Coming Soon!

[FrediFizzx]

@gill1109 Oh BTW, you forgot to point out where my explanation is wrong so you must accept that it is absolutely correct. But even without that explanation the statistics shoot you down. Out of 1 million trials there are probably less than 20 events with the flips associated with a change in b. That is pretty statistically insignificant! That means there are 999,980 events that are completely local no matter what. You are finished. Done! Time to get over it and move on!

PS. And another BTW, I've got my own non-local model that is way better and more simple than John Reed's strawman example. It was the inspiration for the local model.
.

[gill1109]

Joy has written that the purpose of his work is to challenge conventional understandings of metaphysical concepts (like “locality” and “realism”).

I have not written any such thing. Only an illiterate fool who cannot read plain English would read what I have written and claim what you are claiming. Stop spreading lies about my work.

For anyone else who is interested, please read Section II of my following paper to find out what I have actually written: https://ieeexplore.ieee.org/stamp/stamp ... er=9418997
.

Thanks for the link, Joy! Here are your actual words:

My work on the subject [2]–[6] does not challenge this mathematical core, if it is viewed as a piece of mathematics. What it challenges are the metaphysical conclusions regarding locality and realism derived from that mathematical core. My work thus draws a sharp distinction between the mathematical core of Bell’s theorem and the metaphysical conclusions derived from it.

Now, regarding Fred's programming:

@gill1109 Oh BTW, you forgot to point out where my explanation is wrong so you must accept that it is absolutely correct. But even without that explanation the statistics shoot you down. Out of 1 million trials there are probably less than 20 events with the flips associated with a change in b. That is pretty statistically insignificant! That means there are 999,980 events that are completely local no matter what. You are finished. Done! Time to get over it and move on!

PS. And another BTW, I've got my own non-local model that is way better and more simple than John Reed's strawman example. It was the inspiration for the local model.

Your explanation is an explanation of what your code does. You can say that it is all absolutely local if you like, but I disagree. This is how I investigate the locality of your programming. Here is a piece of R code which reproduces exactly what your code does, once the random selections of various variables are replaced by pre-chosen values. You will notice that I keep the two hidden variables the same, for four "counterfactual" trials, in which the settings vary. See if you can spot the nonlocality. (I need to do some checks, this is not definitive).

Code: Select all

M <- 4
eta <- 3
beta <- 0.3

s <- 100         # Fred's singlet angle "s" in degrees
S <- rep(s, 4) * pi / 180

e <- 20          # Fred's auxiliary hidden variable "e" for Michel's lambda
E <- rep(e, 4) * pi / 180
R <- beta * cos(E / eta)^2  # Fred's lambda

aBits <- c(0, 1, 0, 1)
bBits <- c(0, 0, 1, 1)
aRadians <- aBits * pi / 2
bRadians <- bBits * pi / 2 + pi / 4
aDeg <- aBits * 90
bDeg <- bBits * 90 + 45

C <- abs(cos(aRadians - S)) < R        # TRUE is "fail", FALSE is "good"
A <- ifelse(C, -sign(sin(aRadians - S)), -sign(cos(aRadians - S)))
AA <- -sign(sin(aRadians - S))
D <- abs(cos(bRadians - S)) < R        # TRUE is "fail", FALSE is "good"
B <- ifelse(D, +sign(sin(bRadians - S)), +sign(cos(bRadians - S)))
BB <- +sign(sin(bRadians - S))
A <- ifelse(D & !(A == AA), -A, A)
B <- ifelse(C & !(B == BB), -B, B)
data <- cbind(aDeg, bDeg, A, B)
s                # angle "s"
e                # angle "e", determines "lambda"
data             # settings and outcomes

Here are the results:

Code: Select all

[1]> s                # angle "s"
[1] 100
> e                # angle "e", determines "lambda"
[1] 20
> data             # settings and outcomes
     aDeg bDeg  A  B
[1,]    0   45  1 -1
[2,]   90   45 -1  1
[3,]    0  135  1  1
[4,]   90  135 -1  1
>

Spot the nonlocality?

[Joy Christian]

Joy has written that the purpose of his work is to challenge conventional understandings of metaphysical concepts (like “locality” and “realism”).

I have not written any such thing. Only an illiterate fool who cannot read plain English would read what I have written and claim what you are claiming. Stop spreading lies about my work.

For anyone else who is interested, please read Section II of my following paper to find out what I have actually written: https://ieeexplore.ieee.org/stamp/stamp ... er=9418997
.

Thanks for the link, Joy! Here are your actual words:

My work on the subject [2]–[6] does not challenge this mathematical core, if it is viewed as a piece of mathematics. What it challenges are the metaphysical conclusions regarding locality and realism derived from that mathematical core. My work thus draws a sharp distinction between the mathematical core of Bell’s theorem and the metaphysical conclusions derived from it.

Now, regarding Fred's programming:

@gill1109 Oh BTW, you forgot to point out where my explanation is wrong so you must accept that it is absolutely correct. But even without that explanation the statistics shoot you down. Out of 1 million trials there are probably less than 20 events with the flips associated with a change in b. That is pretty statistically insignificant! That means there are 999,980 events that are completely local no matter what. You are finished. Done! Time to get over it and move on!

PS. And another BTW, I've got my own non-local model that is way better and more simple than John Reed's strawman example. It was the inspiration for the local model.

Your explanation is an explanation of what your code does. You can say that it is all absolutely local if you like, but I disagree. This is how I investigate the locality of your programming. Here is a piece of R code which reproduces exactly what your code does, once the random selections of various variables are replaced by pre-chosen values. You will notice that I keep the two hidden variables the same, for four "counterfactual" trials, in which the settings vary. See if you can spot the nonlocality. (I need to do some checks, this is not definitive).

Code: Select all

M <- 4
eta <- 3
beta <- 0.3

s <- 100         # Fred's singlet angle "s" in degrees
S <- rep(s, 4) * pi / 180

e <- 20          # Fred's auxiliary hidden variable "e" for Michel's lambda
E <- rep(e, 4) * pi / 180
R <- beta * cos(E / eta)^2  # Fred's lambda

aBits <- c(0, 1, 0, 1)
bBits <- c(0, 0, 1, 1)
aRadians <- aBits * pi / 2
bRadians <- bBits * pi / 2 + pi / 4
aDeg <- aBits * 90
bDeg <- bBits * 90 + 45

C <- abs(cos(aRadians - S)) < R        # TRUE is "fail", FALSE is "good"
A <- ifelse(C, -sign(sin(aRadians - S)), -sign(cos(aRadians - S)))
AA <- -sign(sin(aRadians - S))
D <- abs(cos(bRadians - S)) < R        # TRUE is "fail", FALSE is "good"
B <- ifelse(D, +sign(sin(bRadians - S)), +sign(cos(bRadians - S)))
BB <- +sign(sin(bRadians - S))
A <- ifelse(D & !(A == AA), -A, A)
B <- ifelse(C & !(B == BB), -B, B)
data <- cbind(aDeg, bDeg, A, B)
s                # angle "s"
e                # angle "e", determines "lambda"
data             # settings and outcomes

Here are the results:

Code: Select all

[1]> s                # angle "s"
[1] 100
> e                # angle "e", determines "lambda"
[1] 20
> data             # settings and outcomes
     aDeg bDeg  A  B
[1,]    0   45  1 -1
[2,]   90   45 -1  1
[3,]    0  135  1  1
[4,]   90  135 -1  1
>

Spot the nonlocality?

Your argument is nonsensical. Neither quantum mechanics nor experiments have anything meaningful to say about what happens in one trial or four, n = 1 or n = 4. The purpose of a local-realistic model is to reproduce statistical predictions of quantum mechanics in a large n limit, which the model presented in our paper does spectacularly.

The fact that our model is local is quite manifest from the analytical prescription for the measurement results A(a, h) and B(b, h) presented in our paper. See equations (13) to (26). One has to be quite ignorant of Einstein and Bell's definition of locality to believe the strawman you have created and then knocked down triumphantly. That is the kind of deception I have seen from you time and again during the past ten years. It does not surprise me anymore.

And quoting me journalistically from my papers will only convince dishonest Bell-believers. It will not impress honest scientists because by now everyone knows your game. See the recent posts by "local" and Justo. It is not only me and Fred who have deeply negative opinions about you.
.

[FrediFizzx]

@gill1109 Yep, more freakin' nonsense. I'm going to start deleting those posts if you keep it up. Worst case, there are 20 events out of a million where a change in b might coincide with a quaternion sign flip. Completely statistically insignificant. 999,980 events that are completely local. Go ahead, check that out with your R code. And no strawmen allowed. Your finished. Time to get over it and move on!
.

[gill1109]

Your argument is nonsensical. Neither quantum mechanics nor experiments have anything meaningful to say about what happens in one trial or four, n = 1 or n = 4. The purpose of a local-realistic model is to reproduce statistical predictions of quantum mechanics in a large n limit, which the model presented in our paper does spectacularly.

The fact that our model is local is quite manifest from the analytical prescription for the measurement results A(a, h) and B(b, h) presented in our paper. See equations (13) to (26). One has to be quite ignorant of Einstein and Bell's definition of locality to believe the strawman you have created and then knocked down triumphantly. That is the kind of deception I have seen from you time and again during the past ten years. It does not surprise me anymore.

And quoting me journalistically from my papers will only convince dishonest Bell-believers. It will not impress honest scientists because by now everyone knows your game. See the recent posts by "local" and Justo. It is not only me and Fred who have deeply negative opinions about you.
.

Dear Joy,

Are you telling me that you didn't actually mean what you wrote, but just put it into the paper in order to get published?

I don't think my quote from your latest published paper was "journalistic". If you didn't actually mean what you wrote, you shouldn't have written it.

Of course, it is easy to write a single computer program which for very large N reproduces the predictions of quantum mechanics. So what? You and Fred have to do more. The program should implement your own analytic prescription.

In my little example I actually performed an impossible experiment: I not only took N = 1, but I also actually performed four counterfactual runs of the same N = 1 experiment; namely, with each of the four pairs of CHSH settings. I kept the two hidden variables in Fred's code fixed: they were identically the same in four counterfactually different runs of the same N = 1 experiment. If the code correctly implemented a model of the Bell type A = A(a, lambda), B = B(b, lambda), then the outcome A could not change (with lambda and "a" fixed) when "b" is changed. But that's what happens! Fred's implementation of the so-called quaternionic spin flip is actually non-local. At least, in the flatland which is a classical digital computer.

This was just a test of whether or not Fred's computer program satisfies your analytical specifications.

I still have to check that Fred's most recent notebook still gives the same result with the same values of hidden variables and same settings and same N. (It's a moving target!).

@gill1109 Yep, more freakin' nonsense. I'm going to start deleting those posts if you keep it up. Worst case, there are 20 events out of a million where a change in b might coincide with a quaternion sign flip. Completely statistically insignificant. 999,980 events that are completely local. Go ahead, check that out with your R code. And no strawmen allowed. Your finished. Time to get over it and move on!
.

Actually, Fred, it 's a whole lot more than "at most 20 in a million". I will go ahead and check and let you know.

[FrediFizzx]

@gill1109 I actually messed up on the original percentage. Went the wrong way. For 10,000 trials I get 2 possible cases out of 10 runs. So, that is 0.2/10,000 = 0.00002 or 0.0000002 percent. That gives you 20 events for a million trials. Anyways, 1,000,000 x 0.00002 = 20.
.

[gill1109]

@gill1109 I actually messed up on the original percentage. Went the wrong way. For 10,000 trials I get 2 possible cases out of 10 runs. So, that is 0.2/10,000 = 0.00002 or 0.0000002 percent. That gives you 20 events for a million trials. Anyways, 1,000,000 x 0.00002 = 20.
.

Yes, but in order to determine whether or not a flip occurs, Alice and Bob need to communicate with one another. This paper: https://www.pnas.org/content/118/17/e2020569118 shows that Alice and Bob would need to communicate with one another a bit more than 40% of the trials in order to "fix" the observed statistics generated from a local realistic set-up by changing outcomes in some trials.

"A bit more than 40%" is actually the fraction 0.4142136... = (2 sqrt 2 - 2)/2

Violations of locality and free choice are equivalent resources in Bell experiments
Pawel Blasiak, Emmanuel M. Pothos, James M. Yearsley, Christoph Gallus, and Ewa Borsuk

PNAS April 27, 2021 118 (17) e2020569118; https://doi.org/10.1073/pnas.2020569118
Edited by Anthony Leggett, University of Illinois at Urbana–Champaign, Urbana, IL, and approved March 11, 2021 (received for review October 1, 2020)

Significance: Faced with a violation of Bell inequalities, a committed realist might pursue an explanation of the observed correlations on the basis of violations of the locality or free choice (sometimes called measurement independence) assumptions. The question of whether it is better to abandon (partially or completely) locality or free choice has been strongly debated since the inception of Bell inequalities, with ardent supporters on either side. We offer a comprehensive treatment that allows a comparison of both assumptions on an equal footing, demonstrating a deep interchangeability. This both advances the foundational debate and provides quantitative answers regarding the weight of each assumption for causal (or realist) explanations of observed correlations.

Abstract: Bell inequalities rest on three fundamental assumptions: realism, locality, and free choice, which lead to nontrivial constraints on correlations in very simple experiments. If we retain realism, then violation of the inequalities implies that at least one of the remaining two assumptions must fail, which can have profound consequences for the causal explanation of the experiment. We investigate the extent to which a given assumption needs to be relaxed for the other to hold at all costs, based on the observation that a violation need not occur on every experimental trial, even when describing correlations violating Bell inequalities. How often this needs to be the case determines the degree of, respectively, locality or free choice in the observed experimental behavior. Despite their disparate character, we show that both assumptions are equally costly. Namely, the resources required to explain the experimental statistics (measured by the frequency of causal interventions of either sort) are exactly the same. Furthermore, we compute such defined measures of locality and free choice for any nonsignaling statistics in a Bell experiment with binary settings, showing that it is directly related to the amount of violation of the so-called Clauser–Horne–Shimony–Holt inequalities. This result is theory independent as it refers directly to the experimental statistics. Additionally, we show how the local fraction results for quantum-mechanical frameworks with infinite number of settings translate into analogous statements for the measure of free choice we introduce. Thus, concerning statistics, causal explanations resorting to either locality or free choice violations are fully interchangeable.

********

More than 40% seems like an awful lot. In my own little experiments I found that with Fred's model, only very roughly 20% of trials were non-local in the sense that A = A(a, lambda) and B = B(b, lambda) is violated. But Blasiak et al. have a stricter criterion. They ask: how often would Alice and Bob need to communicate in order to change the data of individual trials in order to fix the statistics. My experiment says that they would have to communicate *at least* on about 20% of the trials - but of course, they could communicate, only to discover that no flip is needed. So if half the communications led to no flip, the total amount of communications could be about 40%.

A simple way of looking at this is as follows. If they communicated every time they could get S = 4 . Without communication they can get S = 2. They would like to get S = 2 sqrt 2. So they communicate on a fraction p of the trials and want to arrange that (1 - p) * 2 + p * 4 = 2 sqrt 2. Solving for p we find p = (2 sqrt 2 - 2)/2 = 0.4142136... The paper of Blasiak et al. shows that this is actually the best you can do.

[FrediFizzx]

@gill1109 Stop with all the nonsense and program up the actual calculation for our model! Ask questions if you need help.
.

[gill1109]

@gill1109 Stop with all the nonsense and program up the actual calculation for our model! Ask questions if you need help.
.

That was not nonsense. The authors of that paper are very smart guys, and the journal PNAS is a very prestigious journal of the US Academy of Sciences. I've checked their maths. It looks correct to me. Thanks, but please first re-code your model in fast Python or R and I would be only too glad to help with a big computer experiment.

You see, I'm afraid that if I use the formulas which John Reed found you'll accuse me of using a straw-man argument. So what would be the point of making the effort?

But Fred, your model is fantastic, it sheds new light on the Blasiak et al. results, and suggests exciting new questions to investigate. It seems to me that your model is minimally non-local in another interesting sense. Not in terms of communication per trial as the experiment proceeds, but in terms of post-experiment fixing of individual outcomes. My conjecture is that the answer will be one quarter of 2 sqrt 2 minus 2.

[Joy Christian]

The authors of that paper are very smart guys...

Anyone who believes in Bell's junk theorem, or that nonlocality is inevitable on its basis, cannot possibly be "smart guys", at least in my eyes and the eyes of some participants of this forum.
.

[gill1109]

The authors of that paper are very smart guys...

Anyone who believes in Bell's junk theorem, or that nonlocality is inevitable on its basis, cannot possibly be "smart guys", at least in my eyes and the eyes of some participants of this forum.
.

I participate on this forum precisely because about half the participants are pro and half are anti Bell. You learn unmeasurably more by talking to people who think differently from you, than by talking to people who think the same. Fred has uncovered a really interesting phenomenon and as a mathematician I plan to study it in more depth and relate it to other recent mathematical developments.

My work on the subject does not challenge this mathematical core, if it is viewed as a piece of mathematics. What it challenges are the metaphysical conclusions regarding locality and realism derived from that mathematical core.

What we are now investigating is the quantification of the amount by which "locality" and/or "realism" should be adapted so as to accommodate quantum mechanical predictions (and hence to match laboratory reality).

[gill1109]

Here is a new Mathematica notebook which incorporates the core of Fred's recent Mathematica code for a CHSH experiment. I have adapted it to perform 4 counterfactual repetitions of a single trial with the same values of the two hidden variables in each trial. In the four counterfactual repetitions, the setting pairs simply run through the four CHSH setting pairs. https://www.wolframcloud.com/obj/gill1109/Published/newCS-11-CHSH-paper-RDG-cfg.nb
At the end of the notebook I print out the settings and the outcomes. Note that the first two trials have a = 0, then a = 90, while b =45 (fixed). But Bob's outcome changed sign!

@gill1109 It is nonsense. In our model the sign flips are roughly about 5 percent of the total events. Of that 5 percent a small fraction are associated with a change of b. Of course you CAN'T use a strawman to calculate this. What? You haven't programmed the actual model yet? You should be asking questions instead of spewing nonsense.
.

I haven't programmed anything new. I took your program, changed nothing in the outcome calculation part or the matching part, Fred. All I did was fix the hidden variables and the settings of four trials, at the start of the program.

My Mathematica trial has just about finished now, so that is all for the time being. I may be able to use a university licensed version soon.

[FrediFizzx]

@gill1109 It is nonsense. In our model the sign flips are roughly about 5 percent of the total events. Of that 5 percent a small fraction are associated with a change of b. Of course you CAN'T use a strawman to calculate this. What? You haven't programmed the actual model yet? You should be asking questions instead of spewing nonsense.
.

[FrediFizzx]

@gill1109 Your tests are all statistically insignificant. Try again with more trials. Plus the change in sign is due to quaternion sign flips.
.

[FrediFizzx]

@gill1109 Your tests are all statistically insignificant. Try again with more trials. Plus the change in sign is due to quaternion sign flips.

We are not sure why a mathematical statistician doesn't seem to understand that yet.
.

[Heinera]

@gill1109 Your tests are all statistically insignificant. Try again with more trials. Plus the change in sign is due to quaternion sign flips.
.

This has nothing to do with statistical significance. It is a method to check if you have faithfully implemented the functions A(a, lambda) and B(b, lambda) in your program. If we keep lambda fixed and change only b, the test runs by Richard and jreed shows that sometimes the outcome of A(a, lambda) also changes in your program. One instance of this is enough to prove that there is something fishy with your code.

It seems that "quaternionic sign flips" is just a code word for "spooky action at a distance."

[FrediFizzx]

@gill1109 Your tests are all statistically insignificant. Try again with more trials. Plus the change in sign is due to quaternion sign flips.
.

This has nothing to do with statistical significance. It is a method to check if you have faithfully implemented the functions A(a, lambda) and B(b, lambda) in your program. If we keep lambda fixed and changes only b, the test runs by Richard and jreed shows that sometimes the outcome of A(a, lambda) also changes in your program. One instance of this is enough to prove that there is something fishy about your code.

It seems that "quaternionic sign flips" is just a code word for "spooky action at a distance."

Well, we certainly don't expect you to understand statistical significance since you haven't shown much brilliance about Bell's junk physics theory. 999,980 events are completely local out of a million trials! 20 are due to emulating quaternion sign flips. Gill knows what we are talking about. He just doesn't want to cop to it. Plus Joy has already successfully shot down the junk theory via product calculations using both geometric algebra and quaternions.
.

[gill1109]

@gill1109 Your tests are all statistically insignificant. Try again with more trials. Plus the change in sign is due to quaternion sign flips.
.

This has nothing to do with statistical significance. It is a method to check if you have faithfully implemented the functions A(a, lambda) and B(b, lambda) in your program. If we keep lambda fixed and change only b, the test runs by Richard and jreed shows that sometimes the outcome of A(a, lambda) also changes in your program. One instance of this is enough to prove that there is something fishy with your code.

It seems that "quaternionic sign flips" is just a code word for "spooky action at a distance."

Heinera is absolutely right.

Fred himself (or was it Joy?) is already calling it “mysterious”.

I think 20 is more like 20%.

[FrediFizzx]

@gill1109 More freakin' NONSENSE! Heine is rarely correct and is totally wrong here. I already told you that the total number of sign flips is about 5 percent so no way you can get 20 percent. Pay freakin' attention! So, the percentage of a change of b involved with a sign flip is a tiny fraction of the 5 percent. Time to get over it and move on!
.

[Joy Christian]

@gill1109 More freakin' NONSENSE! Heine is rarely correct and is totally wrong here. I already told you that the total number of sign flips is about 5 percent so no way you can get 20 percent. Pay freakin' attention! So, the percentage of a change of b involved with a sign flip is a tiny fraction of the 5 percent. Time to get over it and move on!

Moreover, "the percentage of a change of b involved with a sign flip" is accidental, not systematic. One does not have to be a "mathematical statistician" to understand that.

Twenty quaternionic sign flips in a million trials accidentally coincide with a change in the setting b. Big deal.
.