What did Einstein mean by “spooky action at a distance”?

(backreaction.blogspot.com)

126 points | by nsoonhui 1055 days ago

15 comments

  • eternalban 1054 days ago
    This is a great 1985 Danish documentary (it's in English) that details Einstein & Bohr's fundamental disagreement regarding the interpretation of QM, and Alain Aspect's 1980-82 EPR experiments showing the violation of Bell's Inequalities. Both the historic elements and Aspect's groundbreaking experimental setup are detailed.

    It's a little gem featuring interviews with a young Alain Aspect, John Bell, David Bohm, Abner Shimony, John Wheeler, & a very excited Holger Bech Nielsen. It is very accessible yet hugely informative, and somehow (with that soundtrack) enchanting.

    Quantum Entanglement - Atomic Physics and Reality: https://www.youtube.com/watch?v=BFvJOZ51tmc

    https://en.wikipedia.org/wiki/Alain_Aspect

    https://en.wikipedia.org/wiki/Aspect%27s_experiment

    • kergonath 1054 days ago
      There are several good seminars by Alain Aspect on YouTube as well, ranging from highly technical to accessible vulgarisation.
  • emrah 1054 days ago
    Entanglement isn't like how it's described in the post at all. Entangled particles are correlated across spacetime in a way that goes beyond "instantly knowing there is red in one envelope when blue is in the other".

    First of all, we don't know which color is in which envelope at the start. And revealing of one instantly determines the other physically regardless of whether it came up red or blue. So entanglement is "spooky action at a distance".

    This property doesn't constitute FTL communication though since we can't encode any information this way, we don't know what the particles will reveal when measured. We just know that measuring one will affect the other particle no matter how far apart they are physically.

    So I imagine Einstein meant both.

    • DebtDeflation 1054 days ago
      >Entanglement isn't like how it's described in the post at all. Entangled particles are correlated across spacetime in a way that goes beyond "instantly knowing there is red in one envelope when blue is in the other".

      When I watched this on YouTube yesterday, I knew that as soon as she used the sock example that someone would get salty about it without listening to the rest of what she says. She addresses this right towards the end:

      "Yes, that explains the case for the socks. But in quantum mechanics, that explanation does not work. If you think that really it was decided already which spin went into which direction when they were emitted, that will not create sufficiently strong correlations. It’s just incompatible with observations."

      >So I imagine Einstein meant both

      I think her view is that when Einstein said "spooky action at a distance" he was referring SPECIFICALLY to the instantaneous effect of measurement on the wave function. Entanglement, then, is just an example of this where we can increase the distance in the lab to experimentally confirm the implication.

    • zelphirkalt 1054 days ago
      You state this as a fact, but there seems to be an interpretation of entanglement in what you wrote, so I will ask a question:

      Sticking to the metapher, how do you know, that looking at the content of one envelope will _affect_ the content of the other envelope, in contrast to you _discovering_ the other envelope's content to be the other color?

      Asked differently: Why do you necessitate that viewing one envelope's content _affects_ the other, rather than merely allowing to conclude the other, because some law of the universe makes it happen to be always the other color?

      To me this seems really an interpretation, rather than a fact, that one can simply put there without dispute.

      • ahelwer 1054 days ago
        This is something you can experimentally test. First you should know that "red ball or blue ball" is a simplification. Staying with the metaphor, if you open one envelope differently you might get a purple ball and so know the other envelope contains a green ball. What we're actually dealing with are vectors on the complex unit sphere instead of ball colors, and opening the envelope is choosing two perpendicular axes along which to measure the state, causing probabilistic collapse to one or the other axis.

        Since you can choose which axes you measure, there are all sorts of tricky experiments you can do (called Bell tests) to test the theory that you're just discovering a pre-existing value vs. causing collapse to those values. The pre-existing value idea is called a local hidden variable, the existence of which has been conclusively experimentally disproven.

        I wrote a post on this you might find useful, explaining a Bell test called the CHSH game that computer scientists will find intuitive. All you need is your rememberance of the unit circle from high school path! https://ahelwer.ca/post/2018-12-07-chsh/

        • PoignardAzur 1054 days ago
          I've read your article, and it helped me understand the shape of the proof, but I still don't understand it.

          If no information is transmitted between the people observing the quantum particles, how is it that you can improve their "performance" at the CHSH game above 75%?

          Like, what is the fundamental unit of "information"/coordination the players are extracting from their particles/qbits?

          Actually, while I'm being confused, what even is a qbit? From the descriptions, it sounds like it's not just an information, where two qbits with the same vector representation are equivalent, and it's more like a physical concept. Is it a particle? Or, like, it it some sort of global variable of the universe that stores some information that is undetermined until you "collapse" the global variable?

          • ahelwer 1054 days ago
            A qbit is a mathematical model of some quantum phenomena that exhibits superposition & entanglement, like particle spin. The model seems to match the behavior rules of spin (and other quantum phenomena) as far as measurement and manipulation are concerned. Physicists generally avoid the question of whether nature really internally keeps track of the spin of a single particle with a complex 2-vector; instead we just say the model matches the experimental behavior as far as we have actually found, so the model is useful as a reasoning tool. Sorry for the sort of abstract answer.

            When we talk about information being communicated we're talking about someone choosing a bit that is then somehow reproduced with some fidelity somewhere else. Quantum entanglement doesn't allow such straightfoward possibilities. Instead it offers correlations that are stronger than would be possible with purely classical means. If this answer is frustrating because it doesn't match your intuition of what communication is - welcome to quantum mechanics! You just kind of have to learn the rules of the systems and get used to them. What you can do with entanglement (quite a lot!) is more important than what is "actually" happening under the hood.

        • zelphirkalt 1054 days ago
          Thanks for the link. It seems to involve quite a bit more than unit circles though ;)

          I have one note, which I hope you don't take the wrong way: I think it would be great, if you could put the slides as PDF on your page (viewable with browser's PDF.js for example), rather than on slideshare.). Then the slides you link to for understanding the braket notation would be as accessible as the original page you link to.

          • ahelwer 1054 days ago
            Oh sure I'll do that.
      • emrah 1054 days ago
        > Why do you necessitate that viewing one envelope's content _affects_ the other, rather than merely allowing to conclude the other, because some law of the universe makes it happen to be always the other color?

        Right, I'm not the one necessitating it. That's how quantum mechanics works for entangled particles. You can read more on this if you look up "EPR paradox".

        • zelphirkalt 1054 days ago
          I read most of https://en.wikipedia.org/wiki/EPR_paradox. To me it is still unclear, how all this proves, that it must be one particle or some observation of its attributes affecting the entangled other particle or observation of one of those attributes of it. The wiki article always says things like "X has shown in experiment Y, that Z.", but the details are what matters, if I am to really understand it. I guess without investing years for the mathematical background and physics background, it is hopeless, to try and understand it in an afternoon. My mind will always ask the next "but why?" question and I will be left with the feeling of not understanding it properly.

          Perhaps my idea how it might happen is too much of a metaphysical one? Here is what I think:

          I imagine entangled particles to be something like 2 sides of the same coin. Something we can observe in the known "world" (all the things we know about or can look at in a lab or experiments), but which has an actual character, somewhere/when, that we cannot grasp yet. Perhaps like an object in front of 2 light sources, throwing 2 shadows, and we observe the shadows, instead of the real object. But the shadows for us are all we currently can observe at our level of scientific understanding or with our technology.

          It is unclear to me, which definition of locality that object would be in or not in, reading that wiki page, so I am trying to not use that word.

          The problem is, how one could prove or disprove such a thing, because it is basically an "escape hatch", with which one can always say, that there is something, we do not yet see. From an "unknown object, whose shadow we can perceive" one also cannot draw many predictions, I think, so it might not be immediately useful to assume such. My guess is, that physicists stick to theories one can prove or disprove things about. Perhaps I will ask a knowledgeable coworker to explain to me, why my idea cannot be how it really is. Thanks for the pointers, I can also mention those in my questions.

          • ahelwer 1054 days ago
            There is absolutely no way to understand quantum mechanics without learning the math. All metaphors are faulty and will lead you to incorrect conclusions. Fortunately the advent of quantum computing has made the math quite easy - it's just linear algebra of the type one would pick up within the first week of an undergraduate course. I made a video aimed at computer scientists which you might find useful: https://www.youtube.com/watch?v=F_Riqjdh2oM
            • qbutt 1054 days ago
              This talk is excellent. Thank you for that.

              I think an issue with communicating the lack of information transmission is that the statement usually fails to acknowledge two aspects:

              1) We are unable to reliably force any one entangled particle to a specified state prior to measurement. ( You mention this more passively in a gp >"First of all, we don't know which color is in which envelope at the start." )

              2) After measuring we have now entangled our particle to our measurement. So, if after our measurement we change the state of our particle the previously entangled particle fails to reliably correlate.

              Information is transmitted; it's useful information that we are unable to communicate.

              If I give you instructions saying if the envelope contains this then do this or if it contains that then do that, you can use the information from the particle's entanglement to make that decision, but because of 2) you can only use it to make that decision reliably once, and because of 1) those two outcomes will have to be equally desirable for this form of communication; making it useless. Just roll a dice.

              • zelphirkalt 1054 days ago
                Couldn't one argue, that in some situations coordination via previously agreed upon criteria is worth more than which one of two decisions is made?

                If A and B agree upon, that B will know what A is doing in either case of discovery, then B needs only to look at their observation and will know what A's observation is (the opposite). Then B can know how A will act, because A decides what to do based on what they observe. Then B can act in a coordinated way. Even though it is only a roll of a die, it can still be helpful.

                For Sci-Fi example: Some alien spaceship fleet (bare with me!) wants to attack another civilization from 2 sides across wast distances. They split up waaay before that and each keeps an entangled particle. They agree upon which observation means what detail in the first fleet's attack plan. Depending on that, the second fleet will act differently on the other side. But the first fleet only decide which of the 2 strategies to use, when they are already split up, based on the observation of the entangled particle. The second fleet has to act accordingly, or the strategy will fail.

                Is this impossible? Or is this simply not seen as useful information in some specific definition?

                • qbutt 1053 days ago
                  > Or is this simply not seen as useful information in some specific definition?

                  You are adding another conditional but nested conditionals still all rely on the first.

                  We are getting 1 bit of data:0|1; from this entrangled pair.

                  Fleet1 has Strategy0 in case their measurement reads 0.

                  Strategy0 requires Fleet2 to act using Strategy2 for the highest probability of success.

                  Fleet1 also has Strategy1 in case their measurement reads 1.

                  If Strategy1 also requires Fleet2 to act using Strategy2 for highest probability of success then Fleet1's measurement is useless as Fleet2 can just use Strategy2 in both events.

                  If instead Strategy1 requires a new Strategy3, and even though you could communicate which of these strategies to choose faster than we know how to communicate that electromagnetically, our 1 bit of information is still useless because you would still have to prepare for both eventualities.

                  Both pairs: Strategies02 and Strategies13; would equally have to have the same probability of success because if one is favoured over another then using this technique to decide your fleet movements could result in your allowing the possibility for the measurement to order you in a strategy you know is less likely to be successful.

                  This is what I was trying to get at by using the word 'desirable'.

                  Fleet2 may appear to be waiting on Fleet1 for orders, but in fact both fleets are waiting on their orders from the first measurement.

                • ahelwer 1053 days ago
                  You can accomplish the same thing with a PRNG initialized to the same seed.
      • simcop2387 1054 days ago
        Part of this is because of tests like the ones involved with Bells theorem. Minute physics does a better job explaining it then I can. https://m.youtube.com/watch?v=zcqZHYo7ONs&vl=en
    • russdill 1054 days ago
      The envelope description doesn't allow the core of the issue to be expressed. Just allowing for a single measurement of a single property is completely uninteresting.

      Being able to measure spin at arbitrary angles is the best description, but I don't know any good analogies to it, especially ones involving envelopes.

      • k__ 1054 days ago
        Yes.

        The envelope example always seem obvious.

        But when particle physicists talk about the problem, it sounds like it isn't that obvious.

        • eigenket 1054 days ago
          I don't think many physicists really think of entanglement as a problem now. Its kinda weird, but its also completely consistent, doesn't lead to any wild paradoxes and turns out to be very useful in a variety of scenarios.

          I think there are people who are concerned with the foundations of quantum mechanics more generally, and they might be concerned with entanglement as an emergent feature of the weirdness in the foundations they don't like, but most of the work I see these days is on interesting ways to use entanglement.

          • tarsinge 1054 days ago
            That’s a good point and a lot of incomprehensions come from that, in classical physics you can develop an intuitive understanding of the how with the foundation, but in QM it’s distinct. Physicist usually only answer on the how because that’s where the work is (“entanglement works like this, here is the model”) but there is interest in the foundation too, even if it’s a bit philosophical.
    • headmelted 1054 days ago
      I was trying to think of a way to explain this to a family member a while back who had a background in mechanical engineering.

      The best example I could think of to analogise it were meshed gears.

      There is no messaging happening between them. When one rotates, the other rotates in the opposite direction at the same exact time because they are interlinked. They are two parts of one connected set.

      That entangled particles are interlinked in this way shouldn’t be surprising. Having what to us seem to be arbitrary distances between the particles could simply mean that they’re interlinked outside of the three spatial dimensions we’re used to perceiving the universe in - but this last part is just my own understanding and I’m quite happy for someone who knows better to tell me if (and why) it’s wrong.

      • TeMPOraL 1054 days ago
        > There is no messaging happening between [meshed gears].

        Isn't there? Atoms rub against atoms, electric fields interact and transfer forces. The gears communicate by touching, but it happens so fast that it looks instantaneous. This can be experimentally verified by weakening one of the teeth of one gear, and driving some load with that gear set - eventually, the weakened tooth will break off.

        • rleigh 1054 days ago
          There is messaging through physical contact. It occurs at the speed of sound as opposed to the speed of light.
          • redis_mlc 1054 days ago
            No, physical ("intermolecular") forces are electromagnetic, and the force between atoms is transmitted at the speed of light.

            Vibration in gears travels at the speed of sound for the material that the gears are made of, but that is not what separates the gears physically from merging at an atomic level.

        • headmelted 1054 days ago
          If there’s a delay, it’s not perceptible to an onlooker, which is probably good enough for an analogy in this case I hope?

          It would need to be, given that the gears aren’t going to reverse time and break apart into light beams just because I looked at them.

          Quantum physics doesn’t fit very well into simple examples.

    • IncRnd 1054 days ago
      Well, Einstein prefaced that exact portion by saying, "To see what I mean, forget all about quantum mechanics for a moment." You misunderstood what he wrote. He did not describe entanglement when discussing the red and blue pair of socks. He was creating an analogy to build his argument that he later transferred to "spooky action at a distance".
  • choeger 1055 days ago
    Wait a second. Does that mean the nonlocality was introduced in QM just because distance simply didn't occur in the model? There was no prior theory that predicted you could ignore distance? Later experiments then confirmed that yes, distance does indeed not matter for the wave function collapse? That's ... incredibly unlikely, right? I mean, one simply wouldn't expect a model to simplify things and then learn that the simplification reflects reality.
    • JumpCrisscross 1055 days ago
      > one simply wouldn't expect a model to simplify things and then learn that the simplification reflects reality

      Isn't this a refrain in the history of science? Observations inspiring models that illuminate a deeper, truer representation of reality?

      • georgeecollins 1054 days ago
        It's a conceptual fallacy that the underlying mechanics of a phenomena are simpler than the observed data. It's appealing to our way of thinking and aesthetics to believe that the rules that govern are simple. But it may or may not be the case.
        • throwamon 1054 days ago
          Phenomena is the plural of phenomenon.
    • dandanua 1054 days ago
      Standard QM indeed doesn't say anything about distances. But this doesn't automatically imply that QM is non-local. QM is not a theory of everything, it's just some subset of rules that we know about the Universe. Though, it's hard to unite QM with our classical understanding of locality.
    • GistNoesis 1054 days ago
      The non-locality of QM was introduced because of the Fourier transform which is non-local. Mainly because it's the easy way to calculate by hand when you have no computer in the 1920s.

      [Tinfoil hat warning :] Since then it's a self-reinforcing confirmation bias to avoid the obvious locality of the real world mainly because of geopolitical reasons in the 1960s. /[Tinfoil hat warning :]

      • DSingularity 1054 days ago
        Your second statement, what do you mean? I’m not sure I’m understanding what you wrote.
        • GistNoesis 1054 days ago
          Some people are flat-earther, I'm a Bell-theorem denier.
          • eigenket 1054 days ago
            What do you think is happening in the experiments in 2015 in Delft, Vienna and Boulder and Rosenfeld et al. from 2017 (or any other Bell test you like)?
            • wizeman 1054 days ago
              I'm not a physicist so I could be wrong, but I think Bell's theorem assumes there is only one outcome for each measurement.

              If there can be many outcomes, like the Many Worlds Interpretation suggests, then I think there is the possibility that the "spooky action at a distance" is a result of purely local interactions.

              • ahelwer 1054 days ago
                Unfortunately you are indeed wrong. Bell's theorem works no matter the outcome of the measurement. We just see that some outcomes are more likely to occur than others. People in other universes would see different outcomes, but the overall distribution would be the same. I guess there could be some hideously improbable universe where every Bell test experiment ever performed lead to results not exceeding classical correlations, but that sure as hell isn't the universe we live in.
                • wizeman 1054 days ago
                  While what you say makes sense superficially, I think you are not considering the possibility that the worlds interact with each other.

                  Notice that in the double-slit experiment, the wave function at different places interacts with itself so as to lower the probability of a photon or electron arriving at a place where it was likelier to arrive with just one slit open.

                  Informally speaking, that could be seen as an example of different "worlds" (i.e. different places where the photon could be observed to exist) interacting with each other locally so as to effectively lower the probability of observing the photon arriving at some place where it would be likelier to arrive had the different worlds not have interacted with each other. Taking the analogy a bit too far, this seems like if two different worlds interfere with each other so as to cancel each other out. Again, the photon/electron that was likely to arrive at some place, will not likely arrive anymore, and this is purely due to the the wave function interfering with itself.

                  Wouldn't that mean that there is the possibility that the observed correlations in Bell's theorem could be the result of this phenomenon and not necessarily mean that there is non-local behavior?

                  That is, couldn't the observed correlations be the result of the different worlds interfering with each other locally, e.g. when the light cones of the different measurements intersect each other, so that some worlds cancel each other out and we don't get to observe them, leading to the correlations that we do indeed observe?

                  Bell's theorem, as far as I understand, only takes into account what we can observe, but could be falsified if there are "worlds" that we cannot observe, i.e. that are temporally created but stop existing when they cancel each other out by local interactions.

                  Although, admittedly, I could be completely wrong because I don't understand enough of this to be sure of what I'm saying :)

            • GistNoesis 1054 days ago
              Statistics done by physicists based upon a wrong model is the root cause.

              The exact flaw changes from paper to paper. They are plenty of ways to get something wrong when you are not doing statistics carefully.

              It often comes down to the precise definition of what the experimenter for what a measurement is.

              One usual mishap is "involuntary post-sampling", by wrongfully ignoring some blips or discarding particle that don't fit the mold.

              The more complex the experiment trying to close a loop-hole the more probable the bug slips in.

              • eigenket 1054 days ago
                To clarify, have you actually seen any of these errors in the papers, or are you just hoping/believing/assuming they are there?
                • GistNoesis 1054 days ago
                  >Ah yeah, I went to skim the Pitowsky reference. I forgot that people call (what I was taught as) Bonferroni inequalities Boole's inequalities.

                  Bonferroni inequalities are also on the wikipedia page of the Boole's inequality https://en.wikipedia.org/wiki/Boole%27s_inequality

                  >I'm still not sure what probability distribution (or measure) you want to be defined in that paper. Experimentalists generally don't assume any probability distribution in advance, I am not sure why they would need to in this case.

                  That's the thing, if you try to model it correctly you realize that what you were doing was not the thing to do, and that there is just nothing to see here, apart from yourself asking the wrong question.

                  Take a zone of space (the detector volume), and define the discrete event "blip" as 1 if somewhere inside the detector volume at instant t0 the field value is greater than 1, or zero elsewhere.

                  That's all you need : this mean you just expressed an event as a conditional probability. Therefore when you have multiple such events, you are expected to encounter some Boole's/Bell type inegalities (provided that there is information conservation in your model).

                  Experimentalists are doing circular reasoning. By not assuming an underlying model, they don't realize that the answer they get is due to implicitly defining events as conditional probabilities which don't span a consistent Filtration_(probability_theory), then they try to ask the wrong question, getting an answer which confirm their biases.

                  QM is a not a model of the world, but a model of what we can observe from the world based on what we observe from the world.

                  Bell's inequalities tell nothing about how the world works but only tell you that your way of observing the world is wrong.

                  • eigenket 1054 days ago
                    You're missing something I think (or we're talking past each other) because I can write down probability distributions which explicitly break Bell's inequalities, whereas I can't write down probability distributions which break Boole's inequalities.

                    This is because Boole-type inequalities are derived just by assuming the basic rules of probability distributions, while Bell-type inequalities are derived by assuming some local hidden variable model (or similar locality assumption).

                    Consider the Popescu-Rorhlich (PR) box probability distribution where P(a,b | x,y) equals 1/2 if x*y = a xor b and zero otherwise.

                    This probability distribution explicitly breaks Bell's inequality, but it does not break any of the Bonferroni / Boole inequalities. This is precisely because although it obeys all of the axioms of probability distributions (which are the assumptions required for Bonferroni / Boole) it does not obey the extra axioms of locality that Bell's inequality requires.

                    I'm still not sure what your issue is with the experimental papers. You acknowledge I think that the real detectors they have in the lab actually do "blip", right? And you can take these observed "blips" and use them to compute a CHSH value?

                    • GistNoesis 1054 days ago
                      Violations of Boole's inequalities only happen when you do something bad with probabilities, you just need to be a little creative.

                      Experimenters observe CHSH > 2 because they are likely doing something bad with probabilities : the same bad things with probabilities you need to do to violate Boole's inequalities.

                      The measurement process of the pairs by the detectors implicitly do some additional filtering on the hidden_state. Typically when you define what your detection threshold for registering a blip is, or what you consider is a coincident pair by filtering based upon the time of measurement.

                      When you try to apply Boole's inequality to something like

                      P( anything | measurement1, measurement2, 0< hidden_state < 1),

                      P( anything | measurement1, 0< hidden_state < 1),

                      P( anything | measurement2, 0< hidden_state < 1)

                      The chain rule doesn't apply the same way. Because hidden_state is not a specific value but a range of values : The joint probability is not factoring as a product of conditionals anymore.

                      You have some incompatible sets of events and you can observe CHSH type violation for your Boole's inequalities.

                      • eigenket 1054 days ago
                        > Violations of Boole's inequalities only happen when you do something bad with probabilities.

                        I sort of agree with this, if by "something bad" you mean not obeying the axioms of probabilities.

                        > Experimenters observe CHSH > 2 because they are likely doing something bad with probabilities.

                        I disagree with this completely. Note that the probability distribution I gave just now (which is an honest probability distribution) violates CHSH. It obtains a CHSH score of 4, indeed. I am not doing anything "bad" with my probabilities to obtain this score, I'm just providing a probability distribution that does not obey Bell's additional locality assumption.

                        I ask you specifically, what is "bad" about the distribution

                        P(a,b| x,y) = 1/2 if ab = x xor y and zero otherwise

                        which makes it obtain a CHSH score of 4?

                        • GistNoesis 1053 days ago
                          I think you are missing the point I'm trying to make.

                          You can violate CHSH by giving a non-local distribution (That's what happening with the distribution you provide).

                          But there is more than one way to skin a cat.

                          You can also violate them using the mathematical trick as I suggested above. This trick to replicate the result seems a lot more plausible than non-locality.

                          > not obeying the axioms of probabilities.

                          That's exactly what I blame the experimenters for.

                          By ignoring that they are conditioning on the hidden variable, they are treating what is a "conditional probability" exactly the same way as it if was a "standard probability". Which is a big no-no from the math standpoint.

                          • eigenket 1051 days ago
                            Could you give an example of exactly what you think they're doing wrong? I've read a few of these papers and I've not seen any such issues.
                • GistNoesis 1054 days ago
                  Replying here due to max comment depth tree, to explain in more detail.

                  About the link between boole's and bell's inequalities : There are tons of paper : A quick search for example https://arxiv.org/pdf/quant-ph/0406004.pdf (I just skimmed it for a few seconds but it's probably talking about the right thing)

                  In the paper you suggest https://arxiv.org/abs/1508.05949, there is nothing which looks like a probability definition.

                  Assuming the world is using continuous fields, defining the probability for what register as a discrete event like a "blip" must be done carefully, it's not a trivial thing.

                  Otherwise when you are taking joint probabilities and condition probability, it just tells you that those are questions you can't ask.

                  • eigenket 1054 days ago
                    Ah yeah, I went to skim the Pitowsky reference. I forgot that people call (what I was taught as) Bonferroni inequalities Boole's inequalities.

                    I'm still not sure what probability distribution (or measure) you want to be defined in that paper. Experimentalists generally don't assume any probability distribution in advance, I am not sure why they would need to in this case.

                • GistNoesis 1054 days ago
                  Those paper don't have rigorous mathematical definition of what "probability measure" they use.

                  From a mathematician perspective, the error is "incompatible probability measure" for events, that's why probability laws don't apply.

                  Bell's theorem is exactly a well known mathematical theorem "Boole's inequality" which basically tells you, you are doing the math wrong when defining your measure.

                  • eigenket 1054 days ago
                    If by Boole's inequality you mean the thing that I know as the union bound, then it is far from obvious to me that it is exactly the same as Bell's theorem. In fact they seem to me to be completely different. Boole's inequality also does not tell you "you're doing the math wrong when defining your measure", it tells you that the probability that either A or B happens is not greater than the sum of the probability of A happening and the probability of B happening.

                    I'm not sure what you mean in the first part of your comment, could you indicate specifically what probability measure is defined incorrectly (or not defined) in (e.g.) this paper?

                    https://arxiv.org/abs/1508.05949

          • Der_Einzige 1054 days ago
            I always knew bells theorem would instill extreme amounts of butthurt in a certain group of people if they actually knew what it meant.

            Everytime I listen to somebody raffle off about determinism or the law of identity I remind them that bells theorem (there's real actual randomness) and it's corresponding no-cloning theorems proved objectively that the universe has real randomness, and that truly "cloning" an object (e.g. A=A) is literally impossible.

            Sorry that reality doesn't work the way that you want homie. Bells theorem is my favorite theorem because of this level of butthurt that it instills...

          • myWindoonn 1053 days ago
            To take a different path from the existing subthread: Do you also doubt the Kochen-Specker lemma? Quantum contextuality cannot be dodged just by denying Bell's version; you also have to show that Kochen and Specker got their linear algebra wrong.
  • ziofill 1054 days ago
    Quantum physicist here. If two particles are entangled, the very value of the entangled properties does not exist locally (if it did it would be a hidden variable, with crazy consequences, but there are even cases where pre-existing values are forbidden algebraically, like for the GHZ state). Entanglement only gives us a promise of joint behaviour like "if the particles are in the Bell singlet state and we measure them with the same measurement settings, the results are going to be opposite" (notice that the promise is only about the values being opposite, not about what they will be locally). I believe that what Einstein meant by spooky action is exactly about this: that a particle seems to know which settings have been used for the measurement of the other, because if the settings are different then the promise can be broken.
    • jgrowl 1054 days ago
      I'll be that guy and say for completeness that while local hidden variables are ruled out, global variables are not via something like Bohmian mechanics.
      • ziofill 1054 days ago
        True. However each alternative comes with a lot of baggage, and there is no easy way out: 1) it could be that locality doesn't hold (as in Bohmian mechanics), but then causality takes a big hit because we must accept that it can also work backward in time. 2) it could be that realism doesn't hold (which is the more commonly accepted evil) i.e. that there is no cause for the particular value emerging from the measurement of an entangled property. True randomness. (And yet, it satisfies the required correlations).

        After years of losing sleep over it, I like to think that entangled properties are somehow 'part of the same whole', which is independent of space-time, and that space-time is a construct that emerges from the particular way decoherence seems to choose position and momentum as preferred bases.

    • pmoriarty 1054 days ago
      "the very value of the entangled properties does not exist locally"

      Does physics have a definition for existence?

      • ziofill 1054 days ago
        Yes! (and good question) In this case we use Einstein's definition of local hidden variable (which he called "element of reality"):

        "If, without in any way disturbing a system, we can predict with certainty (i.e., with probability equal to unity) the value of a physical quantity, then there exists an element of reality corresponding to that quantity."

        • pmoriarty 1054 days ago
          If I may infer that this statement has been interpreted as not a conditional as it is stated verbatim (in the form A -> C where A is the antecedent and C is the consequent), but instead is a biconditional (in the form A <-> C), then it seems to be stating that to exist is to be certainly predictable.

          That's a curious definition of existence.

          And now that I think about it a bit more, it seems like it's not really talking about what existence is, but rather the consequences of existence and the consequences of certain prediction. That's not quite the same as a definition of existence, is it?

          • ziofill 1054 days ago
            What the statement says is that if a property is certainly predictable, then it exists. It doesn't say that all existing things are certainly predictable.
            • pmoriarty 1054 days ago
              But that doesn't say what it means by existence.

              It leaves that part unsaid.

              To go a bit deeper in to this, it's not just saying certainly predictable properties are certainly predictable. It wants to say something else about them: that they exist. What does it mean by that?

              The statement as it stands is more about certainly predictable properties than it is about existence itself.

              • ziofill 1054 days ago
                I think I understand your point: we don't define existence per se. But here we only need to define the existence of the value of a property. And we define it by saying that if the we can predict the value before measuring it, then the value "exists". Take it as an axiom.

                When I said "the very value of the entangled properties does not exist locally" in my first post I meant that there is no way to predict the value (or of consistently assigning a value to the property) before performing a measurement.

      • javajosh 1054 days ago
        The EPR paradox is about precisely this: https://en.wikipedia.org/wiki/EPR_paradox
  • paraknight 1055 days ago
    What did she mean by the quantum correlations being stronger than the non-quantum ones? She lost me there
    • tsimionescu 1055 days ago
      Bell's theorem has proven that quantum entangled particles do not behave exactly like the socks. To understand it, let's imagine the sock experiment, but with 2 differences: one of the socks is red and fluffy, the other is blue and coarse; and you can't see exactly how red and how fluffy the sock you got is, you can only measure a combination of the two.

      So, you open your envelope and set your device to measure if the sock is 0.2 red and 0.8 soft (so, either blue or soft), and you get True. If the person who got the other sock set their device to 0.8 red and 0.2 soft, the classical probability that they get True would be 0.5 I believe.

      But if you perform this experiment with quantum entangled properties, you would see a greater value. Further, Bell has shown that one of 3 things must hold for quantum socks:

      - either they are neither blue/red nor coarse/fluffy until the measurement; after the measurement, your sock gets some value, and the other sock implicitly gets the opposite value, but there is no real "state change" since they didn't have any state to begin with (we reject realism)

      - they "tell" each other instantly in what state they were found (we reject locality)

      - the fact that you were always going to receive the red sock also caused you to choose 0.2 red and 0.8 soft, so what may look like a coincidence is in fact pre-determined (superdeterminism)

      • monkeydreams 1054 days ago
        Do you mind if I clarify something about this that I have never understood?

        Taking your example of the socks, with their varying and inverse properties, why do we need "spooky action at a distance" to describe the fact that one can test the properties of their local sock and therefore know the properties of the distant, entangled sock? If we now that the total value of each property will equal 1 (as in your case) then measuring the local sock will tell us the values of the distant sock.

        I am not sure what I am missing, only that I am missing something here.

        • eigenket 1054 days ago
          So you can describe the set of possible joint probability distributions for the socks, and you can describe the set of possible probability distributions you can get out of quantum mechanics in a "Bell-type experiment". It turns out the set of probability distributions from quantum mechanics is strictly larger.

          Fundamentally the answer is that you can't break a Bell inequality with the socks, but you can with some entangled particles.

          The socks analogy kinda fails to capture the interesting stuff here, and I'm not sure if any classical analogy really works, but I'm gonna try to summarise the setup.

          You have two boxes (imagine they're in different rooms, being operated by different people), call them A and B. Each Box allows you to input a single bit, we'll call A's bit x and B's bit y. Each box outputs a single bit in response to the bit you give it, call the bit output by A a, and the bit output by B b. You can then form a probability distribution

          P(a,b | x, y)

          The joint probability distribution of a and b conditioned on the inputs x and y. Quantum mechanics predicts a particular form for this distribution, and in particular it does not depend where the boxes are, and when (in which order) the bits are input. Based on special relativity therefore we would expect that the distribution of a conditioned on x should be independent of the distribution of b conditioned on y. In maths this means that

          P(a,b | x, y) = sum_i P(i) P(a|x, i) P(b|y,i)

          Basically we're allowing the output bits to depend on some shared randomness P(i), but we're not allowing a to depend on y and we're not allowing b to depend on x. This is reasonable because we have the freedom to put x in a after we receive b from y, or vice-versa, or have them separated by arbitrary amounts of space, or whatever we like.

          It turns out that the probability distribution P(a,b|x,y) that quantum mechanics predicts is not of this form. The probability distributions we observe in reality, when we actually do the experiment are also not of this form. This seems to mean that somehow, even if they're separated by arbitrary amounts of space, or done in whatever order you like, or whatever the output bit a does somehow depend on what input you give box B and vice-versa.

      • throwaway316943 1054 days ago
        Are the misses caused by the measurement equipment or do you really sometimes get true on both ends? If so, why?
        • tsimionescu 1054 days ago
          The sock metaphor is already stretched a little thin, and this problem doesn't happen with classical quantities/objects as far as we know.

          Basically what is happening is in some ways similar to what gives quantum computing its extra power compared to classical computers: as far as we can tell, quantities at the microscopic level are in fact complex numbers, and so combine in ways that give different values from the real-valued quantities we are used to from the classical world.

      • goldenkey 1055 days ago
        Or.... consciousness collapses the wave function and solipsism is the nature of reality. Observables are just very large unevaluated expressions that collapse via lambda reduction ala becoming a tautology. As you gain more information, more alpha and beta reductions become possible, and expressions simplify into static values instead of multivariate/dependent expression trees. The universe is a first person JIT experience.

        Search for Wigner's friend. Relational quantum mechanics is the only logical and consistent epistemology for what all this means.

        Effectively, nothing ever changes, just extra information allows for simplification of observables backed by large expressions.

        The universe is inherently lazy, and it's like LISP. PG come at me!

        • tsimionescu 1055 days ago
          Well, if you go for pure solipsism, that is, your own consciousness is the only thing which exists in the universe, that can explain any phenomenon you want.

          But objective collapse due to anyone's consciousness does not solve anything - my observation of this particle still causes infinite speed effects on your particle, so the universe is anyway non-local.

          • goldenkey 1054 days ago
            It's only solipsism in the sense that my wavefunctions do not have to agree with yours.

            There is a correlation between us, still. So it isn't like our wavefunctions are going to be radically different. We both are entangled since the Big Bang and even after that many times over as shared information exchanges between us.

            https://en.wikipedia.org/wiki/Wigner%27s_friend

            • tsimionescu 1054 days ago
              The article you cite provides various alternatives to this world view. It even contains the fun observation that the experiment has actually been performed, but with photons instead of observers - showing just how important consciousness really is in physical theories.

              In general, I believe the most obvious interpretation of consciousness is simply as a computational process, so looking at consciousness as a key to QM is incompatible with my beliefs.

              Furthermore, given the classical-ness of the world we see everyday, I have huge problems believing theories like MWI or these consciousness-based ones that claim the world is entirely quantum everywhere and it's just some kind of accident that it seems so classical to us (contrast to relativity which has a very satisfying answer for why we don't notice the difference from classical mechanics).

              • goldenkey 1054 days ago
                > I have huge problems believing

                The universe is indifferent to your persuasions. The double slit experiment has been successfully performed on large molecules of 810 atoms. Quantum mechanics is fundamentally about knowledge and information after interactions. You seem to be a physics guy but something tells me you aren't aware that every single interaction causes entanglement. Entanglement is simply a constraint on interaction products, given the input particles. It is just easier to see its effects on small objects since they are less likely to be perturbed.

                https://arxiv.org/abs/1310.8343

                • tsimionescu 1054 days ago
                  What you're describing is one self-consistent possibility. But there are other self-consistent possibilities, and all have an equally-good claim to being 'how the universe really works'. Some people are really adamant that MWI is the one. Some are adamant that Copenhagen is the only sound one. It may even turn out that Pilot wave theory is true and none of this quantum weirdness is really all that different from the classical world.

                  I think we just don't know enough yet to choose any one interpretation. I hope that at some point we'll find a way to address the measurement problem - either finding that the classical world also behaves quantically or that there is some clear limit between them or that the quantum world behaves classically (least likely).

                  But until then, we're all just guessing.

                  Edit: 810 atoms is great, just ~6.023*10^21 more to go and we'll have something that is truly in our range of advertising.

        • kergonath 1054 days ago
          > consciousness collapses the wave function and solipsism is the nature of reality

          This would assume that somehow the atoms in our brain work differently than the atoms anywhere else. There is nothing special about consciousness.

          Honestly, the rest of your post is a bunch of mumbo jumbo, it’s difficult to see what you are trying to say.

          • goldenkey 1054 days ago
            > This would assume that somehow the atoms in our brain work differently than the atoms anywhere else. There is nothing special about consciousness.

            You assume wrongly. The implication is that all wavefunctions have a corresponding consciousness. Since the universe is a nested set of wavefunctions, it's essentially all consciousness.

            There is nothing special about consciousness, it's the unified field, it's all there actually is.

            Honestly, the rigidity of your thinking holds you back from seeing any deeper truth.

            https://youtu.be/LPhgDfT4Zpc

            • kergonath 1054 days ago
              I have some sympathy for your thinking, but it is not borne of any understanding of the physical concepts you mention.

              > You assume wrongly. The implication is that all wavefunctions have a corresponding consciousness. Since the universe is a nested set of wavefunctions, it's essentially all consciousness.

              First, a wavefunction is a way of making sense of (some of) the concepts of quantum mechanics. It has a specific meaning, and scientists to this day argue about whether it is necessary or actually a thing.

              When it is used, it is a scalar field that is used to describe the probability of some experimental result, i.e. a mathematical tool. It has no link whatsoever with consciousness, which is the result of electric signals going up and down a group of neurons bunched together.

              You sound like you want to redefine “consciousness” as “existence”. In that case there is some tautological beauty in what you say, but it is not as enlightening as you seem to think it is.

              > There is nothing special about consciousness, it's the unified field, it's all there actually is.

              What the hell does that even mean? You use scientific-sounding words without seeming to grasp their meaning.

              I mean, surrealism is fun, and metaphysics can be interesting, but don’t confuse them with a deep understanding of physical truths.

              > https://youtu.be/LPhgDfT4Zpc

              I need to point out that you are citing a crackpot and that there was absolutely nothing scientific in what I have seen in this video. But yeah, it probably is my mental rigidity talking.

              • goldenkey 1053 days ago
                There is no difference between physicality and the information that backs it. Only a philistine would try to claim that.

                https://en.wikipedia.org/wiki/Universal_wavefunction

                > Since the universal validity of the state function description is asserted, one can regard the state functions themselves as the fundamental entities, and one can even consider the state function of the entire universe. In this sense this theory can be called the theory of the "universal wave function," since all of physics is presumed to follow from this function alone.[7]

                You appeal to a trivially naive interpretation of what the wavefunction represents, yet claim that I or Hagelin are the crackpots. Don't you know that all modern science requires unobservables? Yet you want to constrain the wavefunction to being a mathematical tool for measurement only because that was its intended purpose and conceptual creation. Narrow mindset.

                https://en.wikipedia.org/wiki/Unobservable

                > You use scientific-sounding words without seeming to grasp their meaning.

                You think you understand quantum physics when no one does. Furthermore, you don't even understand wavefunctions except by their textbook definition...

                • tsimionescu 1053 days ago
                  > There is no difference between physicality and the information that backs it. Only a philistine would try to claim that.

                  The map is not the territory. A ball is a ball and its motion is its motion, while Newton's equations of motion for that ball are something else.

                  While it is possible that the MWI is right and the state of the entire universe could be described by a single, universal wavefunction, that wouldn't mean that the universe is that wavefunction, anymore than Maxwell's equations are the electromagnetic field.

                  Now, there is an interesting question that indeed no one knows the answer to for now: is the wavefunction a direct description of a physical object, or is it just a statistical tool for predicting possible positions of the real objects (waves and particles existing in some duality)?

                  Also, you shouldn't forget that none of the existing interpretations of QM actually solve the measurement problem, not even MWI. Here is Sabine Hossenfelder's explanation of this [0]:

                  > In the many worlds interpretation, if you set up a detector for a measurement, then the detector will also split into several universes. Therefore, if you just ask “what will the detector measure”, then the answer is “The detector will measure anything that’s possible with probability 1.”

                  > This, of course, is not what we observe. We observe only one measurement outcome. The many worlds people explain this as follows. Of course you are not supposed to calculate the probability for each branch of the detector. Because when we say detector, we don’t mean all detector branches together. You should only evaluate the probability relative to the detector in one specific branch at a time.

                  > That sounds reasonable. Indeed, it is reasonable. It is just as reasonable as the measurement postulate. In fact, it is logically entirely equivalent to the measurement postulate. The measurement postulate says: Update probability at measurement to 100%. The detector definition in many worlds says: The “Detector” is by definition only the thing in one branch. Now evaluate probabilities relative to this, which gives you 100% in each branch. Same thing.

                  > And because it’s the same thing you already know that you cannot derive this detector definition from the Schrödinger equation. It’s not possible. What the many worlds people are now trying instead is to derive this postulate from rational choice theory. But of course that brings back in macroscopic terms, like actors who make decisions and so on. In other words, this reference to knowledge is equally in conflict with reductionism as is the Copenhagen interpretation.

                  [0] http://backreaction.blogspot.com/2019/09/the-trouble-with-ma...

                  • goldenkey 1053 days ago
                    You are equating equations, Maxwell's equations, with the electromagnetic field. This is a folly akin to equating an image created in Photoshop _with_ Photoshop itself.

                    The wavefunction of the universe is the state of it, and all conscious observers _ever_ do, is measure states. Interaction is all there is, just relationships. The state is isomorphic to the reality. The only difference is the reality has some process running on it, to advance the arrow of time.

                    This is not confusing the map for the territory. Every physicist you ask will tell you that a particle is the matrix it is represented by. Now, we may not have those matrixes fully complete or accurate, but the backing of reality IS information. And no, information and numbers do not have to have a reality onto themselves - they are just the medium of comparison, otherwise any description is impossible. All descriptions that humans use are metaphors based on past experience.

            • ShamelessC 1054 days ago
              There are many theories in this space to consider and we can't feasibly consider them all. Some amount of rigidity is of course required to even begin probing into something new.

              Since humans have finite lifespans, we all develop some heuristics (probably heavily biased ones) to serve as filters for what is efficient to consider.

              As such, one of my (again, heavily biased) filters are any such claims that meditation could have an outsized impact on the level of deviance/crime in the Washington DC.

              " We present the published results of a National Demonstration Project-in which 4,000 advanced meditators markedly reduced violent crime in Washington, DC."

              I simply don't have time to consider this as my personal biases are telling me to ignore it and that it is bad science.

              Furthermore, the researcher here very frequently makes use of highly theoretical and even unproven concepts such as string theory and a particular variant of the multiple universes idea which has also not been proven. Both are presented misleadingly as concrete fact rather than assumptions made. I dislike being misled and distrust the researcher moreso because of that.

              I assume your biases are quite different from mine. I assume some level of experimentation with hallucinogenic drugs may be involved? No judgement - I've just seen lots of similar appeals to philosophical interpretations of physics from other LSD users I know.

            • tsimionescu 1053 days ago
              > There is nothing special about consciousness, it's the unified field, it's all there actually is.

              That just begs the question of what consciousness actually is. You can define consciousness as being the wave function, but then you still have to actually check if what 'consciousness' colloquially means has any resemblance to that.

              For example, when you tell me something is conscious, I expect that I can teach that thing certain things, like how to sing or how to fetch. Since a photon has a wavefunction, by your definition I would expect to be able to teach a photon to do things.

              Similarly, I expect conscious things to be able to interact with the world and have feelings about it. Do you beleive a photon admires an electron more than a proton, or anything similar?

              Until you can show any experiments where anything that is known to be described by a wavefunction is shown to have all of the properties we associate with the word 'consciousness' in everyday usage, your assertion that all wavefunctions have a corresponding 'consciousness' doesn't really mean anything. I could just as easily say that all wavefunctions have an associated music, and that music is all there actually is. It's a definition, so it's not wrong, but it's not illuminating.

              • goldenkey 1053 days ago
                > For example, when you tell me something is conscious, I expect that I can teach that thing certain things, like how to sing or how to fetch. Since a photon has a wavefunction, by your definition I would expect to be able to teach a photon to do things.

                The possible futures for a photon, given the laws of physics, are much simpler than for you or I. Its wavefunction is known. The usual (classical) Maxwell field is the quantum wave function for a single photon.

                > Similarly, I expect conscious things to be able to interact with the world and have feelings about it. Do you beleive a photon admires an electron more than a proton, or anything similar?

                You are confusing consciousness with reactiveness. Deterministic rules can cause reactions. Your refrigerator for example, reacts to temperature changes. It however, is not exercising free will or any kind of quantum randomness to do so.

                > Until you can show any experiments where anything that is known to be described by a wavefunction is shown to have all of the properties we associate with the word 'consciousness' in everyday usage, your assertion that all wavefunctions have a corresponding 'consciousness' doesn't really mean anything. I could just as easily say that all wavefunctions have an associated music, and that music is all there actually is. It's a definition, so it's not wrong, but it's not illuminating.

                There is no rule-based construct for random number generation except for PRNG[1]. It follows that, if true randomness is exhibited by the universe, one that is not generated by rules, then it must be the leftover ingredient. Consciousness is the only leftover ingredient. Everything else we know about physics is deterministic and rule-based. Now, you know very well, things like this are likely not able to be proven given the hard consciousness problem and the other proximal issues. So I'm making an argument. Some things in life are arguments, not proofs. Like political views. Get used to it instead of being cliche and resorting to Newton's Flaming Laser Sword.

                [1] https://en.wikipedia.org/wiki/Pseudorandom_number_generator

                • tsimionescu 1053 days ago
                  Consciousness has a specific meaning, even though we can't formally define it so far. That includes reactiveness, feelings, curiosity and other traits, which elementary particles obviously do not possess. So, it is obvious that elementary particles are not conscious.

                  Now, you could say that consciousness is a quantity that humans possess a lot of, while photons possess a tiny little amount of, like the Integrated Information Theory people claim. But there is no proof (or even indication, in my opinion) that the IIT 'consciousness' actually captures what we mean by human consciousness.

                  > It follows that, if true randomness is exhibited by the universe, one that is not generated by rules, then it must be the leftover ingredient. Consciousness is the only leftover ingredient. Everything else we know about physics is deterministic and rule-based.

                  Randomness is not necessarily a part of QM. For example, in MWI, there is no randomness in the universe - it is perfectly deterministic, as every state of the wave function is realized. In pilot-wave theory, the only randomness is classical - errors in knowing the initial state with absolute precision.

                  Even if CI or other fundamental randomness interpretations of QM are correct, there is no obvious connection between consciousness and randomness. There are other unknowns in the universe - you could just as easily say that dark matter is consciousness or dark matter is the source of randomness, or dark energy etc.

                  And only since you briefly mentioned it, free will (as in conscious choice affecting the next state of the universe) is incompatible with QM, as QM is a linear theory, where the next state is uniquely determined by the previous state, especially in MWI.

                  • goldenkey 1052 days ago
                    Essential consciousness is the self appearance of free will. MWI still has randomness - how do you think you reside in one universe versus a different one? You keep sidestepping with erroneous statements.

                    Rules are not a source of real randomness and there is a consensus among physicists that quantum randomness is actually true randomness. Even if you put the source of randomness into some other thing like dark matter, you cannot explain its construct..because all you have is rules. You must forfeit and concede to nature a black box RNG library. This is clearly ludicrous. The more obvious and consistent view is that conscious beings (that feel they have willful decisions) are the manifestation of nature's RNG for quantum mechanics

                    IIT is an entirely different beast related to self attention, a different kind of definition for consciousness, that has little to do with will. Unrelated...

                    QM is unitary, but it is one to many. It's a giant branching rooted tree. This is irrelevant, it is not deterministic like general relativity.

                    Please step back in the element. These refutations are becoming exhausting.

                    • tsimionescu 1052 days ago
                      No, MWI has no randomness. All versions of you exist and are as real. The moment after the Big Bang, the fact that we would be having this conversation in the year 2021 on Earth was certain. This is an absolute requirement from the linearity of the Schrodinger's equation, which means that every state is uniquely determined by the states that came before. If you could measure the state of the universe with enough precision, you could tell today what I will feel tomorrow. And sure, each copy of you may experience a different comment, but they are indistinguishable, and all possible comments will be experienced.

                      In CI there is indeed true randomness, since only one 'branch' of the wavefunction is actually happening.

                      • goldenkey 1052 days ago
                        The randomness in MWI is observer-centric. It's trivially obvious. You reside in one world of many. The theory doesn't explain why your consciousness follows a specific bifurcation. The supposed copies of you are irrelevant.
                        • tsimionescu 1050 days ago
                          The theory does not allow for 'your consciousness' to exist as a meaningful concept that experiences more than one moment.

                          Say you are going to flip a quantum coin. Before you flip it, given MWI, you know that there will be two copies of you, each experiencing one of the outcomes. After you flip it, one copy sees tails, the other heads. Each copy is perfectly identical except for the new observation. Each may ask itself 'what were the chances that I am seeing heads/tails?', and with MWI in mind, they should answer 1: the chance that the copy-that-sees-heads sees heads is 1, and same for copy-who-sees-tails seeing tails, since there is no other way to identify them as separate entities.

                          This also corresponds to the mathematics of QM: if you want to give proper predictions, you need to 'update' the wavefunciton to have value 1 for the observed outcome and value 0 for any other outcome. Doing that for 'your branch' is precisely how MWI works.

                          • goldenkey 1049 days ago
                            Ffs, I understand the linearity of the Shrodinger equation under a model which MW entails. However, you are sidestepping the randomness involved in the choice of world that you reside in. You can dress it up all you want, but your consciousness is bound to one world of the many worlds. And that is defacto random.
    • lopuhin 1055 days ago
      I think the second paragraph in the wiki page about Bell's theorem explains that: https://en.wikipedia.org/wiki/Bell%27s_theorem
  • lamontcg 1054 days ago
    > a lot of people too cheerfully declare that Einstein was wrong about quantum mechanics.

    Einstein Podolsky and Rosen put down the scientific and theoretical foundation for testing enganglement in Quantum Mechanics.

    Treating it like just a horse race where Einstein was "wrong" is missing the contribution they made by designing the experiment (refined by Bell and tested by Aspect).

    What is interesting is that Bohr was "correct" but his contribution is mostly a bunch of untestable philosophical nonsense.

    The better contribution to science here is Einstein even though nature decided that his hunch was wrong.

    And there's a lesson in here somewhere about human nature and how we value people who just make lucky guesses that pay out and we call people who change their mind after more information comes in "flip-floppers".

  • throwaway32341 1055 days ago
    I'm curious if Einstein would've been satisfied with the theory of superdeterminism and how it answers the "spooky action at a distance" expression.
  • ergocoder 1054 days ago
    Is there an experiment that shows the entangled particles are NOT set before they are separated?

    It seems to me that it's fairly easy to explain why one particle correlates to the other. Both of their properties are set before they are separated. We just didn't measure those properties beforehand.

    I've read and watched a lot of quantum physics on youtube. This kind of refusal is always elusive, as in nobody has an experiment to negate this explanation (or maybe I miss it).

    Wave-particle duality is another one that doesn't feel real, especially when we fire one electron and we say "electron interfere with itself". It seems we think of it as a wave simply because the electron is too small for us to see. And that's it. Somehow we come up with this convoluted explanation with catchy name "wave-particle duality".

    This one is also elusive. Is there an experiment that refuses this explanation of how wave-particle duality isn't real?

  • tempestn 1055 days ago
    I've really got to think many worlds is the complete version, although I have no idea how that would be proven. Did Einstein get a chance to consider that framing? It seems like it would be up his ally as an explanation for the apparent measurement update.
    • Myto 1054 days ago
      Unfortunately no, Everett presented the idea in 1957, Einstein died in 1955.
    • eigenket 1054 days ago
      I am generally positive towards the mwi/Everetian approach, but it should be emphasised that we really don't have a satisfactory understanding of what "measurement" is in the Everetian approach, and in particular I don't think it is at all clear how to reproduce the Born rule.
  • stareatgoats 1055 days ago
    People with an interest in cosmology and bleeding edge physics can do worse than follow Sabine Hossenfelders blog or Youtube channel. Maybe everyone already does, but personally I discovered it just a few months ago.

    She is the kind of specialist that we expected to find more of in the early days of the internet: deep knowledge and an interest in and capability for explaining things clearly to non-experts. We need more people like her, if only to be able to evaluate if she is right or not in some key questions (like her opposition to still more powerful particle accelerators).

    In this case I find her explanation of Einstein's famous phrase regarding quantum mechanics quite rational, which is a novel experience for me when it comes to quantum mechanics.

    • jasonwatkinspdx 1055 days ago
      I'd add a disclaimer that Hossenfelder has some very strongly held negative views on how resources in the physics community should be allocated, particularly theoretical particle physics. This stack overflow answer is a pretty charitable explanation of why the overwhelming majority of the physics world disagrees with her on these specific meta issues: https://www.quora.com/How-do-theoretical-particle-physicists...
      • tsimionescu 1055 days ago
        To be fair, it seems that that blog post and the Quora responses are older than her book, where she clarified her position significantly more, though your claim about the majority of particle physicists disagreeing with her is likely still correct.

        The important difference between now and when the answer was written is that she has given her own alternatives: she believes that only theories that resolve an explicit contradiction or open problem in current physics should be considered, such as quantum gravity or dark matter or the measurement problem; while theories that simply fix 'ugliness' should be ignored, such as grand unification.

        Furthermore, she believes that theories with open parameters whose range exceeds possible near future possible colliders should not be tested, even if the lower bounds on the parameter could be testable - this is true for supersimmetry or WIMPs. Since ruling out supersimmetry entirely requires a planet-sized accelerator, there's no reason to keep building slightly larger ones just in case.

        Finally, she makes a concrete recommendation: shifting focus to astronomy as a better way to observe high energy particle physics occurring in the universe, investing in much more advanced telescopes instead of bigger particle colliders.

        • irthomasthomas 1054 days ago
          " she makes a concrete recommendation: shifting focus to astronomy as a better way to observe high energy particle physics "

          Each starlink launch makes this task harder and more expensive. They have chosen a military satellite grid over scientific research :(

          • ghusbands 1054 days ago
            Inside the atmosphere is a terrible place to use telescopes, anyway, so we should use those launches to also get more telescopes into space.
            • jasonwatkinspdx 1054 days ago
              This is overly simplistic. Adaptive optics solves a lot atmospheric issues. There are also entire categories of telescopes that it simply makes more sense to keep on earth. Space telescopes are cool and all, but making a space version of something isn't necessarily better, even in a future with rapidly declining launch costs.
      • sampo 1054 days ago
        > Hossenfelder has some very strongly held negative views on how resources in the physics community should be allocated

        If you advocate that A (for example, a large telescope) should be funded instead of B (for example, a particle collider), because you estimate A to be more fruitful, it's not a negative view. It's just a view.

        You must be pretty strongly in the particle collider camp, if you view funding telescopes as a negative view.

      • bawolff 1054 days ago
        Unless her views are unfairly colouring her presentation of info in her vlogs, it seems a bit unfair for that to need a disclaimer.

        People can have strong opinions and still present the issue fairly. I'd even go as far to say, you're probably not a very good expert if every view you have exactly mirrors the status quo.

      • RedShift1 1055 days ago
        Why the disclaimer?
        • jasonwatkinspdx 1054 days ago
          The bulk of her audience are folks without enough exposure to the physics world to understand that while she is very much an expert on physics itself, her meta views on how the community should conduct themselves are fringe.
    • TeeMassive 1055 days ago
      I've been a huge fan of Sabine for a while. Not only her explanations are rational, well informed and well presented, but she also has very nice dresses!
  • amelius 1054 days ago
    Why didn't he apply his "equivalence principle"? If two particles behave in the same way, they must be the same particle!
  • hyperpallium2 1054 days ago
    Aren't fields - magnetic, electric, gravitional etc - also spooky action at a distance? EDIT Thanks all!
    • dogma1138 1054 days ago
      They are just action at a distance, which was used much earlier than QM and GR to depict basically anything that didn’t require physical contact like gravity and electromagnetism.

      “Distance” is still a factor in fields hence locality, the action isn’t instantaneous and is governed by the universal speed limit and the amplitude is also affected by distance.

      Entanglement works instantly and regardless of distance hence the spooky part.

    • fsh 1054 days ago
      No, these fields are all local. Disturbances propagate at the speed of light and an event in one place cannot cause an effect in another place instantaneously.
    • pantalaimon 1054 days ago
      No they are limited by the speed of light
  • ZeroGravitas 1054 days ago
    I'm no expert, but I thought that Einstein was pointing out the flaws of the Copanhagen interpretation of quantum physics.

    Sean Carrol gave a talk about his new book and his claim is that quantum mechanics is pretty straightforward, and all the spooky weirdness is only needed if you try to ignore the implication that the data supports the-many world interpretation.

    https://youtu.be/F6FR08VylO4

    i.e. similar to the sock example given, when you open the envelope you find out what universe you are in, the blue sock or red sock universe and that tells you information about the other sock in your universe. But another you is finding the other color sock in a different universe.

    • TheOtherHobbes 1054 days ago
      MWI people keep making these strong claims and ignoring the very credible criticism of them.

      https://www.quantamagazine.org/why-the-many-worlds-interpret...

      • ZeroGravitas 1054 days ago
        That's a very philosophical rather than physics-based rebuttal.

        He appears to be saying it can't be true because it would make life/and or the concept of self meaningless.

        Whereas I would say that a theory that suggests we are meaningless temporary blips in an infinite universe is probably a much greater indication that it is true than otherwise.

        • fsh 1054 days ago
          Most interpretations of QM make identical predictions for the outcome of any possible experiment. This makes the field fundamentally philosophical rather than scientific.
      • AgentME 1054 days ago
        I previously found this article and wrote a bunch of notes about it that I never got around to posting before. It was too long to be an HN comment so I've just cleaned it up and posted it here: https://macil.tech/2021/05/09/bad-mwi-criticism/.

        I'm not a big fan of the article, it mostly just defines self in a nonsense way and then treats it as a contradiction with MWI despite also laying out more sensible answers.

    • jiggawatts 1054 days ago
      But physics is so much more mysterious if everybody carefully refuses to acknowledge the mere existence of the clear and obvious answer.