China has taken over Science

T. Greer argues China is the future of science:

The scientific publisher Nature makes a similar catalog on a slightly more granular level, looking at specific fields of science. According to Nature’s most recent rankings, 18 of the top 25 most productive research institutes in the physical sciences, 19 of the top 20 in geosciences, and a full 25 out of 25 in chemistry are Chinese. Only in the biosciences do American scientists still have a lead—but even on that list three of the top ten are Chinese.6

The kicker is, none of that was true even just a decade ago.

The QWERTY Keyboard was not Designed to be Slow

I need to watch Youtube with an AI LLM to fact check it.

Physicist Brian Keating posted a cosmology interview, and started making a false analogy about keyboard. My AI advises:

While Brian Keating is an astrophysicist and not a historian of technology, he is repeating a very common popular myth. The claim that the QWERTY layout was designed specifically to "slow down" typists is technically incorrect, though it is based on a grain of truth regarding mechanical limitations.

According to historical research into the origins of the typewriter:

The "Jamming" Problem: In early mechanical typewriters, if a typist hit two keys that were physically close to each other in rapid succession, the "typebars" (the hammers that hit the paper) would collide and jam.

The Real Purpose of QWERTY: The layout was designed to separate frequently used letter pairs (like "TH" or "ST") so their physical hammers were on opposite sides of the machine. This allowed the hammers to fall back into place before the next one rose, actually allowing the user to type faster without the machine seizing up.

The "Slow Down" Misconception: The myth likely arose because the layout is not "ergonomically optimal" for modern electronic keyboards. Since we no longer have mechanical hammers to worry about, the QWERTY layout seems unnecessarily difficult compared to alternatives like the Dvorak Simplified Keyboard.

In the video, Keating uses this as an analogy for "technological lock-in", arguing that even if a system is no longer the most efficient, it becomes so widespread that it is nearly impossible to replace—much like he fears is happening with current AI architectures.

The truth is that the qwerty keyboard is not significantly slower than the Dvorak. It just reduces jamming on a mechanical typewriter.

I am wary when some expert tries to make a point using some ridiculous myth that was debunked decades ago. He could have just googled it, and seen that he was wrong.

In the video, Keating shares another classic bit of "technological lock-in" lore, claiming that modern railroad gauges (the distance between the tracks) can be traced back to the width of a Roman chariot, which was designed to fit two horses' butts.

While it’s a great story for a podcast, historians and engineers generally categorize this as a "spurious correlation"—it's mostly a myth, though it has a tiny thread of logic.

The Fact vs. The Fiction

• The Claim: Roman chariots created ruts in European roads. Wagons were built to fit those ruts so they wouldn't break their wheels. Early tramways used those same wagon dimensions, and English trains followed suit, eventually leading to the Standard Gauge (4 feet 8.5 inches) used in the U.S. and much of Europe.

• The Reality: While the standard gauge is roughly the width of an ancient wagon, the specific measurement of 4' 8.5" was actually popularized by George Stephenson, the "Father of Railways." He chose it because it was the width of the existing pre-steam coal tramways in North England where he worked.

• The "Horse's Butt" Part: There is no evidence that Romans measured horses' rear ends to build chariots. Chariots were built to be wide enough to hold two people and stable enough not to tip over. The fact that two horses fit in front of that width is a matter of practical spacing, not a design blueprint that dictated the Space Shuttle's dimensions (another common layer of this myth).

Why Keating Uses It

Keating’s broader point at isn't necessarily about Roman history; he's using it to explain Path Dependence. He argues that once a standard is set—whether it's QWERTY keyboards, railroad widths, or GPU-based AI architectures—it becomes "locked in" because the cost of changing the entire infrastructure is too high, even if a better way exists.

He is probably also wrong with his predictions about AI architectures. He also compares AI to a cockroach, and I think his point is that our AIs could suffer a technological lock-in at a sub-cockroach lever.

I assume that Keating is more accurate when he talks about cosmology experiments. But he says this:

0:30 If you said there's one galaxy, you're stupid. If you said there's one planet, you're stupid. If you said there's one. So why say there's one universe?

I am sticking to one universe. Maybe I am stupid.

WSJ Hypes Quantum Computers

new WSJ article has a glowing account of the future of quantum computing:

Microsoft, International Business Machines, Google and a host of other tech companies are in a race to disrupt the nature of computing.

Collectively in the U.S., these companies have spent hundreds of millions in the past few years to develop a new type of computer—known as a quantum computer—that leverages the principles of quantum physics to solve problems far beyond the capabilities of today’s best supercomputers.

And the companies say they could do it in the next two to five years.

When this point is reached, some problems that would take a traditional computer more than trillions of years to solve could take a quantum computer mere minutes, changing business as usual for industries involved with financial trading, shipping logistics, pharmaceuticals, scientific discovery, data encryption, insurance, internet delivery and more.

No, none of this is going to happen in five years.

It gives the usual explanation, even though Dr. Quantum Supremacy hates it:

A quantum computer, however — because of entangled qubits’ ability to calculate many probabilities at once — can evaluate all options simultaneously.

He says this is misleading, because if that were true, then the quantum computer could solve NP complete problems like the traveling salesman problem.

The article is paywalled. It has some nice graphics, if you can get it.

Dr. Bee Pushes Spooky Action and Superdeterminism

Sabine Hossenfelder posted a new video on causality.

I want to talk today about an issue that bothers me a lot, it’s that most physicists believe the speed of light is an ultimate, absolute, and impossible to overcome limit. Indeed, I think it is THE biggest mistake that physicists are making, that our entire species is making. ... 11:45 I am making this video to save mankind from its biggest mistake

Causality is the idea that if events A and B are separated, then A can only affect B by some sort of signal or interaction being transmitted from A to B. Relativity realizes this by putting a non-euclidean geometry on spacetime, so that everything propagates at the speed of light or slower.

She goes on the explain how she believes in the arrow of time, but that faster-than-light signals ought to be possible. She says we should not be constrained by relativity, because we know it is wrong.

2:14 I don’t know any physicist who thinks that Einstein’s theories are ultimately correct because they don’t include quantum effects, we are missing a theory of quantum gravity. So why should the limitation of the non-quantum theory continue to hold when we know it ultimately isn’t correct?

No, we are very sure the non-euclidean local structure of spacetime is correct. Her hypothetical quantum gravity only applies to the first second of the big bang and the center of a black hole, both far outside any scientific observation.

Next she talks about entanglement. Just where you might expect her to say that entanglement proves spooky action at a distance, she correctly explains that it does not.

This is the supposed non-locality. You make 9:51 a measurement here and you infer information about the properties of something elsewhere. This is nonlocal in some sense, but there is no information actually traveling anywhere, it’s just that you learned something about what is going on elsewhere. And this is the only way in which quantum physics is non-local. ... Because the measurement doesn’t affect the probability of either of the measurement outcomes.

This point trips up many physicists. The measurement does affect your knowledge, and hence how you calculate your expectations, but does not violate the causality being limited by the speed of light.

Then she goes off the rails.

11:53 It’s like this. Suppose that you think the collapse of the wavefunction is NOT a physical process. That actually the particles did have some specific state before you measured them, you just don’t know which. This is what’s called a “hidden variables theory”. In the simplest case the hidden variable just directly encodes which side is up. And the only way to make a hidden variables theory compatible with Einstein’s theory is by using what has been called “superdeterminism”.

Superdeterminism is the only local explanation for all the observations of quantum physics. This is why I think it’s trivially correct, and I am waiting for the rest of physicists to wake up to that. Which means that now we are properly crossing over into the range where most physicists would try to disagree with me, meaning they would be wrong. Physicists have mostly dismissed superdeterminism because of social reinforcement. The majority doesn’t even know how it works, they never thought about it, they just heard someone else say it’s wrong and they decided to believe this.

When she assumes that the particles had some "specific state", she means that the state could be completely described by hidden variables. Such hidden variables contradict quantum mechanics, and that is why no one believes in them.

When she says superdeterminism is the only local explanation for quantum observations, what she really means is that it is the only local hidden variable explanations. That is just a fancy way of saying that quantum mechanics is not a classical theory. Everyone has agreed to that for a century.

If you did believe in a hidden variable theory, as well as local causality, then she is correct that you are led to superdeterminism. Physicists dismiss it because it is philosophically absurd, anti-science, and no one knows how it would work. I have criticized it several times, such as here.

Another new video lists 10 Theories That Suggest Our Universe Is Not Real. It describes simulations and Boltzman brains, but superdeterminism is not respectable enough to make the list.

Quantum Crypto Wins Turing Award

Quanta magazine announces:

Bennett and Brassard have now been named the winners of the A.M. Turing Award, one of the highest honors in computing, for “their essential role in establishing the foundations of quantum information science and transforming secure communication and computing.” The award comes with a $1 million prize.

No, they did not transform secure communications. Their work has no practical applications.

Scott Aaronson says:

This is the first-ever Turing Award specifically for quantum stuff (though previous Turing Award winners, including Andy Yao, Leslie Valiant, and Avi Wigderson, have had quantum among their interests).

As a practical proposal, BB84 is already technologically feasible but has struggled to find an economic niche, in a world where conventional public-key encryption already solves much the same problem using only the standard Internet—and where, even after scalable quantum computers become able to break many of our current encryption schemes, post-quantum encryption (again running on the standard Internet) stands ready to replace those schemes. Nevertheless, as an idea, BB84 has already been transformative, playing a central role in the birth of quantum information science itself.

The BB84 protocol suffers several technical flaws.

The most important thing in secure communications is authentication. This is currently done with digital signatures and certificates, using RSA or ECDSA. This underlies everything. It is the most important part of ubiquitous protocols like https and ssh. But the quantum crypto cannot do it. That makes it useless for anything serious.

The next fatal defect is that it depends on hardware quirks. You have to have analog equipment that may or may not have the required precision, and may have analog vulnerabilities. This makes is subject to hardware attacks.

This makes is vastly inferior to the math-based crypto methods, because the math is not subject to hardware attacks. If a digital crypto device outputs bits that do not have the desired voltage or frequency, no info is leaked.

Another flaw is that the main BB84 security guarantee is that an attacker can probably be detected, so that transmission can be terminated. This has no value. In today's internet, systems get attacked all the time, and no one wants to shut down a communication because it is being attacked. Conventional cryptosystems are designed to be immune to such attacks.

Another flaw is that the internet is run on millions of routers. Using quantum crypto requires that all those routers be quantum computer routers. The quantum router has not even been invented, and even if it is possible, it will never be economical or have the necessary throughput.

All this has been known for decades, and that is why no one uses it, except for a few research demo projects.

New Survey Article on Many-Worlds

I regularly bash many-worlds theory as unscientific and nonsensical, but it is useful to see how it is defended by its proponents. Here is a new encyclopedia article on the subject.

The Everett or Many Worlds interpretation is claimed to be the only realist interpretation that can recover the empirical success of quantum theory in its relativistic and non-relativistic variants, its advocates suggest that it does so without any additions to the physics.

Yes, it claims that, but it has never recovered any empirical success. None.

Probability within Everettian theories is strongly contested and it’s un- clear whether the many distinct resolutions in the literature are mutually incompatible and thus undermine one another ...

EQM [Everettian Quantum Mechanics, ie many-worlds theory] is taken to face two major problems: the preferred basis problem and the probability problem. The preferred basis problem concerns how the universal wavefunction is decomposed, leading to different classical-like branches. ...

The second major problem is the probability problem, which asks how probability can make sense in a deterministic theory where all possible outcomes occur.

Yes, those are two big problems. They cannot define the branches or the probabilities.

The article also discusses other approaches like Bohmian mechanics, and how they do not work either. The obvious inference is that the textbook Copenhagen interpretation of 1930 works better than any of these more modern ideas.

Has Quantum Supremacy been Achieved?

Dominik Hangleiter writes in a new paper:

Recently, I gave a couple of perspective talks on quantum advantage, one at the annual retreat of the CIQC and one at a recent KITP programme. I started off by polling the audience on who believed quantum advantage had been achieved. Just this one, simple question.

The audience was mostly experimental and theoretical physicists with a few CS theory folks sprinkled in. I was sure that these audiences would be overwhelmingly convinced of the successful demonstration of quantum advantage. After all, more than half a decade has passed since the first experimental claim [AAB+19] of “quantum supremacy” as John Preskill called the idea “to perform tasks with controlled quantum systems going beyond what can be achieved with ordinary digital computers” ...

I could not have been more wrong: In both talks, less than half of the people in the audience thought that quantum advantage had been achieved.

After several pages describing the experiments, he concludes:

I hope that I could convince you that quantum advantage has been achieved. There are some open loopholes, but if you are happy with physics-level experimental evidence, then you should be convinced that the RCS experiments of the past years have demonstrated quantum advantage.

No. The whole point of quantum supremacy, aka advantage, is to do an experiment that convincingly demonstrates that super-Turing computers are possible. If most of the experts have not been convinced, then the principle has not be demonstrated.

I am a skeptic, and will be hard to convince. But they have not even convinced the experts who work in the field.

But read the paper, and make up your own mind.

Scott Aaronson seems to be still on the fence. His recent postings have been either about how he hates Pres. Trump, or how he agrees with Trump's pro-Israel foreign policy.

Science Papers are now mainly read by AI LLMs

Alexander Kustov, with Claude AI assistance, writes Academics Need to Wake Up on AI and Part II:

1. AI can already do social science research better than most professors.

2. The academic paper is a dead format walking.

3. The commercial journal system may not survive this.

4. Academics hold AI to absurd double standards.

In particular: Most papers are already mostly read by AI, not humans. Your primary audience is increasingly LLMs.

Physics Lifetime: 1820 to 1970

Popular tweet:

Many people don't understand just how brutal diminishing returns in theoretical physics were.

Physics barely existed before 1820. After 1970, there was essentially nothing left to discover.

In 1819 there were probably less than 100 full-time paid physicists in the whole world.

By 2026 there are probably about a million physicists across academia and industry, and that number was already huge in the 1970s when physics sort of "ended" with QCD and electroweak unification.

A small, brave band of gentlemen-scholars and amateurs worked out the most important parts of physical law in the 1800s. People doing it as a hobby!

Today, vast armies of professionals equipped with supercomputers toil away in the quantum gravity dungeon, unable to make progress.

Diminishing returns are brutal.

my point is that the low hanging fruits of physics were all picked in a brief window from about 1820 to 1970.

Before that, it was difficult to get anything done at all, there was no funding

Quantum Computing and National Security

The Wash. Times reports:

One of the most advanced technologies intersecting with U.S. national security today is quantum computing. Quantum has arrived in 2026, and how it ultimately gets implemented will impact America’s standing in great power geopolitical competition, especially with U.S. adversaries. National Security Editor Guy Taylor sits down with industry leaders at “Qubits26 Quantum Realized,” a conference hosted by D-Wave Quantum, for a wide-ranging discussion on what quantum computing is and how it stands to change the world.

The rest of the article is paywalled, and I do not need to read it. It is all a scam. Quantum computers will not affect national security.

Good Wlll Hunting

SciAm reprints an article on Why mathematicians hate Good Will Hunting.

With the award ceremony for the Oscars this month, many people are thinking back on past winners—including Good Will Hunting. It’s worth taking a closer look at the blackboard in a film that, in 1997, took nine nominations and won for both original screenplay and actor in a supporting role. ...

But I still think the filmmakers chose this particular math problem poorly, even for a Hollywood film.

No, that is not why mathematicians hate the movie.

The hero is a fictional exceptionally talented math prodigy. Supposedly he enjoys math so much that he gets a job as an MIT janitor, and eavesdrop on the research there.

But he never spends any of his free time doing math. Instead he goes drinking with his non-math buddies, and getting into fights. In the end, he decides that math is for losers, and he abandons a wonderful math opportunity in favor of chasing a girlfriend.

Nobody gets that good at math unless he enjoys it very much. The movie fails to portray that at all.