B What sort of an experiment can refute QM or QFTs?

[PeterDonis]

I guess you cannot falsify a dogma which states that the world is not classic.

You don't have to adopt any "dogma" about this. Classical physics was tested by experiment and experiment found it didn't work. That's why we have QM: because classical physics was making wrong predictions about the phenomena experimentalists were seeing at the atomic and subatomic scale in the late 19th and early 20th centuries.

 

[physika]

Yes but candidate theories might take decades or even centuries to get to that point.

maybe not so long, GRW one of the candidates.

.

​

 

[gentzen]

How would you falsify a tested quantum computer? What sort of scientific twist would it take?

Third law of thermodynamics and the scaling of quantum computers

One twist could be the important role of temperature when it comes to increasing the number of qubits.

 

[MathematicalPhysicist]

BTW I read the following in Joseph Polichnski's wiki bio:

In July 2012, Polchinski, with two of his students, James Sully and Ahmed Almheiri, and fellow string theorist Donald Marolf at the University of California, Santa Barbara (UCSB), published a paper[12] whose calculations about black hole radiation suggested that either general relativity's equivalence principle is wrong, or else a key tenet of quantum mechanics is incorrect.

was this problem in his paper resolved eventually?

 

[CoolMint]

Third law of thermodynamics and the scaling of quantum computers

One twist could be the important role of temperature when it comes to increasing the number of qubits.

This, as far as I can see, has been resolved to some extent by Google in their 2018 machine. The first computer in the world was as big as a room and is a million times less powerful than a low budget smartphone of today, so time will bring better and more practical solutions.

People take issue with QT because they cannot live without knowing. Without the cushion of the familiar classicality and objects in time and space.
I have no issue with everything being mostly the result of electromagnetism. One day, gravity too will be described in quantum terms - the remaining question will be - how did we come to the conclusion that we knew what we had assumed? If anything, the history of physics and science show that every knowledge is provisional, pending falsification.

 

[PeterDonis]

was this problem in his paper resolved eventually?

The whole area of research of which that paper was an example, basically how to reconcile quantum field theory with GR (which is not limited to the case of black holes although black holes do highlight a number of key issues), is still open and unresolved.

 

[physika]

BTW I read the following in Joseph Polichnski's wiki bio:

was this problem in his paper resolved eventually?

now summed up in the hairy vs not hairy dispute.

https://journals.aps.org/prl/abstract/10.1103/PhysRevLett.128.111301

https://journals.aps.org/prl/abstract/10.1103/PhysRevLett.127.055101

 

[gentzen]

This, as far as I can see, has been resolved to some extent by Google in their 2018 machine.

That was a link to a paper from 2022, so if their concern would have already been resolved in 2018, then the paper would be invalid to some extent. The paper references Google's results, and I see the following quotes from that paper against that background:

We argue that, in order to go beyond NISQ devices, substantial efforts are needed to improve not only the fidelity of single gates, but also the quality of initialization of multi-qubit registers. Thus, with this work, our aim is to help improving the design of quantum protocols and devices that properly takes into account fundamental thermodynamic constraints, hitherto neglected.

Accordingly, in order to move beyond NISQ devices, we need to prepare pure states with increasing fidelity by properly taking into account also the needed resources, at least at the energetic level. In doing this, quantum state initialization could need to be improved at a faster rate than the one at which the size of quantum computers increases.

Now I am unsure about the level you want to discuss such a paper, and also unsure about the allowed level of discussion in a B(eginner) level thread. Anyway, here is my opinion about the objections against scalable quantum computing from that paper:

I thought about similar "temperature based objections" against scalable quantum computing in the past, but the open question for my approach was whether quantum error correction could allow to limit the relevant size of the states to be considered. This paper might circumvent this question by focusing on the initial state instead. But maybe the question is still there, "conservation of difficulty", I currently simply don't know.

 

[Fra]

I assume most people when they refer to Quantum Theory, they sort of referring to QFT.

So my question boils down to, what sort of an experiment could potentially refute QFTs (its plural because there are QF theories)?

When pondering about this, I think the falsification paradigm is a bit blunt and uselss. As long as one considers QFT as effective theories, I don't see an obvious way how to falsify it as you can probably always tweak the parameters, enlarge the state spaces etc. But if one tries to ask, how the effective theories FORM, one may be required to step outside the QFT paradigms. In this sense, I doubt there will be a single experiment that refutes QFT, I think it will be if an alternative paradigm demonstrates it's superioirty in solving some open questions. Superior maybe in the sense or computation complexity, compact representation etc. Ie. I think it's not sufficient for a theory to be consistent, it must also be solvable or computable by it's host agent - otherwise it is useless. It's in these sense I think the current paradigms may need rethinking. Ie. it is not enough that something is solvable in principle, if it is not solvable by the resources at hand, then what is it's value, and rational from a naturalness perspective?

/Fredrik

 

[bob012345]

I'm not sure if this would qualify but in the spirit of this paper and talk;

https://www.researchgate.net/publication/224775736_There_are_no_particles_there_are_only_fields
https://www.preposterousuniverse.com/blog/2020/05/19/the-biggest-ideas-in-the-universe-9-fields/
https://www.symmetrymagazine.org/article/july-2013/real-talk-everything-is-made-of-fields

If an experiment could show there are particles such as the electron, distinguishing it from fields, would that violate the modern notion of QFT?

 

[MathematicalPhysicist]

When pondering about this, I think the falsification paradigm is a bit blunt and uselss. As long as one considers QFT as effective theories, I don't see an obvious way how to falsify it as you can probably always tweak the parameters, enlarge the state spaces etc. But if one tries to ask, how the effective theories FORM, one may be required to step outside the QFT paradigms. In this sense, I doubt there will be a single experiment that refutes QFT, I think it will be if an alternative paradigm demonstrates it's superioirty in solving some open questions. Superior maybe in the sense or computation complexity, compact representation etc. Ie. I think it's not sufficient for a theory to be consistent, it must also be solvable or computable by it's host agent - otherwise it is useless. It's in these sense I think the current paradigms may need rethinking. Ie. it is not enough that something is solvable in principle, if it is not solvable by the resources at hand, then what is it's value, and rational from a naturalness perspective?

/Fredrik

Well, I think we need both notions.
The ultimate notion of solvability, even if it's not feasible. and the more mundane doable testable theories.

 

[Fra]

Well, I think we need both notions.
The ultimate notion of solvability, even if it's not feasible. and the more mundane doable testable theories.

What I was thinking of was to put a stronger focus of fitness of a theory. I think the modern view on QM as a kind of information theory; where the expectations reflects the information we have about the system, is merely a first step towards a bigger revolution where both the encoded information and also the information processing is consider in it's physical processes, and this is then something that has to be modeled on the boundary, or the "observer side" of things. Here I think the often reductionist minded views on information theory may fail, and with it the fundamental QFT paradigms. I think we still have mote steps to take here, and I can perhaps imagine that the "trouble with QFT" as it stands, is that it will lead to problems such as unreaonable fine tuning, and beeing meaningfull only in a context of unlimited information capacity. I consider this a serious sign as suboptimal and not very fit.

/Fredrik

 

[Son Goku]

Run a loophole free CHSH test and see a violation of the Tsirelson bound. All quantum theories obey that bound.
Very hard to imagine Tsirelson exceeding frameworks being true given the rest of physics, but that's one example.

 

[zdcyclops]

Fair point but since I have not seen a consensus or a peer-reviewed resolution to these Schrodinger's cat questions about what exists between observations/interactions, Phinds and MathematicalPhysicist should probably not be asking them in this particular forum.

Objects exist even when they are not " looked at. " In quantum systems the only way to get information about quantum particles is to make a measurement or to look at them.

 

[vanhees71]

That's even true within classical physics. I never understood, why "the moon" shouldn't be there if nobody looks at it within quantum mechanics. There are fundamental conservation laws that tell us that the moon must be there when nobody looks, given the fact that it was there, when somebody has looked before.

 

[martinbn]

That's even true within classical physics. I never understood, why "the moon" shouldn't be there if nobody looks at it within quantum mechanics. There are fundamental conservation laws that tell us that the moon must be there when nobody looks, given the fact that it was there, when somebody has looked before.

If you prepare a spin one half system in the state |up>+|down> along a given axis, without measuring it what is the spin of the system along that axis?

 

[vanhees71]

It's indetermined. All the preparation implies are the probabilities to find the possible values of any measured observable.

 

[martinbn]

It's indetermined. All the preparation implies are the probabilities to find the possible values of any measured observable.

Then why do you have a problem with the same statement about the moon?! If the moon is not in an eigenstate of the position observables, then it is not there, it is not here, nor over there. You cannot say that it is in any specific location.

 

[CoolMint]

Objects exist even when they are not " looked at. " In quantum systems the only way to get information about quantum particles is to make a measurement or to look at them.

A quick summary of the key points:

Quantum theory is well established and grounded in experimental facts
It is a theory of measurements.
Who/what/when does the measurements is far less established and far less clear. Lots of assumptions and theories attempt to explain this weird situation with varying successes.

 

[vanhees71]

Then why do you have a problem with the same statement about the moon?! If the moon is not in an eigenstate of the position observables, then it is not there, it is not here, nor over there. You cannot say that it is in any specific location.

Then nothing would be ever anywhere, because the position eigenstates are not square integrable. It's as with all observables: Given a state (a "preparation") there are only probabilities for the outcome of measurements of any observable, which holds also for position. Position is never determined although the probability distribution for position can be arbitrarily sharply peaked around some value. Whether or not you can resolve the corresponding fluctuations depends on the resolution of your detector.

 

[martinbn]

Then nothing would be ever anywhere, because the position eigenstates are not square integrable. It's as with all observables: Given a state (a "preparation") there are only probabilities for the outcome of measurements of any observable, which holds also for position. Position is never determined although the probability distribution for position can be arbitrarily sharply peaked around some value. Whether or not you can resolve the corresponding fluctuations depends on the resolution of your detector.

Yes, in contrast to classical physics, where every single thing is there at any given time. Quantum mechanics is different, the moon is not there if you don't measure that observable. Einstein did have a problem with that, but you do you?

 

[Paul Colby]

Yes, in contrast to classical physics, where every single thing is there at any given time. Quantum mechanics is different, the moon is not there if you don't measure that observable. Einstein did have a problem with that, but you do you?

I didn't know the moon not there was an allowed transition?

 

[vanhees71]

As I said, I don't understand the conclusion that something is "not there", only because its position vector has no determined value. I also don't understand, why it is sometimes claimed (usually in popular-science writing) that a particle can be "at two places at ones", only because it's somehow prepared in a state, where its wave function peaks around two (or more) different regions. As soon as you accept that there is "irreducible randomness" in nature and that QT describes right that, such obviously meaningless paradoxa vanish into nothing.

 

[martinbn]

As I said, I don't understand the conclusion that something is "not there", only because its position vector has no determined value.

No, it is not a conclusion, it is the meaning of "not being there". Shirley when Einstein says "God doesn't play dice." you are not thinking of a deity and actual dice.

I also don't understand, why it is sometimes claimed (usually in popular-science writing) that a particle can be "at two places at ones", only because it's somehow prepared in a state, where its wave function peaks around two (or more) different regions.

It is an example of bad use of expressions that can be (actually always are) misleading to anyone who reads only popular science. The one above is not so bad and was just part of the way Einstein phrased his thoughts.

As soon as you accept that there is "irreducible randomness" in nature and that QT describes right that, such obviously meaningless paradoxa vanish into nothing.

I suppose Einstein never accepted that.

 

[vanhees71]

Again, why shouldn't the moon not be there only because its position is more or less well known to me?

Of course Einstein was far from being as naive as these bon mots being quoted all the time without also quoting them in the context with his very clear and detailed criticism against the QT and the implications concerning "irreducible randomness". This also leads to the wrong picture, also sometimes errorneously claimed in the pop-sci literature, that Einstein hadn't understood QT.

 

[martinbn]

Again, why shouldn't the moon not be there only because its position is more or less well known to me?

If it is there, where is that there? What are the coordinates? You agreed that if you haven't measured them those observables do not have values. Just like the spin one half system, if it is in the state |up>+|down> what is the value of the spin along that axis?

Of course Einstein was far from being as naive as these bon mots being quoted all the time without also quoting them in the context with his very clear and detailed criticism against the QT and the implications concerning "irreducible randomness". This also leads to the wrong picture, also sometimes errorneously claimed in the pop-sci literature, that Einstein hadn't understood QT.

 

[Demystifier]

That's even true within classical physics. I never understood, why "the moon" shouldn't be there if nobody looks at it within quantum mechanics. There are fundamental conservation laws that tell us that the moon must be there when nobody looks, given the fact that it was there, when somebody has looked before.

Because the Moon has only a few conserved quantities: energy, momentum, angular momentum and charge. Many other objects, which are not Moon, can have the same values of these conserved quantities. Why is Moon not transformed into one of those other objects?

 

[martinbn]

Again, why shouldn't the moon not be there only because its position is more or less well known to me?

Let me try this way. Suppose we have two disjoint regions of space, suppose $\psi$ and $\psi'$ are the wave functions that describe a particle being in each region (ignore mathematical subtleties). Suppose now that the particle is in a superposition of those two states. Where is the particle? Can you say there and point to a point in space, or just one of the two regions? If not, which I assume is your answer, then why is it hard for you to say that the particle isn't there?

 

[PeterDonis]

If the moon is not in an eigenstate of the position observables, then it is not there

I have always understood the claim that the moon is "not there" when nobody looks to mean, not that the moon has no specific position, but that it doesn't "exist" when nobody looks. That is the claim I take @vanhees71 and others to be arguing against.

 

[martinbn]

I have always understood the claim that the moon is "not there" when nobody looks to mean, not that the moon has no specific position, but that it doesn't "exist" when nobody looks. That is the claim I take @vanhees71 and others to be arguing against.

It is possible that it means exactlky that. Is there any way to know the context and what exactly Einstein meant? More importantly is there any reference that supports that QM states that?