What I don't quite get about entanglement...

[imsmooth]

I have watched videos on Bell's Theorem and quantum entanglement. What I don't quite get is:

if a pair of particles are created and separated and then we measure one we already know the other particle has the opposite orientation. How do we know for sure that the first particle we measure does not have a pre-defined orientation? That seems to be the easiest answer.

 

[PeterDonis]

I have watched videos

Videos are generally not good sources to learn from. Textbooks are better. There are plenty of good QM textbooks available that explain entanglement and Bell's Theorem; my personal preference is Ballentine, but there are others as well.

How do we know for sure that the first particle we measure does not have a pre-defined orientation?

Because such a model is ruled out by the fact that we have experimentally measured violations of the Bell inequalities.

 

[imsmooth]

I have read how we see 50% difference in measurements when one expects 55.6%, supporting that there are no hidden variables. I still feel that there is still something wrong with the measurements rather than accepting faster-than-light transfer of information. We just don't know where there error lies.

 

[PeterDonis]

I have read

Where? Please give a specific reference.

how we see 50% difference in measurements when one expects 55.6%, supporting that there are no hidden variables.

I'm not sure what experiment you are referring to, but the violations of the Bell inequalities that have been observed are larger than this. The violation can be by as much as a factor of $\sqrt{2}$.

I still feel that there is still something wrong with the measurements than faster-than-light transfer of information.

There is no FTL transfer of information in QM. Bell inequality violations cannot be used to transfer information faster than light.

 

[imsmooth]

Information is not transfered FTL, but do not the particles assume opposite spins faster than information can travel?

 

[Hornbein]

Information is not transfered FTL, but do not the particles assume opposite spins faster than information can travel?

By "no information" they mean, you can't use entanglement to beat the stock market or otherwise communicate anything that is useful. They call this sort of thing "nonlocality." Even Albert Einstein felt nonlocality was too weird, so you have plenty of company. As far as I know the mechanism is completely a mystery. Susskind of Stanford has at least one proposal -- a wormhole connects the particles -- but such hasn't caught on.

 

[PeterDonis]

do not the particles assume opposite spins faster than information can travel?

No. QM doesn't work that way.

 

[Demystifier]

I still feel that there is still something wrong with the measurements rather than accepting faster-than-light transfer of information. We just don't know where there error lies.

Many smart physicists initially had "feelings" similar to yours. But when studied more carefully, they convinced themselves that there were no errors, except in their "feelings". Nature is not consistent with the assumption of local realism; either realism or locality (or both) is wrong.

In particular, if realism is true, then locality is wrong. Non-locality (with realism assumed true) means that some kind of "information" travels instantaneously, but this is not a kind of information one can control and use for communication. This "information" is hidden.

Alternatively, if locality is true, then there is no any hidden "information" that travels instantaneously, but then realism is wrong, meaning that systems do not have any pre-existing properties (values of observables) before measurements.

 

[martinbn]

Many smart physicists initially had "feelings" similar to yours. But when studied more carefully, they convinced themselves that there were no errors, except in their "feelings". Nature is not consistent with the assumption of local realism; either realism or locality (or both) is wrong.

In particular, if realism is true, then locality is wrong. Non-locality (with realism assumed true) means that some kind of "information" travels instantaneously, but this is not a kind of information one can control and use for communication. This "information" is hidden.

Alternatively, if locality is true, then there is no any hidden "information" that travels instantaneously, but then realism is wrong, meaning that systems do not have any pre-existing properties (values of observables) before measurements.

May be it is worth pointing out the QM is non-realistic. Especially when it comes to non-commuting observables, for example spin along different axes. That was known well before Bell.

 

[DrChinese]

a. I have watched videos on Bell's Theorem and quantum entanglement. What I don't quite get is:

if a pair of particles are created and separated and then we measure one we already know the other particle has the opposite orientation. How do we know for sure that the first particle we measure does not have a pre-defined orientation? That seems to be the easiest answer.

b. .I have read how we see 50% difference in measurements when one expects 55.6%, supporting that there are no hidden variables. I still feel that there is still something wrong with the measurements rather than accepting faster-than-light transfer of information. We just don't know where there error lies.

a. As already mentioned, Bell's theorem and related experiments demonstrate conclusively that there cannot be a predetermined orientation - until there is some kind of nonlocality. Some people equate nonlocality with faster-than-light (FTL), and it is possible there is action at a distance.

b. The important thing to understand that while you may not think the difference is much, perhaps due to experimental error: that cannot be the case. Most Bell tests measure using a formula called CHSH, used to compute a value usually labeled "S". The angle settings for this formula are selected to highlight the issues involved. Bell tests have been performed to extremely high precision, such as this one in which the accuracy is to 30 standard deviations:

Violation of Bell's inequality under strict Einstein locality conditions (1998)

i) Your pre-defined/predetermined spin hypothesis (also known as "local realism"): S cannot exceed 2.
ii) Ideal quantum predictions: S=2.82, which is higher than i) by a factor of √2 as @PeterDonis mentioned.
iii) Actual experiment: S = 2.73±0.02, which fits the quantum prediction well - but is way outside the local realistic bound.

The science of entanglement has progressed substantially since Bell. It is not even necessary for a pair of photons (such as measured in the reference) to have ever even interacted in the past to become entangled. They can be created from different sources and later become entangled remotely, which makes it difficult to imagine them as starting out with predefined spin (polarization).

 

[FallenStarFeatures]

I have read how we see 50% difference in measurements when one expects 55.6%, supporting that there are no hidden variables. I still feel that there is still something wrong with the measurements rather than accepting faster-than-light transfer of information. We just don't know where there error lies.

Sorry, but your conclusions are incorrect. Experiments confirming violations of Bell's inequalities consistent with quantum mechanics rule out local realism and support non-local hidden variable theories. These experiments do not require FTL transmission of information. No one has identified any significant error in these measurements.

 

[martinbn]

Sorry, but your conclusions are incorrect. Experiments confirming violations of Bell's inequalities consistent with quantum mechanics rule out local realism and support non-local hidden variable theories. These experiments do not require FTL transmission of information. No one has identified any significant error in these measurements.

I don't think that they support non-local hidden variables. It is perfectly possible that there are no hidden variables of any kind. They do show that there are no local hidden variables.

 

[FallenStarFeatures]

I don't think that they support non-local hidden variables. It is perfectly possible that there are no hidden variables of any kind. They do show that there are no local hidden variables.

Bell used David Bohm's particle-spin version of the EPR paradox to formulate Bell's Inequalities, which he used to demonstrate the incompatibility of local hidden variable theories (of the type that Einstein had proposed) with the predictions of quantum theory. In that paper, Bell drew attention to the fact that Bohm's own non-local hidden variable theory was explicitly vindicated by this proof:

"There have been attempts to show that even without such a separability or locality requirement no "hidden variable" interpretation of quantum mechanics is possible. These attempts have been examined elsewhere and found wanting. Moreover, a hidden variable interpretation of elementary quantum theory has been explicitly constructed. That particular interpretation has indeed a grossly non-local structure. This is characteristic, according to the result to be proved here, of any such theory which reproduces exactly the quantum mechanical predictions."

https://cds.cern.ch/record/111654/files/vol1p195-200_001.pdf

 

[martinbn]

Bell used David Bohm's particle-spin version of the EPR paradox to formulate Bell's Inequalities, which he used to demonstrate the incompatibility of local hidden variable theories (of the type that Einstein had proposed) with the predictions of quantum theory. In that paper, Bell drew attention to the fact that Bohm's own non-local hidden variable theory was explicitly vindicated by this proof:

"There have been attempts to show that even without such a separability or locality requirement no "hidden variable" interpretation of quantum mechanics is possible. These attempts have been examined elsewhere and found wanting. Moreover, a hidden variable interpretation of elementary quantum theory has been explicitly constructed. That particular interpretation has indeed a grossly non-local structure. This is characteristic, according to the result to be proved here, of any such theory which reproduces exactly the quantum mechanical predictions."

https://cds.cern.ch/record/111654/files/vol1p195-200_001.pdf

He says that it is possible. But there is a difference between possible and supported by experiments. I only objected to your statement that experiments support non-local hidden variables as if there are no other possibilities.

 

[PeroK]

I don't think that they support non-local hidden variables. It is perfectly possible that there are no hidden variables of any kind. They do show that there are no local hidden variables.

I was disappointed to learn that the Bohmians have no expectation of ever being able to conduct an experiment to reveal the inner workings that the interpretation is based on.

In that paper, Bell drew attention to the fact that Bohm's own non-local hidden variable theory was explicitly vindicated by this proof:

Every interpretation of QM is vindicated by the experiments testing Bell's Theorem. Bohmian Mechanics is not especially vindicated.

 

[Fra]

How do we know for sure that the first particle we measure does not have a pre-defined orientation? That seems to be the easiest answer.

a. As already mentioned, Bell's theorem and related experiments demonstrate conclusively that there cannot be a predetermined orientation - until there is some kind of nonlocality. Some people equate nonlocality with faster-than-light (FTL), and it is possible there is action at a distance.

I think it's worth noting that there can be "pre-determination", as to explain the correlation in two different ways, with any "action at distance at all".

1) What is normally meant is a pre-determiniation which is treated as a simple ignorance of the physicsist. Meaning that the physical interactions happens as per known mechanisms partitioned into the hidden variable, and the actual outcome is random selection among them. This is the form assumed in bells theorem and it is the kind of pre-determination that is ruled out. Unless you expose some IMO quite pathological loophole ideas or pathological action at distance.

2) The other "pre-determiniation" that is entirely compatible with QM, is that there is some variable that is created and correlated from the preparation procecure, that determines the correlation, but the outcome is not determined by simple averaging simple actions, but from actions that account for the uncertainty. Ie. the uncertainty of the hidden variable, affects the interactions themselves, not just the physicists predictions. This forms is allowed still by QM, but we do not understand it's mechanism. Some don't care, and just settle with that we have a computational model. Some think there is a value to find the mechanism. This form does not require any action at distance, and the "pre-determined" value here can possibly be subjective, in the sense that, a key in these experiments i ISOLATION. Which means that it is fully possible that noone except hte particles themselves "know" this hidden value, and that this influences the interactions themselves. This means that the interacting environment itself, such as detectors etc, also are ignorant to this value. Thus it seems at least within reason, that this kind of intrinsic uncertainty does affect the interaction itself - unlike the Bell type of hidden variables.

Except for those that still seek odd loopholes to the "original" type (1), with (2) I think the options are to seek this deeper understanding or just settle with that the model works, and thinkg that there is no benefit in understandin the mechanism.

/Fredrik

 

[zoltrix]

as far as I know the true mistery is the wave-particle duality
entanglement is a direct consequence

 

[Fra]

as far as I know the true mistery is the wave-particle duality

But what does that mean? that duality is not really understood, because the "wave" is not a "classical wave" on par with a pressure wave for example. So we still do not understand.

/Fredrik

 

[PeterDonis]

What is normally meant is a pre-determiniation which is treated as a simple ignorance of the physicsist.

The other "pre-determiniation" that is entirely compatible with QM, is that there is some variable that is created and correlated from the preparation procecure, that determines the correlation, but the outcome is not determined by simple averaging simple actions, but from actions that account for the uncertainty.

Do you have references for this?

 

[Fra]

Do you have references for this?

I think referring to normal quantum mechanics is enough here? I am not here claiming a particular form or understanding of this pre-determination that determines the correlation, except its determined by the preparation procedure (and of course the choices of the detector settings). So don't see that this is controversial. All I did was putting this in an unusual scentence to be more suggestive. And the non simple averaging i refer to is of course the normal "superposition in QM". We know this is right, but do we "understand" it, is the question.

/Fredrik

 

[DrChinese]

I think it's worth noting that there can be "pre-determination", as to explain the correlation in two different ways, with any "action at distance at all".

1) What is normally meant is a pre-determiniation which is treated as a simple ignorance of the physicsist. Meaning that the physical interactions happens as per known mechanisms partitioned into the hidden variable, and the actual outcome is random selection among them. This is the form assumed in bells theorem and it is the kind of pre-determination that is ruled out. Unless you expose some IMO quite pathological loophole ideas or pathological action at distance.

2) The other "pre-determiniation" that is entirely compatible with QM, is that there is some variable that is created and correlated from the preparation procecure, that determines the correlation, but the outcome is not determined by simple averaging simple actions, but from actions that account for the uncertainty. Ie. the uncertainty of the hidden variable, affects the interactions themselves, not just the physicists predictions. This forms is allowed still by QM, but we do not understand it's mechanism…

I’m glad @PeterDonis asked for a reference on the above. Almost everything you say goes against generally accepted science as of today. What you were talking about is the state of the art in 1984, and there has been much change since then.

So no, just referring to QM in general is not going to do it. I can supply references that show that entangled particles can be created from independent sources far away from each other, never existing in a common light cone. They can be entangled at any time; mid flight, or before or even after detection. And there’s nothing statistical involved, they display perfect correlations at any common angle setting. This is basically a demonstration of the 1935 EPR concept, but with a twist they never imagined. Namely, that the elements of reality are demonstrated far away from each other and without mutual interaction of an any kind.

Further, in the late 80s, we saw the advent of the GHZ theorem. Its QM prediction is exactly the opposite of any realistic hypothesis that lacks non-locality.

This is 2024, and we cannot ignore the amazing science of the last 30 years or so on this fascinating subject.

 

[FallenStarFeatures]

I was disappointed to learn that the Bohmians have no expectation of ever being able to conduct an experiment to reveal the inner workings that the interpretation is based on.

Have a citation to share? I'd be curious to ponder which "inner workings" you're referring to.

Every interpretation of QM is vindicated by the experiments testing Bell's Theorem. Bohmian Mechanics is not especially vindicated.

Yes, the confirmation of Bell's Theorem was welcomed by all QM interpretations, except perhaps local realist interpretations which subsequently resorted to superdeterminism, local interpretations forced to give up on realism, and realist interpretations obliged to embrace non-localism. Bohmian Mechanics accepted the resounding silence in stride, as its inherent non-local realism was never in doubt.

There was another historical context, however, in which the confirmation of Bell's Theorem indeed brought vindication to non-local hidden variable interpretations such as Bohmian Mechanics. An early version of the theory, proposed by Louis de Broglie in 1927, had been rejected by virtually all of the early pioneers of quantum mechanics. John von Neumann subsequently published a proof ruling out all hidden variable theories as inconsistent with QM. Errors in von Neumann's assumptions were pointed out by Greta Hermann, but were not widely acknowledged. As a result, Bohm's reformulation of pilot wave theory in 1951 was largely dismissed until Bell spotlighted it in his 1964 paper on Einstein's EPR paradox.

With the confirmation of Bell's Theorem in Nobel Prize-winning experiments by John Clauser, Alain Aspect, and Anton Zeilinger in 2015, one might expect the case in support of non-local hidden variable theories would finally be acknowledged. However, not even the Nobel Prize Committee took care to make that non-local distinction its 2022 press release, reiterating von Neumann's erroneous claim while honoring Bell's groundbreaking work that had six decades earlier conclusively demolished it.

https://www.nobelprize.org/prizes/physics/2022/press-release/

 

[Fra]

I think we have confusion again. Perhaps I have expressed something in an unclear way...

I’m glad @PeterDonis asked for a reference on the above. Almost everything you say goes against generally accepted science as of today. What you were talking about is the state of the art in 1984, and there has been much change since then.

Not sure I follow you. Do you mean 1964, guess it was a typo? IF so, I think you must missed my point, which may very well be my fault.

So no, just referring to QM in general is not going to do it. I can supply references that show that entangled particles can be created from independent sources far away from each other, never existing in a common light cone. They can be entangled at any time; mid flight, or before or even after detection. And there’s nothing statistical involved, they display perfect correlations at any common angle setting. This is basically a demonstration of the 1935 EPR concept, but with a twist they never imagined. Namely, that the elements of reality are demonstrated far away from each other and without mutual interaction of an any kind.

Further, in the late 80s, we saw the advent of the GHZ theorem. Its QM prediction is exactly the opposite of any realistic hypothesis that lacks non-locality.

This is 2024, and we cannot ignore the amazing science of the last 30 years or so on this fascinating subject.

Nothing of what I tried to convey, contradicts what we know. Neither does your entaglement swapping. Indeed particles that was never in contact can be MADE to be entangled (via mechanisms that has been thourouglhy discussed in other threads, so no need to rehash)

It sound like you think I am advocating a bell type hidden variable theory by looking for loopholes as many did before 1964?? I'll leave it there, otherwise this will turn into an interpretation discussion, whichwasn't the intention.

What i was asked to give reference to, really refers to that preparation procedures (including pair production, key swapping or whatever) by construction implies the correlation - it is determined. That is not the same as that the single final measuements are determined, because they are also modulated by settings, and are random. The "variable" I mention is not a bell-type variable, that I thought was obvious, it's some other strange thing (ie. non classical). But I refrain from dressing it.

/Fredrik

 

[PeterDonis]

I think referring to normal quantum mechanics is enough here?

No, it's not. What you are posting looks like your personal viewpoint on QM. It does not look like just repeating what is in standard QM textbooks or peer-reviewed papers. But if you think it is, then, as I said, you need to give a reference to support that.

 

[PeroK]

Have a citation to share? I'd be curious to ponder which "inner workings" you're referring to.

"Inner workings" was a shorthand for well-defined particle trajectories.

In Bohmian mechanics, a particle has a definite position at all times. In the double slit, the particle has a well-defined trajectory that goes through a specific slit. Personally, I would only subscribe to Bohmian Mechanics if there were some hope of experimentally demonstrating that.

Moreover, these well-defined particle trajectories can be extended to QFT. Here's a recent paper on this:

https://arxiv.org/abs/quant-ph/0303156

"

 

[DrChinese]

A. The other "pre-determination" that is entirely compatible with QM, is that there is some variable that is created and correlated from the preparation procedure, that determines the correlation, but the outcome is not determined by simple averaging simple actions, but from actions that account for the uncertainty.

B. Not sure I follow you. Do you mean 1964, guess it was a typo? IF so, I think you must missed my point, which may very well be my fault. ... Nothing of what I tried to convey, contradicts what we know. Neither does your entanglement swapping.

C. Indeed particles that was never in contact can be MADE to be entangled (via mechanisms that has been thoroughly discussed in other threads, so no need to rehash)

What i was asked to give reference to, really refers to that preparation procedures (including pair production, key swapping or whatever) by construction implies the correlation - it is determined.

A. This is not standard science today, which is why I (and @PeterDonis) are asking for references. There is no predetermination at the time of creation (preparation), because the particles are created in a superposition. This is orthodox QM. It is quantum nonlocality that is responsible for the correlations, via an unknown mechanism which otherwise describes expected outcomes accurately. Zeilinger, 1999: "...quantum teleportation and quantum computation are direct applications of quantum entanglement, the concept underlying quantum nonlocality...". Or GHZ: Experimental test of quantum nonlocality in three-photon Greenberger–Horne–Zeilinger entanglement.

And please note an important detail often overlooked, as you have in your statement: QM predicts the violation of a Bell Inequality in entanglement tests, which is (as you say) statistical. But it also predicts "perfect correlations" which must be accounted for as well! Predetermination cannot account for those, and the nail in the coffin on that was GHZ (cited above) in which each and every run produces precisely the opposite result of a predetermination prediction. "...actions that account for the uncertainty"? There is no uncertainty in these.

B. I meant 1984. That was 40 years ago, right after the ground-breaking works of Aspect et al. Note that there have been 3 (actually of course many more) substantial developments that wreck the ideas you are espousing:

i. 1998: The experiment of Weihs et al demonstrates and extends the Aspect concept (which used time varying analyzers) to change measurement settings midflight faster than there could be any type communication affecting the Alice/Bob outcomes. A Bell test excluding sub-lightspeed/c explanations for the results.

ii. 2001 (cited above): Experimental confirmation of the GHZ Theorem, which shows there cannot be predefined outcomes using a completely different technique than Bell. And without statistics!

iii. 2008: Experimental demonstration of remote entanglement swapping of photons with no common past (no prior interaction with any 3rd element either). This shows that perfect correlations (what I call "EPR with a twist") must be quantum nonlocal. In his wildest imagination, Einstein could never have anticipated this result.

C. No one knows the mechanisms for the above cited experiments. Interpretations of QM attempt to supply that.

So the answer to the OP's question is: There can be no predetermination/predefinition of orientations during preparation/creation of entangled particles because it is ruled out experimentally on many levels, as predicted by garden variety QM - and everything since.

 

[DrChinese]

Here's a recent paper on this:

https://arxiv.org/abs/quant-ph/0303156

I have no issue with the paper itself, even if I don't agree with it (since I am not a Bohmian realist).

But it still made me smile to see a paper from 21 years ago labeled "recent".

(And to be fair, I consider almost any papers from the last 30 years to be recent.)

 

[FallenStarFeatures]

In Bohmian mechanics, a particle has a definite position at all times. In the double slit, the particle has a well-defined trajectory that goes through a specific slit. Personally, I would only subscribe to Bohmian Mechanics if there were some hope of experimentally demonstrating that.

Thanks for the link to the 2004 paper by Durr, Goldstein, Tumulka, and Zanghi. While I don't have enough background in QFT to comment, it led me back to a subsequent 2009 paper by the latter three authors:

https://arxiv.org/pdf/0912.2666

This paper shows how to use Bohmian trajectories to analyse the problem you referred to, the "well-defined" trajectory of a particle in the Double-Slit experiment. What I like about this approach is how you don't have to "subscribe to Bohmian Mechanics" to make use of the practical tools it provides. Section 4 explains how to calculate the Quantum Potential Q, which quantifies the influence of wave function amplitudes in configuration space on the Hamilton-Jacobi representation of the particle's motion in 3D physical space:
$$Q = - \frac {ℏ^2}{2m} \frac {∇^2 R}{R}$$
where m is particle mass, R is the amplitude of the wave function, and $∇^2$ is its Laplacian second derivative. In practical terms, the smaller the particle mass and/or probability amplitude, the more negative influence extreme amplitude gradients will have on the particle's motion. So in close vicinity to the edge of a slit, where wave function amplitude varies dramatically, this predicts huge shifts in probability density in 3D physical space, i.e. quantum interference patterns. Of course, the path of a particular particle will be sensitive to its initial position, and its precise trajectory subject to quantum probability density. But we can run the experiment and measure how closely our calculations do in fact correspond to observed results. Here's an example of what you would see:

https://www.researchgate.net/figure...two-slits-is-considered-as-the_fig1_368485221

TL;DR: It works, regardless of whether I decide to believe in non-local hidden variables.

 

[Fra]

So the answer to the OP's question is: There can be no predetermination/predefinition of orientations during preparation/creation of entangled particles because it is ruled out experimentally on many levels, as predicted by garden variety QM - and everything since.

Agreed. But I think this is this the misunderstanding of what i tried to say. I never suggested said that the orientations that was pre-determined, the orientation is manifestly a relation to the environment, and this relation is not in place as the entangled pair is explicitly ISOLATED from interacting with the environment. If the isolation is broken, the entanglement is immediately lost. It is their _correlation_ that is pre-determined; this is a relation between the two entangled parts only. This is all that is needed to explain the correlation, we aren't asked to explain the outcomes, because they are random. My only simply point was, we do not need to pre-determined the absolute outcomes, in order to determine the correlation as Reichenbach's Common Cause Principle.

There is no reference to the environment, thus the notion of orientation beyond the relative correlation is not known or even defined. As soon as the relation to the environment is established, the isolation is broken and so is entanglement. I think the detail of isolation is critical, and it is also what makes the relation to the environment undefined. For me I think this is a logical conclusion from standard QM. If this simple but not often pointed out observation needs a reference of someone making same basic point in a paper? then I give up. The simplest solution is to to main moderators delete my previous post if it doesn't makes sense in relation to forum rules?

What i try to do here is rather to demystify while still beeing clear what isnt known. Sometimea i get a feeling many wants to fuel the mystery.

/Fredrik

 

[DrChinese]

Agreed. But I think this is this the misunderstanding of what i tried to say. I never suggested said that the orientations that was pre-determined, the orientation is manifestly a relation to the environment, and this relation is not in place as the entangled pair is explicitly ISOLATED from interacting with the environment. If the isolation is broken, the entanglement is immediately lost. It is their _correlation_ that is pre-determined; this is a relation between the two entangled parts only. …

It may seem like I’m trying to pick apart everything you were saying, but I’m really not. Recent* experiments rule out many of the ideas being posted here, so I try to point those out.

It is possible to entangle distant photons after they are created. Prior to their creation, they have no relationship nor correlation whatsoever. So the correlation you mention cannot be predetermined, even as a relationship and not specific values.

Also, I’m not sure how you can factor in the environment in outcomes. Some elements of the environment do interact with photons, but don’t have any observable effect. Other elements of the environment, such as a polarizing beam splitter, do have an effect. But it is difficult to envision that interaction as determining the outcome due to the measurement apparatus. After all, Alice and Bob are distant and the environment always cancels out. We know that because of perfect correlations.

* recent meaning within the last 20 or 30 years lol.