A New Theorem Maps Out the Limits of Quantum Physics
quantamagazine.org
quantamagazine.org
The key part of the argument that is (I feel) a bit obscured to make it sound more shocking, is the fact that "observers" here includes "observers who are themselves in a superposition state".
Edit: the closing paragraph of the article is worse:
>In which case, taking the position that an observation is subjective and valid only for a given observer — and that there’s no “view from nowhere” of the type provided by classical physics — may be a necessary and radical first step.
I think that if you remember that they only proved this for "a given observer" that is itself in a superposition, it is a lot less "radical".
PS I wrote another more elaborate comment about this research here https://news.ycombinator.com/item?id=24322949
Seems easy to reject, it's just an extension of what Einstein's Relativity already showed (time is not objective), which is an extension of what Newton already formalized (velocity is not objective).
*Yes, quantum mechanics is kind of debatable.. But if I understand it correctly, the actual theory is deterministic. It's just the interpretation of the measurements that can lead to some debate. However they can also be viewed in a deterministic framework, as the link you cite shows.
No, it isn't. Quantum randomness is of a fundamentally different character than chaos. Bell's theorem (and concommitant experiments) shows that it is not possible for the results of quantum experiments to depend on information that exists in our universe before the experiment was actually conducted. So if the results of quantum experiments are deterministic, the information that determines their outcomes can only exist outside of our universe.
EDIT: Also, as Gerard 't Hooft points out[1], if you have a deterministic theory and you don't know the initial state, that is, you feed it probabilities, then you get probabilities on the other end. In this light there is nothing special about the quantum randomness. It's just that you don't know the initial state of the system.
It's not just that you don't know, it's that you cannot know. The initial state includes the state of the wave function, and you can't know that because of the no-cloning theorem.
Maybe I’m splitting hairs here, but surely there’s a difference between “not knowable in advance” and “doesn’t exist”?
The uncertainty principle makes this inherent, but the uncertainty principle itself is a consequence of the wave nature of particles.
Heisenberg says that we cannot know both position and velocity with arbitrary precision.
This is inherent to any wave-based system. BUT, this only means we cannot know (as in theoretically prevented) all of the variables to arbitrary enough precision to accurately predict the outcome.
It doesn’t mean, however, that there aren’t any initial conditions even prior to measurement.
Nor does Bell’s inequality doesn’t negate this. Note Lso that non-local does not imply that causality is broken (you cannot transmit information FTL via decoherence).
In fact, one of the more interesting (and unexplored) possibilities is that the boolean-logic law of excluded middle is wrong.
This is because Bell’s derivation is pure arithmetic and logic. It’s the one bit of QM that any student can follow.
Lest this is handwaved away, know that there are entire branches of constructivist mathematics that do just this.
This is the EPR argument.
> Bell’s inequality doesn’t negate this.
Bell shows that there is no measurement you can make, even in principle, that will give you the information you need to predict the outcome of a quantum experiment.
You can, if you wish, insist that those initial conditions exist notwithstanding our inability to measure them even in principle. But you could equally well insist that the outcomes of quantum experiments are determined by an invisible pink unicorn. Both hypotheses are equally unfalsifiable (if QM is correct).
I have actually coined the term IPU (Invisible Pink Unicorn) as an intentionally derisive description of hypothetical constructs that cannot be measured even in principle. Many QM interpretations contain IPUs. Bohmiam particle positions, for example, are an IPU.
It is also odd to say that position cannot be measured. We can tell in an experiment whether something ended up over there or over here. It would be reasonable to then try to have a theory that correlates the position measurements with something that has a position. Now it is not necessarily the case that there has to be such a thing, but it seems like a reasonable first step.
We can even see trails of particles in cloud chambers and the like. Why is that an IPU?
I will grant that it does not have to be the case that the only possible explanation is that of particles with position. But it certainly seems like if there is such a theory (and, of course, there is), then it would seem reasonable to consider it as quite plausible.
It also helps to ask you what is real in your theory. Are wave functions real? They certainly can't be measured in their entirety. Are operators the real thing? We don't measure them, but rather get something close to their eigenvectors/eigenvalues. Are those real?
Many worlds is the closest version with nothing added, but even that requires some kind of mass density function to make explicit connection with our lived experience. While it doesn't add too much in the way of extra mathematical structure in the theory (integrate over the wave function in a certain way: https://arxiv.org/abs/0903.2211 ), the implication in terms of what it says reality is actually like certainly involves a heck of a lot of IPUs.
That's a good question. The real answer is that no one actually knows. I think this is actually the biggest mystery in QM. But let me start with this, because I didn't make myself clear:
> It is also odd to say that position cannot be measured.
When I said that Bohmian positions are an IPU I did not intend that to mean that particle positions can't be measured. Obviously they can. The IPU-ness of Bohmian positions has to do with their ontological status, not their epistemic status. On Bohm's theory, a particle position considered along some axis is a real (in the mathematical sense) value, which is to say, it contains an infinite amount of information. But this information cannot be accessed in the same way that information stored in (say) a book can. I can open a book, even a book with an infinite number of pages, to any page and start reading it, and having read any page, I can go back and read that same page again. The information stored in Bohmian positions doesn't work that way. The laws of physics somehow conspire to hide all that information so that it can only be accessed serially and non-repeatably. The first time you measure a particle's position you get the most significant bits of its position. Those are then lost forever. You can never measure them again. The next time you measure a particle's position you get the next most significant bits of what that particle's position originally was, and so on. But you can never go back and do a second experiment to verify that the result you got for any of your measurements was actually correct and not a result of experimental error.
So the much-vaunted determinacy of Bohmian mechanics is not a reflection of the determinacy of the underlying metaphysical reality. It is really nothing more than a rhetorical trick. All the randomness is still there, it's just "pre-computed" and stored in particle positions in a way that it can only be accessed so that the world behaves exactly as if it were "really random" (whatever that means).
This same kind of trick is made manifest in a thought experiment [https://www.mathpages.com/rr/s9-07/9-07.htm] proposed by Kevin Brown. He points out that, if pi is normal (which is almost certainly is) then all of the results of all experiments ever conducted could be produced by a "cosmic Turing machine" computing the digits of pi. (See the two paragraphs beginning with "Even worse, there need be no simple rule of any kind relating the events of a deterministic universe.") Bohmian positions have exactly the same ontological status as the cosmic Turing machine. Only the window-dressing is different.
> It also helps to ask you what is real in your theory. Are wave functions real?
See http://blog.rongarret.info/2015/02/31-flavors-of-ontology.ht... for my answer to this.
> Many worlds ... involves a heck of a lot of IPUs.
Yep. That's why I'm not a big fan of the MWI either. See:
http://blog.rongarret.info/2019/07/the-trouble-with-many-wor...
We don’t know whether a non-local interpretation will lead to different, correct predictions.
One of the persistent hurdles is that gravity refuses to be folded into QM (or really, QCD).
Do you absolutely know whether hidden variable couldn’t be the missing factor?
Also—because I see you didn’t catch this—it’s possible Bell’s proof is incorrect because it assumes classical logic applies in QM.
In any case we’re done.
Yes, we do. An interpretation cannot lead to different predictions. If it did, it would not be an interpretation, it would be a new theory.
> In any case we’re done.
Seems so.
I'm not talking about determinism, I'm talking about Superdeterminism, which is akin to saying that coins will always land heads-up because people are compelled to avoid flipping coins in situations where it will land on tails. Please see my link.
But superdeterminism is nothing like that. It's determinism not based on past states, but based on future states, where the universe prevents anyone from making decisions that would cause them to learn a certain type of information. How would that work?
That doesn't necessarily mean signals going back in time. It could mean that the initial condition of the universe was set, such that nobody would do X. But that's still a way of the past (initial condition) being determined by the future, even if the future was just predicted via determinism. It makes physics teleological.
A common criticism of superdeterminism is that it eliminates falsifiability from science. E.g. to quote physicist Nicolas Gisin, "If we did not have free will, we could never decide to test a scientific theory. We could live in a world where objects tend to fly up in the air but be programmed to look only when they are in the process of falling." https://en.wikiquote.org/wiki/Superdeterminism
Fair enough I guess, but I'm not convinced it's a good answer to say "I think a theory with property X is the answer, and you can't criticize it because I have no actual theory with property X." I think it makes more sense to give little credence to superdeterminism until someone comes up with a plausible superdeterministic theory.
To make a dumb example, you measure some system and always observe outcome A. You might conclude that A is a property of the system you measured, but it just so happens that your decision to do the experiment is correlated with the system, so it always has A when you measure it, but it has other properties at other times.
"Consciousness"/"Observer" etc are too high level concepts to matter there.
I don't need to choose the experiment directly, I can make a complicated arrangement that then picks the experiment to be executed. (Let's say, I make a random number based on the current wind speed and I make a mechanical apparatus that throws a die and choose the experiment based on that). Try changing that experimental freedom.
(If you can't tell, I fully subscribe to the Many World interpretation. AFAICT it's, by quite a long way, the most parsimonious explanation in terms of extra "theoretical stuff", even if it is "wasteful" in terms of the amount of actual stuff it posits.)
MWI just posits that the measurement apparatus exists in a single world, while the particles exist in many worlds. But it can't explain this basic fact, like any other interpretation. What we need is an actual new theory that can actually measure what a measurement is (at what precise point the Born rule must be applied, or how can we do without the Born rule).
Their are some hidden variables which means it’s deterministic and there is no amplitudes involved. The universe has some sort of random number generator which determines the outcome. They don’t actually collapse, which is what the many worlds theory essentially posits.
That's not consistent with classical physics, and it is not what MWI actually posits. The universe branching that MWI posits is not really different from wave function collapse, and decoherence also doesn't really solve things. You still fundamentally have a single position for any classical system, but multiple positions with different amplitudes for quantum systems, and some mysterious threshold where you pass from one to the other.
Probability also can't really explain this, since the classical model also applies to single particles after they have interacted with a macroscopic system.
MWI considers that observers inside one branch can't interact or notice observers inside other branches, and this is why we perceive the world as if objects have unique definite positions. But this still doesn't hold up for quantum systems, which do in fact perceive and can interact with all of the other "worlds", including interacting with themselves in other worlds such as in the single particle double-slit experiment. So MWI doesn't really get away from the duality in any rigorous way.
You can’t experimentally differentiate between MWI and single universe wave function collapse.
Apparently this has become easier to swallow after all the research that has gone into QM during the past century.
There is another interpretation that can also be considered as close as possible to no-added sugars, the "ensemble interpretation". In "ensemble interpretation" the formulation of QM is understood to be applicable to ensembles of similarly prepared systems. In this interpretation QM is not really applicable to a single system by construction.
In MWI you accept it can be applied to a single system but you have to pay as a price that the system can branch into many different ones simultaneously (the "many worlds").
It's a kind of pick your own poison situation.
TL;DR; most probably you can't be sure what another person saw when they made a quantum measurement, because the person itself is a quantum system and therefore can be in a state of superposition of measuring A and measuring B.
The "instantaneous" aspect comes up in entanglement and measurement. Say you start off with an entangled pair of electrons A and B in your lab, and measure electron A. Measurement is an interaction, so you become entangled with the electron. By knowing the initial entangled state and your reading of electron A's state, you can know stuff about the state of electron B at the time of the measurement, even if electron B has traveled a very long distance in the meantime.
I personally find this point of view nicer than assuming that I somehow collapsed electron A's state and caused an instantaneous remote collapse of B's state.
Observer A observes one entangled particle at a spin up. At the same time, Observer B observes particle B at a spin down.
Observer A's observation of a spin up means observer B must see a spin down. However, the two observers are restricted at the speed of light to communicate this to each other.
It would initially seem like there is a instantaneous causal impact based on observer A's observation. But that's not the case.
I tend to think it was predestined what each observer see, but that leads to all sorts of philosophical questions.
I hope I did this topic justice, I find it quite interesting but have no background in it.
I suppose I should have asked what predictions SD actually makes.
http://backreaction.blogspot.com/2019/07/the-forgotten-solut...
Superdeterminism on the other hand is not even a theory. It cannot be falsified by design.
To actually collect information, you need to transmit data via some classical method i.e. the hou-c always wins
The fact is, there are theories of QM that do not assume that entanglement happens faster than light. The Many Worlds theory is one that has no need for such hypothesis. And more generally, since you need to send the results via a classical (no faster than light) communication method, there's no way to be sure that the entanglement has happened faster than light.
And here is some experimental data
It's probably not done because it's too hard, but the day they do this and actually include it in any equations that try to describe a system, I suspect much clarity will ensue.