5 ms·

A Private View of Quantum Reality

40 points by pskotarczak 11y ago

stared 11y ago
QM interpretation is a subtle thing and QBism makes a lot of sense. Yet (IMHO) it trades practical interpretation for some (nomen omen) purity. In QM you can describe a quantum state with a wavefunction (to incorporate quantum ignorance) or a density matrix (to incorporate both quantum and classical ignorance). (A for classical ignorance only, one would just use probability vector). States than can be described with just a wavefuctions are called pure states. And many-world interpretation revolves around having _only_ pure states (so, while being unsettling, does not add other assumptions and is, well... more pure). In QBsim pure states are not singled out. Having said that, if you are coming to QM, QBism will protect you against a lot of BS. :) And BTW: I remember my meeting with Chris Fuchs. It was a very illuminating story how did he turned from a (sinful) frequentist to an (enlightened) bayesianist. And how we should focus on knowing and not knowing things, rather that consider reality as an abstract thing. A --bit-- lot of shameless self-advertisement: - Hydrogen ion and classical vs quantum not-knowing: https://johncarlosbaez.wordpress.com/2015/03/13/quantum-superposition/ https://johncarlosbaez.wordpress.com/2015/03/13/quantum-supe... - My quantum game - I want _show_ how does quantum mechanics work http://quantumgame.io/ http://quantumgame.io/ (just a sign-up list; I will release alpha this Sept) - in my PhD thesis I coined qubism (for a plotting scheme for quantum states), only to realize later that a similar word is already taken; chapter 3 of http://arxiv.org/abs/1412.6796 http://arxiv.org/abs/1412.6796: "The name qubism (inspired by Cubism, the art movement) should not be confused with QBsim (quantum Bayesianism)"
- platz 11y ago
  QB still seems like Berkely to me, but just with a dose of math.
  - stared 11y ago
    In QM math is essential. Most of things I had learnt about QM before learning its basic, well, mechanics were actually building false intuitions. Without math, many-world interpretation is a ridiculous absurd, and an Ockham's nightmare. With - well, it's pure QM _without_ adding additional assumptions.
    - amelius 11y ago
      > ...and an Ockham's nightmare... Just never forget that Ockham's razor has absolutely no basis in science. Also, what one person finds a simpler or more elegant theory, another person might not; in that sense it is quite arbitrary.
      - pdonis 11y ago
        > what one person finds a simpler or more elegant theory, another person might not; in that sense it is quite arbitrary But one can quantify the complexity of a theory as the number of bits required to compute all of its predictions. In that sense, the MWI is the least complex version of QM, because, as stared pointed out, it's just the basic math of QM with nothing added on. All of the other interpretations add something on, which means requiring more bits for computation.
      - jasonwatkinspdx 11y ago
        > what one person finds a simpler or more elegant theory, another person might not; in that sense it is quite arbitrary There are ways of making it formal and precise. See section 3 in this chapter from Judea Pearl's book: http://bayes.cs.ucla.edu/BOOK-99/ch2.pdf http://bayes.cs.ucla.edu/BOOK-99/ch2.pdf or for a more lightweight discussion that happens to be in my tab stack atm, see https://probmods.org/occam's-razor.html https://probmods.org/occam's-razor.html I understand people think they're being helpful with the knee jerk "that's not science" response to someone asserting Occam, since the principle is often abused/overused, but the academic perspective on this has been much richer for decades now. It's not fair to just sweep that away with a ba-humbug.
      - nazgulnarsil 11y ago
        Has a basis in information theory. But it's funny how you would assert that so confidently. What lead you to being that confident? https://en.wikipedia.org/wiki/Kolmogorov_complexity https://en.wikipedia.org/wiki/Kolmogorov_complexity
        amelius 11y ago
Animats 11y ago
This is one of those things where it's really hard to tell whether it's a major step forward or utter bullshit. The question to ask, of course, is "how could this be tested experimentally?". It it's testable, it's a lot more important than if it isn't. The proposed theory would seem to indicate that it's possible to get the real world into a state similar to an out-of-sync network game. Some games can get into a state where the player states have diverged, but the players are still connected. Is this article claiming that the real world can do that? That should be testable.
- lisper 11y ago
  A good litmus test for an article like this is whether it talks about the Bell inequalities. If it doesn't (and this one doesn't) it's very likely bullshit.
  - gus_massa 11y ago
    I agree with both sentences, but this articles at least say something about the Bell inequalities: > My fellow QBists and I instead think that what Bell’s theorem really indicates is that the outcomes of measurements are experiences, not revelations of something that’s already there. I studied the Bell inequalities at the university, but I have no cue about what this sentence means.
    - lisper 11y ago
      Makes two of us. Sounds like BS to me.
c3d 11y ago
I think its a step in the right direction, but I believe one can go even further. It is not necessary to assume an observer or a conscious entity (the dog problem in the article), if one remarks that physical entities that have the "collapse of the wave function" property make good measurement instruments. In other words, knowledge is gained more rapidly with a good measurement instrument, and that is what defines the "collapse", not consciousness. In that interpretation, QM is seen as the only way to do physics when you represent your knowledge of a system using probabilities. Measurement instruments are a choice among all the physical systems, and what we consider good physical instruments lead to QM "axioms". This is elaborated here: http://cc3d.free.fr/tim.pdf http://cc3d.free.fr/tim.pdf.
- darkmighty 11y ago
  This kind of argument works for classical thermodynamics too, where observing a particle in a gas collapses it's velocity distribution; where if we take a confined gas we can localize it (confining it within a tight space), only to grow uncertain about the momentum of constituents, or expand it, growing uncertain about the position. In other words, for any probabilistic physical framework it seems to me you are going to need to put a magical 'sampling' or 'observation' somewhere. Making this explicit is one of the reasons I like Bohm's formulation: https://en.wikipedia.org/wiki/De_Broglie%E2%80%93Bohm_theory#Relation_to_the_Born_Rule https://en.wikipedia.org/wiki/De_Broglie%E2%80%93Bohm_theory...
beefman 11y ago
Original source: https://www.quantamagazine.org/20150604-quantum-bayesianism-qbism/ https://www.quantamagazine.org/20150604-quantum-bayesianism-...
DennisP 11y ago
I guess I don't get it. How is this different from a hidden-variable theory? How does it explain the two-slit experiment?
- nazgulnarsil 11y ago
  Yeah, this seems to raise more questions than it answers. Anyone have a link to a better, more technical explanation?
  - gsteinb88 11y ago
    Chris Fuchs' writings are pretty good: http://perimeterinstitute.ca/personal/cfuchs/ http://perimeterinstitute.ca/personal/cfuchs/ I mentioned this in another comment, but I particularly like "Quantum Mechanics as Quantum Information, Mostly"
- altrego99 11y ago
  Yep this is hidden value theorem. I think either the author messed up the concept badly, or what is more likely, the proponent of this 'QBism' doesn't get it. He appears to think spooky action is just lack of information - something like "if I open the box and see white marble, you must have the black one". Possibly haven't read or doesn't understand EPR thought experiment (which shows how by choosing to make a certain kind of observation, what he and thus the other guy sees). Looks like pseudoscience to me.
  - gsteinb88 11y ago
    Yeah, no, Chris Fuchs is definitely not doing pseudoscience. Please don't make judgements like that based on popular science writeups (though, this is actually one of the better ones I've seen on interpretations of QM, even if the figure is horribly misleading/wrong). More generally, QBism is not a hidden variable theorem in any sense. What the article glosses over is that QBism does still require a modification to standard probabilities that (when combined with Baysian/information theoretic reasoning) gives you the measurement probabilities you actually see in the lab. Chris Fuchs' writeups are pretty fantastic: http://perimeterinstitute.ca/personal/cfuchs/ http://perimeterinstitute.ca/personal/cfuchs/ Specifically, "Quantum Mechanics as Quantum Information, Mostly" is a short and fun introduction (okay, 32 pages, but pretty easy reading).
DonGateley 11y ago
Luboš Motl, arch conservative, thinks it's essentially correct and is exactly what the founders tried to convey in the first place. As always, of course, Lumo gets it and almost no one else does. :-) http://motls.blogspot.com/2015/06/is-quantum-reality-personal.html http://motls.blogspot.com/2015/06/is-quantum-reality-persona...
norea-armozel 11y ago
It seems like an odd approach, but it could explain consciousness better since it assumes that each measured outcome is private information of the observer and each observer makes up the difference in regards to having differently measured outcomes. But there's something troubling about that for me. Does it mean when someone measured the health of Christopher Lee he didn't die? But for another observer who watched his vitals he did die? Are they in the same world? Or does this require the many world's interpretation to remain intact? I hope I'm not stretching QBism beyond its intended goals, but that's the sort of conclusion I can see being made from it.

40 points by pskotarczak 11y ago

QM interpretation is a subtle thing and QBism makes a lot of sense. Yet (IMHO) it trades practical interpretation for some (nomen omen) purity. In QM you can describe a quantum state with a wavefunction (to incorporate quantum ignorance) or a density matrix (to incorporate both quantum and classical ignorance). (A for classical ignorance only, one would just use probability vector). States than can be described with just a wavefuctions are called pure states. And many-world interpretation revolves around having _only_ pure states (so, while being unsettling, does not add other assumptions and is, well... more pure). In QBsim pure states are not singled out. Having said that, if you are coming to QM, QBism will protect you against a lot of BS. :) And BTW: I remember my meeting with Chris Fuchs. It was a very illuminating story how did he turned from a (sinful) frequentist to an (enlightened) bayesianist. And how we should focus on knowing and not knowing things, rather that consider reality as an abstract thing. A --bit-- lot of shameless self-advertisement: - Hydrogen ion and classical vs quantum not-knowing: https://johncarlosbaez.wordpress.com/2015/03/13/quantum-superposition/ https://johncarlosbaez.wordpress.com/2015/03/13/quantum-supe... - My quantum game - I want _show_ how does quantum mechanics work http://quantumgame.io/ http://quantumgame.io/ (just a sign-up list; I will release alpha this Sept) - in my PhD thesis I coined qubism (for a plotting scheme for quantum states), only to realize later that a similar word is already taken; chapter 3 of http://arxiv.org/abs/1412.6796 http://arxiv.org/abs/1412.6796: "The name qubism (inspired by Cubism, the art movement) should not be confused with QBsim (quantum Bayesianism)"

platz 11y ago
QB still seems like Berkely to me, but just with a dose of math.
- stared 11y ago
  In QM math is essential. Most of things I had learnt about QM before learning its basic, well, mechanics were actually building false intuitions. Without math, many-world interpretation is a ridiculous absurd, and an Ockham's nightmare. With - well, it's pure QM _without_ adding additional assumptions.
  - amelius 11y ago
    > ...and an Ockham's nightmare... Just never forget that Ockham's razor has absolutely no basis in science. Also, what one person finds a simpler or more elegant theory, another person might not; in that sense it is quite arbitrary.
    - pdonis 11y ago
      > what one person finds a simpler or more elegant theory, another person might not; in that sense it is quite arbitrary But one can quantify the complexity of a theory as the number of bits required to compute all of its predictions. In that sense, the MWI is the least complex version of QM, because, as stared pointed out, it's just the basic math of QM with nothing added on. All of the other interpretations add something on, which means requiring more bits for computation.
    - jasonwatkinspdx 11y ago
      > what one person finds a simpler or more elegant theory, another person might not; in that sense it is quite arbitrary There are ways of making it formal and precise. See section 3 in this chapter from Judea Pearl's book: http://bayes.cs.ucla.edu/BOOK-99/ch2.pdf http://bayes.cs.ucla.edu/BOOK-99/ch2.pdf or for a more lightweight discussion that happens to be in my tab stack atm, see https://probmods.org/occam's-razor.html https://probmods.org/occam's-razor.html I understand people think they're being helpful with the knee jerk "that's not science" response to someone asserting Occam, since the principle is often abused/overused, but the academic perspective on this has been much richer for decades now. It's not fair to just sweep that away with a ba-humbug.
    - nazgulnarsil 11y ago
      Has a basis in information theory. But it's funny how you would assert that so confidently. What lead you to being that confident? https://en.wikipedia.org/wiki/Kolmogorov_complexity https://en.wikipedia.org/wiki/Kolmogorov_complexity
      - amelius 11y ago

Animats 11y ago

This is one of those things where it's really hard to tell whether it's a major step forward or utter bullshit. The question to ask, of course, is "how could this be tested experimentally?". It it's testable, it's a lot more important than if it isn't. The proposed theory would seem to indicate that it's possible to get the real world into a state similar to an out-of-sync network game. Some games can get into a state where the player states have diverged, but the players are still connected. Is this article claiming that the real world can do that? That should be testable.

lisper 11y ago
A good litmus test for an article like this is whether it talks about the Bell inequalities. If it doesn't (and this one doesn't) it's very likely bullshit.
- gus_massa 11y ago
  I agree with both sentences, but this articles at least say something about the Bell inequalities: > My fellow QBists and I instead think that what Bell’s theorem really indicates is that the outcomes of measurements are experiences, not revelations of something that’s already there. I studied the Bell inequalities at the university, but I have no cue about what this sentence means.
  - lisper 11y ago
    Makes two of us. Sounds like BS to me.

c3d 11y ago

I think its a step in the right direction, but I believe one can go even further. It is not necessary to assume an observer or a conscious entity (the dog problem in the article), if one remarks that physical entities that have the "collapse of the wave function" property make good measurement instruments. In other words, knowledge is gained more rapidly with a good measurement instrument, and that is what defines the "collapse", not consciousness. In that interpretation, QM is seen as the only way to do physics when you represent your knowledge of a system using probabilities. Measurement instruments are a choice among all the physical systems, and what we consider good physical instruments lead to QM "axioms". This is elaborated here: http://cc3d.free.fr/tim.pdf http://cc3d.free.fr/tim.pdf.

darkmighty 11y ago
This kind of argument works for classical thermodynamics too, where observing a particle in a gas collapses it's velocity distribution; where if we take a confined gas we can localize it (confining it within a tight space), only to grow uncertain about the momentum of constituents, or expand it, growing uncertain about the position. In other words, for any probabilistic physical framework it seems to me you are going to need to put a magical 'sampling' or 'observation' somewhere. Making this explicit is one of the reasons I like Bohm's formulation: https://en.wikipedia.org/wiki/De_Broglie%E2%80%93Bohm_theory#Relation_to_the_Born_Rule https://en.wikipedia.org/wiki/De_Broglie%E2%80%93Bohm_theory...

beefman 11y ago

Original source: https://www.quantamagazine.org/20150604-quantum-bayesianism-qbism/ https://www.quantamagazine.org/20150604-quantum-bayesianism-...

DennisP 11y ago

I guess I don't get it. How is this different from a hidden-variable theory? How does it explain the two-slit experiment?

nazgulnarsil 11y ago
Yeah, this seems to raise more questions than it answers. Anyone have a link to a better, more technical explanation?
- gsteinb88 11y ago
  Chris Fuchs' writings are pretty good: http://perimeterinstitute.ca/personal/cfuchs/ http://perimeterinstitute.ca/personal/cfuchs/ I mentioned this in another comment, but I particularly like "Quantum Mechanics as Quantum Information, Mostly"
altrego99 11y ago
Yep this is hidden value theorem. I think either the author messed up the concept badly, or what is more likely, the proponent of this 'QBism' doesn't get it. He appears to think spooky action is just lack of information - something like "if I open the box and see white marble, you must have the black one". Possibly haven't read or doesn't understand EPR thought experiment (which shows how by choosing to make a certain kind of observation, what he and thus the other guy sees). Looks like pseudoscience to me.
- gsteinb88 11y ago
  Yeah, no, Chris Fuchs is definitely not doing pseudoscience. Please don't make judgements like that based on popular science writeups (though, this is actually one of the better ones I've seen on interpretations of QM, even if the figure is horribly misleading/wrong). More generally, QBism is not a hidden variable theorem in any sense. What the article glosses over is that QBism does still require a modification to standard probabilities that (when combined with Baysian/information theoretic reasoning) gives you the measurement probabilities you actually see in the lab. Chris Fuchs' writeups are pretty fantastic: http://perimeterinstitute.ca/personal/cfuchs/ http://perimeterinstitute.ca/personal/cfuchs/ Specifically, "Quantum Mechanics as Quantum Information, Mostly" is a short and fun introduction (okay, 32 pages, but pretty easy reading).

DonGateley 11y ago

Luboš Motl, arch conservative, thinks it's essentially correct and is exactly what the founders tried to convey in the first place. As always, of course, Lumo gets it and almost no one else does. :-) http://motls.blogspot.com/2015/06/is-quantum-reality-personal.html http://motls.blogspot.com/2015/06/is-quantum-reality-persona...

norea-armozel 11y ago

It seems like an odd approach, but it could explain consciousness better since it assumes that each measured outcome is private information of the observer and each observer makes up the difference in regards to having differently measured outcomes. But there's something troubling about that for me. Does it mean when someone measured the health of Christopher Lee he didn't die? But for another observer who watched his vitals he did die? Are they in the same world? Or does this require the many world's interpretation to remain intact? I hope I'm not stretching QBism beyond its intended goals, but that's the sort of conclusion I can see being made from it.