Maybe it's just because I have increased the frequency with which I write about the foundations of quantum mechanics and people send me various stuff. But the amount of anti-quantum misconceptions that I have been exposed to in recent weeks was higher than ever before.
At the end, almost all if not all of these people just aren't willing or able to even consider the possibility that the assumptions of classical physics are wrong. It must sound surprising to an open-minded student who hasn't faced any serious obstacles when he was learning quantum mechanics. But when you actually interact with these anti-quantum zealots and you see how they react to some extremely simple, rock-solid explanations of yours, you simply have to conclude that their IQ is probably below 70. The degree of stupidity they are willing to promote in order to defend the indefensible is shocking.
I have discussed the totally wrong 2015 paper by Mark Alford that has claimed that nonlocality had to exist basically even in quantum field theory. In previous cases, these people were just sloppy. They used an unphysical definition of nonlocality – they always conflated it with classicality/realism in some way.
But Alford actually didn't. He defined locality exactly in the right way that is usable even if you teach quantum field theory. Probabilities of measurements in the region R must be fully calculable out of the knowledge in the past light cone of R. And even though Alford persistently claims that it's not true, in quantum field theory, this principle of locality is exactly true. Despite very detailed explanations including lots of simple enough formulae (which many undergraduates are learning in first lectures of QFT courses), he remains clueless. He is just incapable of understanding that the space-like separated commutators of fields vanish; and why it implies that spacelike-separated events or decisions can't influence the probabilities predicted for a given region.
To talk to these people is apparently a much more hopeless waste of time than attempts to teach quantum mechanics to a kitty, something I was actually far more successful with.
Alford's crackpot preprint has no citations according to arXiv or SPIRES but... one according to Google Scholar. It's a January 2016 book by Jean Bricmont, Making Sense of Quantum Mechanics. So I needed to see what's inside the book and finally I succeeded.
It's basically a pro-Bohmian manifesto repeating many of the same stupidities that you can find in Alford's article. Nonlocality is there regardless of anything. Well, no. The amount of nonlocality in QFT – our ultimate description of all non-gravitational phenomena in Nature – is exactly zero. At least, I wrote a one-star Amazon review of that book. Sadly, Bricmont is a guy who co-wrote some texts with Alan Sokal against postmodern philosophers. That admirable "activism" can't prevent him from being completely wrong about the foundations of modern physics, however.
Florin Moldoveanu is confused about quantum mechanics most of the time as well but he wrote a review of the GHZM experiment which looks totally OK; see my Reference Frame review of GHZM and David Mermin's Reference Frame review of GHZM or Coleman's Your Face lecture and compare.
The initial state is a simple entangled state of three spins, a superposition of up-up-up and down-down-down. If the three experimenters measure either \(j_x=\pm 1/2\) or \(j_y=\pm 1/2\) (8 different possibilities how the three people choose either \(j_x\) or \(j_y\)), they get various correlations and rules. The funny thing is that any local realist theory predicts certain correlations between the signs. And when you make the experiment, you get exactly the opposite signs than predicted by any local realist theory.
A strength of GHZM relatively to Bell's inequality (and relatively to the double slit experiment) is that you don't need any continuous numbers, any inequalities. Local realist theories predict one sign; quantum mechanics predicts the other sign. Both do so with the 100% certainty. And experiments confirm the quantum mechanical prediction.
While Moldoveanu is mostly confused, this blog post was OK. But a crackpot named Andrei came to please us with his thoughts that local realism had to be right, anyway. He said that there were just two options: 1) local realism holds and something else, or 2) local realism holds and something different. I told him that it's like efforts to explain Darwin's theory about the origin of species to a creationist who only allows two options: 1) God created the species in one week, and the other, brave, heretical possibility 2) God created the species in two weeks and he was also allowed to work remotely during his one-week business trip to Hell.
Great. You see that the creationist seems to believe that he is very open-minded. He allows not only one week for the Creation. He generously allows the two-week theory, too! That's a proof of his immense creativity and open-mindedness. He may even get in trouble for his tolerance towards this huge, one-week deviation from Genesis. But you know, to understand Darwin's theory, one has to adjust his opinions about more fundamental questions than whether it took one week or two weeks to create the species. Exactly analogously, to understand quantum mechanics and the phenomena for which quantum mechanics is needed, a believer in local realism has to change something more fundamental than some questions about the detailed character of the classical degrees of freedom described by his favorite local realist theory. He has to abandon realism as such.
People like Andrei are just absolutely incapable of understanding that both of their "options" are wrong. It seems that some brain defect prevents them from following the simple proposition "the correct laws of physics disagree with local realism" even at the level of the human language. Something collapses about their brains whenever you're assuming that they may finally learn the essence and they are just absolutely screwed. It seems that they can't even understand that you are saying "something is wrong" when you are saying it. They must misunderstand the very meaning of the words "no" and "not" because whenever you say "no" or "not", they behave just like if you haven't said anything. Not even an infinitesimal amount of their effort to understand the correct theory – or what you are saying about it – has ever been observed. They simply look brain-dead whenever the discussion gets to the level of quantum mechanics. Andrei's inability to understand that both candidate hypotheses in a list of two hypotheses may be incorrect was what made me pick the subtitle of this blog post.
(Andrei wrote numerous other comments making it clear that he probably doesn't have a clue about physics, even classical physics, so maybe these frustrating exchanges aren't a good example showing some people's specific problems with quantum mechanics.)
Alford, Bricmont, Andrei, ... and the list goes on and on and on. Today in the morning, someone sent me a link to a preprint by authors named Frauchiger and Renner in Switzerland. They claim to have proven that there have to be many worlds. There can't exist a theory that produces the same results as quantum mechanics, that is logically consistent, and that contains one world. Well, quantum mechanics itself is a clear counterexample. It agrees with itself, it is logically consistent, and it describes one world. How could these people have overlooked this counterexample? ;-)
You may try to read the paper and try to find the argument. I couldn't succeed and I don't recommend you to waste your time. It starts by a story.
A girl was doing experiments in the past. Eve is grinding meat. ("Eva mele maso" is one of the first sentences Czech children learned to read and write.) She measured a spin. She wrote notes. She was careful. She was not too careful. She forgot to write something. She forgot the result of her measurement. Someone told her.And so on. You read this stuff for several pages and expect the important argument about physics to suddenly emerge out of this sequence of simple sentences appropriate for retarded kids in the kindergarten. But when you start to lose the patience, the presentation stops and they tell you it's right that you haven't learned anything.
So you may triple your patience and keep on reading. And you learn things like:
Physics is fun. But one may also talk. One can say stories. Stories may say something about physics. The stories may be incomplete. Stories may have a plot. A story may be written by a writer. He may be helped by a script writer. A script is a story as written by the script writer. Here is a story about a cat and a cake. The cat may eat the cake. Here are the four pictures. A cat is frowning. The cake is not there. An arrow points up. A cat is smiling. The cake is there. An arrow points down. There is a clock above the cake. There is a clock above the missing cake. The happy cat may be in a different world than the sad cat.You keep on reading this additional irrelevant stuff to learn about the great idea that is supposed to change the foundations of quantum physics discovered by Heisenberg and pals but I think that at some point, you will lose your patience, too.
There just can't possibly be anything intelligent for a physicist in the paper. The authors clearly have the IQ of a pumpkin. They are not capable of finding anything interesting in science. They are probably even incapable of understanding that Heisenberg's discovery of quantum mechanics was "slightly more intellectually demanding" than to draw a picture of a cat and say a childish story about it. They must believe that if they don't understand quantum mechanics in terms of pictures of cats and cakes, it must be due to their superior intelligence relatively to Heisenberg. They are barely capable of understanding words such as "story" and a "plot" but they don't see why this is not an adequate background for doing physics. You are (and I am) an idiot because you tried (and I tried) to read several pages of this preprint although it was pretty clear that nothing valuable would have come out of it.
At the end, it's very likely – and the abstract and conclusions strengthen this belief – that their "argument" against quantum mechanics is that "the differences between two observers' description of a situation imply an inconsistency". But they don't. They only imply an inconsistency between how Nature works and how anti-quantum zealots believe that Nature "should" work. It's a basic point of QM that each observer has his own measurements and the "statements about Nature" that result from them. What must be consistent and what is consistent about QM is the axiomatic system of propositions that one observer does. But two observers' systems can't be "reconciled" into one shared system. If that were possible, QM wouldn't have to talk (and physicists wouldn't talk) about observers at all; QM would be replaced by a classical theory i.e. a theory not dependent on any observers. Well, whenever classical physics is a good approximation for the statements, the perspectives of the observers can be (approximately) reconciled into one. In general, when the quantum effects are strong, it's not possible. By definition, QM does describe Nature from a perspective that does depend on the observer and this fact doesn't imply any internal inconsistency.
How is it possible that these people can't see that quantum mechanics is a counterexample to their claim? How can they overlook the theory that underlies modern physics? Don't they think that if the observer-dependence of QM implied an inconsistency, someone (Dirac? Feynman?) would have already noticed this serious problem? They seem to believe that the QM predictions for the experiments are right, so how the theory achieving this amazing result could be inconsistent? The basic knowledge from first 5 lectures of an undergraduate quantum mechanics course must be enough for a student with IQ above 80 to point out that quantum mechanics is a physical theory that has all the properties they claim to be incompatible. So what the problem can possibly be with these people?
All these people are irrational in a certain way I simply can't empathize with at all. It's some circuit in their skulls that always turns their brains off completely when the critical ideas are coming. An apparent problem is that they are not even capable of understanding the simple statement that
quantum mechanics is a theory of physics.It is incredible but it must be so. By now, I think that I have collected a sufficient amount of evidence that the anti-quantum zealots are this hopeless.
Look e.g at Page 2 of the Swiss paper:
Main result (informal version)If you at least realize that quantum mechanics – or quantum theory, whatever term you like – is a physical theory, you must see that the text above is utterly illogical. Why would you talk about physical theories \(T\) that are "compliant with quantum theory"? If theory \(T\) forbids exactly what quantum theory forbids, then \(T\) is pretty clearly exactly equivalent to quantum mechanics. It is quantum mechanics. If two theories are inequivalent, there has to exist some difference in their prediction. This difference may always be reformulated as a statement that a certain result is impossible according to quantum mechanics but possible according to the other theory.
There cannot exist a physical theory \(T\) that has all of the following properties:
(QT) Compliance with quantum theory: \(T\) forbids all measurement results that are forbidden by standard quantum theory (and this condition holds even if the measured system is large enough to contain itself an experimenter).
(SW), (SC) Single-world and Self-consistency
But there's something in these people's minds that totally prevents them from realizing that quantum mechanics – or "standard quantum theory" – is a theory at all. That it is a candidate hypothesis that may run and that may be a counterexample to many statements such as the bold statement in their paper. On one hand, they clearly realize that quantum mechanics implies something – in the quote above, they talk about things that are "forbidden by the standard quantum theory". But the fact that quantum mechanics implies things about physics isn't enough for them to realize that it is a physical theory.
These two confused Swiss people were two of the reasons why I picked the title. But they're not the only ones who indicate that anti-quantum zealots generally don't understand the very simple statement that quantum mechanics is a theory in physics. If you studied some of the exchanges on a Stack Exchange page that is (also) dedicated to the many worlds, you would find some extremely bizarre responses by a user named Dirk Bruere.
On the page, someone asked how the "many worlds interpretation" describes the split into two worlds. Obviously, no coherent answer exists. There is no consistent body of knowledge here. DeWitt, a key popularizer of Everett's interpretation, added certain things that Everett considered "bullšit" (Susskind has the original manuscript where Everett added this word). There's no agreement on whether the worlds split at all, when they split. More seriously, there can't be any justifiable and consistent set of rules of this kind – there can't be any consistent theory – that would answer such questions. It's just like trying to figure out the detailed acts of God during the second week of creation. They're detailed questions about a theory that is wrong at a much more fundamental level so it's meaningless to answer.
But I want to focus on this Dirk Bruere and his usage of the word "theory":
@LubošMotl It's [MWI is] not a theory, it's an interpretation.The word "interpretation" is surely popular in this anti-quantum movement (Heisenberg coined the phrase "Copenhagen interpretation" and he was sorry before his book was even released because he correctly predicted that anti-quantum zealots would abuse the phrase to claim that there can also be "other interpretations" which is just plain wrong) but if one wants a system of rules that answer detailed questions about phenomena that this set of explanations considers "physical", then sorry, but the set of rules and ideas that produce such answers must be called a theory. In quantum mechanics – the theory for which people like Heisenberg, Born, and others got their Nobel prizes – no world is splitting. So an explanation of the same phenomena where the worlds are splitting is clearly a different, competing theory. It can't be an interpretation of quantum mechanics i.e. of what Heisenberg et al. have done. What they have done doesn't require any interpretation by different people. They have explained their work themselves and it's basically unambiguous. And Heisenberg, Bohr, and others were specifically unambiguous about the conclusion that all the critical claims about quantum mechanics and all the work on the "other interpretations" were bullšit. What sort of extra "interpretation" does this statement need? Bullšit is some stinky brown object. If you really don't understand what it is and you need some extra "interpretations", it's necessary to push the bullšit to your throat. Strč prst skrz krk if you need to remove the bullšit again.
OK, so far, the difference between "interpretations" and "theories" may be just a matter of a preferred jargon assuming that the physical content of the competing theory is sufficiently compatible with the content of quantum mechanics. But Dirk Bruere's crusade against the very simple observation that "quantum mechanics is a theory" went much further:
@LubošMotl QM does not say anything beyond the mathematics. Perhaps you would care to comment on the Born Rule in this context – Dirk Bruere 2 days agoWow. Quantum mechanics doesn't say anything beyond the mathematics, we learn. One must be so incredibly stupid to write something of the sort that it simply leaves me breathless.
The reality is that all the beef, all the relevant content of "quantum mechanics" is about physics. That's why quantum mechanics is taught to physics students, not mathematics students. In mathematics, one studies linear vector spaces, linear operators, spectra, eigenvalues etc. but these mathematical insights don't "automatically" have any relationship with Nature.
This relationship with Nature only begins when these mathematical objects are connected with objects in Nature, with aspects of our experiments and observations. Only when this is done, the linear algebra from the previous paragraph becomes quantum mechanics, a theory in physics. For example, the complex probability amplitudes (coordinates of complex vectors etc.) have the interpretation due to Born. If you square their absolute value, you get the probability of the corresponding outcome of an experiment.
Quantum mechanics is not the mathematics itself but the set of statements that connect some mathematical objects with Nature and aspects of our observations! The "interpretation" of these mathematical objects (and this sentence is a rare example where the word "interpretation" is actually fine) is what quantum mechanics is all about! What quantum mechanics (or its heart) "is"? It's the set of rules, the universal postulates of quantum mechanics. If I copy a paragraph from my review of Bricmont's book:
As the founders have realized already in the mid 1920s, quantum mechanics – the correct theory (or framework) to describe Nature – is (and has to be) a set of rules to predict probabilities of observations by an observer from the (subjective) knowledge about the previous observations by the observer. It cannot be made independent of an observer. It cannot be "objectified". Observations always influence the observed system and this influence can't be removed. All observations are always associated with linear operators on a complex Hilbert space. Most pairs of operators refuse to commute with each other; the nonzero commutators are the source of the uncertainty principle and all of its consequences. A priori possible results of a measurement are always given by the spectrum of the observable (e.g. the eigenvalues of the operator). Every measurement always collapses the state vector (wave function) to an eigenstate of the measured operator corresponding to the measured eigenvalue, it always acts as a projection. The squared absolute values of the complex probability amplitudes always determine the probability that the measurement produces a corresponding eigenvalue (Born's rule). The probabilistic character of the predictions is fundamental, not due to "pseudo-random generators" (or hidden variables), and the unavoidable presence of probabilities strictly between 0 and 100 percent follows from the uncertainty principle. The evolution in time (or other transformations) are always given by unitary operators. The allowed states are always elements of a linear vector space – complex superpositions of vectors psi,phi are always as allowed as the vectors psi,phi themselves. The superposition of two states phi,psi always means "one *or* the other", not "and", as Bricmont and others incorrectly say all the time. Composite systems are always described by Hilbert spaces that are tensor products of Hilbert spaces for the parts (subsystems). Most of the vectors in such Hilbert spaces (of the composite system) are always entangled i.e. they cannot be written as tensor products of vectors from the two subsystems' spaces (only as a superposition of several terms of this type).This is the paragraph that should replace all the thousands of pages of books by anti-quantum zealots. It's not only more concise; it's also correct. One may pedagogically explain how it works. One may give lots of examples. One may give proofs why some competing or "too simple" alternative theories would fail to work. But (the core of) quantum mechanics is the paragraph above, not any detailed mathematical elaboration disconnected from the observations.
Note that all the rules quoted in the paragraph above have some connection to Nature, experiments, or observations. None of the sentences is "just about mathematics". If you have only understood the mathematics of the linear operators etc., you have understood zero percent of the physical theory known as quantum mechanics.
Without all this frustrating experience (exchanges with Dirk Bruere), it would sound unbelievable and implausible to me. But thanks to this experience, the lesson is clear. These people just can't possibly understand the difference between mathematics and physics. They can't understand the difference between the rules that Nature around us actually obeys; and arbitrary axioms that mathematicians invent and play with. They can't understand why some people major in mathematics and others in physics.
Bruere's comment "just the mathematics" may be perhaps interpreted as some extreme reformulation of "shut up and calculate" or physicists' opposition to philosophers. But at Bruere's level of the discourse, it is self-evident that he has thrown the whole baby out with the bath water. You can't eliminate the connections between the mathematical apparatus and the observations from a physical theory (or any theory in natural sciences). Some metaphysics or conceptual foundations is always needed. What physicists mean when they criticize philosophers is that just like the mathematical apparatus, these connections must be picked by physicists who are directed by the empirical evidence, rational arguments, and the scientific method in general – not by philosophers who follow the unscientific, philosophical method.
But it's still true that some connection with the observations and Nature always has to exist. If those things didn't exist in quantum mechanics, then quantum mechanics wouldn't belong to science at all. Needless to say, quantum mechanics describes all these connections that are needed to do science – predictions and explanations of past observations – very clearly and unambiguously. And quantum mechanics authoritatively says that statements that can't be formulated in the ways allowed by quantum mechanics are scientifically meaningless. If these anti-quantum zealots have ever heard a few lectures on quantum mechanics, they must have heard what the rules are. They are just trying to fool themselves (and everyone else) into thinking that these rules don't exist at all. That quantum mechanics is a tabula rasa with some linear algebra whose connection with the physical world is arbitrary, up to your taste.
Well, the connection is not arbitrary at all. There only exists one correct way to make all these connections between the mathematical objects on one side; and the aspects of our real-life observations on the other side. Quantum mechanics tells us what the right connections are; the right connections are what is called quantum mechanics, they are what quantum mechanics is all about.
In many cases, I disagree with other people but I can imagine what could lead them to their wrong thinking. Maybe, I was confused in a similar way at some moment. This gives me some empathy. But in some extreme situations, I just can't find any empathy at all. I have never had psychological problems with physics or quantum mechanics that could compare to the problems of those people. I understood these basic relationships and differences between mathematics and physics when I was 6. Then there's the general insight that physics flourishes on the sweet spot of a balance between "boring cold facts and mechanical work" on one side and "the hot world of ambitious and philosophical ideas" on the other side.
I saw that sweet spot well before I was 10, too. I just can't possibly understand how someone may be so confused about the very existence of physics as a science, like Dirk Bruere seems to be. On one hand, he is a fan of the writing of long meaningless philosophical tirades about the "many worlds" and similar stuff. On the other hand, he isn't able to see – or willing to admit – that quantum mechanics says something about our actual observations and the "physical interpretation" of the mathematical objects in it at all. He only sees either "just the pure mathematics" or the "unconstrained philosophical speculations" but nothing in between. But physics is in between. It's a science (with its method) that actually allowed us to be basically certain about many previously unanticipated and far-reaching, seemingly philosophical statements about Nature while maintaining a level of rigor that is comparable to the rigor in mathematics (which has no guaranteed connections to anything particular in Nature, by definition of mathematics).
Maybe these limitations are just several (many) people's idiosyncrasies. Maybe they prove that something is fundamentally wrong about the whole education system because none of these people could have possibly understood what it means to think rationally or scientifically, like a physicist.
Why the observer dependence isn't a contradiction again
I've said almost the same things many times but I always try to change the perspective in a complementary way. Imagine that an external observer A (e.g. "Wigner") observes a system that includes the observer B (Wigner's friend). In the middle, B believes that he made an observation. But A didn't know the result so he describes the system including B in terms of superpositions.
In classical physics, A could also be ignorant about the result gotten by B which only B feels and everyone would know there is no problem. What's the essence of the "potential problem" in quantum mechanics? The essence is that if B "really measures" some value of an observable \(L=\lambda_j\) in the middle, it means that he will calculate the probabilities of histories in terms of the summation of probabilities including well-defined "classical" states of \(L\). The "collapse induced by B" will change the predictions.
However, the description used by A is different. A sums the probability amplitudes for the intermediate state because no collapse to \(L=\lambda_j\) takes place from A's viewpoint. The difference between the calculation using probabilities and probability amplitudes are exactly the mixed terms, i.e. the quantum interference (see at least the last part of this text on EPR in positronium if you don't understand this statement).
By not intervening into the experiment, A is capable of a more accurate description of the whole situation than B. B, by its claiming that he "feels" a particular \(L=\lambda_j\), is basically neglecting the mixed terms (quantum interference). It's a good approximation if the different values \(L=\lambda_j\) decohere from each other accurately enough. When they don't, it's a bad approximation and the difference between the future predictions of A and B can always be blamed on B's sloppiness. He thinks that he feels some result of the measurement \(L=\lambda_j\) as if it were classical, but A knows that this isn't accurate. By keeping track of the possible quantum interference between the various options \(L=\lambda_i\), A will be able to predict accurate interference patterns that could be affected by this interference.
B won't be able to make these very fine predictions correctly but this may be blamed on his sloppiness – he basically assumes that some things are classical (and quantum interference is absent) even though they are not (and it is not). Any observation requires some degree of this sloppiness or dirtiness (although, in macroscopic measurements, the unavoidable errors are super-insanely tiny and unmeasurable in practice – and it may be even science-fiction practice in most cases). The work of experimenters "is" dirty, it "does" imply some unavoidable pollution of the system, and quantum mechanics makes the origin of this clear and it allows us to quantify it.
So every time we declare some observation to be a "classical fact", we are in principle introducing some small errors – we become unable to correctly describe some interference effects between the different \(L=\lambda_j\) basis vectors. On the other hand, this doesn't mean any "intrinsic finite inaccuracy" of quantum mechanics. By taking more decohered, and therefore more reliable, apparatuses, we may reduce this unavoidable error resulting from the measurements as accurately as we want. That's enough to claim that quantum mechanics may (and does) predict some results with an arbitrarily good accuracy.
This is clearly a totally consistent system of principles by which a theory, QM, describes Nature. All the complaints are either totally incoherent or equivalent to the complaint that QM is not equivalent to classical physics – where the observer doesn't play any important role. QM is not equivalent to classical physics but it doesn't mean that there's something inconsistent or incomplete about it.
The latest section of this blog post was also meant to answer a would-be deep question that Scott Aaronson asked in an interview with the John End-of-Science über-crackpot Horgan.
And Deutsch asked the question, which I’m sure resonates with you: how could one ever experimentally test the Many-Worlds picture?My section has addressed exactly this would-be deep Deutsch-Aaronson setup. The answer is that the experimenter who does all the experiments with the superior accuracy will be able to observe the effect of the quantum interference (i.e. mixed terms) between the different \(L=\lambda_j\) states of the B's brain. So this superior experiment will of course see that at the fundamental level, there's no collapse.
Here Deutsch had the following thought: suppose you could do a quantum-mechanical interference experiment on yourself. That is, rather than sending a photon or a buckyball or whatever through Slit A with some amplitude and through Slit B with some other amplitude, suppose you could do the same thing with your own brain. And suppose you could then cause the two parallel “branches” of your experience to come back together and interfere. In that case, it seems you could no longer describe your experience using the traditional Copenhagen interpretation, according to which “the buck stops”—the wave of amplitudes probabilistically collapses to a definite outcome—somewhere between the system you’re measuring and your own consciousness. For where could you put the “collapse” in this case? You can’t have Bohr and Heisenberg’s famous divide between “the observer” and the “the quantum system” if the observer is the quantum system!
Now, your brain is such a big, hot, wet object, with so many uncontrolled degrees of freedom coupled to the external environment, that even a hyper-advanced civilization of the far future might never be able to do the experiment I just described. But, OK, what if we could build an artificially-intelligent computer,...
The human who is an "object" in this experiment may describe the situation in terms of a collapse and this will lead him to make wrong predictions of the patterns resulting from the interference of different \(L=\lambda_j\) states of his brain. But it's also clear whom this error should be attributed to: it should be attributed to B who "believes that he has perceived a classical fact" although he really couldn't because the classical description was just an approximation, and someone else was (by assumption) able to measure its errors.
So this Deutsch-Aaronson would-be deep experiment is in no sense deeper or more mysterious than a small animal whose thinking is occasionally messed up by the quantum tunneling and that still trusts his eyes and reasoning. Well, the animal shouldn't trust its eyes and brain so much because quantum mechanics shows that certain aspects of its feelings are measurably non-classical (and the rate of errors in his reasoning is nonzero), so any interpretation of any variables as being "classical facts" leads to measurable errors. Someone else may always do the experimental work or its description more accurately than the small animal.
The actual conclusion from the Deutsch-Aaronson experiment if you performed it (and you really can, just choose any quantum variable \(L=\lambda_j\) in your brain that you "sort of" feel but there's still an important interference between the outcomes, there's nothing mysterious about it) is nothing else than the insight that "human eyes and nerves and brain cells aren't observing, transmitting, and processing the information as flawlessly as flawless classical computers would". A big deal? Not at all. When you assume that some cells or gadgets XY act perfectly according to the rules of classical physics but they don't (and for a finite-size brain, they demonstrably don't), it will lead to errors and the error may be easily attributed to the wrong assumption of classicality! All these errors may be minimized by using larger and more reliable apparatuses.
(Also note that I haven't mentioned "many worlds" anywhere in my answer because it is a totally ill-defined and stupid ideology that has nothing to do with the physical description of these phenomena, the description known as "quantum mechanics" and often insulted by the synonym "Copenhagen Interpretation".)