Does the EPR experiment imply QM is incomplete?

[kurt101]

I know this has been discussed in so many ways on this forum, but it is hard for me to separate interpretations, fact, speculation, and inaccuracies. You can probably just skip the next two paragraphs that I describe the EPR/Aspect experiment and go right to my question.

In an EPR experiment like the one Alain Aspect performed with entangled photons being measured by respective polarizers, Quantum Mechanics treats the experiment as a system where the probability for the polarization of the photons leaving the polarizers is determined to be cosine squared of the difference between the polarizer angles.

Say one of the polarizers, call it polarizer B, is far away relative to polarizer A and the photon source. Say it is so far away that by the time photon A hits polarizer A, there is no possible way by instant communication (faster than the speed of light) that photon B, heading towards polarizer B has any idea of what it might interact with in the future.

QUESTION:
So if all the components in the system have no idea what photon B will interact with in the future, how can QM math give the correct result?

The only two logical possibilities that come to mind are:

1) That QM happens to give the correct probability, but the reason it does is because there is an underlying mathematical algorithm that describes how photon A collapses some of its state to photon B when it interacts with polarizer A.
2) That QM works because the system is actually everything in the entire universe and somehow every particle in the universe knows the state of every other particle in the universe and can follow the rules of QM math..

At least in my eyes, #1 seems the simplest, most likely explanation, and it is easy enough to simulate and get the correct result in this specific scenario with linear polarized photons. Would #1 be called a collapse or Copenhagen interpretation of QM? Is this the most popular interpretation by QM experts? Do most QM experts who subscribe to this interpretation think there is a deeper realistic (i.e. based on rules of math) theory that gives an explanation for the collapse that is not just some random probability generator? If not, are there any strong reasons why this interpretation is a dead end?

#2 seems unlikely to me. QM has very simple rules, you add probability amplitudes when a state can change in multiple ways, but otherwise you multiply; then you square the result for the probability. These simple rules seem to have a local basis and so I have a hard time equating these rules with the state of the entire universe. Is there a name for this type of interpretation? How popular is this interpretation with QM experts?

Thanks.

 

[bobob]

You are making this way too complicated. Quantum mechanics gives the correct answer with no need to invoke anything more than that. The photons are anti-correlated, so there is no cause and effect relationship. Actually, I think wat confuses most people is the relativistic aspect in not understanding simultaneity. Since thye photons are spacelike separated, their measurements cannot be time ordered, so if one were to interpret quantum mechanics to give the epr experiment some cause and effect mechanism, either that interpretation of quantum mechanics or relativity would be wrong. There doe not need to be any deeper reality and I think the belief that there must be is an attempt to understand the quantum world in classical terms, even if the classical terms are weird. In addition, I think it misses the point.

Yeah, it's basically the way most physicists (at least the ones I have met) interpret quantum mechanics. It's a local theory.

 

[PeterDonis]

1) That QM happens to give the correct probability, but the reason it does is because there is an underlying mathematical algorithm that describes how photon A collapses some of its state to photon B when it interacts with polarizer A.

The fact that the Bell inequalities are violated in actual experiments means that no such algorithm is possible that does not use inputs from spacelike separated events. In other words, any such algorithm would have to include information propagating faster than light.

 

[Boing3000]

There doe not need to be any deeper reality

There is a deeper reality, that's why we do use laboratory and not (negative)wishful thinking.

I think the belief that there must be is an attempt to...

... do science, and not just gazing at equations and be quite satisfied by them, and believing that they ARE reality instead of a (just another)VIEW of reality.
But there is a believe that there exist a proof of impossibility concerning some better theory. I have yet to see one (and then understand why we haven't closed every physics university)

 

[DrClaude]

To cite from F. Laloë, Do We Really Understand Quantum Mechanics?

A general way to express the Bell theorem in logical terms is to state that the following system of three assumptions (which could be called the EPR assumptions) is self-contradictory:

1. validity of their notion of “elements of reality”
2. locality
3. the predictions of quantum mechanics are always correct.

Up to now, point 3 holds, so one of the two other assumptions must be let go of.

 

[Boing3000]

no such algorithm is possible that does not use inputs from spacelike separated events.

That's very well put. I would add that, in such form, those algorithms absolutely belong to the "classical" category. There is definitively nothing "weird" about them.

In other words, any such algorithm would have to include information propagating faster than light.

I think you are contradicting your previous sentence. Those algorithm cannot possibly factor-in any type of causality or change in space-like separation.

 

[bobob]

There is a deeper reality, that's why we do use laboratory and not (negative)wishful thinking.

It's not negative wishful thinking. I'm pointing out that if you take the uncertainty relations seriously, there is physics in that alone, without trying to circumvent it.

... do science, and not just gazing at equations and be quite satisfied by them, and believing that they ARE reality instead of a (just another)VIEW of reality.

No, I just don't believe that the microscopic reality can be described in terms we can have any physical experience with and that trying to do so is just wrong headed. Particles are not just little things with little masses that move along well defined (if bizarre) trajectories. The physics is in what they are not what you want them be. If the epr photons represent a cause and effect phenomena, then relativity is wrong. Take your pick.

 

[Boing3000]

I'm pointing out that if you take the uncertainty relations seriously, there is physics in that alone

I haven't seen anybody, especially the OP, not taking it seriously. Quite the opposite. I have always though that such fundamental principle were the most interesting ones to investigate, to gain better and better understanding. Not just dismiss it by "it's just the way it is"

without trying to circumvent it.

I may have a vocabulary issue here. I think it's fine to try to circumvent other deep principle of "reality" like gravity, by gaining knowledge enough to build bridge,shopper or rocket (in more an more refined/efficient ways).

No, I just don't believe that the microscopic reality can be described in terms we can have any physical experience

No "terms" in physics connect to "physical experience". They connect to rational inference. A real number is no more physical than a complex number. A Hilbert Space is barely more abstract than a Galilean FoR. People deal in probabilities all the time.
I find curious peoples that are trying to stop that process of "improving" their understanding, by whatever additional connection with experience they make (string, filed, particle, dice and what not).
And yet I wouldn't call it "wrong headed". There are certainly plenty of reasons to consider physics as complete.

Particles are not just little things with little masses that move along well defined (if bizarre) trajectories.

And the OP never suggested that. So why bring it ?

If the epr photons represent a cause and effect phenomena, then relativity is wrong. Take your pick.

Bell has provide a way to actually exclude cause and effect. So it does not seem than QM is the problem here.
Nor is relativity, which is not at all disproved by any EPR experiment (nor Bell's proof).

The overwhelming evidences indicates that non-locality is a definitive feature of the universe. And yet, so is locality. They are not mutually exclusive.

 

[PeterDonis]

I think you are contradicting your previous sentence.

No, I'm not. An algorithm that calculates what happens at event A using inputs from event B, which is spacelike separated from A, is implicitly assuming that information can travel from event B to event A, which would require information to travel faster than light.

in such form, those algorithms absolutely belong to the "classical" category

I'm not sure what you mean by this.

Those algorithm cannot possibly factor-in any type of causality or change in space-like separation.

I have no idea what you mean by this.

 

[Boing3000]

No, I'm not. An algorithm that calculates what happens at event A using inputs from event B, which is spacelike separated from A

That's more specific, so let's be even more specific (that is: representing what entanglement is): An algorithm can compute what happens at event A (or event B) using input that are NOT connected to any other event (That's after all what the wave function is).
So any other type of input can do, it just have to be unique (in case of entanglement) and non-local(isable) (as per Bell)

, which would require information to travel faster than light.

No. That is an additional statement that does not follow. It would be true only if a change occurred at B (and only B).

I'm not sure what you mean by this.

I mean the space where every algorithm live (the UMA architecture on top a the RAM) is the simplest thing there is. A 1 dimensional "space" with not even a notion of distance. A basic set that everybody can wraps his head around.

I have no idea what you mean by this.

It means "there is a burden of implementation". Anyone that pretend that "signaling info" is identical to "sharing info" has to actually prove it by providing an "implementation" (provide some solvable formula).
In this case you will need to add to the "algorithm" new data like positions(including time), some FoR, and trying to compute that signal FTL propagation (maybe there is a formula for tachyon?), and solve the causality issues (that are not due to FTL, but only the addition of a time coordinate).

Algorithm can't do that because they never evolve an "entire state" but only the atomic step. And none of those step can be swapped (they are not commutative)

 

[kurt101]

The fact that the Bell inequalities are violated in actual experiments means that no such algorithm is possible that does not use inputs from spacelike separated events. In other words, any such algorithm would have to include information propagating faster than light.

Certainly! If I did not say it in my original post, I completely meant to. Whatever algorithm happens appears to happen instantly (FTL) over any distance. It still confuses me when "collapse" is used, is this what everyone means by "collapse"?

 

[kurt101]

You are making this way too complicated. Quantum mechanics gives the correct answer with no need to invoke anything more than that. The photons are anti-correlated, so there is no cause and effect relationship.

I am just using math based logic that all of physics including Quantum mechanics is based on. I don't see how using math based logic is making things more complicated. QM or QFT does not explain everything and is probabilistic in nature and if I understand QFT perturbation theory to some extent (I don't really) it is more of a tool than a theory. If you believe the universe rules are based on math, at least at the level we are discussing, then I would expect the natural instinct when encountering a probabilistic theory is to understand the behavior of the instances that make up the probability. I assume that is the first thing physicists did when QM was first formalized. And there are many examples of instances that appear to be exactly deterministic like orthogonal polarization behavior. If you think we should stop looking for a deeper theory now, I would like to understand what strong reasons you have for this are? I obviously don't have the knowledge of QM that most have on this forum and so I am very interested in understanding these strong reasons (if they exist).

Actually, I think wat confuses most people is the relativistic aspect in not understanding simultaneity. Since thye photons are spacelike separated, their measurements cannot be time ordered, so if one were to interpret quantum mechanics to give the epr experiment some cause and effect mechanism, either that interpretation of quantum mechanics or relativity would be wrong.

I don't see how entanglement and FTL effects would invalidate relativity. As far as I know, their domains do not conflict. The only thing that I can see invalidated at this point are some interpretations and non-math definitions of "time".

There doe not need to be any deeper reality and I think the belief that there must be is an attempt to understand the quantum world in classical terms, even if the classical terms are weird. In addition, I think it misses the point.

I don't see how math based logic translates to classical thinking.

Yeah, it's basically the way most physicists (at least the ones I have met) interpret quantum mechanics. It's a local theory.

Good to know, thanks.

 

[Boing3000]

Whatever algorithm happens appears to happen instantly (FTL) over any distance.

There is no such algorithm. Something that happens instantly exclude the possibility for something to travel (going from place to place at any speed, going smoothly from any value to another value) from somewhere to elsewhere, because everybody (any FoR for any observer) observe a ZERO "distance" between those value. Happily this is the default behavior for computer that run algorithm, a (unique)value V is simply shared/referenced in X other values.

The only algorithmic solution for otherwise different object value to be kept synchronized (at all time, always), is to add two reference (in each value V1 and V2) to one another. Then to use a locking mechanism (which is identical to a causality violation barrier)) rendering this two value indistinguishable from a unique value (so why bother ?)

Other even more complicated scheme may be tried by adding reference to A in B and vice versa, and trying to signal between them (whatever the speed you manage to simulate). In such case, you'll immediately encounter algorithmic singularities (like infinite recursion), and even more complex causality synchronization problem (running the algorithm in parallel will crash the simulation)

Actually the problem is darn simple when observed in the most basic an classical logic framework.

1. A is independent of B at any and all (space)time. A is not even "aware" of B
2. B is independent of A at any and all (space)time. B is not even "aware" of A
3. A and B have a reference to the same V (not a copy) (whose location don't matter(non-locality), at any and all (space)time)
4. Whenever A or B need to test/update V (atomically) its reference to V is severed and replaced by a new value (call it collapse if you want)

This guaranties a perfect correlation (again not FLT, because there is never a speed of change involved), and a strict impossibility to A to signal/affect/action to B (and vice versa)

 

[DrChinese]

Actually the problem is darn simple when observed in the most basic an classical logic framework.

1. A is independent of B at any and all (space)time. A is not even "aware" of B
2. B is independent of A at any and all (space)time. B is not even "aware" of A
3. A and B have a reference to the same V (not a copy) (whose location don't matter(non-locality), at any and all (space)time)
4. Whenever A or B need to test/update V (atomically) its reference to V is severed and replaced by a new value (call it collapse if you want)

This guaranties a perfect correlation (again not FLT, because there is never a speed of change involved), and a strict impossibility to A to signal/affect/action to B (and vice versa)

I don't follow. First, you say V is non-local (and location doesn't matter). OK, that part works. So you are suggesting:

1. When A tests V to get a value, is A sending something to V? (I.e. A is testing polarization at 45 degrees, and is requesting V to provide the answer for that measurement basis.)
2. And then I guess V is sending something back, the "answer"?
3. And then I guess V updates itself to say "At 45 degrees, the answer is "[+]" and I must save that so I can answer B consistently at a later time. And by the way, I cut off my connection to A.
4. And when B tests V to get an answer, B sends its measurement setting and V returns something "consistent" with what it told A. If B also tests polarization at 45 degrees, the result must agree with what was told to A. If B tests at some other angle, then V uses a Cos^2(theta) rule (or whatever is appropriate for the entanglement) to get the correct QM prediction.
5. And then V cuts off its connection to B as well. Now it is not connected to anything. Does it still exist? Because it is otherwise isolated from the rest of the universe.

So if the above is what you are suggesting: It's non-local. And A and B are in partial communication with each other. V is simply the intermediary switching station for their communication. V is in FTL communication with both A and B. And obviously: A's request affects B's outcome, in contradiction to your last statement.

On the other hand, if I describe the mechanism incorrectly, please correct me.

 

[PeterDonis]

An algorithm can compute what happens at event A (or event B) using input that are NOT connected to any other event (That's after all what the wave function is).

No, that's not what the wave function is. The wave function contains information about the entire system, which includes subsystems that are measured at events spacelike separated from each other. The wave function is not local to a particular measurement event.

That is an additional statement that does not follow. It would be true only if a change occurred at B (and only B).

How would you tell whether a "change" occurred at B and only B?

 

[DrChinese]

1. Since the photons are spacelike separated, their measurements cannot be time ordered, so if one were to interpret quantum mechanics to give the epr experiment some cause and effect mechanism, either that interpretation of quantum mechanics or relativity would be wrong.

2. It's a local theory.

1. Of course you can order the measurements of Alice and Bob in all reference frames. You measure them in the same inertial frame at the same place. However, you delay one of the measurements. Add a meter of fiber and voila, one precedes the other.

2. That's strictly an interpretation dependent comment. Entangled state correlations lack a spacetime limitation or dependency. Obviously, distance between measurements can be made arbitrarily large. Further, entanglement can exist between particles that have never existed in a common light cone. Not sure how you get Einstein locality out of that, although certainly some interpretations do anyway.

 

[DarMM]

I mightn't explain this well!

Would #1 be called a collapse or Copenhagen interpretation of QM? Is this the most popular interpretation by QM experts? Do most QM experts who subscribe to this interpretation think there is a deeper realistic (i.e. based on rules of math) theory that gives an explanation for the collapse that is not just some random probability generator? If not, are there any strong reasons why this interpretation is a dead end?

Firstly "collapse" need not be a physical process. Some interpretations see it as epistemic. In these interpretations the wavefunction is your set of predictions for the system. However when you make an observation you update your set of predictions in light of that observation. It's like rolling a dice. You initially have ##p(E) = 1/6## for each number, but when you look at the dice after the roll this collapses to ##p(E) = 1## for one number and ##p(E) = 0## for the rest.

As for the realist question, this is really a question as to whether quantum probabilities reflect your ignorance of the properties of the particles or not.

Realist Interpretations:
In some interpretations (called "Realist" in Quantum Foundations) the probabilities come about because you don't know something about the particles or system you are studying or something about their interactions with you.

In Bohmian Mechanics it's because you don't know the exact position of the particles.

In Many Worlds it's because they will split you into multiple copies of yourself, each copy corresponding to one of the multiple states the particle occupies and in advance you don't know which one you will experience/which copy you will be.

In the Transactional Interpretation (which has retrocausal signals) its because you aren't aware of the configuration of the signals from the future. In any case there is something mathematically modellable related to the particles that you don't have full certainty of.

The general critique of these interpretations is that they require fine tuning. Bohmian Mechanics has faster-than-light signals, but for us to never see them or use them requires our ignorance of the particles' positions to mask the signals precisely. In Many-Worlds a symmetry called Operational Time Symmetry is violated (an observed symmetry in QM) and so one needs the initial state of the multiverse to be fine tuned to mask the breaking of this symmetry. The Transactional Interpretation explicitly has communication with the past, so you need some sort of effect on our equipment that masks this.

Participatory Realist Interpretations:
The other class of interpretations are called "AntiRealist" or "Participatory Realist". In these interpretations the probabilities aren't related to some ignorance of the system under study's properties. Rather the probabilities are a way for an "agent" to organise their interactions with the system. There can be several versions of this.

In Copenhagen the ultimate constituents of reality are incomprehensible/ineffable. The best a classical agent like us can do is organise our dealings with such ineffable objects with the probability calculus known as Quantum Mechanics. The uncertainty relations tell us how valid it is to reason with a given classical concept (e.g. position or momentum). Heisenberg would have said that when a concept applies very well (low uncertainty), such as how the wave concept applies very well in atomic orbitals, then the ineffable stuff is somehow "more like" that concept in that situation. Bohr didn't even allow this.

In QBism quantum mechanics is just a way an agent organises their probabilities (understood as beliefs) in a world where interactions "create a new fact". The world wasn't completely created in the Big Bang, but rather new elements/facts of reality emerge into the universe with each quantum interaction. They are random in the sense that they aren't determined by anything prior, being entirely new.

So you can see the common element is "Reality is weird in some way, QM is the framework an agent must use to navigate such a world" rather than "QM's probabilities come from this or that aspect of the particles". The phrase "Participatory" comes about because the observation result cannot be understood in any way that removes the observer. In QBism they create result. In Copenhagen the result takes place in our classical mental framework, the observer defines the space of results.

The general critique of these interpretations is that they in some sense go against the usual notion of a scientific understanding of phenomena, depending on who you ask.

 

[kurt101]

There is no such algorithm. Something that happens instantly exclude the possibility for something to travel (going from place to place at any speed, going smoothly from any value to another value)

Agree, that is what I meant by instant

Actually the problem is darn simple when observed in the most basic an classical logic framework.

1. A is independent of B at any and all (space)time. A is not even "aware" of B
2. B is independent of A at any and all (space)time. B is not even "aware" of A
3. A and B have a reference to the same V (not a copy) (whose location don't matter(non-locality), at any and all (space)time)
4. Whenever A or B need to test/update V (atomically) its reference to V is severed and replaced by a new value (call it collapse if you want)

This guaranties a perfect correlation (again not FLT, because there is never a speed of change involved), and a strict impossibility to A to signal/affect/action to B (and vice versa)

Yes, this is the algorithm I more or less have in mind and have discovered in my own simulation attempts. As far as I know, for linear polarization it seems to hold up in every case. All of the amazing experiments like "delayed quantum eraser" seem to follow this algorithm precisely. At least from a "realist" view point, this is the only way the experiments make sense to me. Do you know of any experiments it does not hold up?

In step 4 of your algorithm, by "test/update" you mean an interaction with the polarizer? And the reference to the shared V is severed after the interaction (and thus photon B inherits the result of the interaction between A and polarizer A)?

I was thinking about attempting to add circular polarization to the algorithm and try to model the probabilities when going between linear and circular polarization which introduces complex numbers with the QM math.

 

[Boing3000]

1. When A tests V to get a value, is A sending something to V? (I.e. A is testing polarization at 45 degrees, and is requesting V to provide the answer for that measurement basis.)

In the algorithm, the process is trivial (and happily non-local). A(or B) probe/read/peek the value V (V is a placeholder for a value). Then A will decide for itself what it will do with value V (for example compare it to a local polarization angle, then maybe change V's value (that is what severe the link). The trick is that probing/modify can only happens once and atomicaly. After that A will have a new reference to a copy to an ##V_{after}## while B is still referencing the old (but updated and equal to ##V_{after}##)

2. And then I guess V is sending something back, the "answer"?

Kind of, but V is something much simpler that have no decision making to do. It is just a Value placeholder (here an angle)

3. And then I guess V updates itself to say "At 45 degrees, the answer is "[+]" and I must save that so I can answer B consistently at a later time. And by the way, I cut off my connection to A.

Ho no. That part is A's (or B's) business only. Again V is totally unaware of A or B. Being read_and_modified by "prober" is its only functionality.

4. And when B tests V to get an answer, B sends its measurement setting and V returns something "consistent" with what it told A. If B also tests polarization at 45 degrees, the result must agree with what was told to A. If B tests at some other angle, then V uses a Cos^2(theta) rule (or whatever is appropriate for the entanglement) to get the correct QM prediction.

In a simpler way, because B's reason to have a correlated answer (whatever angle B is going to use) is that A have left a new value in V (corresponding to A's "engeinvalue" (here 45 degree)) before severing its link with V.

5. And then V cuts off its connection to B as well.

More the other way around. Its it kind of important. V is truly unaware that any other object have a link to it.

Now it is not connected to anything. Does it still exist? Because it is otherwise isolated from the rest of the universe.

A very good question. Happily, when writing algorithm we don't have to bother with such menial tasks. Implementation wise it will be literally garbage collected.

But if I try to connect this pure classic-simple information base simulation of QM result, to some physical quantity, I think V is definitely connected to conserved quantities. The fact that this spin is "nowhere" begs the question of what the local volume of space at A and B actually spin value is (between measure).

Likewise I have always wondered what happens if you put a polarizer before the cristal. If the incident photon have a ##V_{in}##, the way a ##V_{out}## is created not only create entanglement (a unique ##V_{out}##) but a random ##V_{out}## (even if ##V_{in}## is known).

So if the above is what you are suggesting: It's non-local. And A and B are in partial communication with each other.

That's a way to put it. But what is really important is that there is no path between A and B (nor is away from V). There is ##A{\rightarrow}V{\leftarrow}B## the arrow are unidirectional (a pointer or reference)

V is simply the intermediary switching station for their communication

Correct. But no train is ever leaving V on its own accord. There is no arrow away from V

V is in FTL communication with both A and B.

I wish I could let this "FTL" TLA pass without reacting like crazy... but I must. There are two fundamental ways it is not FLT.

1)There is no speed, no continuous evolution of values (like position). Algorithm are discrete, for all purpose, their are also quantized. Their implementation also happens to be quantized. The computer hops from (classical) state to state is discrete step. There is nothing in between those step. No distance, nothing. Some V is something then is something else. B point to ##V## then to ##V_{after}##. There is no intermediate step against which to compute a difference/delta.
Concerning that simulation, the only critical step that need to be "discrete protected" to perfectly match QM (for example same angle at A and B always match) is: [ A(or B) read V, roll the dice against V, update V, clone V] there are actual operation that allows for that

But I don't think instantaneous is a correct way either ("instant" refers to much to "time" for my taste), but we are going to lack correct words ("value update" is not sexy). Values (information in algorithm) always update at the same speed of 1 cycle per cycle. Algorithm are totally timeless (everything is instantaneous, no distance), values (at addresses) are spaceless (always here, no distance).
Inventing(implementing/simulating/making up) a notion of time within the 1 dimensional distance-less "space" of algorithm, is a very complicate business that is never perfect (actually the values(informations) themselves are quantized, not only the evolution)

2)There is no communication simply because there is no "wire" between V and "the not V". The "wires/arrows" are unidirectional. This is of utmost importance.

And obviously: 's request affects B's outcome, in contradiction to your last statement.

Absolutely not. You get that "feeling" because as an observer you see A,B,V. You have arrow pointing to them. But those arrow don't allow updated. You are a helpless observer.
Actually I even contest the fact that B have been affected. What is affected is your sequence of your copy (as an outsider) of the the evolution of state. ##[B_{step1},S_{step1}], [B_{step2},S_{step2},...]## You choose to make those copy because you only have pointer to every value in the system, including the arrows.

Implementation-wise no value of B is affected before B "decide" to test_and_update V. This my implementation of DrChinese's challenge. I defy anybody to equip B(or A) with a (local) way to detect any affect on its values

 

[Boing3000]

No, that's not what the wave function is.

I suppose that you are talking of the QM equation there, but QM equation can also be used by computer. See below.

The wave function contains information about the entire system, which includes subsystems that are measured at events spacelike separated from each other.

Which is equivalent to what I said

input that are NOT connected to any other event

The wave function can be queried/peek at anytime by any entity in the algorithm. The wave function is a non-local object with no dependency whatsoever on the actual included subsystem. The information as been encoded at the start of the simulation, the wave function do not live in real space.

The wave function is not local to a particular measurement event

My words, nearly exactly (see above)

How would you tell whether a "change" occurred at B and only B?

You cannot (I have developed the point in my response to DrChinese).
The simpler(trivial) version is that B don't know its initial value and it can peek it only once. So there is no way for it to detect any change.

That's my understanding of Bell's proof, and the very reason he used the word "instantaneous".

 

[Boing3000]

Do you know of any experiments it does not hold up?

Probably a lot (if not most). The problem is to map an algorithm with observable in a lab. A "polarization angle" is quite easy to map (I don't even bother with direction of the plane), but I am quite sure there are much more complicate phenomena that just would escape "real value".
Right now I am interested about experiment about a full tree-four-... pure entangled state. The algorithm fail at this point, it alows for easy instantaneous communication.

In step 4 of your algorithm, by "test/update" you mean an interaction with the polarizer? And the reference to the shared V is severed after the interaction (and thus photon B inherits the result of the interaction between A and polarizer A)?

Precisely.

I was thinking about attempting to add circular polarization to the algorithm and try to model the probabilities when going between linear and circular polarization which introduces complex numbers with the QM math.

Go for it.
I myself will stay away from complex numbers. For me there is no point to recover the wavefunction by ... encoding a wavefunction

 

[kurt101]

Probably a lot (if not most). The problem is to map an algorithm with observable in a lab. A "polarization angle" is quite easy to map (I don't even bother with direction of the plane), but I am quite sure there are much more complicate phenomena that just would escape "real value".

To be clear, I didn't mean any experiment, but experiments that involve linear polarization both with and without entanglement. And I don't mean to question whether this algorithm explains all phenomenon which it obviously does not. But I am wondering if it is a true aspect of reality and whether the algorithm is present in all interactions between particles.

Right now I am interested about experiment about a full tree-four-... pure entangled state. The algorithm fail at this point, it alows for easy instantaneous communication.

So the experiment with "full tree-four-... pure entangled state" is one where the algorithm violates reality? I don't see how you could communicate anything instantly with this algorithm. For one thing, even if polarizer B knows what polarizer A orientation is, it is always going to look like random distributed data to polarizer B and you are not going to find a pattern in truly random data. If somehow you knew before hand what the pattern was going to be then I would say you could use it to instantly communicate with this algorithm. However the only way to learn the pattern is by testing it and when you test it the collapse has already happened and you can only forward the result of the collapse to the other end by means of communication bound by the speed of light.

Go for it.
I myself will stay away from complex numbers. For me there is no point to recover the wavefunction by ... encoding a wavefunction

That is not what I mean to do. I want to understand how the aspect of complex numbers in QM could come out of this algorithm. Complex numbers represent a rotation in another dimension. A probability amplitude that is a complex number or vector of complex numbers must be multiplied by its conjugate to get the QM probability. When calculating the probability for two independent, but possible states, you must add these probabilities amplitudes together before multiplying by the conjugate. The probability of these two possible states are thus linked even though they may be separated in space. The algorithm is the same way, the two photon states are linked even though the photons are separated in space. However why do complex numbers come into play and why multiply by conjugate (opposite rotation)? I figure by trying to model this, it might give insight.

 

[PeterDonis]

I suppose that you are talking of the QM equation there

I'm talking about exactly what I said: the wave function.

Which is equivalent to what I said

Huh? You said:

An algorithm can compute what happens at event A (or event B) using input that are NOT connected to any other event (That's after all what the wave function is).

That is saying that the wave function is an algorithm that can compute what happens at event A using input that is only connected to event A. And that is false: the wave function "computes" what happens at event A using input from other events as well as event A.

My words, nearly exactly (see above)

In post #20, yes. But in post #20 you are contradicting what you said in post #10.

You cannot

In which case you cannot say that a change did not occur at B, so you cannot say (assuming your own claim in post #10 is true) that information did not travel faster than light.

That's my understanding of Bell's proof

I don't see how anything you are saying relates to Bell's proof at all.

B don't know its initial value and it can peek it only once. So there is no way for it to detect any change.

I have no idea what you are talking about here.

the very reason he used the word "instantaneous".

Where did Bell use the word "instantaneous"? I've read his original paper on Bell's theorem and that word does not occur anywhere.

 

[DrChinese]

That's a way to put it. But what is really important is that there is no path between A and B (nor is away from V). There is ##A{\rightarrow}V{\leftarrow}B## the arrow are unidirectional (a pointer or reference)

Correct. But no train is ever leaving V on its own accord. There is no arrow away from V...

You must be able to see what is wrong with your argument now. Why have V at all? If there is no back reaction from V to A, or from V to B, then the communication *to* V is completely superfluous. It doesn't change anything!

And your action diagram clearly implies a *separable* function for the results at A and the results at B. Such functions are strictly ruled out by Bell.

 

[Boing3000]

You must be able to see what is wrong with your argument now. Why have V at all? If there is no back reaction from V to A, or from V to B, then the communication *to* V is completely superfluous. It doesn't change anything!

V is just a placeholder for the hidden variable, for A (and B, if entangled). V is not superfluous because particle without spin don't exist.
A can read and write V, so maybe there is a misunderstanding of "back reaction". But V never change A nor B. It is always the unique hidden value (maybe unknown) of the system that reference it.

What is superfluous is the clonage/copy of V after its update by A's interaction, to severe the link to the previous V.
That clonage is not there to actually change anything into the final result of the EPR simulation, it is there to avoid A to be retested later to again probe the original V (because B can have modify it in between, allowing for trivial communication).

And your action diagram clearly implies a *separable* function for the results at A and the results at B. Such functions are strictly ruled out by Bell.

I don't think it is an action diagram. The arrows represent the link, the pointer. If you have the address of someone you can go there, talk to him make up an info, give it to the person, come back with that info. The person is still unaware of where you come from (it is unidirectional) nor how many other people have his address (this is irrelevant for FAPP).

There is no separability, and that's why the simulation works fine.

 

[Boing3000]

That is saying that the wave function is an algorithm that can compute what happens at event A using input that is only connected to event A.

Correct, it actually compute everything that happens everywhere, that is all events. The problem with the wave function is not its "non-locality". The problem is it only spew out probability of happening.

And that is false: the wave function "computes" what happens at event A using input from other events as well as event A.

That's interesting. The wave function picks up new information after the preparation and before measurement ? I've never heard of that...

In post #20, yes. But in post #20 you are contradicting what you said in post #10.

I don't think it is possible. Post #20 is a copy-paste from post #10.

In which case you cannot say that a change did not occur at B, so you cannot say (assuming your own claim in post #10 is true) that information did not travel faster than light.

I certainly can. I only make positive claim. People that pretend that B can "change" within this algorithm must prove it.
It is kind of impossible, since V is "use once and forget", at every (locally situated) site.
Actually V which is the placeholder for the hidden variable, is "nowhere", is non-local. And never move anywhere.

I don't see how anything you are saying relates to Bell's proof at all.

Have you ever heard of "hidden variable", "non-locality", "entanglement" and "correlation" ?

I have no idea what you are talking about here.

The idea of "differentiation".

Where did Bell use the word "instantaneous"? I've read his original paper on Bell's theorem and that word does not occur anywhere.

Visit Dr Chinese site page 5 (VI conclusion)

 

[DrChinese]

V is just a placeholder for the hidden variable, for A (and B, if entangled). V is not superfluous because particle without spin don't exist. A can read and write V, so maybe there is a misunderstanding of "back reaction". But V never change A nor B. It is always the unique hidden value (maybe unknown) of the system that reference it. ...

There is no separability, ....

There are no hidden variables! That's the whole point of Bell. There is no set of hidden variables V that can drive A and B to the QM predictions WITHOUT there being communication between the parties.

What you are saying is different than what your diagram indicates. Clearly: A<--->V<--->B and therefore A and B are in communication. That communication is FTL, because otherwise there are Bell experiments that would give different results if it wasn't.

The DrChinese challenge is to give me values for outcomes at 3 angles (0, 120, 240 degrees) for A and B, and I will choose which angles of those to measure. That is, the measurement settings are not know when V is populated. On the other hand, if you have the V function learn the measurement setting for A before B is measured, then we are back to FTL communication. You seem to be stretching it both ways. There cannot be complete independence between A and B, and having an intermediary V does not accomplish that.

 

[DrChinese]

Let me put it another way. When Alice measures A at a specific angle, say 138 degrees, and B is later measured far away... too far away for a conventional signal to reach it from Alice...

B needs to know 2 pieces of information from A: the angle (138 degrees) and the outcome (+ or -). That's it, and there is no *less* information that can be given to B so that the stats work out.

 

[DrChinese]

Have you ever heard of "hidden variable", "non-locality", "entanglement" and "correlation" ?

You shouldn't use that tone with an expert Moderator such as PeterDonis.

The fact is, you seem to be ignoring Bell. I know you have read the material, but you are skipping the essential elements - and that is what PeterDonis is referencing.

 

[Boing3000]

There are no hidden variables! That's the whole point of Bell. There is no set of hidden variables V that can drive A and B to the QM predictions WITHOUT there being communication between the parties.

That's incorrect. There is no local hidden variable. And that's the whole point of Bell.
There is a set(of one element) of hidden variable that can easily reproduce QM prediction. But those variable must be non-local.

What you are saying is different than what your diagram indicates. Clearly: A<--->V<--->B and therefore A and B are in communication.

That's not my diagram. My diagram is ##A{\rightarrow}V{\leftarrow}B##. A cannot reach B in any way.

That communication is FTL, because otherwise there are Bell experiments that would give different results if it wasn't.

There is no communication AT ALL. So, there is no need to bring FTL.

The DrChinese challenge is to give me values for outcomes at 3 angles (0, 120, 240 degrees) for A and B, and I will choose which angles of those to measure. That is, the measurement settings are not know when V is populated.

So, have you taken the challenge and did you succeed ? You can enter any angles you want, and the number of photon.

On the other hand, if you have the V function learn the measurement setting for A before B is measured, then we are back to FTL communication.

There is no such function. A and B behave independently. They happens to share the hidden variable before measurement. That's straightforward.

You seem to be stretching it both ways.

What does it even mean ? What the problem to represent logic and process with (darn simple) algorithm ?

There cannot be complete independence between A and B, and having an intermediary V does not accomplish that.

There isn't complete independence, because V accomplish that. The dependence just can NOT be guessed by A nor B. (unless they compare their result)

 

[DrChinese]

that's why the simulation works fine.

It doesn't work too bad, but there is the one problem where it produces the wrong values for about half the photon angle settings - at least the ones I tried.

0 and 0: 100%
0 and 30: about 75%
0 and 60: about 75% OOPS should be 25%
0 and 90: 100% OOPS should be 0%.

Just saying...

 

[Boing3000]

The fact is, you seem to be ignoring Bell. I know you have read the material, but you are skipping the essential elements - and that is what PeterDonis is referencing.

I ignore it so much that I have actually translated the logic of its proof in functional code.

You see, there is no progress to be made if you keep thinking that Bell's proof is wrong in anyway, or that I am trying to prove it wrong.

 

[Boing3000]

It doesn't work too bad, but there is the one problem where it produces the wrong values for about half the photon angle settings - at least the ones I tried.

0 and 0: 100%
0 and 30: about 75%
0 and 60: about 75% OOPS should be 25%
0 and 90: 100% OOPS should be 0%.

Just saying...

The correlation is 100% at 90°.
I think you are mixing up correlation with identity. In other word cos(delta) with 1/2 + (cos(delta))^2

Edit: Ho, and I just see your last line. So much for the tone...

 

[DrChinese]

That's incorrect. There is no local hidden variable. And that's the whole point of Bell.
There is a set(of one element) of hidden variable that can easily reproduce QM prediction. But those variable must be non-local.
...
There isn't complete independence, because V accomplish that. The dependence just can NOT be guessed by A nor B. (unless they compare their result)

No one is arguing that non-local hidden variables won't work as an algorithm. The issue is that the measurement setting of either A or B... and the outcome at that setting... must be part of the algorithm. You are in denial that must occur.

1. Assume A and B are randomly given a "+" or a "-" initial value (same for both). To be specific, it is "+".
2. A is measured first at 138 and yields value of "+", its initial value, and needs nothing else.
3. But B, which is also "+", needs to know about the 138. That value is transmitted to B via FTL means.
4. B can now give answers for any measurement setting simply by applying the Cos^2(theta) function.

All good. There is FTL communication in this model.

 

[DrChinese]

The correlation is 100% at 90°.

Photons don't produce those statistics, my friend. It's COS^2(theta).

If you're not sure, try taking a couple of polarized lenses and crossing them. 0% light transmitted.