Watering down the Definition of Quantum Supremacy

Scott Aaronson admits:

If your point is just that quantum supremacy claims depend for their credibility on people trying hard to refute them and failing, then I vehemently agree! So you and I should both be glad that this is exactly what’s happening right now.

Regarding your question: no, I would not bet that Google’s Sycamore chip can’t be spoofed by a classical computer in 10 years — or right now, for that matter!

In other words, this Google quantum supremacy means nothing except that Google has made a machine that no one has bothered to simulate yet. And he has no confidence that it cannot be done.

Even if it cannot be done, it is still subject to the teapot supremacy argument.

I thought that quantum supremacy meant demonstrating decisively a computational ability superior to ordinary Turing machine computers. Aaronson seems to think that it just means doing something complicated that critics have not yet matched.

Here is how John Preskill originally used the term:

we hope to hasten the day when well controlled quantum systems can perform tasks surpassing what can be done in the classical world. One way to achieve such "quantum supremacy" would be to run an algorithm on a quantum computer which solves a problem with a super-polynomial speedup relative to classical computers ...

We therefore hope to hasten the onset of the era ofquantum supremacy, when wewill be able to perform tasks with controlled quantum systems going beyond whatcan be achieved with ordinary digital computers. ...

I have emphasized the goal of quantum supremacy (super-classical behavior of con-trollable quantum systems) as the driving force behind the quest for a quantumcomputer ...

I have emphasized the goal of quantum supremacy (super-classical behavior of con-trollable quantum systems) as the driving force behind the quest for a quantumcomputer, and the idea of quantum error correction as the basis for our hope thatscalable quantum computing will be achievable.

Aaronson is like Dr. Fauci admitting that the coronavirus may have come from the Wuhan lab. He is preparing for research that could be profoundly embarrassing to the entire field.

Update: Gil Kalai responds:

Hi Scott, regarding the analogy between Google’s Sycamore and IBM’s Deep Blue, here are three differences
1. In the Sycamore case, the researchers largely invented a new game to play
2. In the Sycamore case, the researchers themselves also played the part of Kasparov
3. In the Sycamore case, a huge leap compared to previous efforts is claimed

Aaronson previously wrote:

As I’ve said in dozens of my talks, the application of QC that brings me the most joy is refuting Gil Kalai, Leonid Levin, and all the others who said that quantum speedups were impossible in our world.

What makes quantum mechanics mysterious?

Everybody says quantum mechanics is mysterious, but what is the mystery?

Entanglement. A lot of smart physicists have made a big deal out of this, but it is really not mysterious unless combined with some other effect. I recently posted about this.

Double-slit experiment. We get a similar interference pattern for any wave phenomenon. We would expect it for light, even with no QM.

Uncertainty principle. Not too surprising, once you assume a wave nature of matter.

Superposition. Like the half-dead half-alive cat. Just a useful fiction.

Many worlds. This is just a fantasy. You can have the same fantasy about classical theories, if you wish.

Identical particles. It is strange that all electrons are the same. It makes the world very simple, compared to the alternatives.

Discrete energy levels. This is truly one of the nice features of QM, but not usually described as mysterious.

Probabilities. Scott Aaronson says the essence is negative probabilities. I say there is no such thing. QM has regular probabilities, and so does every other theory.

Nonlocality. QM has no true nonlocality, in the sense doing something in one place causing action at a distance elsewhere.

Canceling infinities. These are mostly artifacts of extreme assumptions, such as having mass and charge concentrated at a point, having infinite density.

Linearity. Some other theories are linear, such as Maxwells equations.

Super-Turing computation. This would be interesting, if proved. Some say it has been. I don't believe it.

Lack of local hidden variables. Bell made a big deal out of this, but nobody expected local hidden variable anyway.

No counterfactual definiteness. This is closely related to the lack of hidden variables.

So what makes quantum mechanics different from classical mechanics? I say that it is the non-commuting observables and positivist outlook. The above stuff is just not that mysterious.

Google promises to make a Logical Qubit

In my skepticism about quantum computing, I have occasionally remarked that all the claims about research machines with dozens of qubits is exaggerated, because they still have not made a true qubit yet. Here is an explanation of that.

Google announces at its annual I/O conference:

Within the decade, Google aims to build a useful, error-corrected quantum computer. ...

To begin our journey, today we’re unveiling our new Quantum AI campus in Santa Barbara, California. ...

[wild futuristic hype snipped]

To reach this goal, we’re on a journey to build 1,000,000 physical qubits that work in concert inside a room-sized error-corrected quantum computer. That’s a big leap from today’s modestly-sized systems of fewer than 100 qubits.

To get there, we must build the world’s first “quantum transistor” — two error-corrected “logical qubits” performing quantum operations together — and then figure out how to tile hundreds to thousands of them to form the error-corrected quantum computer. That will take years.

To get there, we need to show we can encode one logical qubit — with 1,000 physical qubits.

Got that? Google claims that it has achieved quantum supremacy, but it is still a long way from building a system with even one logical qubit.

Rethinking entanglement of a single particle

Dr. Bee has caused me to rethink entanglement, and reader Ajit sends a paper on Entanglement isn't just for spin

Quantum entanglement occurs not just in discrete systems such as spins, but also in the spatial wave functions of systems with more than one degree of freedom.

It is sometimes said that Einstein discovered entanglement in 1935, and it was immediately recognized as the central defining feature of quantum mechanics. But as the above paper notes, the word was not in common use until about 1987, and did not find its way into textbooks until after that.

As the article explains, entanglement is not some peculiarity of tricky spin experiments. It is a property of all quantum systems.

Entanglement is explained as the thing that makes quantum mechanics nonlocal, and hence the essence of why the theory is non-classical and mysterious.

Paul Dirac one said:

Quantum-mechanically, an interference pattern occurs due to quantum interference of the wavefunction of a photon. The wavefunction of a single photon only interferes with itself. Different photons (for example from different atoms) do not interfere.

This is not an exact quote, but he said something similar.

This is a confusing statement, and I would not take it too literally. But in a similar spirit, I would say that a quantum particle can be entangled with itself.

Entanglement is often introduced by describing creation of a pair of particles with equal and opposite spins. But it is much more common. In any atom with several orbital electrons, those electrons are entangled. Nearby particles usually are. The case of the equal and opposite pair is interesting because that gives distant entanglement, but nearby entanglement occurs all the time.

Consider a stream of particles being fired into a double slit. Each particle is interfering with itself, and is entangled with itself. The interference results in the interference pattern on the screen.

The entanglement results in each particle hitting the screen exactly once. If you purely followed the probabilities, there are many places on the screen where the particle might hit. Those possibilities are entangled. If the particle is detected in one spot, it will not be detected in any other.

You cannot understand the experiment as localized probabilities in each spot of the screen.

Viewed this way, I am not sure the 2-particle entanglement story is any more mysterious than the 1-particle story. Maybe explanations of entanglement should just stick to the 1-particle story, as the essence of the matter.

Update: Reader Ajit suggests that I am confusing entanglement with superposition. Let me explain further. Consider the double-slit experiment with electrons being fired thru a double-slit to a screen, and the screen is divided into ten regions. Shortly before an electron hits the screen, there is an electron-possibility-thing that is about to hit each of the ten regions. Assuming locality, these electron-possibility-things cannot interact with each other. Each one causes an electron-screen detection event to be recorded, or disappears. These electron-possibility-things must be entangled, because each group of ten results in exactly one event, and the other nine disappear. There is a correlation that is hard to explain locally, as seeing what happens to one electron-possibility-thing tells you something about what will happen to the others. You might object that the double-slit phenomenon is observed classically with waves, and we don't call it entanglement. I say that when a single electron is fired, that electron is entangled with itself. The observed interference pattern is the result.

President Joe Biden is Politicizing Science

Lawrence Krauss has a WSJ article attacking Pres. Biden for politicizing science.

The New Scientific Method: Identity Politics
The National Academy of Sciences fights bias by explicitly introducing more of it.

Lubos Motl praises the article.

In particular there is now an aggressive affirmative action program at the National Academy of Sciences, where less competent women and Blacks are being appointed in order to meet diversity quotas.

Biden's White House Coronavirus Response Coordinator is a Democrat political hack named Jeffrey Zients. Donald Trump had an immunology expert in that job.

The authorities are still not telling us the truth about the virus. See this article by a NY Times science reporter on evidence it came from the Wuhan Institute of Virology.

Update: From a CDC official in a press conference, as reported in the NY Times:

DR. WALENSKY: … There’s increasing data that suggests that most of transmission is happening indoors rather than outdoors; less than 10 percent of documented transmission, in many studies, have occurred outdoors.

The paper goes on to explain that the true number is more like 0.1%. Yes, that is less than 10%, but appears to be a distortion attempting to justify outside mask requirements.

Quantum wavefunction is not everything

Reader Ajit argues that I am not following textbook quantum mechanics properly. He has posted Postulates of quantum mechanics, as stated by various authors.

Checking other versions of the postultes, I find:

The state of a system is completely described by a wavefunction

Associated with any particle moving in a conservative field of force is a wave function which determines everything that can be known about the system.

I wonder why this would be stated as a postulate. It is not used by the theory anywhere, and it is not true.

Sometimes it is stated for a single particle, but it cannot be true if the particle is entangled with another. Sometimes it is stated for scalar wave functions, but that cannot be true if the particle has spin.

You can correct those problems by introducing spinor-valued wave functions of several variables, but then you are still ignoring quantum fields and all sorts of other complexities.

Now you might say: Okay, but if use the whole Standard Model, or some bigger unified field that takes into account all possible interactions, and then we construct a wave function of the universe, then that would completely describe the state of the universe.

That would not be quantum mechanics. That would be some theorist's fantasy that has never been carried out.

Quantum mechanics is a theory that takes in some available info, and makes some predictions, but never achieves a complete description of the system. Nobody has any idea how any such complete description would ever be accomplished.

Take a simple example, the Schroedinger Cat. The wavefunction is a superposition of dead and alive states. Is it a complete description of the state of the system? No, of course not. The cat is either dead or alive. You can get a more complete description by opening the door and looking to see if the cat is dead. The wavefunction is most emphatically not giving a complete description.

I don't know why anyone would say that the wavefunction is a complete description of the system. Other physics theories do not start off with a postulate declaring some sort of god-like omniscient. It doesn't make sense to even say something like that.

And yet this postulate is prominent on various lists of postulates for quantum mechanics. I will have to do some further research to find out who is responsible for this silly idea.

This week's Dr. Bee video is on Einstein's spooky action at a distance. She says that the spookiness is the measurement update (ie, collapse of the wavefunction), not entanglement.

Believing that the wave function is a complete description necessarily causes these spooky concerns. Any observation affect distant parts of the wavefunction. If the wavefunction is a complete physical thing, then it is spooky.

Does Quantum AI have Free Will?

A new paper argues that a quantum computer could be conscious, and have free will.

Since I am skeptical that quantum computer will ever achieve quantum supremacy, you probably think that I dismiss this as nuts. Actually I don't.

Turing machines are deterministic, and do not have free will. But I believe humans do. The mechanism is not understood, and may involve quantum mechanics. So maybe a quantum computer can do that, even if cannot factor large numbers.

The London Guardian has a good essay on the arguments about free will. It says:

Harris argues that if we fully grasped the case against free will, it would be difficult to hate other people: how can you hate someone you don’t blame for their actions? Yet love would survive largely unscathed, ...

I personally can’t claim to find the case against free will ultimately persuasive; it’s just at odds with too much else that seems obviously true about life.

I agree with that last sentence. A lot of intellectuals reject free will, but in the process they also reject a lot of things that seem obviously true.

I do not agree with the love/hate analysis. If I believe that someone has no free will, and is merely a preprogrammed robot to do evil things, then sure, that is a good reason to hate him. He would be a sub-human evil nuisance. A puppet of the devil. As for love, try telling your wife that you only love her because the chemicals in your body have made that illusion. Some psychologists say that, and I don't think it helps.

The article says that philosophers have gotten death threats over such issues.

Jerry Coyne endorses most of the essay, but argues:

Contracausal free will is the bedrock of Abrahamic religions, which of course have many adherents.

No. Islam doesn't accept free will. Moslems are always talking about God's will being carried out, as if no one can do anything about it. Jews have mixed views. Catholics believe strongly in free will, and so do many Protestants, but some, such as Calvinists, do not.

A previous Guardian essay by historian Yuval Noah Harari said:

Unfortunately, “free will” isn’t a scientific reality. It is a myth inherited from Christian theology. Theologians developed the idea of “free will” to explain why God is right to punish sinners for their bad choices and reward saints for their good choices. If our choices aren’t made freely, why should God punish or reward us for them? ...

You cannot decide what desires you have. You don’t decide to be introvert or extrovert, easy-going or anxious, gay or straight. Humans make choices – but they are never independent choices. ...

But now the belief in “free will” suddenly becomes dangerous. If governments and corporations succeed in hacking the human animal, the easiest people to manipulate will be those who believe in free will.

He is gay Israeli Jewish atheist. Perhaps he is a slave to his programming, but others are not.

Theologians did not invent free will. The Gospels use phrases like "go and sin no more". This assumes that you can choose to sin, or not sin.

You can decide to be an introvert or extravert. Change is not easy, but people do it.

No, the easiest to manipulate are those who think that they are already slaves.

Coyne argues:

Free will skepticism (sometimes called “hard determinism”). As you must know, this is the view to which I adhere. Though it’s often called “determinism”, with the implication that the laws of physics have already determined the entire future of the universe, including what you will do, that’s not my view. There is, if quantum mechanics be right, a fundamental form of indeterminism that is unpredictable, like when a given atom in a radioactive compound will decay. It’s unclear to what extent this fundamental unpredictability affects our actions or their predictability, but I’m sure it’s played some role in evolution (via mutation) or in the Big Bang (as Sean Carroll tells me). Thus I prefer to use the term “naturalism” rather than “determinism.” But, at any rate, fundamental quantum unpredictability cannot give us free will, for it has nothing to do with either “will” or “freedom”.

I call this argument: Only God has Free Will.

Coyne is an atheist, but he seems to believe in some sort of Spinoza God. Humans have no freedom or free will. We are just puppets being controlled. God is not a predictable robot, and can make choices for us and the world. God even guides evolution of biological species by directing mutations.

The phrase "fundamental quantum unpredictability" means that the human observer can only predict probabilities. It always leaves open the possibility that someone with more info could make a better prediction. If Coyne wants to believe that it is some sort of God making all our choices for us, I guess that possibility is allowed.

For example, a quantum mechanics textbook might say that a uranium atom has a certain probability of radioactive decay in the next hour. And maybe that is all that can be said with the info available. But nowhere will it say that it is impossible to make a better prediction, if the state of the atom could be more precisely determined. As a practical matter, it is hopeless to get the wavefunctions of all the quarks in a uranium nucleus, but the point remains that better info might give a better prediction.

In my opinion, attributing all the decisions in the world to a Spinoza God is contrary to common sense and experience, and does not really solve anything. It is like a turtle argument that atheists like to mock. In fact, I worry about the mental health of anyone who believes that, as it is similar to schizophrenics who say that they are obeying voices in their heads.