Poll about foundations of QM: "experts" disagree on everything

The physics arXiv blog (and, after your humble correspondent, John Preskill) was intrigued by a new preprint by Schlosshauer, Kofler, and Zeilinger:

A Snapshot of Foundational Attitudes Toward Quantum Mechanics

They kind of repeated a 1997 poll organized by Max Tegmark (among 48 participants of a conference at that time) but among 33 participants of a recent quantum-foundations conference.



Off-topic. Greatest snowfall in Jerusalem since 1992. President Peres teaches the Israeli how to keep snowmen from getting cold.

The main result is that there is no consensus on anything. I was actually pleased because I was expecting that there would be a consensus on the wrong answers because that "subdiscipline" of physics is completely screwed; instead, the correct answers were mostly the strongest ones. What did they ask?




Randomness of quantum processes (e.g. decay of nuclei) is:

• 64%: fundamental 
• 48%: irreducible
• 9%: only apparent
• 0%: masking a hidden determinism

The right answers are highlighted by bold fonts. I don't quite understand how "fundamental randomness" may fail to be irreducible – what's the difference – but the difference between the two scores may be due to people's misunderstanding what the subtle word "irreducible" means.

At any rate, in politics, 64% for "fundamental randomness" would be a solid majority.

Do objects have well-defined properties before measurements?

• 52%: yes, in some cases
• 48%: no
• 9%: undecided
• 3%: always

In this case, the right answer has slightly-below-majority of votes. But the "strongest" answer may also be correct – classical properties or properties determined by quantum mechanics to be 100% certain are well-defined before the measurement. Of course, whether you would check "yes, in some cases" as a result depends on what you're ready to include behind "in some cases", namely whether these "some cases" are supposed to differ by the character of the observables or they may depend on the precise values.

At any rate, at least roughly 1/2 of the folks has an idea about this important point which is not too bad.

Einstein's view on quantum mechanics is:

• 64%: wrong
• 12%: will be shown wrong
• 12%: we don't know
• 6%: will be shown right
• 0%: is right

Again, you see a nearly constitutional majority behind the right answer while no one endorsed the "extreme opposite" answer. That looked rather clear. However, what about this one:

Bohr's view on quantum mechanics is:

• 30%: we have to wait
• 27%: wrong
• 21%: correct
• 9%: will be shown right
• 3%: will be shown wrong

The respondents are as split as you can get. I would answer "right" but it's a sufficiently vague question so that other answers seem tolerable to me, too. Bohr's pronouncement often lacked the rigor and accuracy to be completely certain that he meant the right thing. But if I were inclined to vote for a more "neutral" answer, it wouldn't be because we need time to settle the question whether Bohr was right. Time won't help. We already know everything we need to know. What makes the answer a bit fuzzy is that Bohr wasn't always crystal clear. But this won't improve with time.

A similar split appears when it comes to the measurement problem:

• 39%: solved (now or later) in a different way
• 27%: a pseudoproblem
• 27%: none of the above
• 24%: a severe difficulty threatening QM
• 15%: solved by decoherence

The separation of the respondents to groups is almost compatible with their decision to choose the answer randomly! ;-) I've marked two answers as right. Decoherence is an emergent mechanism or calculational framework to determine the boundary between the regimes in which classical physics is a good approximation (but quantum mechanics is still the right underlying description!) and those in which quantum subtleties (especially interference) are important and no classical approximation is acceptable.

But if you don't care where exactly this boundary lies, you may simply assume that the boundary is somewhere. The microscopic world clearly needs the full quantum mechanics; and for some large objects, the classical logic if not physics becomes acceptable. Bohr et al. knew that so they pretty much postulated this dichotomy – one that we may derive more clearly by decoherence today. If you don't care about the location of the boundary and you're OK with the previous answer "no objective reality prior to the measurement", I think that you must conclude that the measurement problem is a pseudoproblem.

What is the message of the violation of Bell's inequalities?

• 64%: local realism is untenable
• 52%: unperformed experiments have no results
• 36%: some notion of nonlocality
• 12%: action at a distance in the physical world
• 6%: let's not jump the gun, take loopholes seriously

It's true that unperformed experiments have no results and it is a part of the explanation why Bell's inequalities are violated in Nature but it is less accurate a description of the reason than the simple and accurate "local realism is untenable" (which was chosen by a large majority). Too bad that "realism" without adjectives apparently wasn't asked about.

Quantum information is:

• 76%: fresh air in quantum foundations
• 27%: we need to wait
• 6%: useful for applications but of no relevance to foundations
• 6%: neither useful nor relevant

This was a close call but I picked that they're useful for applications but of no relevance for foundations. Indeed, (hypothetical) quantum computers only utilize the laws of physics and mechanisms that are well-known from non-computer research of quantum physics. As their functionality depends on the conventional laws of quantum mechanics, they can't be a game-changer.

The answer "fresh air" could also be fine in the sense that the quantum-foundation people may be stuck with rubbish and crackpot non-quantum theories trying to "explain" quantum mechanics – so if they do something that isn't wrong, e.g. quantum algorithms, it's a breath of fresh air. But I am not sure whether this was what was meant by the "fresh air".

We must also wait and see to decide whether they will be constructed and whether it will be soon etc. If something will stop us for a very long time, we could also say that they're not "useful". So all answers have some potential to be OK. Much of it depends on the next question:

When will we have a working and useful quantum computer?

• 9%: within 10 years
• 42%: 10-25 years
• 30%: 25-50 years
• 0%: 50-100 years
• 15%: never

I chose no answer because without a crystal ball and tarot cards, I have no clue. It's conceivable that a gadget of this sort will be ready in 3 years. It may also take 80 or 800 years or people may lose it and never construct one. I am of course convinced that quantum computers are possible in principle and the improvement in the coherence and precision of the components that we need isn't "exponentially severe".

Some unusual complaints against the quantum computation were recently discussed by Scott Aaronson: response to Dyakonov, Zork's blogorhythm

Right interpretation of state vectors:

• 27%: epistemic/informational
• 24%: ontic
• 33%: a mix of epistemic and ontic
• 3%: purely statistical as in ensemble interpretation
• 12%: other

You see that the respondents were split once again but it's partly due to the fact that the answers are vaguely defined and not quite mutually exclusive.

I believe that all the people who answer "ontic" really mean that the wave function is conceptually a set of classical degrees of freedom. That's why this answer is just wrong. All other answers may be partly OK. Epistemic/informational is OK as a "negation" of the wrong "ontic" answer". However, "epistemic" could also lead one to believe that there exist other hidden variables not included in the wave function – i.e. that the wave function is an incomplete description. That's wrong as well. The wave function isn't "ontic" in the classical sense but it's as complete – and, in this sense, as close to "ontic" – as you can get. So you could vote for some "mix" or for "other" which means that you're not quite satisfied with any answer that was given.

I chose not to label the "ensemble interpretation" as correct because the ensemble interpretation makes the claim that only the statistics of the huge repetition of the very same experiment may be predicted by quantum mechanics. This is a very "restricted" or "modest" claim about the powers of quantum mechanics and this modesty is actually wrong. Even if I make 1 million completely different experiments, quantum physics may predict things with a great accuracy.

Imagine that you have 1 million different unstable nuclei (OK, I know that there are not this many isotopes: think about molecules if it's a problem for you) with the lifetime of 10 seconds (for each of them). You observe them for 1 second. Quantum mechanics predicts that 905,000 plus minus 1,000 or so nuclei will remain undecayed (it's not exactly 900,000 because the decrease is exponential, not linear). The relatively small error margin is possible despite the fact that no pair of the nuclei consisted of the same species!

So it's just wrong to say that you need to repeat exactly the same experiment many times. If you want to construct a "nearly certain" proposition – e.g. the proposition that the number of undecayed nuclei in the experiment above is between 900,000 and 910,000 – you may combine the probabilistically known propositions in many creative ways. That's why one shouldn't reduce the probabilistic knowledge just to some particular non-probabilistic one. You could think it's a "safe thing to do". However, you implicitly make statements that quantum mechanics can't achieve certain things – even though it can.

The observer is:

• 39%: a complex quantum system
• 21%: should play no fundamental role whatsoever
• 55%: plays a fundamental role in the application of the formalism but plays no distinguished physical role
• 6%: plays a distinct physical role (soul collapses wave function...)

I think that the first three answers are kind of correct. An observer is certainly just another complex physical system. It's pretty shocking that a majority of the participants of a quantum-foundations conference tries to deny this self-evident fact. Well, some of them may do so because they love some vitalist or spiritist approach similar to the "collapse of waves by the souls".

While the observer is clearly just another complex physical system and plays no distinguished physical role (so the third answer is also OK), it may play a role in the application of the formalism. However, in synergy e.g. with the consistent-histories approach, we may attribute the "subjective choices" to the consistent histories themselves (a framework listing the questions) rather than the observer as an object. For this reason, it looks legitimate to me if someone says that the observer plays no fundamental role whatsoever.

The denial of the "complex quantum system" by 61% of the participants is the most shocking feature of the answers here.

Reconstruction of quantum theory:

• 15%: gives useful insights and has/will supersede the interpretation program
• 45%: gives useful insights but we still need an interpretation
• 30%: cannot solve the quantum foundations
• 27%: will lead to a deeper theory than QM
• 12%: don't know

Reconstruction of quantum theory is a new minor fad that doesn't make any sense. While we may agree with its spirit – similar to "shut up"; "interpretation of quantum mechanics" is a problematic term by itself – everything else it says is some philosophical flapdoodle. The reconstruction program mainly asks "why is it true?" and it always answers "because we derived it". So far so good. But it never cares too much about what we have actually derived! ;-) The degree of support from the participants for this bogus wisdom is stunning. Incidentally, Zeilinger et al. say the same thing.

I also think that the very hypothetical strategy of "reconstruction" is a misunderstanding of the scientific method in general. The right wisdom about Nature can't be directly "reconstructed" by any well-defined procedure. In science, one has to formulate hypotheses at the beginning, and then test them. The "guesswork" (which can't be mechanical) is a necessary part of the process; the creativity and ingenuity of the physicists matters exactly because the "guesswork" matters. This is related to various logical arrows: the evolution of the Universe is a consequence of the laws of physics, not vice versa, so to determine the laws of physics from the observations of the evolution is an "inverse problem" that simply can't have a straightforward solution. It has to be "inferred" by Bayesian inference. That's why the answers always depend on some priors, too, although the dependence may become weak enough once a sufficient amount of evidence is collected and processed.

Favorite interpretation:

• 0%: consistent histories
• 42%: Copenhagen
• 0%: de Broglie-Bohm pilot wave
• 18%: Everett many worlds/minds
• 24%: information-based
• 0%: modal
• 9%: objective collapse, GRW or Penrose
• 6%: quantum Bayesianism
• 6%: relational quantum mechanics
• 0%: ensemble interpretation
• 0%: transactional interpretation
• 12%: other
• 12%: no preferred one

The most correct interpretation, consistent histories, was picked by 0 percent of the participants. What a pack of assholes, politely speaking. ;-) Thank God, at least Copenhagen was picked by respectable 42% (beating many worlds: so Tegmark's claims about the "dominance" of MWI in similar circles have been superseded) and many of the totally silly answers were at 0% or near 0%, too.

How often have you switched interpretation?

• 33%: never
• 21%: once
• 21%: several times
• 21%: no preferred interpretation

I didn't highlight any right answer because this isn't an objective question. It's a question about the events in different individuals' lives. I haven't switched the interpretation in a real sense. At some moment, I didn't quite understand everything so I didn't "claim" anything. Since the times I have an opinion, it's the same opinion.

Does the choice of interpretation depend on philosophical prejudices?

• 58%: a lot
• 27%: a little
• 15%: not at all

Of course, it's mostly about philosophical prejudices. Especially for the believers in the wrong interpretations, it's a matter of quasireligious beliefs they're just not ready to abandon. Zeilinger also say that the influence of the prejudices isn't necessarily a reflection of physicists' being dogmatic; it's due to the absence of experiments that clearly distinguish the different approaches. Maybe. I am not so sure about it because most of the wrong attitudes to QM are quite explicitly incompatible with some experiments or they at least prevent one from writing a theory, a set of rules, that may explain all of them in a unified way.

Superpositions of macro- different states are:

• 67%: possible in principle
• 36%: will eventually be realized
• 12%: in principle impossible
• 6%: impossible because of collapse theory

It's of course possible in principle – linearity is a universal postulate of QM that applies to any states, any objects. For some "small macroscopic" things, it's been realized but the degrees of freedom that are in superposition are very restricted even though they may be delocalized in a material. For completely general and "warm" objects, the superpositions will never be realized in practice. The "collapse theory" (they probably mean GRW/Penrose) is wrong.

In 50 years, conferences on quantum foundations:

• 48%: will still be organized
• 15%: probably no
• 24%: who knows
• 12%: I organize one no matter what

Crystal balls again... It's plausible that all the participants will already be dead so even the last answer may be wrong. They may also change their mind. ;-) I have no problem with such conferences to take place or not to take place. What I have a problem with are the idiotic answers that many of the self-described "experts" offer to many elementary questions. Will it still be so bad – or worse – in 50 years?

The authors also discuss some correlations between the answers, probably kind of obvious ones... TRF readers voted overwhelmingly "NO", just like I did.