, that Bell's theorem holds for all local realistic theories and you and the author of the paper above are cranks.

    Actually Han's paper is not a preprint, but published. It seems to have taken you about 5 minutes to digest Han's paper. I think you should have another look at it

    von Neumann had a nice proof in 1936, and Bell shot it down in 1966, not because von Neumann's math was wrong, but because his assumption 4 was wrong.

    I agree, there is nothing wrong with Bell's math. Ironically Bell's error is in also in his assumption: the two main ones are 1 locality and 2 spin is the usual quantum spin (outcomes +/- 1). I say that the second assumption is wrong, not the first, and that is why the inequalities are violated by QM.

    You have not given me the mechanism for how angular momentum is conserved between two space like separated entangled spins. That is, of course, because you cannot.

    So until you can, it is your polemics only, and until I get my statistical program working, my theory will remain in the realm of crackpots.

    lumidek
    Dear Bryan, I have never heard about the journal called "Adv. Std. Theor. Phys." and indeed, if you Google search for this bizarre acronym
    http://www.google.cz/search?q=%22Adv.+Std.+Theor.+Phys.%22

    you may check that it only appears on 7 pages and each one of them is linked to Mr or Ms Han. But even if it were published in a journal that also prints non-crackpot stuff, it would be totally irrelevant because this paper would still be crackpottery. I know that crackpots are obsessed with getting their papers published and a small percentage of them inevitably succeeds but this still doesn't make them right. It proves that statistics happens.


    Your comment about the "wrong assumption" of spin having two values shows that you have no clue what you're talking about once again. An ordinary Stern-Gerlach or similar device simply shows that there are just two allowed values of the spin, and the challenge is to understand the entangled particles etc. later measured by something like the Stern-Gerlach apparatus. 


    If you don't get two possible outcomes of your measurement, your theory is ruled out as an explanation of this setup well before one even starts with the first sentence explaining why it's wrong. You're just randomly changing the rules of the game. Your brain is just clearly incapable to focus on the very particular correlation in a very particular experiment that Bell is discussing.
    Well, concerning the endless mathematical stuff I understand and refer you to: Phys Rev E 51, 5151-5154 (1995).

    Concerning the obsessions. I accept you will not allow my obsession a platform, for after all I only took the poison out of mr Motls remarks and tried to correct his reading of my paper.

    But what about the abusive treatment of a serious researcher like Joy Christian? That is an obsession mr Vongehr. Like it or not you demonstrated a compulsive need to look better than Joy. It was needless it was abusive it was silly. And whether you allow me this-or-that is of no concern as long as you and your lot behave like this. You demonstrated that you are unable to understand dr Christian's argumentation and you block mine. The latter I understand the former is like a stain all over your list.

    vongehr
    You demonstrated that you are unable to understand dr Christian's argumentation
    Dear Han, thank you for providing me with another opportunity to point out that Joy Christian himself confirmed my understanding his argumentation, as the linked emails confirm. Thank you for your interest in my blog and welcome.
    Mr Vongehr, thanks.

    Indeed I know that Gregor Weihs urged Joy Christian to come with a computer model. But, for an old crackpot, please tell me why you think Joy's topological argument is wrong? Not the stuff that one cannot find hyperspheres in the laboratory and also not the I-have-mails but just why one may not apply this line of reasoning to spin states. In simple terms of course.

    I think it is the same as saying that my set theory is outlandish math. Mr Motl tries to make a point by claiming the need of superluminal communication in order to construct the sets. But if that is true, then Bell's formulation of the correlation in terms of LHV distribution is not local at all. Please do note that in my paper equations (14) and (15) warrant strict locality. That is, one LHV at instrument A the other at B, a joint uniform distribution and -like Bell did himself- an integration over the joint probability space.

    Uniform probability density, elementary set theory, elementary integration ... where is the outlandish mathematics in ASTP 4(20) 495 (2010) ?

    Please do note that there is (implicit) use of functions that project in {-1,1}. These are generalized functions.Read Gell'fands work on that. One *may* wonder whether or not Bell's use of the absolute signs to obtain his inequality are without any trouble at all. One *may* insist on an independent mathematical proof whithout use of absolute signs (Schwarz inequality).

    Note: (d/dx)sign(x-y) <> (d/dx){(x-y)/|x-y|} when we also use |x-y| = (x-y) sign(x-y). See for yourself.

    Now Bell's integral form of the correlation can be solved along the route of partial integration where (d/dx)sign(x-y) type of evaluations is introduced in the integral. Then (d/dx)sign(x-y) <> (d/dx){(x-y)/|x-y|} makes this ambiguous. Note that a single point in the Riemannian integration makes no difference. So mr Motls reference to the five step proof *perhaps* overlooks the ambiguity.

    I must admit, perhaps too much math in this post but it is not difficult to grasp. I think, around 1900, already Lebesgue noticed this ambiguity. I had the luck to follow dr J van der Linden's course on mathematical physics in Leiden.

    Now you and your friends may act as though Lebesgue did not exist, or deny work is necessary to disambiguate Bell's oh-so-simple statements but *where* is your proof that all is well in CHSH country ???

    You need the old crackpots there. So please be somewhat more serious when a public trial for crackpottery is set up. You might overlook the point of the argument.

    Forgive me but I will not post much because of the fact that I have to earn my living.

    Han Geurdes,

    I can know that anyone who claims to violate Bell's inequality with a local theory is wrong for the same reason that I know that anyone who claims to square the circle is wrong. It has been proven impossible with a very simple proof. I do not need to read about complex theories about hyperspheres in order to know this.

    Say I have a red and a green sock and I select one at random and give it to you without looking. When I look at my sock I instantly know what color your sock is no matter how far away you are. That's because the color of my sock encodes information about the color of your sock.

    The thing is you don't have to use color. You could use smell, spinning socks, different sizes of socks, hyper-spherical socks or even a 27 dimensional sock populated by five dimensional hyper-fleas who encode information about the other sock in Sanskrit written with the blood of virgins. How the information is encoded is entirely irrelevant to Bell's theorem.

    vongehr
    Dear Han, please stop immediately inundating this comment thread and my email with mile long crazy program code!
    1) They are automatically stuck in the spam filters - so nobody will look at them anyway.
    2) I know programming - that is not how working code looks like - ever.
    3) The Quantum Crackpot Randi Challenge tells you how to become famous by putting your working program out there as a downloadable game (Java app or something) onto the net so that any friends Alice and Bob can download it and see Bell's uncertainty be violated (which is proven impossible, but here you go). It is NOT about replacing mystery maths by mystery code when dumping unsolicited garbage into busy people's mail. Your code will make you famous by being a succesful game on the internet, thus bypassing establishment physics and evil peer review. Me agreeing with your code would not help you the slightest little bit.
    Sasha,
    The Java program I wrote is a minimal implementation, and my focus was on simplicity to make sure anyone can understand it (...anyone with basic programming skills :-). It has a NetBeans app with a simple gui to put in the parameters - how about you send me an email and I will reply with both the screen shot and the code if you like.
    Currently, it all runs in one VM. If we want to convert it to a multiplayer game, fine, but that will add another layer of complexity in my opinion and it might not be easy to run anymore and verify. Currently the code is so simple that I think anyone can see that no "cheating" is going on...
    Also, the program has the option to only use the predefined angles, or all angles (for the plot). To me at least it helps to visualize the whole curve to understand what is going on depending on what function is used.
    To change the code, the easiest would be if the contestant would also have NetBeans as editor (open source). Or else, if that is too much to ask for, I could probably make it so that the contestant can put in the formula in a text window that would be compiled on the fly... but obviously it would be easiest if the person would just edit that one class and change the formula...
    Chantal

    vongehr
    NetBeans app with a simple gui in one VM.
    OK, that is already over my head. We need something that two friends, Alice and Bob, who have never programmed, can download from a webpage and play. Only like so can we claim that the modified program can possibly become popular fast on the internet, thus bypassing "establishment science" (e.g. peer review).
    We should first just provide that, because this already shows that anybody with a local model could do something just like that following our example. Then the next step later on could be either:
    1) To let crackpots with not too much more effort modify the hidden variables and measurement modules with their local model without having to host it on their own site. But maybe that is not necessary.
    2) Write a program that cheats (via establishing internet connection at inappropriate times) in order to show that non-locality is what can violate Bell's inequality. This would be just there in order to make clear as to what crackpot local models are supposed to do without cheating. After that, lay people who cannot understand quantum physics will still be able to judge the Quantum Crackpot Randi Challenge, and that is the aim.
    the program has the option to only use the predefined angles, or all angles (for the plot).
    As you can see from the email exchange with Joy, it is common crackpot strategy to shift the discussion towards angles that do not violate Bell's inequality. No other angles but those three (a=0 or 1 for Alice and b=1 or 2 for Bob) may be in the program. Whoever introduces more angles is automatically disqualified.
    the easiest would be if the contestant would also have NetBeans as editor (open source). Or else, if that is too much to ask for, I could probably make it so that the contestant can put in the formula in a text window that would be compiled on the fly... but obviously it would be easiest if the person would just edit that one class and change the formula...
    I will look at NetBeans and see how easy it is to download. Lets keep discussing right here so others can follow and contribute. I am a late adopter and better stay out of the discussion about platform details and such. I suspect, looking at wolfram alpha and the like, that if we find the right person with experience, with a little bit of coaching from me about the vital points and your code as a starting point, the whole could be finished in a day.
    I suggest you throw out the option for other angles and then we talk again - what do you think?
    BTW: you can email me via my profile page here and why not you have your own profile here on Science2.0.
    The program can be started with a url... but then it must be on some server obviously that has a tomcat installation at least (I can install it, but I have no server accessible online).
    It can also be started from a command line if downloaded

    Netbeans can be used to change the code (but you don't need it to run it!)

    I don't have a huge amount of time to rewrite it to have it run on 2 computers or 2 VMs... I don't see what this really adds in terms of the challenge than if it just runs the way it is.

    Why don't you take a look at it and try it. Since I don't have a server set up, I could either:
    - email you an executable that you can run as it is , no netbeans or other required
    - email yoy the source, then you can run it and change it with netbeans.

    RE angles: I can make it so that you can't change the angles if that is an issue. The plot can't really hurt to see in addition to the angles, so I wouldn't remove it (why?).

    Chantal

    vongehr
    have it run on 2 computers or 2 VMs... I don't see what this really adds in terms of the challenge than if it just runs the way it is.
    If it runs on two computers far apart and they do not communicate while doing the measurement on the 8000 photons, and if the Bell inequality would be violated by it, that would be a classical system (not just a simulation(!) but a real system) that violates the Bell inequality with local variables. This is the very aspect that makes the Quantum Crackpot Randi Challenge so compelling! This would make anybody famous in an instant, yet any local model would be able to provide this.
    In other words, if you run it on only one computer, people must look much further into the code and then we have similar nonsense discussions as one has now about mystery maths. If the host site provides the hidden variables and sends them to Alice and Bob, that would be the best

    sure, email me the executable and the source.

    The plot can't really hurt to see in addition to the angles, so I wouldn't remove it (why?).
    I will have a look but suspect the plot may just confuse people about the vital. All we need is the three values and the result, namely whether or not the Bell inequality is violated. You seem to be a little confused about the main aim. Do not think you can via a little game explain quantum physics to people. The aim of the program is to put a James Randi type challenge out in order to have an effective tool in the fight against pseudo-science. I will better write another article to explain once more precisely what I mean by Randi type challenge.
    vongehr
    Similar to chat-roulette, Alice finds the site that explains the vital points about what the significance of the game is and she clicks a button called "play". Instructions tell her how to invite her friend Bob or to connect with a waiting Bob that previously found the site. Then the website pops up a window and tells them something like "8000 classically entangled photon pairs have been generated with hidden variables and are stored only at the host site. For the next step, there will be no interaction via the internet. Please click to generate your 8000 random choices of the two measurement angles."
    After Alice clicked, it goes: "Your angle choices are displayed below and only stored on your computer. Feel free to save it in spread sheets or via screen shots if you like. Click to receive your 8000 entangled photons."
    After that: "Below is the table of your angles in the first column, the hidden variables of the photons you received are next to that, and the measurement results are in the final column. Feel free to check that all measurement results are correct. Neither our website nor Bob's computer knows your angles or the results. Please click to inquire if Bob is ready to exchange his results with you."
    After that:"The table has been updated with Bobs angles, hidden variables, and measurement results. The results for the important measures of coincidences is as follows (feel free to check it via counting): N1,1(1,0) = 476, N ...          Since N + N > N, Bell's inequality is satisfied as it must be without the non-locality of quantum physics. Thank you for playing."
    Hank
    It's a JAVA IDE. we can put it on a server easily enough.  It's still unclear if it will do anything meaningful though, right?
    vongehr
    The point of Randi type challenges is not that they can be successfully completed (they can't!!!) or do anything meaningful, but simply that they are out there at all! Being out there, skeptics can easily debunk that establishment science is to be blamed for the crackpot's local model not getting recognition. Discussions about mystery maths can be refused.
    Halliday
    Sascha:

    I'm a little uncomfortable with allowing Alice's and Bob's computers to communicate (more or less) directly, especially with allowing the communication of the hidden variables between the two.

    Now, I'm not certain whether chat-roulette ever establishes a direct connection between the chatting partners.  However, this can all certainly be handled through a server.  (This also helps avoid potential difficulties in handling firewalls, NAT, etc.)

    May I make a suggestion?  There should be three separate pieces of code, four separate processes (Alice and Bob both run the same code, the only code that does not run on a server).
    1. We have the client code that runs on the Alice and Bob computers.
    2. We have the host process that generates the 8000 "photons", complete with their hidden variables, and transmits these hidden variables to the Alice and Bob computers.
    3. Instead of having Alice and Bob computers try to communicate directly, we have a "correlation" process either running on the host computer, or another host.  This process only receives the measurement results (no hidden variables) from Alice and Bob computers, and sends the correlation and each other's results back to Alice and Bob.
    Furthermore, the "correlation" process can be written once, without ever having to know anything about any hidden variables.  It is the coincidence counter, or the third party check on the results.

    The code for the client and "photon" generating processes are the only ones that need know anything about the hidden variables.  However, in both cases, the only code that needs to know anything about these hidden variables can be highly modularized and separated from the rest of the code.

    So even the client and "photon" generating code can be written once, with only a single module in each that has to know anything about the hidden variables.  Since whatever the nature of the hidden variables, I'm certain we should all be able to agree that XML is sufficiently general that this can be used as the encoded form of any such hidden variables.

    This allows all the communication portions of the client and generation codes to be written once, never allowing the hidden variable understanding embedded module to have any external communication.

    For the "photon" generation code, all the communication, storage, etc. needs are all coded only once, without having to know anything about the hidden variable generating code.  The only thing it has to know about the hidden variable generating code is how to tell it to generate the "photons" (a predetermined application programming interface:  Something like a "generate" call), and that the hidden variables are encoded within XML, as an output of the generation module.

    For the client code, again all communication, storage, user interface, etc. are all coded only once, without having to know anything about the hidden variable "detection" code.  The only thing it has to know about the hidden variable "detection" code is how to tell it what the polarization choices are, pass it the XML of the hidden variables, and receive the detected results.  Again, this is simply a predetermined application programming interface:  Something like a "detection" function call where one passes in the XML for the hidden variable "photon", and the angle choice, and then receives the "detection" result as an output.

    I would suggest that the "generation" and "detection" code application programming interface should be on a "photon" per "photon" basis, rather than some predefined number like 8000.  This will allow for flexibility in changing the number any time one may wish (such as for performance reasons).  The "generation" code always generates a pair of "photons" as a pair of XML objects.  The "detection" code always receives a single XML object and a single angle, and outputs a single measurement result.

    All the more difficult code, with user interfaces, storage and retrieval capability, network communications, etc. are only written once, completely independent of any hidden variable "generation" and/or "detection" modules.  They simply encode the communication with the "generation" and/or "detection" modules using a predetermined application programming interface.

    Does this approach sound satisfactory?

    David

    P.S.  Unfortunately, my programming forte is in scientific analysis and simulation codes (actual computations).  I don't usually have to directly handle network communication, and usually have only bare bones command line like user interfaces.  So while I would certainly be able to code the "generation" and "detection" modules (given good understanding of a particular form of hidden variable model), I would not be the best person for the job of coding the external, model agnostic portions of the code.
    vongehr
    Nice to be taken seriously for once. Yes, best there should be no communication and four computers, one for the generation of the pairs, Alice, Bob, and Charlotte who calculates the Bell relation - my writing about two computers is a compromise so people are not too scared about those details and get started at all. Once it would start rolling (my internet programming sucks - I use an old version of mathematica), those involved would soon find out that using four computers etc is not a big modification in the end.

    Quick one: Above Java IDE was proposed, you say XML (are they somehow the same/similar), yet I found criticism, and it appears to me as if almost everybody has another one. Which one we should focus on as the likely best (in view of people collaborating etc) and regardless of personal preferences?
    I would suggest that the "generation" and "detection" code application programming interface should be on a "photon" per "photon" basis, rather than some predefined number like 8000.  This will allow for flexibility in changing the number any time one may wish
    People are not going to play something 8000 times or even just 800 times. Smaller numbers have a high probability to "violate" Bell's inequality simply due to the statistics of small numbers.
    So while I would certainly be able to code the "generation" and "detection" modules (given good understanding of a particular form of hidden variable model)
    We only need to make it work for the simplest [Every EPR pair is a string of six numbers (A0,A1,A2)(B0,B1,B2) where the Ai are randomly 0 or 1 and the Bi are the opposite of Ai, e.g. (1,0,1)(0,1,0)]. It will already be understood by most lay people that such a program is so simple compared to world of warcraft, it would not be a problem for anybody with a working local model to modify the open source code and win the Nobel prize - conclusion being that there is no such model suppressed by establishment science. Moreover, once it is out there in the simplest version as a game, geeks will like it and make better versions. This is the beauty of the plan: The internet is by now so large, good versions will evolve and be selected from the one we put out without us needing to further bother about it at all. The result is not hampered in the slightest, on the contrary, no Bell violating program ever becoming successful proves that there is none.
    Halliday
    Sascha:

    You ask:
    Quick one: Above Java IDE was proposed, you say XML (are they somehow the same/similar), yet I found criticism, and it appears to me as if almost everybody has another one. Which one we should focus on as the likely best (in view of people collaborating etc) and regardless of personal preferences?
    First, Java and XML are distinct.  Java is a programming language (and the Java IDE [Integrated Development Environment] is a particular set of tools for writing Java programs), while XML is an extensible means for encoding information.

    The criticism is valid, at least to a point, especially since, in this case, we would be using XML to encode and exchange (potentially highly structured) "data between applications".  However, Java, and JavaScript, have built in capabilities for handling XML.

    Since XML is, essentially, just a potentially very long text string, it is very flexible, and easily passed along by code that doesn't have to understand what it encodes.

    Java produces reasonably good systems, especially on host servers (it is often used, along with other languages, for generating dynamic content).  XML seems like a reasonably flexible system for encoding and exchanging the arbitrary information that various hidden variable models may demand (otherwise, the proponents can simply claim that our system is too simplistic to handle their hidden variable needs).  The fact that Java (and JavaScript) talk well with XML is an added reason for using them together.

    The fact is, once we have worked out this client server system once, in whatever language(s), we can then codify it into a network protocol, that can subsequently be implemented in any of a number of other languages.

    David

    P.S.  Java would work quite well on the server side (though Hank may have other suggestions, if this is to be hosted on his system).  However, having to actually download and install a Java application might be a higher hurdle than necessary.

    On the client side (Alice and Bob), it may be reasonable to encode this using more Web based technology:  JavaScript, HTML 5, XHTML, etc.

    The whole thing can be encoded in generally available technology, so the client will simply run within Alice's and Bob's browsers (at least so long as we stick with Web standards, and don't fall prey to special features of non-standard's compliant browsers).
    vongehr
    This all makes me think I definitively do not want to get into all this, XML and Java. And in a few years it will be probably all outdated again. I stick with quantum mechanics - never changes. ;-)
    What I am saying is, if you can handle it or know somebody who can, or if anybody follows these comments who can, I will do the bare physics algorithm and ensuring that it is a "Randi type" challenge, you do the web implementation. I will post the core QCRC in mathematica form soon (I hope the one I posted already explains the sort of hidden variables we need, but I will also post a more "physical" version where there actually are angles and spin directions in the code).
    Halliday
    Sascha:

    When I suggested that the application programming interface of the internal "generation" and "detection" modules should be on a "photon" by "photon" basis, you responded with
    People are not going to play something 8000 times or even just 800 times. Smaller numbers have a high probability to "violate" Bell's inequality simply due to the statistics of small numbers.
    I fully understand the statistical reason for large numbers.  Unfortunately, you appear to have misunderstood what I was recommending.

    I was not at all trying to suggest that Alice and Bob punch some "give me a photon" button 800 to 8000 times, or more.  No, this "photon" by "photon" suggestion, for the application programming interface, is purely internal.

    When Alice or Bob (or whomever) asks for the set of "photons", the process on the host will call the "generation" module 800 to 8000 or more times, in order to obtain the "photon" pairs (of course, these pairs could have been generated only once [for a given version of the hidden variable modules], and stored on the host).  Then on both Alice's and Bob's client processes, once they receive the set of "photons", they feed these "photons", one at a time, along with the corresponding angle, to the "detection" module.

    The "photon" by "photon" aspect is only internal, at the code level, not at the user level.  The intent is to simplify the application programming interface, while making it as general as reasonably possible.

    David
    vongehr
    My misunderstanding derived from "This will allow for flexibility in changing the number any time one may wish". So, actually, I am very confused now by what you mean. I programmed the classical core yesterday (I will post it tomorrow), and it takes 0.8 seconds to finish the whole program with 10000 photon pairs. So, I do not understand where the advantage of photon by photon is or who has time to change the number. Maybe I am missing something about coding - let me be very clear: I am not an expert on computers at all. I'm like one of those paper and pen fans (remember those?).
    Halliday
    Sascha:

    First, while the computational efficiency of the trivial hidden variable model is certainly high enough, I, at least, cannot guarantee anything close to that efficiency for any "real" hidden variable model (such as Christian's, for instance).

    Second, why not allow for different numbers of photons, for flexibility, for whatever purpose may be necessary?  (Maybe we will want to use more photons, for better statistics.)

    Third, if we can create an application programming interface that will allow for the "creation" and "detection" of vectors of "photons" (and angles, in the "detection" case) of arbitrary length then this issue is mute, since we could simply receive or pass in arbitrary length vectors of such.

    Mathematica can certainly handle arbitrary length vectors of whatever data we wish to use, both as inputs and outputs.  However, not all other programming systems allow such flexibility.  (I know Java has the ability to define vectors/arrays of objects, but I neither remember how flexible such is, nor do I know if JavaScript shares Java's flexibility in this regard.  It's just been too long since I have programmed in Java.)

    However, I'm certain that practically any programming system will allow for multiple calls on a module/function/subroutine with scalar arguments and return values.  Since these multiple calls can be accomplished within an inner loop of the respective processes, we have all the flexibility we may desire.

    Besides, there is still the issue of having to communicate these multiple "photons" with their (potentially arbitrarily complex) hidden variables over the Internet.  Again, one can send data streams of multiple XML objects, or one has to find a way to encode vectors of these for such exchanges.  The latter adds an additional layer of encoding/decoding.  (Certainly not impossible, but an additional complication nonetheless.)

    So this was simply a forward looking suggestion.

    We could certainly delay the final decision on multiple-scalar calls vs. single vector calls to the point when someone is actually trying to code all the portions outside of the inner "generation" and "detection" modules.

    However, I strongly recommend against fixing the number of "photons" at this early design stage.  I strongly recommend creating a system where the actual number is a mere parameter of the system (preferably either determined by the "generation" server, or as a request determined by the clients), rather than something that is "hard coded" from the beginning.

    David
    vongehr
    while the computational efficiency of the trivial hidden variable model is certainly high enough, I, at least, cannot guarantee anything close to that efficiency for any "real" hidden variable model (such as Christian's, for instance).
    Joy's is actually simpler than the one I programmed already, since he has 100% anti-correlation at arbitrary angles. Moreover, a crackpot would need to claim that his "co-domain" is so finely fractal that no computer memory can possibly reach the size necessary to just predict the experiments in their finite experimental accuracy. At that point, it is their problem entirely. Most crackpot models are trivial compared to what is done already with online games. You seem to not have kept up with online multi-player games. Me neither, but what I have seen is mind blowing.
    Second, why not allow for different numbers of photons, for flexibility, for whatever purpose may be necessary?  (Maybe we will want to use more photons, for better statistics.)
    8000 EPR pairs gives better stats than the actual experimental observations considering experimental errors. There is no point going higher. Any more than absolutely necessary freedom (like different angles) gives crackpots opportunity to deceive with nonsense. Actually, we probably should fix the number at 8000 or even less, period.
    Halliday
    Sascha:

    If you're really going to be so "paranoid" about "give[ing] crackpots opportunity to deceive with nonsense" (by giving them "Any more than absolutely necessary freedom (like different angles)"), even to the point of "fix[ing] the number [of 'photon' pairs] at 8000 or even less, period", then I would suggest perhaps an even stronger reason for making the application programming interface for the "generation" and "detection" modules a one-"photon"(or pair)-at-a-time, rather than all-at-once:  This will disallow the "crackpots [the] opportunity to deceive with nonsense" by taking away the unnecessary freedom of being able to choose "detection" responses based upon the totality of the "photons", rather than the merely necessary "freedom" of choosing responses on a per "photon" basis.

    David
    vongehr
    1. Freedom to choose detection responses based upon the totality of the photons is necessary. The Hidden variables must allow not to be random, for example a fixed table of 8000 EPR pairs. That would not violate the Bell inequality either.
    2. One reason to do it on a pair by pair basis:  the opportunity to change the hidden variables of the next pair according to the previous outcome. It should be allowed, too, and will also not violate the Bell inequality. The problem is that the constant communication (8000 times on and off) allows for easier cheating type "non-locality" (and maybe requires very reliable internet connections). I suggest that the first web version plainly goes with the 8000-in-one approach and then, if strongly desired, offers a pair-by-pair version. After all, the point is to show that such a program is easy and works as a multiplayer application at all, while the crackpots are to modify it, not us.
    Halliday
    Sascha:

    I'm not at all certain whether I understand you here (so I'm even more uncertain whether you have understood me  ;)  ).

    You say "Freedom to choose detection responses based upon the totality of the photons is necessary."  Yet the example you appear to give, of "a fixed table of 8000 EPR pairs", has absolutely no need "to choose detection responses based upon the totality of the photons".

    So what are you saying?

    On the other hand, if we allow the "generation" and "detection" modules (the inner, hidden variable aware portions of the code—not the full pair generation, or Alice/Bob processes) to generate and act upon "the totality of the photons", then we violate the conditions of the Bell Theorem:  Instead of having some number (8000, say) separate entangled pairs, with only single individual responses (from which we build the needed statistics), we are allowing a formulation of, effectively, a single entangled pair, with far higher order (28000 times higher) freedom of "measurement" and "response".

    As you have already stated:  "Smaller numbers have a high probability to 'violate' Bell's inequality simply due to the statistics of small numbers."

    (Of course, even in this scenario, where the effective output of the "measurement" is of far higher order than that of Bell's Theorem, and his inequality found therein, this higher order output is still going to be fed into the statistics as if they were independent observations.  So I certainly cannot guarantee that one can violate Bell's inequality even in this way, but we would certainly be changing the ground-rules to the point of no longer fitting within the constraints of Bell's Theorem.  Hence, there are no guarantees that Bell's inequality cannot be violated.)

    David
    vongehr
    Oh, I get it, you mean "responses based upon the totality of the photons" that have already been detected, not the totality of 8000. OK, forget the first part of the reply then - a fixed table can also be had on a pair-by-pair basis.
    we violate the conditions of the Bell Theorem:  Instead ...
    I am not sure why this would violate the conditions. Say there are already 200 pairs randomly measured and the results violate the Bell inequality a little by accident. How could the next choice of hidden variables possibly drive the violation up?* The next pair still does not know the next angle settings, and so the next 200 are likely to just bring the little violation that was already achieved back to where it should be, namely non-violation by a large margin.
    So I certainly cannot guarantee that one can violate Bell's inequality even in this way
    Maybe this means the same as I just wrote(?) - not 100% sure.
    but we would certainly be changing the ground-rules to the point of no longer fitting within the constraints of Bell's Theorem.
    I do not see that yet.

    *If there were such a problem, and you can show why, than you should write it up and publish, because if that is the case, the Aspect experiments that have closed the communication loophole would actually not have closed it. The experiment would have to generate all pairs and send them before selecting any angles, needing times t and distances x=c t that no lab can possibly provide yet. Come to think of it - this may be a good reason to actually let it go pair-by-pair. However, I think this is missing the problem. I know of no crackpots with such intricate "theories", secondly, would such conspiracies actually be acceptable to those who want to defend mechanical box-lever-pulley realism? I do not see that a photon pair that looks at what Bob and Alice measured already in order to deceive them is anything other than a god that plants fossils. If you go down that far into the sewers of Crackpotonia, you can never exclude that there is not Jesus setting it all up where the measurements past light cones intersect - hardly classical realism as Newton taught it.
    Halliday
    Sascha:

    Now I know for certain that you have not completely understood what I have said, at least in this instance.

    I'm very sorry for any lack of clarity that may have contributed to any confusion.

    First, let me be perfectly clear:  For the sake of this discussion I am only referring to what the internal, hidden variable code of the "generation" and "detection" modules/subroutines/functions are allowed to have as available information.  I'm not at all concerned about what all the wrapper code does.*

    There is no potential problem with either the "detection" or "generation" code keeping track of "photons"/pairs as they are received/generated.  Any real system could, potentially, be keeping such a "running tally".  This in no way violates the conditions of Bell's Theorem, since Bell's Theorem allows any events within the past light cone to affect a present event.

    The potential problem that does violate the conditions of Bell's Theorem is for both the "detection" and "generation" code to act upon the totality (all 8000, for instance) en masse.  Then the system has information from the future light cone at its disposal, especially when the "detection" code is also given access to all the angles (on one client) at the same time.

    Allowing the internal, hidden variable code to act upon the totality—all "photons"/pairs—as a single whole, effectively converts the "experiment" into one where the system has future knowledge, and/or one where the totality of "photons"/pairs (all 8000) becomes, effectively, a single pair of far higher order complexity than is applicable for Bell's inequality.

    However, like I tried to point out above, while this is a violation of the conditions of Bell's Theorem—so the results of Bell's Theorem (such as his inequality) do not apply—I certainly cannot guarantee that this leaves a "loophole" sufficient to violate Bell's inequality in the way we are applying it even in this case.

    However, do you want to take that chance?  If you are being so adamant about fixing the number of "photon" pairs, and disallowing any angles beyond the ones you have stipulated—apparently all in "fear" of giving "crackpots opportunity to deceive with nonsense" (by allowing them "Any more than absolutely necessary freedom (like different angles)")—then why allow them this Bell Theorem violating "freedom"?

    Like I say, why take that chance when the solution is so utterly simple:  Simply run the internal, hidden variable code, for the "detection" and "generation" modules/subroutines/functions, sequentially on a per "photon"-angle and per "photon" pair basis, independent of how the wrapper code packages, deals with, and communicates such.

    David

    *  While the wrapper code certainly does many things that are in technical violation of the conditions of Bell's Theorem, that code is entirely within our control, and will do nothing with this potentially violating information that is inconsistent with the intent of Bell's Theorem.

    For instance, while the client wrapper code will have calculated the detection angles either before or after receiving "photons" (I know your original description says this happens before, but it makes no difference for this discussion), it will treat them as if each angle choice is independent of the corresponding "photon" (calculated "in flight", if you will).

    Similarly, while the wrapper code may package up the "photon" pairs, and store them in a file for later retrieval, and may even use these prepackaged "photon" pairs for the actual pairs sent to the client processes, the entire system of wrapper code will be treating these pairs as sequential (one pair at a time).

    Furthermore, while the wrapper code for the clients may receive their "photons" en masse, it will, likewise, be treating the "photons" as if they were received sequentially.
    vongehr
    The potential problem that does violate the conditions of Bell's Theorem is for both the "detection" and "generation" code to act upon the totality (all 8000, for instance) en masse. Then the system has information from the future light cone at its disposal,
    Then your issue actually is as I thought the first time around: You do not like the hidden variables (HD) to be one fixed 8000-entry table. But it may be - there is no violation of Bell(?).
    especially when the "detection" code is also given access to all the angles (on one client) at the same time.
    Alice does not know Bob's angles. She may know all her future angles and knows all her past angles anyway, so she may adjust the next measurement according to that. Still, as long as Bob's are not known, Bell cannot be violated (correct?). Or do you mean that Alice may cheat by changing the angles after the HD were received? I addressed this in the code (read "the angles may not be changed"). To enforce that, the wrapper code should send the angles to the host server immediately after calculation. Would that solve the problem or am I still not getting it?
    Halliday
    Sascha:

    I had been wondering how an intelligent person such as yourself was repeatedly missing my point, on this matter—going back and forth between concept 0 and concept 1, while missing my point at concept 0.5772...—then I saw your response under your "official" challenge post, here.

    I suggest you look again at the footnote in my message above.  I have no problem with storing and using "one fixed 8000-entry table", so long as that table is generated and passed on to the internal, hidden variable "detection" code in such a way as to disallow access to future information when calculating earlier results.

    Furthermore, the entire reason for placing the "(on one client)" parenthetical modifier to the phrase "access to all the angles" was to ward off the potential for that phase to be misconstrued as implying that the angles from both clients would even be available, let alone passed to the inner, hidden variable aware "detection" code.

    No, at no time have I or ever would I suggest that either of the client processes should even have access to the other client's angles before calculating the results of the measurements.

    Perhaps you should refresh your memory by reviewing my original proposal.

    As far as the apparent answer to why an intelligent person such as yourself is "still not getting it", I hope that addressing your response under your "official" challenge post, here, will help clear up your apparent "blind spot".

    However, for now, it may help for you to recognize that my rewrite #1 of your Mathematica "generation" code is an example of how this should not be implemented, while my rewrite #2 is an example of an acceptable implementation.  Note:  The "HiddenVariableGenerator" module/subroutine/function is an example of what I have been referring to as the inner, hidden variable aware "generation" module/subroutine/function.

    David
    vongehr
    I have no problem with storing and using "one fixed 8000-entry table", so long as that table is generated and passed on to the internal, hidden variable "detection" code in such a way as to disallow access to future information when calculating earlier results.
    That is what I mean with that you do not like the HD to be a 8000-fixed table. The difference you say I have a "blind spot" toward is that they may interact with all the local angles. Why is that a problem? Bell's proof is concerned with the non-locality between Alice and Bob. If Alice has no access to Bob's angles, there is no way to get the correlation correct.

    I understand now that by "inner code" you mean basically the to be modified code, and "outer code" is the handling of the modules and how they communicate. Thank you for the clarification.
    When Sascha and Lubos both agree you know your theory is in deep trouble on many levels.

    Dammit I was so looking forward to understanding QM with high school maths finally

    You rained on my parade ... no christmas cards from me guys.

    dear sascha, lubos,
    i've found the attack of the crackpots highly annoying, since i believe i could learn a lot from your blog. i hope you won't stop in your quest to enlighten those of us who really wanna learn something. even so, though unwittingly perhaps, i've come to learn something about the psychology of crackpots, which i think is best conveyed by this catherine tate "i can do that" skit:

    http://www.youtube.com/watch?v=XY66ZJ0TFUI

    keep up the good work,
    gabor

    I see that that usual suspects have gathered here to blog on this exciting and interesting topic. I am working on acquiring a better understanding of the EPR paradox. I am wondering about a question I have about the conclusions drawn from the experimental confirmation of QM concerning the EPR paradox. From what I have read the victory for QM centers around violation of Bell's Inequality. I assume that the hidden variables prediction (conditions for particle properties were established at the creation of the entangled pair) proposed by Einstein was not confirmed after statistical analysis of the data. My question is what if a certain percentage of the measured particles were not entangled pairs, wouldn't this poison the results of the experiments. If there were non entangled pairs in the experiment how could we know which ones were not entangled and more importantly which side of the argument would this statistical glitch favor by such a situation.

    vongehr
    if a certain percentage of the measured particles were not entangled pairs
    then the Bell inequality would be the less violated the more non-entangled pairs there are (until it is not violated at all, in case there are only non-entangled pairs).
    Thank you for the reply Sascha

    So my next question is this. Are we confident that all of the measurements are of entangled pairs? Maybe this is not an issue. I don't know. I would like to hear from anyone who has knowledge or experience in this matter. I am basically wanting to be sure about the connectness issue. I have a concern that the experiment is flawed in this matter. Perhaps we were not careful enough to ensure that the all of the measurements were of entangled pairs. This would then prove Einstein to be right. I would not be sad about that. On the other hand, things in nature then become a little less interesting as this magical superluminal connection is very interesting to me. I am very distrustful of what we think we know when experiments are based in statistical mathematics. Statistical mathematics does not explain the foundations of nature. It just says this is the way nature behaves generally speaking, it does not provide the specifics. It does explain why or how. That is not satisfactory.

    Halliday
    Mark:

    It appears that you have misunderstood Sascha's reply to your question.  To wit, the more non-entangled particles there are within the mix, the less the Bell inequality is violated.  Therefore, we need not make sure we have only entangled pairs, since all we are looking for is whether Bell's inequality is violated, since Local-Hidden-Variables (LHVs) will not violate Bell's inequality at all.

    All the "poisonous" non-entangled particles do is move the results closer to "prov[ing] Einstein to be right" (in your words), rather than "proving" Quantum Mechanics to be "right".  So, we only need to have a sufficient percentage of entangled particles so as to allow the Einstein "disproving" statistics to show through.

    David
    Thanks David and Sascha ...

    Okay think I have got it. The very fact that Bell's inequality is violated proves that there are entangled pairs so the more that it is violated the more sure we are of QM and the quality of entanglement sample. So now that we are reasonably sure that the experiments have proven QM correct in this manner. What about this connectedness issue? Would anyone like to try and explain that particular concept. Just exactly how do we know that a change in the first observed particle actually effects the separated twin. It seems to me that there is only an implication of an actual effect by virtue of the fact that Bell's inequality is violated. Has there ever been an experiment performed that takes a pair of entangled particles and then deliberately changes the state of the first particle to a known polarity and then measures the twin to see if its state flipped states as well?

    vongehr
    With "that a change in the first observed particle" and after that, your question imports dangerous assumptions and is not anymore about the physics involved. Before one of the particles is observed, it is not one single particle, but strictly a pair, so you cannot change it. After it is observed, it is not entangled anymore, so no changing it will change the other particle. You need to let go of the concept that there is some hidden variable, like a spin of each of the single particles, that you could change. There is no such thing before the measurement. It is the measurements that are correlated, not the particles' hidden spins (there are none). The whole particle concept is classical. Particles are what appears to us via measurement. There are many situations where the number of particles is uncertain.
    @Sascha

    Ahh good to see this detail explained.
    "it is not one single particle, but strictly a pair, so you cannot change it" I am aware that QM considers the entangled pair a single entity or system. I am definitely looking for information about how easy it is to break entanglement so if I understand you correctly any type of manipulation with particle A or B breaks the entanglement. Good to know.

    Yes, I agree with what you say about me and my concept of classical notions and concept. I am a believer in a deterministic local reality kind of Universe. Everything is cause and effect. That is entirely my belief and at this point.

    Treating an entangle particle pair as just a mathematical single system pair of something (perhaps a wave) that does not really exist until we look at it is not something I am willing to buy into yet. I would like to attempt to explain the EPR phenomenon with classical techniques. So I accept the Quantum Crackpot Randi Challenge certainly there can be no greater crackpot theory than my SR Theory. So in order to do this there has to be one of three reasons for the explanation as to why Bell’s inequality is violated in EPR experiments.

    Reason for Bell’s inequality violation short list

    1. QM Copenhagen Interpretation is correct. Not likely in my opinion.
    2. There is a real physical super-luminal connection that exists between entangled pairs. A new form of
    spacetime reality that occurs only when entangled pairs exist.
    3. Local Hidden variables is the cause and somehow we have managed to botch up the
    calculations/formulations and by a happy coincidence we managed to prove QM correct not local hidden
    variables.
    These 3 possibilities assume that Bell did not make a mistake in his proof.

    Later in this post I will give anyone out there a challenge or most certainly a challenge to the challenger to hopefully settle this once and for all. Before I do that I want to discuss what I think advocates of QM truly believe. Simply put by QM reasoning and mathematics of the wave function and by conservation of angular momentum, whatever the result for the particle we measure, the other particle must instantaneously be fixed in the complementary spin - without our measuring it. Also by the Copenhagen Interpretation there is no state attributable to either particle until a measurement is made, so neither particle has a value for spin or polarization which is as I understand it the same thing. Tell me if my understanding about what you believe is wrong. I think that is pretty much it.
    Ah, but as many people see it, including myself those particles could be a few light years apart therefore it should take a few years for some signal concerning the state of one particle to reach the other at light speed, which is the limiting velocity of space time. We have got quite a conundrum here. We seem to have some sort of instantaneous connection and that to me is what the entire controversy is all about. According to QM advocates and I am going to try and choose my words carefully here as I know there is a lot of frustration concerning the description of EPR. But according to what QM people believe, for the Non-Measured particle's state to instantly collapse you need non-locality what Einstein called (Spooky action at a distance). For this to happen you would seem to need faster-than-light communication. I know, I know there is no such thing as faster than light communication. You have to admit many physicist's use the same language so don't blame me. I also apologize for using the word particles I had no idea that we don't want use that concept anymore. I see no other way but to use that term in this discussion.
    Anyway back to the problem. The problem as I see it is the Copenhagen Interpretation, which by the way is over 85 years old! It says that we have a pair of particles that cannot be treated as separate, their states are entangled in such a way that the sheer act of measurement or observation on one fixes the state of the other. The entire problem boils down to the following small phrase "fixes the state of the other". How does that work exactly? I am not going to accept the idea that because we have an equation that describes this phenomenon perfectly means that we have a physical cause and effect explanation. We all know that we do not have that. In my opinion the Universe does not obey mathematical formulations. Nature obeys the laws of forces which impart action and affect upon real objects. In other words events in reality happen because of causes and effects of objects in motion and interaction with each other. All things are affected by real physical forces. All objects do have position and momentum for every moment in time whether we observe them or not. The trick is to get the one and only set of equations that actually describe the forces of nature. We are never going to get those equations using Quantum Theory. Even now as I speak recent experiments have revealed that it may possible to determine both the path that an individual particle takes and still have that particle exhibit wavelike characteristics. http://www.scientificamerican.com/blog/post.cfm?id=what-does-the-new-dou... Eventually the uncertainty principle will fall. Other experiments using single isolated atoms have indicated that the governing process of the very small may be due to deterministic chaos and not random events as QM says. All we are going to need is a little more time and effort and we will be able to undo Bohr’s interpretations of reality.

    So finally to eliminate reason #2 the super-luminal connection listed in the “Bell’s Inequality Violation Short List” described above I suggest we do an experiment to determine if there is an actual physical connection between entangled particles. I call the mysterious connection the EPR filament. According to Super Relativity Theory space is an elastic solid. This elastic solid possess a high tension. This tension determines how fast light (Electromagnetic radiation) can travel in this medium. According to Super Relativity in the case of entangled particles an electromagnetic field filament might exist between the two particles as they recede from each other. As they separate the tension in this filament increases space’s tension dramatically therefore allowing a region where electromagnetic force can be propagated at super-luminal speeds. The challenge is then design an experiment to detect the possible existence of the EPR filament. My idea for doing this is to perform the EPR experiment as normal except for one key difference. Do not attempt to break the entanglement until the particle passes by the device that will cut or modify the filament. If a real connection exists then breaking the entanglement should effect a change in the correlation numbers at the detector. So, simply put, we need to wait until photon “A” passes by the filter or device that is designed to break the entanglement. According to QM Theory there should be no effect. Why because we have not touched or modified or observed the particle. But, if there is a physical filament connection then interfering with the filament should still cause a break in the correlation.

    The more I study this problem the more I think that in fact there may not really be a super-luminal connection. We should still do the experiment I suggest just to be sure. More than likely as I continue my studies it is becoming my belief that we have somehow botched the calculations and hidden variables (initial conditions) is the cause of the EPR phenomenon. Ah, but if my EPR filament idea works out, oh that would be cool.

    @ Chantal

    I will check out your suggestion and investigate Robert Close www.classicalmatter.com
    It sounds like he and I have the same belief.

    vongehr

    (perhaps a wave) that does not really exist until we look at it is not something I am willing to buy into yet.
    Neither does H. D. Zeh for example, yet he does understand the problem (see below).
    Reason for Bell’s inequality violation short list ... 3. Local Hidden variables is the cause and somehow we have managed to botch up the calculations ... These 3 possibilities assume that Bell did not make a mistake in his proof.
    So he botched up the calculations but did not make a mistake. Good to know this is possible within classical worlds - a good example for that they are ultimately a lot more weird than quantum worlds. ;-)

    About the long rest of your comment: I bail out at this point. There are just too many wrong assumptions in how you state the problem. You will never solve it anyway, classically, quantum, or in entirely new ways, if you do not understand the problem from the get go.
    @ Sasha
    No, Bell did not botch up his proof. The QM calculations for the expected results of EPR are possibly wrong. I understand the problem just fine. Hey, I do appreciate your time though. I have a simple request if you don't mind. Other than the computer model/game that is requested by the challenge. I am curious from your point of view. What if anything bothers you about the EPR paradox? Perhaps nothing. Just would like to hear your opinion. Then I will just watch this discussion after that. It is interesting.

    Take care all...

    Halliday
    Mark:

    In what way could "The QM calculations for the expected results of EPR [be] possibly wrong"?

    You do recognize that "The QM calculations for the expected results of EPR" match with the experimental observations of "EPR" type experiments, don't you?

    So how?

    David
    @ David

    "The QM calculations for the expected results of EPR [be] possibly wrong"

    I was taking a shot in the dark with that statement. Although as you know in the past many if not most particle discoveries attributed to the success of QM where actually made accidentally. This is what I have read in other documents on the web. In these accidental discoveries then the math to explain the discover was retro-fit to explain the discovery. See http://techie-buzz.com/science/eureka-moment-new-particle-discovered-at-...
    Then QM claims another success story. I just thought that perhaps the math that QM predicts was perhaps not so much a prediction but a post-diction. Then over time the post-diction becomes a prediction in the minds of people. Once again, I do not know this for sure as it is speculation on my part. A historian in the EPR topic may know the answer to this question I have. As you know I am very distrustful of the physics community where QM is concerned. Some physicists, not all of them look at Quantum Theory through rose colored glasses and refuse to attack it. I, like Einstein am always looking to be on the attack. I like that about him. I don't have Einstein's gift, knowledge or skill but I do have his heart and share his love for the pursuit of the truth. I have a fundamental philosophical difference with Quantum Theory and the Copenhagen Interpretation. Even now as physicists pursue the Higgs I see them on their blogs modifing the formulations to predict the range of energies where they might expect to find the particle. Why are they doing that? They are guessing, because QM really does not make a precise prediction. As the energies ranges are eliminated they recalculate to compensate for the failure to predict accurately the first time. Most physicists now are afraid to think out of the box. Even Sasha sees nothing wrong or controversial with the EPR paradox. His statement above confirms this. That sadly is the problem. Reduced simply to this statement.

    What bother's me is that in experiment after experiment Bell's Inequality is not violated, but instantaneous communication, or "spooky at a distance", seems to occur. If you rule out instantaneous communication, Bell's Inequality is violated. Seriously that does not bother you?

    Look at it this way. Even Bell's Theorem has assumptions in it.
    1. Logic is completely valid
    2. There must be a reality that is separate from its observation. Ugh!
    3. Finally, no information can travel faster than light. (This is locality)

    @David and Sasha
    So let me ask you first if you agree with me on the 3 assumptions and then if you do let me ask you and Sasha which assumption is wrong? The results of EPR indicate one of these assumptions must be wrong.

    While thus far Einstein has not been proven right in this debate in my opinion he has at least had a moral victory of sort. My challenge to you and Sasha is to pick which assumption is wrong. I know as well as you, anyone you pick will leave you in a whole lot of hurt. So really you guys don't see a problem with the QM explanation of EPR? :)

    Halliday
    Mark:

    First, you appear to have a very selective concept of the experimental "successes" of QM.  Eh, so it goes.  ;)
    • The EPR was most certainly a QM prediction first:  E., P., and R. first used QM (as then formulated) to predict what QM would say would be the "spooky" result for measurements on entangled pairs of particles, like photons.  It was far later that anyone was able to perform any experiment even coming close to an EPR like experiment.
    • The prediction of ant-matter came from Dirac's equation for Fermions.  It was something he and others noticed lurking within the mathematics of his equation.  I expect it seemed rather odd to them, and probably even more so to others when they first learned of such a preposterous prediction from this "fringe" equation.  However, once experimenters actually looked for such, anti-matter was indeed found.
    • Murray Gell-Mann discovered a number of subatomic particles on paper, by noticing missing pieces in the patterns he expected from SU(3) symmetry.  It was only later that experimentalists found matching particles.
    • There are multiple additional examples.  Far too many for me to enumerate them all from memory.

    Second, you all the more appear to have quite the misunderstanding of the nature of the "prediction" and the present (and past) searches for the so called Higgs particle.

    The so called Higgs Mechanism is very general.  There are myriad specific formulations, each with whole ranges of possible masses.*

    The fact is, other than anti-particles of already known particles, we have no specific way (at this time) to predict what actual masses will be.  At best, we can constrain the masses based upon the lack of observation (up to any given point in time), and the masses of other known particles that would interact with said hypothesized particle.

    The case of the Higgs particle is simply another example of this general situation.

    David

    *  Actually, the "mass" parameter of any QM "particle" is the principle "free" parameter of our models/theories.  The "mass problem" (or, more properly, the problem of the "mass spectrum") is a known problem and issue.  However, it has nothing whatever to do with the question of predictions vs. post-dictions of QM.
    Halliday
    Mark:

    I don't know what you mean by your assertion that "What bother's me is that in experiment after experiment Bell's Inequality is not violated".  Of course Bell's Inequality is violated "in experiment after experiment"!  If that were not so, QM would have been either thrown out, or been highly modified long ago.  ;)

    Even more puzzling is your assertion that (only) "If you rule out instantaneous communication, Bell's Inequality is violated."

    How are experiments truly "rul[ing] out instantaneous communication"?

    Or are you arguing something completely different than what your words/sentences appear to be indicating?

    David
    Sorry not the experiments, the interpretation of the experiments. In other words that followers and believers of QM do not really believe that superluminal communication is at work. They have ruled it out as an explanation. Then and only then in my opionion is Inequality violated. It is possible to have higher correlations if we have super luminal communication as an mechanism. Let me explain in another way. In the case of EPR experiments it has been shown that Quantun Mechaincal theory's predictions show correlations that are stronger than could be explained by local hidden variables. In order for this to work this requires super luminal communication but techincally speaking QM does not say that there is Super Luminal communication.
    I guess you don't understand that point. Perhaps it is best to leave it there.

    Halliday
    Mark:

    It appears that you need to take a gander at my latest reply to your post here.

    As I point out, there, there are two assumptions/conditions in Bell's Theorem.  So any violation of Bell's Inequality (the "then" portion of the theorem) implies that one or both of the assumptions/conditions (the "if" portion of the theorem) must not be true.

    Since the assumptions/conditions are:
    1. Reality (that microscopic objects have real properties determining the outcomes of quantum mechanical measurements), and
    2. Locality (that reality is not influenced by measurements performed simultaneously at a large distance).
    Then all that is necessary is that one of these is violated (or both).

    You appear to have a problem, of some kind, with the fact that (as you say) the "followers and believers of QM do not really believe that superluminal communication is at work."  In other words, assumption/condition #2, above, is not what is violated (untrue).

    Yes, it is quite true that many (most?) "followers and believers of QM" (as you put it) assert that the problem is with the first assumption/condition (the "hidden variable" assumption/condition, if you will:  that microscopic objects have real properties determining the outcomes of quantum mechanical [or other] measurements).

    So why do you appear to insist that it must be #2 (as you say "Then and only then in my opionion is Inequality violated.")?  Why can it not be #1 (especially since that is, essentially, the one you attached an "Ugh!", seemingly in commentary)?

    David
    Halliday
    Mark:

    You claim:
    Look at it this way. Even Bell's Theorem has assumptions in it.
    1. Logic is completely valid
    2. There must be a reality that is separate from its observation. Ugh!
    3. Finally, no information can travel faster than light. (This is locality)
    Now, Mark, I thought we had already discussed, at considerable length, the nature of mathematical theorems and proofs thereof.

    All theorems are of the If A then B form.  A embodies the assumptions, conditions, etc., for B to be true.  Violate any portion of A, and the theorem can say nothing about B.  (Of course, there are also theorems that can be formulated as If and only if A then B, which is equivalent to [If A then B] and [If not A then not B].  However, that's really just an extension to the simple single If ... Then case.)

    So, as a theorem, Bell's theorem certainly makes no additional "Logic is completely valid" assumption.  It is simply a mathematical theorem, just as with any other mathematical theorem.  All of mathematics "assumes" that "Logic is completely valid" within the particular realm of a given mathematical context.

    As for the other two "assumptions", they are only applicable insofar as the actual Bell Theorem states them to be.

    Now, Bell's Theorem has the following assumption/conditions:
    1. Reality (that microscopic objects have real properties determining the outcomes of quantum mechanical measurements), and
    2. Locality (that reality is not influenced by measurements performed simultaneously at a large distance).
    The second assumption/condition is to be interpreted as including your "no information can travel faster than light".  It also includes the assumption/condition that changes in the detection apparatuses cannot influence the characteristics of the particle headed for the other apparatus.

    Of course, the first assumption/condition is akin to your "There must be a reality that is separate from its observation."

    From your "Ugh!" commentary on this condition, it appears you do not agree that this is an appropriate condition for actual reality.  Is that the case?

    Now, I hope you recognize that these assumptions/conditions, in Bell's Theorem, do not, necessarily, reflect Bell's concept of what is or is not the nature of actual reality.  They simply codify Bell's interpretation of Einstein's concept of "local realism", as Einstein expressed in papers such as the EPR paper.

    What Bell's Theorem states is that any hidden variable theory that adheres to these assumptions/conditions will produce statistics that satisfy Bell's Inequality.  So any system that violates Bell's Inequality must have violated one or both of these assumptions/conditions.  (If you have a correct statement of the form If A then B, then you do know that [not B] implies [not A], since otherwise you should have had B.)

    Does this help clarify?

    David
    vongehr
    I was taking a shot in the dark with that statement. ... I just thought that perhaps the math that QM predicts was perhaps ... Once again, I do not know this for sure as it is speculation on my part. ... I don't have Einstein's gift, knowledge or skill but I do have his heart
    Einsteins heart was about not shooting around in the dark.
    vongehr
    What if anything bothers you about the EPR paradox? Perhaps nothing.
    Nothing. ;-)
    Yeah, the guy is very entertaining.

    He starts with a model which has lots of fancy stuff in it but at the end of the day is in perfect agreement with Bell.
    But then he says, oh we need to normalize it by dividing it by standard deviation. Well, that's ok, if you just compute it directly the sigma comes out as 1 as it should and it changes nothing. But that would be too easy, where is the challenge in that?

    Tthe expression in question is a product of a constant and a random variable. So what he does, and I'm not kiddin, he moves the constant outside of the sigma sign like that: sigma(kx) = k sigma(x). Or, by the way, didn't I mention that both k and x are grassman bivector whatsises and a product between them is a geometrical product? So now he's taking a standard deviation of a grassman bivector. Since this does not compute, he just replaces x with |x| and now sigma(|x|) comes out as 1 but the whole thing now is a bivector. And apparently it's quite all right to take a standard deviation of a scalar quantity and get a grassman bivector back, who am I to argue?

    vongehr
    Yes, there is lots of funny math, but it can backfire to argue that way, because he just claims you do not understand it. And indeed, one can find easily funny math also in accepted theories, like throwing around infinities to get whatever works, and such is fine in physics as long as it works, although no mathematician would accept it in a formal proof. In this particular case it is just creating a smoke screen around trivialities, not a genius bending the math a little to get a good physical theory the crinkles of which can be ironed out by later generations.
    "Hint: Topological properties are global (as opposed to local)! ."

    I responded to this naive comment a day or two ago. I don't know if its failure to appear is a posting error on my part, or a delay by the moderator. It hardly matters, though -- you just simply have to clue of what you're talking about.

    Tom

    Dear T H Ray, I assure you that topological properties are - pretty much by definition - global, not local. It's because topology is the part of maths that studies properties of objects that "locally look the same" but that may become qualitatively different because the "locally universal patches are connected to the whole in different ways".

    It seems kind of incredible how someone could disagree about this point by Sascha.

    Dear Lubos,

    Joy's framework does NOT dispute the global meaning of topology -- no continuous function model could (as you affirm, by definition) -- but rather subsumes nonlocal probabilistic measure in an analytical locally real manner. That is what is required (and all that is required) to counterexample Bell's claim that no model of quantum mechanics can be both local and real. I.e., that can obviate the assumption of nonlocality.

    The role that topology plays in this framework is in the limit of physical space (7-sphere) driving physical measures in an apparent four dimensional (3-sphere) spacetime.

    Tom

    Dear T H Ray,

    even if a theory fundamentally describes the world with additional dimensions - you know, this is something that people trained in string theory know quite something about - it doesn't change the fact that realistic observations occur in regions that are localized in the 3 dimensions everyone knows, so implications of the extra dimensions for the observations may always be described in the 3-dimensional language. After all, even with extra dimensions, the vacuum has to have the structure of a Cartesian product of M^{3,1} x X_{hidden} for the observed Poincare symmetry in the 3+1 dimensions to hold.

    The proof of Bell's theorem doesn't prohibit the "local realist" degrees of freedom from having an extra-dimensional interpretation; the proof of Bell's theorem doesn't make any illegitimate assumptions that could be avoided in a local realist theory at all. The proof proves that local realist theories inevitably predict certain correlations to belong to certain intervals. Nature (experimentally) and quantum mechanics (theoretically) [often] leads to predictions that are outside this interval which falsifies all local realist theories.

    Best wishes
    Lubos

    vongehr
    Don't waste your time Lubos. Apparently, Tom thinks that if you replace hidden variables' ones and zeros with spheres and doughnuts, the variables are "topological". LOL! He is right to claim that Bell surely did not think about such alternatives and boy are we glad about that. ;-)

    Not sure whether you know this guy, but I just saw that he is Joy's personal fanboy popping up on all kinds of forums, insulting anybody not buying Joy's nonsense. What a sad existence. There is nothing you can hope to reason with people like that.

    Joy's nonsense is still prominently on the FQXI site - like the first thing you find there. How sad is that? Well, perhaps fitting for a venue headed by Max, who apparently now snapped (Seen the "big snap" on Huffpo? FQXI is going down the crackpot gutter ever further.)
    Great discussion here! Long ago I tried to make a bet with Luigi Accardi that he could not rewrite his programs so as to violate CHSH within a well controlled protocol which would exclude his capitalizing on the detection loophole. I figured out that with N=15 000 photon pairs, the chance was less than 10 to the minus thirty that I would lose the bet (nice application of martingale theory). The bet was for 2000 Euro's, but Luigi never accepted the condition that I would chose the detector settings at random and his measurement computers would always deliver a binary outcome. Later I proposed the same to Hess and Phillip, and later to Han Geurdes. Regarding Joy Christian's model, which is quite another kettle of fish, he fortunately auto-destructed about a year ago by putting a one page paper on arXix, which allows everyone to see his conjuring trick.

    Christian's one pager is 1103.1879, I have analysed it in 1203.1504, and Joy has responded in 1203.2529

    For your consideration:

    Florin Moldoveanu, "Disproof of Joy Christian's 'Disproof of Bell's theorem'"

    Four critical elementary mathematical mistakes in Joy
    Christian's counterexample to Bell's theorem are
    presented. Consequently, Joy Christian's hidden variable model cannot
    reproduce any quantum mechanics results and cannot be used as a
    counterexample to Bell's theorem. The mathematical investigation is
    followed by a short discussion about the possibility to construct
    other hidden variable theories

    http://arxiv.org/abs/1109.0535

    Richard Gill
    Indeed Jim, I later discovered Florin's work. The difference is that I only look at Christian's one page paper and I only look at the maths. That makes everything very simple. 

    Christian's book has now come out. You can download the first 25 pages for free. They contain the same fatal error which I identify.

    http://www.brownwalker.com/book.php?method=ISBN&book=1599425645