6 ms·

In said quote, Einstein advocates hidden variable theory, which was disproven. https://en.m.wikipedia.org/wiki/Bell_test_experiments https://en.m.wikipedia.org/

by wyager 10y ago

In said quote, Einstein advocates hidden variable theory, which was disproven. https://en.m.wikipedia.org/wiki/Bell_test_experiments https://en.m.wikipedia.org/wiki/Bell_test_experiments I would strongly advise people who do not have extensive tuition in Quantum Mechanics to refrain from speculating about the consequences of QM. I have been taking QM classes for several years now, and the most valuable thing I've learned is that almost everyone's understanding of QM is based on bad ad-hoc analogies, and it takes years of study to progress to gradually less ad-hoc analogies. Different mathematical models inform different "interpretations" of QM. If you can't tell the difference between a rigged and non-rigged Hilbert space, you certainly don't have enough background to speculate on classical QM, and I'm not at the point where I could even tell you the background required to speculate on modern QFT.

M_Grey 10y ago
Did you think I was endorsing the contents of either quote? I wasn't.
selimthegrim 10y ago
All three volumes of Weinberg?
aluhut 10y ago
You could replace QM with "magic" and it could be a dialogue by one of those fancy mages in any fantasy story. Do you see a way to "sell" QM to the public in a way that would reflect the diversity of the ideas and the concepts in a better way? Something that could replace what we have now? Does it make any sense at this point to do this at all?
jostylr 10y ago
Local hidden variable theories are not possible, but any local theory is not possible, assuming definite outcomes of experiments (many worlds might be considered local, but there is no definite outcome). Non-local hidden variable theories are quite possible, do quite well in explaining a number of facets of QM, and, in fact, inspired Bell's work. Nevertheless, Einstein certainly was no fan of the non-locality though he did not know of Bell's argument, of course.
- tiglionabbit 10y ago
  Non-locality can't be right though, since locality is defined by the speed of light, and nobody will tell you that sending information faster than the speed of light is possible. Many worlds it is, then.
  - M_Grey 10y ago
    Non-locality doesn't imply that you can transmit information FTL, in fact you can't without classical channels. You can send random bits with entangled particles, but that is the precise opposite of information.
  - wyager 10y ago
    Many worlds isn't the only alternative. The "cleanest" (i.e. least magical) proposed solution I know of is einselection, which gives us emergent behavior of the schroedinger equation that looks kind of like many worlds and kind of like "ideal" non-unitary collapse, but doesn't posit any "unobservable" or "magical" processes unlike many worlds or Copenhagen.
    - jostylr 10y ago
      You have to be clear about what is real in the theory. Many worlds presentations often suffer from this as they often talk about splitting universes, whatever that would mean. A nice version which might coincide with what is in mind for einselection is to take the wave function and integrate it out to get a mass density on three space. This mass density will the various different "worlds" present in it by means of the evolution; any given time will not tell us what belongs to which world, but as time evolves, we can see which pictures correspond to which ones, like fuzzy tv sets. See https://arxiv.org/pdf/0903.2211.pdf https://arxiv.org/pdf/0903.2211.pdf
      - M_Grey 10y ago
        There isn't much to recommend any of the many 'Interpretations' of QM, and if you deal with people involved in using it for their jobs, they tend to adhere to a "shut up and calculate" interpretation. After all, why construct an ontology based on what is clearly not complete at all energy scales? Still, where there are gaps, people will "theorize".
TheOtherHobbes 10y ago
Perhaps the "real" QM is itself observer-dependent. :)
ozy 10y ago
How about this intuition? Photons are massless and travel at light speed. But that is our perspective. From the photons perspective time is standing still. It is "touches" all space/time points on its path at once, and passes its quanta to only one point, depending on some chance function. Entanglement is the same, from our perspective it is spooky action at a distance, or over time. But from the two photons (and thus the qm phenomenon that created them) perspective it is the same instance. That is a consequence of time is relative. As mind bending as that might be. Or is this totally off?
- raattgift 10y ago
  It's wrong, but it's hard to explain precisely why without reference to group theory. Imprecisely why, the issue is that you can't find a global frame in which a massless particle is at rest. In General Relativity that's because there are no global frames at all; in Special Relativity, it's because global frames are Lorentz-invariant; a common tool in General Relativity, the Local Inertial Frame, is also Lorentz-invariant. The spacetime of Special Relativity is Minkowski spacetime (AKA Minkowski space AKA flat space) and at every point in that spacetime the Poincaré group (developed by Minkowski) is the isometry group. The Poincaré group has several subgroups, and here we can focus on the Lorentz group. The Lorentz group has three rotation generators J_[x,y,z] (Cartesian coordinates) or J_[\rho,\theta,\phi] (Spherical coordinates) etc, or in general coordinates, J_[1,2,3] or just J_i. It also has three boost generators along the same axes, K_i. There is a commutation relationship which is satisfied between these generators. (The Poincaré group also has translations. In Cartesian coordinates: P_[x,y,z,t]). For a massive spin-1 particle at rest at a single point in the Lorentz frame, the particle's angular momentum is invariant under rotations on any set of axes, but the direction of the angular momentum is changed. However, the photon is a _massless_ spin-1 particle, by definition under most theories and with excellent experimental support. As noted above, we can't find a Lorentz frame in which such a particle is at rest. However, by careful choice of coordinates, we can pick a single axis along which the whole of the massless spin-1 particle's momentum is found. The four-momentum ends up being invariant under the subgroup generated by [J_1, K_2-J_3, K_3+J_2], which produces a transformation matrix giving the rules for rotations around the particle's momentum [given in detail in Wigner, 1939]. Because this sets up a gauge in which _locally_ the momentum is contained, we can rule out the possibility that a photon is smeared out globally, even when considering a very very low-momentum photon (remembering the Einstein relationship for a massless particle, E = pc = \hbar\omega = hc / \lambda; so we're talking a wavelength comparable to the size of the universe). Intuitively, this is because you can always find (nearby, Lorentz) observers for which the photon's momentum is always contained within a single axis, and those observers are no less privileged than any other. Now, to consider your '"touches" all space/time points on its path at once', we need to do a bit of defining. The worldline is the entire path through the whole block universe spacetime Special Relativity. [http://backreaction.blogspot.co.uk/2008/05/block-universe.html http://backreaction.blogspot.co.uk/2008/05/block-universe.ht...] Every fundamental object has its own worldline. So you're verging on a tautology; the photon is at all its (own) spacetime points along the worldline. However, we can apply some 3+1 formalisms to foliate the block universe of Special Relativity by taking 3d spacelike hypersurfaces along surfaces at a constant _coordinate_ time. In each of these hypersurfaces, there is at each point a probability of interaction between the photon and anything that feels electromagnetism. Because of Lorentz invariance, the exact probability distribution is observer independent. However, as noted above, it is possible to choose observers who will see probabilities extremely close to zero except near a small 3-volume. Conversely, this rules out observers who see identical probabilities everywhere. So in that sense, at any given point, a photon does not "touch" all space at each point in time along its worldline. From this we can address your second last paragraph. A pair of correlated (entangled) photons whose worldlines develop a spacelike separation after entanglement are not in non-Lorentz-invariance-violating contact. A discovery of _any_ (local) Lorentz-invariance violation would be extremely exciting new physics, but there is a mountain of evidence suggesting that doesn't happen in our observable universe. [https://www.wikiwand.com/en/Modern_searches_for_Lorentz_violation https://www.wikiwand.com/en/Modern_searches_for_Lorentz_viol...] That is, if we do a 3+1 formalism and do a foliation on a timelike coordinate across the region(s) where the partial worldlines are spacelike separated, we can find no hypersurface on which the nonzero parts of the probability distributions of the pair of photons overlap.
akvadrako 10y ago
I think your comment is a bit inappropriately arrogant. I follow discussions in recent papers and on physicsforums.com from researchers and professors who have been studying QM for more than several years, yet they constantly disagree on even basic points, so I don't think even a lifetime of education in QM means you have a more intuitive understanding of it's consequences than a reasonably informed layman. The only thing they can agree on is how to calculate the expected results of experiments.† Also, hidden variable theories are not disproven by Bell - just local ones. † See https://www.preposterousuniverse.com/blog/2013/01/17/the-most-embarrassing-graph-in-modern-physics/ https://www.preposterousuniverse.com/blog/2013/01/17/the-mos...
- Koshkin 10y ago
  Parent comment is interesting in that it shows the disconnect between the way people get trained in quantum mechanics (or any hard science, for that matter) and what it takes in order to achieve a real understanding of the ways of Nature. Quantum mechanics is a highly mathematized and highly technical subject, and all this math and all these technicalities could be seen as merely a scuffolding; it is still hard to learn about the building by spending years studying the scuffolding. To me, it seems that the arguments about the basics will continue until the final unification with gravity has been found, because this may result in discovery of important constraints on the ways we should be thinking of these things. In the meantime, I think it maybe only people who apply quantum methods in their daily practice - quantum chemists, specialists in solid state physics etc. - will be the ones to have a better understanding of such matters as whether wave functions are real or not. Books about applied aspects of quantum theory could prove tremendously useful in that regard to everybody. One of the best for beginners, I believe, is Quanta by Atkins.

by wyager 10y ago

M_Grey 10y ago

Did you think I was endorsing the contents of either quote? I wasn't.

selimthegrim 10y ago

All three volumes of Weinberg?

aluhut 10y ago

You could replace QM with "magic" and it could be a dialogue by one of those fancy mages in any fantasy story. Do you see a way to "sell" QM to the public in a way that would reflect the diversity of the ideas and the concepts in a better way? Something that could replace what we have now? Does it make any sense at this point to do this at all?

jostylr 10y ago

Local hidden variable theories are not possible, but any local theory is not possible, assuming definite outcomes of experiments (many worlds might be considered local, but there is no definite outcome). Non-local hidden variable theories are quite possible, do quite well in explaining a number of facets of QM, and, in fact, inspired Bell's work. Nevertheless, Einstein certainly was no fan of the non-locality though he did not know of Bell's argument, of course.

tiglionabbit 10y ago
Non-locality can't be right though, since locality is defined by the speed of light, and nobody will tell you that sending information faster than the speed of light is possible. Many worlds it is, then.
- M_Grey 10y ago
  Non-locality doesn't imply that you can transmit information FTL, in fact you can't without classical channels. You can send random bits with entangled particles, but that is the precise opposite of information.
- wyager 10y ago
  Many worlds isn't the only alternative. The "cleanest" (i.e. least magical) proposed solution I know of is einselection, which gives us emergent behavior of the schroedinger equation that looks kind of like many worlds and kind of like "ideal" non-unitary collapse, but doesn't posit any "unobservable" or "magical" processes unlike many worlds or Copenhagen.
  - jostylr 10y ago
    You have to be clear about what is real in the theory. Many worlds presentations often suffer from this as they often talk about splitting universes, whatever that would mean. A nice version which might coincide with what is in mind for einselection is to take the wave function and integrate it out to get a mass density on three space. This mass density will the various different "worlds" present in it by means of the evolution; any given time will not tell us what belongs to which world, but as time evolves, we can see which pictures correspond to which ones, like fuzzy tv sets. See https://arxiv.org/pdf/0903.2211.pdf https://arxiv.org/pdf/0903.2211.pdf
    - M_Grey 10y ago
      There isn't much to recommend any of the many 'Interpretations' of QM, and if you deal with people involved in using it for their jobs, they tend to adhere to a "shut up and calculate" interpretation. After all, why construct an ontology based on what is clearly not complete at all energy scales? Still, where there are gaps, people will "theorize".

TheOtherHobbes 10y ago

Perhaps the "real" QM is itself observer-dependent. :)

ozy 10y ago

How about this intuition? Photons are massless and travel at light speed. But that is our perspective. From the photons perspective time is standing still. It is "touches" all space/time points on its path at once, and passes its quanta to only one point, depending on some chance function. Entanglement is the same, from our perspective it is spooky action at a distance, or over time. But from the two photons (and thus the qm phenomenon that created them) perspective it is the same instance. That is a consequence of time is relative. As mind bending as that might be. Or is this totally off?

raattgift 10y ago
It's wrong, but it's hard to explain precisely why without reference to group theory. Imprecisely why, the issue is that you can't find a global frame in which a massless particle is at rest. In General Relativity that's because there are no global frames at all; in Special Relativity, it's because global frames are Lorentz-invariant; a common tool in General Relativity, the Local Inertial Frame, is also Lorentz-invariant. The spacetime of Special Relativity is Minkowski spacetime (AKA Minkowski space AKA flat space) and at every point in that spacetime the Poincaré group (developed by Minkowski) is the isometry group. The Poincaré group has several subgroups, and here we can focus on the Lorentz group. The Lorentz group has three rotation generators J_[x,y,z] (Cartesian coordinates) or J_[\rho,\theta,\phi] (Spherical coordinates) etc, or in general coordinates, J_[1,2,3] or just J_i. It also has three boost generators along the same axes, K_i. There is a commutation relationship which is satisfied between these generators. (The Poincaré group also has translations. In Cartesian coordinates: P_[x,y,z,t]). For a massive spin-1 particle at rest at a single point in the Lorentz frame, the particle's angular momentum is invariant under rotations on any set of axes, but the direction of the angular momentum is changed. However, the photon is a _massless_ spin-1 particle, by definition under most theories and with excellent experimental support. As noted above, we can't find a Lorentz frame in which such a particle is at rest. However, by careful choice of coordinates, we can pick a single axis along which the whole of the massless spin-1 particle's momentum is found. The four-momentum ends up being invariant under the subgroup generated by [J_1, K_2-J_3, K_3+J_2], which produces a transformation matrix giving the rules for rotations around the particle's momentum [given in detail in Wigner, 1939]. Because this sets up a gauge in which _locally_ the momentum is contained, we can rule out the possibility that a photon is smeared out globally, even when considering a very very low-momentum photon (remembering the Einstein relationship for a massless particle, E = pc = \hbar\omega = hc / \lambda; so we're talking a wavelength comparable to the size of the universe). Intuitively, this is because you can always find (nearby, Lorentz) observers for which the photon's momentum is always contained within a single axis, and those observers are no less privileged than any other. Now, to consider your '"touches" all space/time points on its path at once', we need to do a bit of defining. The worldline is the entire path through the whole block universe spacetime Special Relativity. [http://backreaction.blogspot.co.uk/2008/05/block-universe.html http://backreaction.blogspot.co.uk/2008/05/block-universe.ht...] Every fundamental object has its own worldline. So you're verging on a tautology; the photon is at all its (own) spacetime points along the worldline. However, we can apply some 3+1 formalisms to foliate the block universe of Special Relativity by taking 3d spacelike hypersurfaces along surfaces at a constant _coordinate_ time. In each of these hypersurfaces, there is at each point a probability of interaction between the photon and anything that feels electromagnetism. Because of Lorentz invariance, the exact probability distribution is observer independent. However, as noted above, it is possible to choose observers who will see probabilities extremely close to zero except near a small 3-volume. Conversely, this rules out observers who see identical probabilities everywhere. So in that sense, at any given point, a photon does not "touch" all space at each point in time along its worldline. From this we can address your second last paragraph. A pair of correlated (entangled) photons whose worldlines develop a spacelike separation after entanglement are not in non-Lorentz-invariance-violating contact. A discovery of _any_ (local) Lorentz-invariance violation would be extremely exciting new physics, but there is a mountain of evidence suggesting that doesn't happen in our observable universe. [https://www.wikiwand.com/en/Modern_searches_for_Lorentz_violation https://www.wikiwand.com/en/Modern_searches_for_Lorentz_viol...] That is, if we do a 3+1 formalism and do a foliation on a timelike coordinate across the region(s) where the partial worldlines are spacelike separated, we can find no hypersurface on which the nonzero parts of the probability distributions of the pair of photons overlap.

akvadrako 10y ago

I think your comment is a bit inappropriately arrogant. I follow discussions in recent papers and on physicsforums.com from researchers and professors who have been studying QM for more than several years, yet they constantly disagree on even basic points, so I don't think even a lifetime of education in QM means you have a more intuitive understanding of it's consequences than a reasonably informed layman. The only thing they can agree on is how to calculate the expected results of experiments.† Also, hidden variable theories are not disproven by Bell - just local ones. † See https://www.preposterousuniverse.com/blog/2013/01/17/the-most-embarrassing-graph-in-modern-physics/ https://www.preposterousuniverse.com/blog/2013/01/17/the-mos...

Koshkin 10y ago
Parent comment is interesting in that it shows the disconnect between the way people get trained in quantum mechanics (or any hard science, for that matter) and what it takes in order to achieve a real understanding of the ways of Nature. Quantum mechanics is a highly mathematized and highly technical subject, and all this math and all these technicalities could be seen as merely a scuffolding; it is still hard to learn about the building by spending years studying the scuffolding. To me, it seems that the arguments about the basics will continue until the final unification with gravity has been found, because this may result in discovery of important constraints on the ways we should be thinking of these things. In the meantime, I think it maybe only people who apply quantum methods in their daily practice - quantum chemists, specialists in solid state physics etc. - will be the ones to have a better understanding of such matters as whether wave functions are real or not. Books about applied aspects of quantum theory could prove tremendously useful in that regard to everybody. One of the best for beginners, I believe, is Quanta by Atkins.