Saturday, August 03, 2019

Love for bureaucracy and non-solutions to non-existent problems with quantum mechanics

Mr Q wrote something that is said often – in numerous variations:
I think that a good question remains: What observable will be measured when you set a given experimental device? If we knew the answer, all questions regarding this theme become nonsensical.
He promises us to eliminate all doubts and problems if we help him to add some extra structure to quantum mechanics: We need a recipe that tells us "what is measured" if we insert the information about the "experimental device". This will surely bring an unprecedented clarity to everything.

But will it?

The more I hear such promises, the more I am aware of their similarity with comments by the lovers of bureaucracy and regulation. Some people have heard that there is "problem" so they just offer their "solutions" – without any checking that there is actually a problem to start with or, if a problem exists, whether their solution actually makes the situation any better.

In this short 2011 film, a babe (model Carolina Bang) needs some paperwork ("Form 036", if you understand the Spanish bureaucracy) to become self-employed but the parasitic male bureaucrat is giving her hard time, while comfortably drinking coffee paid by the taxpayer money. But he has underestimated the strength of his enemy! ;-)

Our legal systems are contaminated by millions of pages of mostly arbitrary, counterproductive, and ludicrous laws and regulations (I don't say that all laws and regulations are bad but most are) and new pages are being added every day. More specifically, tens of thousands if not hundreds of thousands of people semi-professionally fight against climate change, a non-existent problem whose non-existence they didn't bother to discover.

In the comment above, Mr Q wants to solve the "trouble with quantum mechanics" (sadly, a sequence of words first used not by a notorious crackpot, as you might guess, but by Steven Weinberg) by 1) forcing us to precisely define what an experimental device is, 2) demanding that the theory tells us what the device precisely measures.

Let's compare the new framework of Mr Q with the conventional quantum mechanics.

Proper quantum mechanics also has some input and output associated with its predictions. The input that you have to insert is the observables that you measured and their values – or, equivalently, the appropriate eigenstate of such an observable or observables (the initial state). And the output is the probability of the statement that when a measurement of an observable is made in the future, we will get a particular value. This probability is obtained by Born's rule – squaring the complex probability amplitude which is a matrix element of a unitary evolution operator between the initial state and the final state.

The postulates of proper quantum mechanics respect George W. Bush's "don't ask don't tell" policy concerning the precise method how the observable was chosen or measured. When it comes to the choice of the observable (of the Hermitian linear operator on the Hilbert space) that is being measured, either in the initial or the final state, i.e. to the Heisenberg choice, there is some role for the observer's free will.

Only observables whose eigenstates tend to decohere from each other may be measured. But decoherence is an approximate, emergent phenomenon that can't be defined precisely because of various reasons. For example, it depends on a slightly arbitrary separation of the physical system to the "system of interest" and "the environment". In practice, what the observer may observe is largely given. But in principle, there can't be a canonical rule that tells us what an observer or his apparatus "must" measure.

Now, a key yet trivial point: If the quantum mechanical predictions are supposed to be trustworthy, the information about the values of the observables in the initial state has to be trustworthy, too.

If you insert imprecise or invalid data about the initial state to quantum mechanics (either because you insert a slightly wrong or completely wrong observable that was known, or an imprecise or wrong value of the observable), the predictions about the future that quantum mechanics produces will probably be imprecise and invalid as well. It's clear, isn't it? Is it a problem? Of course it is a problem when someone makes wrong predictions. But it is not a problem with the theory – with quantum mechanics. It's a problem with the person or entity that tried to apply quantum mechanics!

The previous two sentences – the attribution of the problem – are the main observation that Mr Q and people who think in a similar way neglect. When they see something like a possible problem, they automatically blame the theory – or the "system", if we think about some analogous cases in politics – and they try to fix this "systemic problem". But the problem isn't systemic at all. It's just a problem with the application of the theory, a kind of a problem that can't be avoided in general. People who try to use a theory may always suck. So even the best theory in the world can land in the hands of people who don't use it properly or who can't apply it.

This is also why Mr Q's proposed solution would solve nothing at all. Again: Proper quantum mechanics works as follows:
Find the observables and their values describing the initial state; calculate the values of observables in the final state.
Mr Q wants to make the procedure more complex:
Find the standardized precise description of an apparatus used to determine the initial state. Calculate which observable is determined by the initial state. Translate some data from the measurement to the standardized language describing the possible results obtained by the apparatus. Do a calculation of the final state from the initial state. Do a similar conversion from the final apparatus' measurement to the final observables etc.
Great. What problem has been solved here? What Mr Q has actually done was to add four absolutely useless layers of bureaucracy that probably can't be overcome in general. Mr Q is just like the Spanish bureaucrat who drinks his coffee and invents a whole sequence of useless impenetrable obstacles to give the blonde a hard time! But this freeloader was hired and encouraged to feel important and useful even though he is not.

What problem could be solved by these extra layers of bureaucracy? As I mentioned, a problem could be that the initial state that you insert into ordinary quantum mechanics could be a wrong one. The person who measured it could have been stupid or dishonest. The apparatus that was used could have been broken, dysfunctional, or hacked by a foreign government or by a Big Tech company. Great. If it is so, does Mr Q's solution actually solve this problem?

Not at all. Instead, it acts four extra interfaces that are vulnerable towards hacking. Instead of asking "is the information about the initial values of observables correct?", you also need to ask questions like: 1) was the description of the apparatus correct? 2) was it applied correctly? 3) wasn't the person who accepted the documentation about the device incompetent? 4) wasn't he bribed? And about 4 more questions about the final state and more.

The solution by Mr Q solves absolutely nothing because it is all about the denial of the fact that for things to work, the people and devices that are used must work honestly, correctly, and accurately.

People like Mr Q – sorry that I mention his nickname all the time, there are really millions of people with this exact thinking – simply think that the ability of the people to submit invalid or deceitful inputs (where the inaccuracy may be blamed on some "free will") may be "cured" by adding some extra layers of formalities. But it just cannot! Whatever is the number of extra layers you insert, it will still be true that the steps needed to create predictions may be done incorrectly. This isn't a problem with the theory. It's a problem with its users – and it just can't be "solved" in general.

So why don't you just leave quantum mechanics simple and as it was understood in Copenhagen? One inserts some information about the initial state; and calculates the probabilities of properties of the final state. That's it. Everything else that would be demanded in general is useless baggage.

It doesn't mean that everything else must be useless baggage in every situation. In particular, the laws of physics – especially quantum mechanics – may be also used to study how a particular device works. A particular device converts some observable to another, communicates the information in some way, creates an entangled state with a pointer which is the "external" description how the measurement device actually works, and so on.

But none of these things – the physical description of the inner workings of a gadget – should be considered a part of the foundations or fundamental laws of physics. There is nothing "fundamental" about them at all. You don't gain any fundamental knowledge by describing the measurement as a sequence of a larger number of steps involving the observer and his apparatus or its parts: the information is just being transmitted in a larger number of steps but you still need some trustworthy information about the initial state to be inserted.

You may describe how a voltmeter works just like you may describe how a superconductor or an elephant works. There is really no difference between the two or three. When you discuss apparatuses as "observed objects", they're objects like any other objects that quantum mechanics is capable of describing. These are just examples of an application of quantum mechanics; they are in no way something that should be added to the basic rules of quantum mechanics! And any effort to add such extra useless baggage is analogous to the pathological addition of useless laws and regulations by the omnipresent rogue lawmakers.

Quantum mechanics has really clean, minimalist rules. And all the efforts to make them more formal are just misguided non-solutions to non-existent problems. I think that every person with a good physics intuition understands why such bureaucratic efforts are counterproductive. In particular, objects in Nature – including things that we call devices – are naturally and sort of continuously connected with their environment and their states are continuously connected with other states of the same or similar objects. There is no natural way to precisely separate them from the environment or digitize or discretize the information about their state or design. So any proposal for such a digitization is clearly making the theory less natural, less fundamentally correct, and harder to be applied. Any transformation of reality into some discrete, paperwork-like information is clearly at most an unnatural, approximate operation – if it is possible at all. There is no room for these operations in the fundamental laws of physics.

Such proposals are driven not by the desire to understand how Nature works (the desire to adjust their views and theories so that they increasingly precisely and universally agree with the observations) – but by the pathologically obsessive love for bureaucracy and the tendency to unjustifiably give a hard time to productive people and their wonderfully powerful theories. They want theories – and people – to be squeezed into their stupid and narrow-minded bureaucratic template and if they can't be, these bureaucracyphiles become combative. But Nature (in the case of physical laws) and freedom-loving people (in the case of the societies) won't give up and they will ultimately defeat and/or destroy the bureaucrats.

No comments:

Post a Comment