Apple says LLMs cannot Reason

Since the Nobel Swedes have apparently decided that Physics is dead, and all the action is in Artificial Intelligence, maybe I will switch to posting about that.

Apple is desperately trying to catch up to its competitors in AI, so it is badmouthing the successes:

A new paper from Apple's artificial intelligence scientists has found that engines based on large language models, such as those from Meta and OpenAI, still lack basic reasoning skills.

The group has proposed a new benchmark, GSM-Symbolic, to help others measure the reasoning capabilities of various large language models (LLMs). Their initial testing reveals that slight changes in the wording of queries can result in significantly different answers, undermining the reliability of the models.

The group investigated the "fragility" of mathematical reasoning by adding contextual information to their queries that a human could understand, but which should not affect the fundamental mathematics of the solution. This resulted in varying answers, which shouldn't happen.

The paper uses this typical example of an LLM failure:

Oliver picks 44 kiwis on Friday. Then he picks 58 kiwis on Saturday. On Sunday, he picks double the number of kiwis he did on Friday, but five of them were a bit smaller than average. How many kiwis does Oliver have?

The paper complains that the LLMs subtract the 5 kiwis, even though the statement about them being smaller should be irrelevant.

No, this is a misguided criticism. If the LLM were strictly logical, it would refuse to answer the question as too vague and imprecise. The problem does not say how many kiwis Oliver started with, or whether he got any from other sources, or if he ate any. Without that info, no answer can be given.

The LLMs work by embedding the problem into a convex meaning space. However sloppy the problem is, it gets precise coordinates in the embedding space.

Whoever formulated this problem seemed to be saying that the 5 kiwis should not be counted. Why else is it phrased that way? Okay, it is phrased that way to be a trick question.

What would you want an LLM to do? To assign meaning in the most direct way, or try to interpret the problem as a trick question?

Apple seems to want to benchmark LLMs on trick questions. No thanks.

For another view, see Apple DROPS AI BOMBSHELL: LLMS CANNOT Reason. This video argues that Apple proved that the LLMs are worse that what people thinks, and not likely to get fixed soon.

Aaronson Posts his Manifesto

Dr. Quantum Supremacy Scott Aaronson has posts

In nearly twenty years of blogging, I’ve unfortunately felt more and more isolated and embattled. It now feels like anything I post earns severe blowback,

It appears that I am blocked from commenting there. The following was rejected, so I repost it here:

(7) Goldbach is not a good example because if it is independent of the axioms then there is no counterexample in N, and hence true in N. There are other arithmetical statements that are independent and we have no reason to say they are true or false. "how did we even get started talking about math?" By accepting what is provable. Mathematicians do not need to have an opinion, when there is no proof.

(15) "Everett interpretation ... to result from scientifically conservative choices." By this I assume you mean you can use the Schroedinger equation to predict probabilities, but then reject those probabilities and interpret them as world-splittings instead. No, this is not scientifically conservative. It makes the theory untestable, and introduces vast unobservable worlds.

Yes, I do think his math philosophy, and his many-worlds beliefs, are untenable. In particular, it is a nutty extreme view to say that many-worlds is scientifically conservative.

Sean M. Carroll says that same thing, and I have criticized him for it. He has explained his views in detail on his podcast, so I am sure he is wrong.

The essence of many-worlds is that it takes a scientific theory that predicts probabilities, denies that probabilities exists, and postulates that all possibilities happen in parallel worlds. It is so crazy that any scientist who advocates it should be disregarded on all scientific issues.

Aaronson tries to make his other positions sound moderate, but some are nutty. Item (1) exhibits Trump Derangement Syndrome. He thinks Trump is a threat to the Enlightenment!

He is entitled to his political views, of course, but nothing he says about Trump makes any sense. Some of his viwes seem more aligned with Trump than Harris, except for his Jewish brainwashing against Trump. I say Jewish because a lot of Jews show the same derangement, even though Trump's policies are likely to be much better for the Jews.

Harris and Walz are the most vacuous candidates I can remember. They have low IQ, have nothing to say about anything, and just babble nonsense when asked questions. Surely anyone smart enough to be a professor can see that they are morons. They just support them out of some sort of leftist tribal loyalty.

The Physics News is no New Physics

Physicists must be annoyed that the Nobel Physics prize was for computer science, with no application to Physics.

Here is a defense:

All this sounds nice for computer science, or for building language translators or self-driving cars, but is it Physics? I’d argue the answer is “yes.” These ideas of network properties, stability, and transitions didn’t spring up de novo, but emerged via classic problems in statistical mechanics, one of the fundamental, core areas of physics. Hopfield’s model was, in fact, a “spin glass” model, originally developed to describe magnetic materials and phase transitions.

They were inspired by biological models of the brain also, but today's neural nets are really designed to do what can be made efficient on consumer gaming chips, and do not have much to do with either physics or biology.

I wonder why these guys even take the calls from Sweden. They insisted on interviewing Hinton in a California hotel room at 2am, and keeping him up all morning. He would still get the money if he had slept in.

It is debatable whether these guys did much that was novel. Ignoring that point, look at all the work that Physicists brag out that did not win. There were surely nominations for cosmic inflation, ER=EPR, dark matter, string theory, quantum gravity, holographic principle, supersymmetry, quantum computing, many-worlds. The Swedes have a pretty good record of resisting fads like these.

Is Physics dead? The last big advanced was the discovery of the Higgs boson about 10 years ago, confirming a theory from 50 years before that. There isn't much new and interesting coming out of Physics anymore. Not Big Physics, anyway. I guess there is still lots of good work being done in material science.

Happy Columbus Day. The discovery of the New World to Europeans in 1492 was one of the most important events in all of human history. I cannot think of any other single event that changed the world so much.

Predicted Nobel for Quantum Computing

Dr. Bee predicts:

Which is why my best bet for this year’s 1:35 Physics Nobel is quantum computing. This is clearly on the bingo card, it’s just a 1:40 matter of time until we check this box. Likely candidates are David Deutsch and Peter Shor.

No, not likely. Not yet, anyway.

Nobel prizes for theoretical advances are only given after experimental confirmation. If someone makes a quantum computer that breaks RSA encryption, then that would be an achievement that the Nobel folks would like to recognize.

IK personally think that they should have given a prize for the Higgs mechanism when it proved so essential to the stunning successes of the Standard Model in the 1970s.

Nope. Nobel does not work that way. Higgs could be confirmed by finding a new particle, and no prize was to be given until that particle is found.

There has been no prize for dark matter, despite dozens of brilliant works describing it. Until someone finds a piece of dark matter in a lab.

3:51 And as my fifth and final guess I have a topic that I think should win 3:56 the Nobel Prize but probably won’t, which is modified Newtonian dynamics. That is because, 4:03 regardless of what you think about the maths, it’s been extremely successful in making predictions. 4:09 It has predicted that the Tully-Fisher relation is valid for all types of galaxies, 4:14 it has predicted the height of the second peak of the CMB, it has predicted that galaxies should 4:20 form as early as the James Webb Telescope has observed. Even if these regularities 4:26 will eventually be explained by something else, modified gravity still predicted them correctly.

Interesting opinion, but those people deny that dark matter is a thing, so I don't think they will win a prize as long as people are looking for dark matter particles.

My fourth bet is a topic that I hope will not win, which is cosmological inflation, 3:15 that’s the idea that the early briefly underwent a phase of 3:19 exponential expansion. Astrophysicists talk about inflation as if it’s settled science, 3:25 but the evidence is sketchier than your friend who always ‘forgets’ his wallet.

It is not settled science, because we do not know what the inflation is. Inflation is not getting a prize.

Alfred Wegener discovered continental drift in 1915, one of the great scientific breakthroughs of all time. He got no prizes or respect, because he could not nail down the mechanism in a way that could be accepted and confirmed by others. Like it or not, that's how it works.

So I say, no prizes for quantum computing, inflation, or dark matter.

Here is a new Bloomberg video: The Race to Harness Quantum Computing's Mind-Bending Power | The Future With Hannah Fry. It says our communications systems will be insecure in 5-7 years. And China may be passing us up. The usual hype.

Update: The Physics prize went to two neural net researchers. Following tradition, they were awoken at 2am this morning.

Dr. Bee reports:

0:00 The 2024 Nobel Prize in physics… did not go to physicists. It went to two computer 0:06 scientists for developing the first neural networks, which became the basis 0:11 of what we now call artificial intelligence. If you still doubt that physics is in crisis, 0:16 the fact that the Nobel Prize in physics goes to computer scientists should make you think. Really, 0:21 I’m just grumpy because all my predictions were wrong. I guess my crystal ball needs 0:26 a software update.

The prize work is decades old, and not the large language models like ChatGPT that have gotten so much excitement. I am tempted to infer that the Nobel committee thought that ChatGPT and other LLMs were the hottest advance in science, so it wanted a prize to recognize that, even if the winners never worked on LLMs.

Merits of the Quantum Positivist Instrumentalist Mindset

New video:

Sean Carroll delves into the baffling and beautiful world of quantum mechanics. ...

Under the Umbrella of classical physics of 10:03 course you know that in the beginning of the 20th century quantum mechanics came along and changed everything now there's 10:10 a puzzle with quantum mechanics quantum mechanics is so profound that even though we've known about it for a 10:16 hundred years professional physicists still don't agree on what quantum 10:21 mechanics actually says and that's very embarrassing. I don't know to me anyway. I 10:26 think that we should know what our best theory of nature actually says but the weird thing is even though we don't 10:33 exactly know what the theory says, we do know what it predicts so it's pushed 20 10:39 and 21st century physicists into this sort of positivist instrumentalist mindset, where they say don't ask me 10:46 what's really going on. I can just tell you what you're going to observe in your experiment and you know what I I hate 10:52 that attitude this is very much not much not my attitude but it is the attitude we're going to take for this talk. I 10:59 wrote a whole another book called something deeply hidden about the philosophical mysteries of quantum mechanics but today we're going to be 11:05 hard-nosed physicists and ask what the Theory actually predicts.

Follow that? Sean M. Carroll says that QM was created as a wonderful theory a century ago. It perfectly predicts experiments. It was created by physicists with a positivist instrumentalist mindset. That is, they focused on the science, and not on the philosophizing.

But Carroll and some other philosophers are unhappy about it, and like weird untestable interpretations. That is all such nonsense, that in teaching QM he has to use that Copenhagen mindset of a century ago.

This is bizarre. It is like a Physics professor saying:

I am going to teach the theory of relativity, as understood by Minkowski, Einstein, and others. It passes all the tests. I am going to teach it even though it does not explain the arrow of time, and I personally think it should. It does not, so the theory is unsatifactory. Since nobody properly explains the arrow of time, I will have to teach it that way.

No, that would be ridiculous. No scientist is going to apologize for a scientific theory not answering some vaguely related metaphysical question.

What is Fundamentally Quantum?

What makes quantum mechanics so strange, and so different from quantum mechanics?

Here is a common opinion:

Following the EPR paper, Erwin Schrödinger wrote a letter to Einstein in German in which he used the word Verschränkung (translated by himself as entanglement) "to describe the correlations between two particles that interact and then separate, as in the EPR experiment."[19] However, Schrödinger had discussed the phenomenon as early as 1932.[20]

Schrödinger shortly thereafter published a seminal paper defining and discussing the notion of "entanglement." In the paper, he recognized the importance of the concept, and stated:[3] "I would not call [entanglement] one but rather the characteristic trait of quantum mechanics, the one that enforces its entire departure from classical lines of thought."

Sometimes indeterminacy or other features are said to be fundamentally quantum.

I believe this is mistaken. I have said so many times on this blog, such as here. The only thing that fundamentally distinguishes quantum from classical mechainis is the noncommutation of observables.

Flavio Del Santo, Nicolas Gisin write a new paper:P

What is fundamentally quantum? We argue that most of the features, problems, and paradoxes -- such as the measurement problem, the Wigner's friend paradox and its proposed solutions, single particle nonlocality, and no-cloning -- allegedly attributed to quantum physics have a clear classical analogue if one is to interpret classical physics as fundamentally indeterministic. What really characterizes quantum physics boils down only to phenomena that involve $\hbar$, i.e., incompatible observables.

I agree with this. Classical mechanics is not really deterministic, if you take into account the fact that the variables can only be known to finite precision. Then classical and quantum mechanics are similarly indeterministic.

THe commutator of position and momentum is ℏ, ie h-bar, Planck's constant. That is the core of the uncentainty principle and everything that is truly quantum. There is no other fundamental quantum mystery.

The authors even point out that many-worlds theory can just as easily be applied to a classical theory. Just assume that all possibilities are real. It also says that there is a classical no-cloning theorem. I will have to think about that one.

How Galileo got the Diurnal Tides Wrong

Galileo was famously prosecuted for his 1632 book onheliocentrism. He had agreed not to advocate heliocentrism, but invited to write a book presenting alternate views. So he wrote the book as a dialog, with the Pope being a foolish character named Simplicio.

One thing I have never understand is how such a brilliant scientist could write this book making his main argument based on a completely bogus theory of tides.

Apparently he was fully aware that the Mediterranean Sea had tides contradicting his theory. But he claimed that the Lisbon tides supported his theory.

That is also wrong, and he was told that it was wrong.

A new paper takes a deep dive into the issue.

Galileo’s argument that the tides of the sea are a product of the motions of the Earth in a heliocentric universe needed diurnal tides to bolster it, because the driving action resulting from those motions would be diurnal. If diurnal tides existed, he could explain away other tidal periods as being a result of the local characteristics of sea basins. Given that when writing the Dialogue Galileo had on hand (thanks to Giovanfrancesco Buonamici) information from Andrés García de Céspedes on diurnal tides occurring in the East Indies, and given the reduced length of the Dialogue’s discussion of tidal periods and its content compared to Galileo’s 1616 discourse on tides to Cardinal Orsini — which included (in error) mention of diurnal tides occurring in the Atlantic at Lisbon — it seems reasonable to suppose that Galileo somehow overlooked adding the Céspedes information to the Dialogue. It seems as reasonable to make this supposition as it is to suppose (as has been done in the absence of awareness of Buonamici’s work) that Galileo clung to an idea in the teeth of adverse evidence, evidence that he even suppressed, leaving himself open to criticisms of his tidal theory in his time, and to the criticisms of modern scholars today.

A diurnal would be one high tide and one low tide per day. That's what Galileo thought, and needed to explain the motion of the Earth. Actually, the tides are caused by gravity, with two high tides and two low tides per day.

You are probably going to say that the Church should not have prosecuted him anyway, because the Church scholars were probably not smart enough to understand the errors in the tide theory. Maybe not, but they were smart enough to recognize that Galileo did not have a compelling argument for the motion of the Earth. It is fine to praise Galileo for his good ideas, but we should also recognize his errors.

It is important to get the Galileo story correct, as it is used as an example so much. India-American activist Sunil Mehta writes:

History offers a chilling precedent. In the 17thcentury, humanity teetered on the brink of intellectual darkness when the church, then the world’s most powerful institution, sought to suppress scientific progress and perpetuate the myth of a geocentric universe. Galileo Galilei, a pioneer of physics, dared to challenge this dogma with evidence showing that the Earth revolved around the sun. Threatened with death, he was forced to recant publicly. But legend has it that despite being compelled to disavow his theories aloud, he muttered under his breath: “And yet it moves!”

Whether or not this story is true, it metaphorically represents the intellectual fight that Galileo and many other brave individuals waged on behalf of science. Thankfully, in the end, darkness was averted and truth prevailed.

What? "Whether or not the story is true ... truth prevailed." And thanks to "brave individuals waged on behalf of science."

No, he is an embarrassment to science. Truth does not prevail by telling falsehoods.

It gets worse.

Today, we stand at a similar crossroads. The world’s most powerful nation, a beacon of intellectual freedom for centuries, is now in danger of coming under the control of a demagogue who wants to build an authoritarian regime on the foundation of misinformation and lies. A plague of intellectual darkness has infected half our nation. Hollow nationalism and misguided xenophobia have taken hold. It is hard to fathom this is happening in the 21st century, but the harsh reality is that a divided country and the quirks of the Electoral College could easily pave the way for an authoritarian future.

What can we do to avert this catastrophe? We must follow the example of Galileo and his followers and fight hard on behalf of truth and facts. The candidates in the upcoming election who are weaponizing misinformation (by making outlandish claims that immigrants are eating pets, for example) need to be soundly defeated.

He has this backwards. I wonder how long he has even been in the USA. The Biden-Harris administration has been much more authoritarian than Trump's, by any measure. The Democrats are the ones trying to restrict the free flow of information. California Democrats just passed a law against political parodies.

The Haitians in Ohio are not even immigrants. They have been allowed to stay temporarily for a couple of years. They are supposed to go back. There is some dispute about what animals they eat.

Whatever your political views, how does this relate to Galileo? The Catholic Church did not seek to suppress scientific truth, and neither does Trump. The whole thing is idiotic.

The Galileo story is told as a story of an authoritarian Church suppressing Galileo's truths. It is all a lie. No truths were supressed. It was not a truth that the Earth moved, because motion is relative, and Galileo had compelling argument for it. There were legitimate scientific arguments for and against. Galileo had a mixture of good and bad arguments.

It is a historical fact that science did best in Europe, under Christian domination. If the Christian authorities were so anti-science, then you might expect science to advance more rapidly away from the Christian authorities. Nope. Science advanced best under Christianity.

Sixth Reason for Quantum Computing Skepticism

Dr. Quantum Supremacy, Scott Aaronson, posted 5 reasons for quantum computing skepticism, and why he thinks they are reduced by recent research. I tried to post another reason, but it did not go through, so I post it here.

That is a good list of reasons, but does not include the main reason for my QC skepticism.

It is plausible that QC should be able to simulate QM reactions. But what is the intuitive argument for the super-Turing computations?

You say that QM uses amplitudes instead of probabilities, and they can be negative and interfere. Fine, but all waves show interference, and you cannot build a QC out of classical waves. There must be some special QM property that makes the magic possible. Possible candidates are: (1) an electron can be in two places at once; (2) two objects can be entangled, and influence each other instantaneously; and (3) a computer can make use of a calculation in a parallel universe.

Yes, these ideas are all magical, and make QC supremacy plausible. The trouble is that they are all dependent on QM interpretations that are not generally accepted. If one of those ideas is really the key to QC, then someone should have already demonstrated it with a Nobel prize-winning experiment. There was a 2022 prize for experiments ruling out local hidden variables, but those experiments just confirmed QM and did not show anything to make QC plausible.

Now a $100 million or more has been pumped into QC research, and yet there is still no convincing proof that QC is real. Fusion power has been a similar disappointment, but at least we have proof that fusing hydrogen into helium releases energy.

Brian Greene is still chasing Einstein's Dream

Peter Woit trashes a Brian Greene Wash. Post op-ed. Woit emphasizes the failures of string theory, but I want to point out how much foolishness is based on chasing Einstein's dreams.

Decades later, string theory continues its march toward Einstein’s dream

Forty years ago this month, the physics community was electrified by a remarkable paper that hinted at the realization of Albert Einstein’s long-held dream: a unified theory of physical reality. The new approach, called string theory, captured the attention of researchers worldwide, as its elegant mathematics offered the potential to reconcile the two most successful yet conflicting frameworks in physics: Einstein’s general theory of relativity, which governs the vast structures of the cosmos where gravity rules, and quantum mechanics, which governs particles populating the subatomic world. ...

The application concerns two papers that Einstein wrote in the spring of 1935. One analyzed “entanglement,” an iconic quality of quantum physics in which the behavior of two distant particles can be so tightly choreographed, it’s as if they have a secret connection bridging the space between them. Einstein famously called this connection “spooky.” ...

Although Einstein published these papers in the span of two months, he considered them completely unrelated. ...

But the duality between Einstein’s two 1935 papers would suggest that quantum mechanics and general relativity are already deeply connected — no need for them to marry — so our challenge will be to fully grasp their intrinsic relationship.

Which would mean that Einstein, without realizing it, may have had the key to unification nearly a century ago. ...

Einstein’s math suggested the big bang, black holes, dark energy and gravitational waves — all consequences that Einstein, who was a cautious revolutionary (human nature, again) considered too exotic to be true.

String theory has produced some interesting math, but no physics. This essay is mostly about Einstein, but he did not even believe in a lot of it. He was not a mathematician, and probably would not have liked string theory.

I am all in favor of discovering new mathematics, but that was not Einstein's dream.

String theory, as Greene describes it, has nothing to do with science. It is disconnected with experiment or explaining natural phenomena.

I believe that Einstein has been a bad influence on Physics. He did some good work, but he seems to be mainly idolized for his supposed dreams about mathematizing a unified field theory. Nothing good has come out of chasing that dream.

Greene is smart enough to know that his essay is nonsense. He is just plugging his book. The new media have no interest in mathematical advances. The math coming out of string theory is hard to explain. To sell physics research, it has to be a new discovery, or it has to be an Einstein essay. He could not get this essay published, if he removed the Einstein nonsense.

Medical Science recognizes Race and Sex

SciAm podcast: In recognition of Prostate Cancer Awareness Month this September, host Rachel Feltman sits down with Alfred Winkler, chief of urology at NewYork-Presbyterian Lower Manhattan Hospital, to discuss proactive steps individuals can take to protect themselves against prostate cancer. Black American men, in particular, face some of the highest rates of the disease in the world, with multiple factors contributing to this elevated risk. This episode explores efforts to raise awareness and promote early screening within this high-risk group. Other articles from the same magazine say Race Is a Social Construct and Human Sex Is Not Binary.

It is funny how our scientific institutions have been taken over by people willing to say obvious wrong things in order to serve misguided ideas about social justice.

Update: A new SciAm podcast today interviews a drag queen about teaching math.

SciAm goes Political Again; gets Trump Derangement

announces:

In the November election, the U.S. faces two futures. In one, the new president offers the country better prospects, relying on science, solid evidence and the willingness to learn from experience. She pushes policies that boost good jobs nationwide by embracing technology and clean energy. She supports education, public health and reproductive rights. She treats the climate crisis as the emergency it is and seeks to mitigate its catastrophic storms, fires and droughts.

In the other future, the new president endangers public health and safety and rejects evidence, preferring instead nonsensical conspiracy fantasies. ...

Only one of these futures will improve the fate of this country and the world. That is why, for only the second time in our magazine’s 179-year history, the editors of Scientific American are endorsing a candidate for president.

Harris and Walz are probably the most incompetent candidates for President in a long time. Harris avoids questions, and in the recent debate and interviews, she just babbled nonsense instead of answering the questions. When asked the simplest questions about what she plans to do or how she differs from Pres. Biden, she cannot answer.

This is science, so I looked for some example of Harris relying on science, following evidence, or learning from experience. No, there are no examples.

Trump touted his pandemic efforts during his first debate with Harris, but in 2020 he encouraged resistance to basic public health measures, spread misinformation about treatments and suggested injections of bleach could cure the disease. By the end of that year about 350,000 people in the U.S. had died of COVID; the current national total is well over a million.

This is all nonsense. Yes, thousands died under Trump, but more died under Biden-Harris. Trump never suggested injections of bleach. I don't think he even used the word "bleach". He merely commented on a scientific presentation on using disinfectants on surfaces to kill covid. Yes, Trump did downplay the shutdowns that health authorities were recommending, but we now know that the shutdowns did more harm than good.

It is hard to get any firm stand by Harris on any issue, except that she is pro-abortion, and against any restrictions on it.

Trump appointed the conservative U.S. Supreme Court justices who overturned Roe v. Wade, removing the constitutional right to a basic health-care procedure. He spreads misinformation about abortion—during the September debate, he said some states support abortion into the ninth month and beyond, calling it “execution after birth.” No state allows this.

Some states do allow abortions in the ninth month, and Trump correctly cited the Virginia governor as favoring it after birth.

The Biden-Harris administration’s 2023 Executive Order on Safe, Secure and Trustworthy Development and Use of Artificial Intelligence requires that AI-based products be safe for consumers and national security.

Really? Has that order made AI safe? Of course not. A science magazine should be able to back up what it says.

People can have political disagreements, and many will prefer Harris for various reasons. But a science magazine should stick to science. If it comments on a political race, it should be able to give evidence that a candidate is better for science.

There is no sign Harris understands any science issue. Reporters confronted her on her flip-flops on fracking, and she was unable to explain her views, and may not even know what fracking is.

New Research in Quantum Fault-tolerance

Dr. Quantum Computing has spent the last couple of years working for OpenAI, and returns to announce:

Let me end by sticking my neck out. If hardware progress continues at the rate we’ve seen for the past year or two, then I find it hard to understand why we won’t have useful fault-tolerant QCs within the next decade. (And now to retreat my neck a bit: the “if” clause in that sentence is important and non-removable!)

I think they created one logical qubit that lasts for a millisecond or so. Or something like that, I did not read the details.

I am just passing this prediction along. Seems doubtful to me.

The Invention of Large Language Models

Most people think that LLMs were invented by Google in 2016, or by OpenAI several years later.

AI expert Andrej Karpathy wrote on his blog:

The Unreasonable Effectiveness of Recurrent Neural Networks

May 21, 2015

The concept of attention is the most interesting recent architectural innovation in neural networks.

He constructs some LLMs, and his results seem pitiful compared to what is done today, but I would say he has a proof of concept.

Google introduced the transformer in Attention Is All You Need, a 2017 paper. As you can see, attention was already a hot idea at the time.

I am not sure who should get credit for inventing LLM. The basic ideas of neural nets go back decades. They got a whole lot smart when gaming GPU chips became fast and widely available, and AI researchers figured out how to use them efficiently.

No Quantum Nonlocal Effects

This is an elementary fact about quantum mechanics.

Dr. Bee explains:

Most importantly, if you do something to one of the pair of entangled particles, 4:46 that does nothing to the other. Say you turn the spin of the particle on the right upside-down 4:53 even though you don’t know what it is. Then the spin of the other particle doesn’t change at all. 4:59 No, it doesn’t. You wouldn’t believe how often I see even physicists get this wrong. I just 5:05 the other day heard a talk from someone who works on quantum computing for heaven’s sake 5:10 who said that if you manipulate one of a pair of entangled particles then that will non-locally 5:16 affect the other. It will not. It’s just that if you make a measurement on one of the particles, 5:23 then that will tell you something about the other. Because they’re correlated. ...

The person who 5:50 makes a measurement on one end can’t tell that a measurement was even done on the other end. 5:56 So: Entanglement is real, we know that. Whether spooky action is real is still a matter of debate, 6:04 but you can’t send information faster than light with either.

She is correct. No one has ever found any nonlocal quantum effects. Just correlations.

It should not be complicated. The same happens classically.

What she says about "spooky action" is a little confusing, so here is her explanation.

2:36 Strange or not, Bohr said that when we measure a particle, 2:40 this superposition “collapses” and suddenly the particle is in only one place. It’s this collapse 2:47 that Einstein referred to as spooky action. Because it would indeed be faster than light. 2:53 The moment you find the particle in one place, you instantaneously know it can’t be elsewhere. 3:00 Einstein disagreed with Bohr. Einstein thought that quantum particles are really only in one 3:06 place and that the sudden update of the wave-function just means that you 3:10 have learned the particle isn’t elsewhere. And his main argument, here it comes, was that by 3:17 claiming the collapse is a physical process, Bohr was introducing a “spooky action at a distance”. 3:26 Even if this spooky action existed though, it couldn’t transfer information. Just because you 3:32 find out what’s going on elsewhere doesn’t mean you sent information there.

When you find a classical particle, you immediately know it cannot be elsewhere. If that is spooky, then classical mechanics is spooky. Regardless, no information or anything else goes faster than light.

Usually she branches into a plug for superdeterminism. Mercifully, she did not this time.

Five Ways to Think About Quantum Supremacy

Aventine reports:

When Google announced that it had achieved quantum supremacy in 2019, the headlines were thrilling.

The world of quantum computing had taken a remarkable step. Google, with its Sycamore quantum processor, had performed a calculation in 200 seconds that, the company claimed in the journal Nature, would take a supercomputer 10,000 years.

This feat, named quantum supremacy by John Preskill, a theoretical physicist, back in 2012, promised to usher in a new world of computing performance. ...

Only it didn’t play out the way Google hoped or expected. ...

And then, earlier this summer, researchers from Shanghai Artificial Intelligence Laboratory in China completed the same task in just 14.22 seconds, driving a final stake through the heart of the Google quantum supremacy claim.

It’s not the only warning sign for the industry. Venture capital investment in the sector has fallen off a cliff, from $2.2 billion globally in 2022 to about $1.2 billion in 2023.

It then reports the opinions of five experts, but none of them say that quantum supremacy has been achieved. The closest is Scott Aaronson who says:

Quantum supremacy can be achieved and then unachieved later. It’s a little bit of a moving target in that sense. But all expect that we’ll eventually get to a place where quantum computers are just routinely doing things that classical computers cannot replicate within thousands of years or millions of years, and at that point there’s no more arguing about it.

Achieved and then unachieved? This is a bit like a mathematician saying something was proved, and then disproved. If it was later disproved, then it was never really proved.

Gil Kalai argues that quantum supremacy is impossible, and explains further here.

ChatGPT is better at English than Tigrinya

From a recent Nature magazine podcast:

0:00 chat GPT has a language problem so my language is called T it's spoken in in 0:07 Ethiopia and also in a Syria ...

how we design 3:38 this technology how it could be used and the impacts it could have in this podcast we are going to 3:45 explore the relationship between llms and languages and ask what must be done 3:50 to ensure that AIs work for 3:56 everyone. I think everybody in the community agrees that we need to democratize AI there should not be 4:02 disproportionate benefit in one language versus the other, right? so we want Fair 4:08 access you know and we want to empower communities in different 4:13 [Music] 4:24 languages

I never heard of this language, so I google it:

Tigrinya is a Semitic language spoken in the Tigray Region of Ethiopia and central Eritrea1.

Actually, ChatGPT does surprisingly well in this language.

No, nobody agrees that ChatGPT should be equally beneficial in all languages. It is impossible. Most of the worthwhile training materials are in English, and English is well-suited to being the world's language of choice.

I post this as an example of how the leading science publication has gone Woke.

Albert explains Weirdness of Quantum Mechanics

David Z. Albert is one of the leading popularizers of quantum mechanics, and is on this podcast:

David Albert is the Frederick E. Woodbridge Professor of Philosophy at Columbia University, director of the Philosophical Foundations of Physics program at Columbia, and a faculty member of the John Bell Institute for the Foundations of Physics. This is David’s eighth appearance on Robinson’s Podcast.

I think he is mostly known for trying to give a philosophical defense of string theory, in the absense of any evidence.

He tries to explain what is weird about QM

24:01 what's often considered striking and unsettling about quantum 24:08 mechanics is that at the beginning of the 20th century people start doing all 24:15 kinds of experiments where unlike in the Newtonian case where we're just taking 24:20 it for granted that yeah there're you know you look at these little dust Moes 24:26 or something like that how they're moving around they themselves presumably consist of billions upon billions of 24:33 these Elementary Point particles which we can't actually see we're sort of taking it for granted that things are 24:38 going to work out in the beginning of the 20th century people fooling around with cathode ray tubes and and stuff 24:46 like that begin to be able to keep track 24:51 of the Motions of individual Elementary particles 24:58 and these particles um um are behaving in ways that are almost 25:08 inconceivably bizarre okay in 25:13 particular people manag to convince themselves by doing lots of experiments 25:19 with these Elementary particles that things like electrons for 25:26 example could be in you you know um it's possible for an electron to be located 25:31 at this point in space and possible for an electron to be located at that point 25:37 in space those are the familiar Newtonian possibilities here or here or 25:42 here or here or here what these experiments at the beginning of the 20th century suggested 25:49 to people and I'm condensing here 30 years or so of wrestling with 25:56 the with the results of these experiments in the beginning of the 20th century there are certain sets of 26:03 experiments so-called interference experiments um double slit experiments 26:10 if you can read about this want to read about this in the literature Neutron interferometry experiments stuff like 26:16 that there's a whole um family of of experiments which 26:23 slowly persuades people that um 26:29 um that apparently there are certain states that 26:34 electrons can be in electrons and neutrons and all elementary particles 26:41 can be in um once again there's a possible state where the electron is at 26:46 Point a there's a possible state where the electron is at point B what people became convinced of is that in addition 26:54 there are possible states of electrons such that the very question is it 27:01 located at Point a or is it located at point B or is it located at neither of 27:07 them fails to make sense okay that um that the question that there are certain 27:15 situations in which the question is the electron in box a or in box b or in 27:22 neither of the boxes is like um is is is is like a 27:32 question of the form um um is the number five Married 27:38 okay or or what is the weight in grams of Catholicism okay or something like 27:45 that um questions that philosophers often refer to as category mistakes okay 27:51 um that there can be situations of a material particle an electron okay where 27:59 asking whether or not it's in box a is somehow like asking what the marital 28:06 status of the number five is or what the weight and grams of Catholicism is okay 28:12 um this strikes people is absolutely bizarre people felt forced to 28:18 conclusions like this because if you tried to tell the story of these 28:23 experiments in a way that used locutions like well at this this point in the story the electron I suppose must have 28:30 been here or at this point in the story the electron must have been there or even if you just insist that at this 28:37 point in the story there must be some place where the electron was because after all it got from here in the 28:43 beginning of the experiment to there at the end of the experiment it must have gone by some particular route either 28:51 through this route or through that route you try to insist on that you try to tell yourself a story about how these 28:59 experiments went okay that's consistent with the results you got you find that 29:05 every particular Claim about which route the electron might have taken somehow collapses into nonsense becomes 29:13 inconsistent with certain of the results of these experiments that you actually did okay so an idea grows 29:21 up um um that it appears to be a feature 29:28 of the way these fundamental particles behave that for every pair of states 29:34 that such a particle could be in that is the state of being located at a and the state of being located at B there's also 29:42 another radically unfamiliar but physically possible state which is 29:47 referred to as the superposition of being located at a and being located at b or the quantum 29:54 mechanical superposition of being located at a or being being located at and being located at B which is which 30:03 which we can we can argue from our experiments is not a case of being 30:08 located at a and not a case of being located at B and not a case of being 30:14 located at both A and B as you often find in the popular literature when they 30:19 try to talk about superposition and that's very bad too and that would reduce the mysteriousness of this way 30:27 below what it actually is what these experiments suggest is that it is wrong 30:33 to say under those circumstances that the electron is located at a and that it's wrong to say that it's located at B 30:40 and that it's wrong to say that it's both at A and B and that it's also wrong 30:45 to say that it is neither at a nor is it at B okay if you think there must always be a 30:53 fact of the matter about where the electron is that exhausts the line iCal 30:58 possibilities okay um um and so people felt 31:04 forced um to acknowledge that what these experiments seem to be screaming at us 31:12 okay is that for any two states that an electron could be in A and B there are 31:18 other physically possible States in in in you know which can't be rightly 31:25 characterized by saying that they're at a can't be rightly characterized by saying it's a b can't be rightly 31:31 characterized by saying it's a both and can't be rightly characterized by saying it's a neither okay and rather you seem 31:41 to be confronted with a situation in which um asking whether the 31:47 electron is at A or B is like asking about the marital status of the number five okay it's just a 31:55 nonsensical question good um

No, this is not weird. The same thing happens in classical mechanics. If you put a ball in a box and shake it up, you do not know where it is. When you open the box, you find it in one position, and not others.

Why is it surprising that electrons behave this way, when classical balls do also?

He goes on to strongly attack Bohr for saying that it does not make sense to talk about the exact location of an election in a box, until you open the box and look at it. Then he praises various alternatives, such as hidden variables and many-worlds.

Bohr was right, and Albert is wrong. The electron is not a particle. If you ask for the exact location of the election, you are basing the question on the faulty premise that the electron is a particle, and it has no answer because it is a meaningless question. As Albert complains, the Physics community accepted Bohr's philosophy about a century ago.

Albert is one of many QM expositors who argue that the textbooks do not make sense so there must be something better. They are just wrong. The textbooks present a perfectly good theory, and all the alternatives have horrible defects.

Sabine Hossenfelder is another one. She just posted a podcast on Can Quantum Physics Explain Consciousness After All? Her answer is No, of course, as she ends up saying Bell's Theorem requires superdeterminism.

She is a free will denier, but see the new paper Decision theory presupposes free will. You have to reject a lot of good science, if you reject free will.

Peebles on Philosophy of Physics

Astrophysicist P. J. E. Peebles writes:

I argue that research in physics operates under an implicit community philosophy, and I offer a definition I think physicists would accept, by and large. I compare this definition to what philosophers, sociologists, and historians of science, with physicists, say we are doing.

He argues that there is an unspoken phylosophy of Physics that most physicists agree with, and they do not necessarily agree with what philosophers say.

Recall Weinberg’s remark that “no one has been able to think of any way to change quantum mechanics in any way that would preserve its successes without leading to logical absurdities.”

I agree with that. People complain about quantum mechanics a lot, but all their alternatives are absurd.

Einstein offered the elegant thought that (in the English translation by Sonja Bargmann 1954)

The supreme task of the physicist is to arrive at those universal elementary laws from which the cosmos can be built up by pure deduction.

I wrote a whole book on how Einstein ruined Physics with this sort of thinking.

People think that Einstein's greatest accomplishment was discovering relativity in 1905 from universal laws. My book shows that relativity was actually developed by others who worked directly from experimental results.

The above essay is a good summary of mainstream physicist thinking.

Cantor and the Theology of Infinity

Kateřina Trlifajová wrote a new paper on infinity:

Discussions surrounding the nature of the infinite in mathematics have been under way for two millennia. Mathematicians, philosophers, and the- ologians have all taken part. The basic question has been whether the infinite exists only in potential or whether it exists in actuality. Only at the end of the 19th century a set theory was created that works with the actual infinite. Ini- tially, this theory was rejected by other mathematicians. The creator behind the theory, the German mathematician Georg Cantor, felt all the more the need to challenge the long tradition that only recognised potential infinite.

The issue turns surprisingly theological, with St. Augustine, St. Thomas Aquinas, and the Pope on the side of Cantor.

Cantor was convinced that the knowledge of infinite numbers had been revealed to him by God, who guided his steps from pure mathematics to an interest in theol- ogy and philosophy so that he could improved a proper understanding of God and nature.

Cantor's papers from 1883 to 1895 were met with skepticism, but by 1897 Mathematicians were on board with his theory of infinities.

Xkcd just had a cartoon about large numbers, explained here.

Weinberg blamed Religion for Scientific Ignorance

Physicist Steve Weinberg is deceased, and said this in a recently released interview:

7:12 expensive uh we waste enormous amounts 7:14 of money on man space flight which have 7:17 has no uh scientific function 7:21 um I worry about that and maybe a more 7:26 important worry is whether or not um 7:30 the forces of uh religious 7:33 zealotry uh which are 7:36 uh so obvious in the Islamic world but 7:40 not entirely absent in the west uh will 7:44 lead to us turning away from science 7:47 because after all the Scientific 7:49 Revolution uhu of course did occur in 7:53 the uh 16th century and 17th century but 7:58 uh there had been a great period of 8:00 scientific Advance before that in 8:02 helenistic times which then came to an 8:05 end uh it was continued for a while in 8:08 the Arab world it disappeared in the 8:10 Christian world and then after the 13th 8:14 century it really uh did not pick up 8:17 again in any real way until the time of 8:20 Galileo uh we may not 8:23 continue with the great 8:26 um great tradition of science uh for 8:29 reasons which are even deeper and more 8:31 frightening than the lack of funding

Weinberg is putting a lot of blame on Christianity here. He praises ancient Greeks because they were pre-Christian, and Galileo because he had a dispute with the Pope.

Other people do say stuff like this, but I do not agree with it.

This "Scientific Revolution" took place in Christian Europe. The scientists were Christians, and acting under Christian sponsorship. Areas dominated by other religions did not make these scientific advances. Is that just a coincidence? I don't think so.

How Michelson-Morley was Crucial for Relativity

Alejandro Cassini, Leonardo Levinas write a new paper:

How the Michelson and Morley experiment was reinterpreted by special relativity

They note how relativity textbooks describe M-M as being crucial for relativity, and discuss whether this is historically accurate. It includes some interesting history, but omits the most important pieces to the puzzle.

Einstein's 1905 paper does not mention M-M. Later on he admitted that M-M was crucial for special relativity, but denied that he paid any attention to it.

FitzGerald, Lorentz, Poincare, and Minkowski all described relativity as a consequence of M-M. This paper does not even mention Poincare or Minkowski. It discusses Einstein a lot.

A paper on the influence of M-M on relativity should primarily be on those who were influenced, not Einstein.

In the years immediately following the M-M experiment, there was no inclination to conclude that the ether was non-existent, nor that the speed of light was constant even though the light source was in motion relative to the ether. Moreover, no one thought that the principle of relativity - the equivalence of all inertial frames of reference for the description of electromagnetic phenomena- would be confirmed. Nor did anyone think the hypothesis that the speed of light was invariant, that is, the same in any inertial frame of reference, would be confirmed.1 What exactly did this experiment confirm or refute?

Not true. In those years, Lorentz and Poincare did say that the speed of light was constant, and Einstein got that postulate from Lorentz. Poincare did say that the principle of relativity was confirmed, and Einstein got that terminology from him.

Lorentz did think that M-M refuted the aether motion theories.

M-M does not actually refute the aether. Belief in the aether is consistent with relativity. Einstein said so himself. M-M just refuted the idea that the Earth had a measurable motion against the aether.

As is well known, FitzGerald in 1889 and independently Lorentz in 1892 proposed a different interpretation. They argued that the experiment refuted the hypothesis that the length of the arms of the instrument remained unchanged when it was in motion relative to the ether, a tacit assumption of the M-M experiment. They then formulated the hypothesis that the length of rigid bodies that move with respect to the ether is not invariant, but rather contracts in the direction of motion

That's right. They saw the M-M as finding that measuring the speed of light did not depend on the frame of reference, and used that to deduce the Lorentz transformation.

It was similar to what Einstein did in 1905, except that Einstein did not use the M-M, but rather what Lorentz had deduced from Maxwell's equations and M-M -- that the speed of light is constant and appears the same in different frames.

At the time Einstein formulated it, it was the only explanation of the M-M experiment that was compatible with all known phenomena about the propagation of light, such as stellar aberration, Fizeau's experiment, and many others.28

No, that was all done by Lorentz in 1895. Lorentz even got a related Nobel Prize in 1902. Einstein did not attempt to explain M-M at all. You might say that Einstein was trying to give a recapitulation of Lorentz's 1895 theory, without mentioning M-M or other experiments.

The contraction hypothesis is usually considered to be purely ad hoc since it was proposed solely to accommodate the result of the M-M experiment in order to save the quiescent ether hypothesis from refutation. This affirmation is debatable and depends on what is understood by the concept of an ad hoc hypothesis.13

Here, "ad hoc" means deduced from M-M or other experiment. A theory is not ad hoc if it abstracts out principles from the experiment. There are some anti-positivist philosophers who consider ad hoc to be a bad thing.

The M-M really was crucial for relativity. Those who discovered the Lorentz transformations and spacetime all said so.

Posulating the Constant Speed of Light

D V Redžić writes in a new paper:

The historical path to special relativity starts from the second postulate introduced by Einstein in 1905 [10]. Immediately after the publication of reference [10],

No, the historical path starts with Maxwell's 1865 theory, and the motion invariance tests of it by Michelson in the 1880s.

A popular interpretation of the 1887 Michelson-Morley experiment was that the speed of light was the same for all observers.

Newton, and all physicists before Einstein (including Voigt, Larmor, Lorentz and Poincar´e [15-18]), took it for granted that there was only one ‘time,’ absolute Newtonian time, for all observers in motion with respect to one another. Einstein was bold enough to venture that each inertial observer has her/his own absolute Einsteinian time.

No. Moving objects had they own "local time" in Lorentz's 1895 relativity theory. Poincare accepted this, and believed that motion affects time. Not sure about Voigt and Larmor, but they gave equantions for time changing; what else could they have thought?

It is amazing that someone could write a commentary on on the historical path of relativity, and act as if it all started and ended with Eeinstein's 1905 paper.

Google Quantum Supremacy is Smashed

NewScientist reports:

Google's claim of quantum supremacy has been completely smashed

Google's Sycamore quantum computer was the first to demonstrate quantum supremacy – solving calculations that would be unfeasible on a classical computer – but now ordinary machines have pulled ahead again

In 2019, Google claimed that its Sycamore quantum computer could perform calculations that would take even the world’s most powerful classical supercomputer 10,000 years to complete – but now it seems that a non-quantum computer crunches the numbers several times faster than Google’s machine, and uses less energy doing so.

Quantum computers have the potential to carry out some kinds of calculations vastly more quickly than classical computers, but are still in their infancy. Google announced in 2019 that Sycamore had achieved “quantum supremacy” – the point at which a quantum computer can…

The rest is paywalled, and I haven't read it, but I don't think I need to. Others have said the Google claim was refuted.

I was skeptical in 2019 for the simple reason that Google did not do something that known to be computationally difficult.. It merely generated some random numbers and then claimed that it would be hard for a classical computer to simulate it.

Then some Chinese researchers showed that it was not so hard to simulate it. Meanwhile the Google team moved on to other problems, rather than defend the integrity of their biggest accomplishment.

Without quantum supremacy, the whole field of quantum computing is a sham. No good can come of it.

David Z. Albert Plugs Bohmian Mechanics

New podcast, with a physicist interviewing a philosopher of physics:

Could physics serve as our best guide to metaphysics? What fundamental metaphysics is best motivated by quantum mechanics? And what’s the deal with the age-old feud between philosophers and physicists?

Here to shed light on all these questions and more is none other than David Z. Albert, professor of philosophy and director of the MA program in The Philosophical Foundations of Physics at Columbia University in New York. David is a prominent American philosopher and physicist widely recognized for his contributions to the philosophy of quantum mechanics and the foundations of physics. He has published four popular books and numerous articles on quantum mechanics.

I have some disagreements, but I was especially struck by this comment at the end:

1:19:40 last question, this one comes from my friend Professor Luke Barnes at Western Sydney. Luke. Yeah, he's a great friend of the show. 1:19:47 He's been on multiple times. he has he makes a controversial claim, as if bohmian mechanics have been proposed. First. 1:19:54 No one would have proposed the Copenhagen interpretation. Right. Your thoughts, sir? I think that's absolutely right. 1:19:59 I think, you know, somebody, somebody had discovered about me in mechanics. 1:20:06 And you imagine before then walking into a room and saying, no, I've got a whole new view. 1:20:12 Okay. it's much more elegant. It respects the symmetry between position 1:20:18 space and momentum space, blah, blah, blah. The only little catch is that you have to give up on the idea 1:20:25 that there's a real external world out there. Okay, I think you would have been laughed out of the room, right?

No, this is bizarre. Albert has written a lot about the philosophy of quantum mechanics, but this comment is so foolish that we should disregard everyone he says on the subject.

Bohmian mechanics is weirdly deterministic and nonlocal. While it has its own cult following, it is nearly useless for practical physics. It was invented to make a theoretical point about hidden variables, but not because it is a satisfactor interpretation.

Earlier, Albert said that quantum logics were too confusing, because he doesn't how to reason about it if regular logic is abandoned. I feel similarly about locality. Give it up, and I am not sure what you mean by experiments anymore, because you cannot isolate any physical processes.

Does E=mc^2 Require Relativity?

Physicist Tony Rothman has a new paper arguing that one can get the famous Einstein mass-energy equivalience E = mc² before Einstein's 1905 paper, and without relativity. In particular, it appears in a 1900 Poincare paper.

Many physicists, for instance, are under the impression that ℰ=m⁢c² can be established by employing the four-vector formalism of special relativity. An early draft of Wikipedia’s page on mass-energy equivalence in fact offered exactly such a “derivation.” Four-vectors, however, are defined in order to be consistent with ℰ=m⁢c²; consequently any argument based on them to prove the relationship is circular. ...

A universal, assumption-free proof of ℰ=m⁢c² is no more attainable than a universal proof of conservation of energy or momentum, and the very idea that all physics can be derived from a master Lagrangian without experimental input must be doomed to failure. For that reason, all demonstrations of mass-energy equivalence rely on specific assumptions and approximations. The closest thing that exists to a general proof of ℰ=m⁢c² is the Laue-Klein theorem [16, 17, 18] of 1911 and 1918, which in essence states that if ℰ=m⁢c² holds for a point mass, then it also holds for an extended closed system, under specified boundary conditions. If radiation can escape to infinity, for example, the boundary conditions are evaded.

Einstein was aware of the inadequacies of his 1905 article and attempted to correct them in six further papers, but as Ohanian argues [19], none is free of errors and inconsistencies. Physicists who have actually read the 1905 paper know that the dubious step is the final one, in which Einstein relies on the Newtonian value for the kinetic energy. ...

Can one arrive at ℰ=m⁢c² in a consistent and plausible manner using only Galilean mechanics and “perhaps Maxwellian” electrodynamics?

Okay, but Maxwellian electrodynamics is a fully relativistic theory, if interpreted correctly. The whole theory of special relativity is mostly a recognition of that fact.

Another Misreading of Bell's Theorem

There is a steady stream of crackpot papers that misrepresent Bell's Theorem. The Wikipedia description is adequate:

Bell's theorem is a term encompassing a number of closely related results in physics, all of which determine that quantum mechanics is incompatible with local hidden-variable theories, given some basic assumptions about the nature of measurement.

It does not say anything about reality, or quantum mechanics, except that we cannot replace QM with a local hidden variable theory.

Here is a new paper that gets it wrong:

Allori, Valia (2024) “Hidden Variables and Bell’s Theorem: Local or Not?”. [Preprint] ...

Equation (2) might misleadingly suggest that Bell’s reasoning only applies to hidden variable theories. This is not the case, as discussed later: Bell has shown that all quantum theories, not just hidden variable ones, must be nonlocal. ...

To summarize the result of the previous section, Bell’s theorem shows that, assuming locality, the perfect (anti)correlations can only be explained by non-contextual hidden variables; however, non-contextual hidden variable theories have been empirically falsified by the violation of Bell’s inequality, when seen as a constrain that such theories need to obey to. Therefore, they only other option to explain the perfect (anti)correlations is to assume that there are nonlocal interactions. ...

Some have argued that Bell’s nonlocality result is unacceptable and have tried to get around it. One possibility which has recently received attention is to reject a hidden assumption called statistical independence. ...

Let’s grant that Bell’s theorem has proven that reality is nonlocal. One theory which respects this theorem is the pilot-wave theory, a hidden variable theory which is explicitly nonlocal. ...

It has been argued that retaining locality would be a desideratum for making quantum mechanics and relativity compatible. However, since locality has to come together with superdeterminism, it is not going to help with much at all.

No. Bell's reasoning does only apply to hidden variable theories. It only gives reasons to accept QM, and reject Bohm's theory and superdeterminism.

The main point of this paper is to argue that superdeterminism is no better than Bohm's theory. As opposed to people like Sabine Hossenfelder who argue for superdeterminism.

A lot of people, like Sean M. Carroll, were hoping that the 2022 Nobel Prize would endorse Bell nonlocality. But it pointedly did not.

The Wikipedia article occasionally has someone inserting text that Bell figured out how to get rid of the hidden variable hypothesis, and apply the theorem to all theories. But that is nonsense, of course.

New video: The 'spooky' side of quantum physics | Tim Maudlin on astonishment and fear in #quantumphysics.

Maudlin admits at 10:30 that you get a similarly spooky and incomplete theory if you tear a dollar bill in two, and send the halves to Alice and Bob. When Alice opens the envelope, she immediately knows what Bob got.

So the entanglement itself is not spooky or surprising. The only surprising part is that QM cannot be completed with local hidden variables. Maybe Maudlin explains that later. Reply

Epicycles are Real

A reader cites Carl Sagan's 1980 Cosmos a personal journey Ep. 3. Harmony of the Worlds (video, transcript).

This little machine shows Ptolemy’s model. The planets were imagined to go around the Earth, attached to perfect crystal spheres — but not attached directly to the spheres, but indirectly, through a kind of off-center wheel. The sphere turns, the little wheel rotates, and — as seen from the Earth — Mars does its loop-the-loop. This model permitted reasonably accurate predictions of planetary motion: where a planet would be on a given day. Certainly good enough predictions for the precision of measurement in Ptolemy’s time and much later. Supported by the church through the Dark Ages, Ptolemy’s model effectively prevented the advance of astronomy for 1,500 years.

Finally, in 1543, a quite different explanation of the apparent motion of the planets was published by a Polish cleric named Nicolaus Copernicus. Its most daring feature was the proposition that the sun, not the Earth, was the center of the universe. The Earth was demoted to just one of the planets. The retrograde (or loop-the-loop) motion happens as the Earth overtakes Mars in its orbit. You can see that, from the standpoint of the Earth, Mars is now going slightly backwards and now it is going in its original direction. This Copernican model worked at least as well as Ptolemy’s crystal spheres, but it annoyed an awful lot of people. The Catholic Church later put Copernicus’s work on its list of forbidden books.

So Ptolemy developed a reasonably accurate model, but it was somehow used by the Catholic Church to keep everyone in the Dark Ages?

No, this is mostly nonsense. The Church only objected to nine sentences in the Copernicus book, and not to publication of the model. Ptolemy did not require crystal spheres.

The Wikipedia article on epicycles is much more sensible:

Epicycles worked very well and were highly accurate, because, as Fourier analysis later showed, any smooth curve can be approximated to arbitrary accuracy with a sufficient number of epicycles. However, they fell out of favor with the discovery that planetary motions were largely elliptical from a heliocentric frame of reference, which led to the discovery that gravity obeying a simple inverse square law could better explain all planetary motions.

The first Ptolemy epicycle for each planet is the orbit of the Earth, and is used to explain the retrograde modtion. People can argue that this epicycle is not real, because Mars never really goes backwards. But Mars does go backwards, as viewed from the Earth, and that is what Ptolemy was modeling.

Subsequent epicycles were used to model variations in speed and direction from uniform circular motion. The Copernicus model did not need the first epicycle, as the Earth had its own orbit, but needed the subseqent epicycles, as deviations from uniform circular motion was observable at the time.

As Sagan later explained, Tycho Brahe made the observations that enabled models superior to Ptolemy's. Astronomy was not held back by epicycles, but by a lack of more precise data. That data came from Tycho, and later the invention of the telescope

Sagan pushes his anti-religion beliefs, but all these advances took places in Christendom, and not elsewhere.

Situational Awareness

The AI world is buzzing with the predictions of Leopold Aschenbrenner. It has some overlap with the AI doomsayers who say AGI superintelligence will take over the world, except that he has inside knowledge from OpenAI, and he has specific arguments how it is all going to happen in the next five years.

Scott Aaronson raves about the analysis.

Interestingly, Aschenbrenner sees his fellow AI researchers as being among the first to be put out of work by the new superintelligence. There is a global arms race to create the best AI, and by 2030, the winner will be doing its own research to improve itself. No humans needed, except to keep mounting Nvidia computers into server farms.

Update: Sabine Hossenfelder posts her critique. She agrees with some of it, but gets off the bus with the wild predictions of scientific progress.

Carroll tries to defend Many-Worlds again

Physicis Sean M. Carroll tries to answer this question:

Joel remok says if we can't access any of the many worlds, what 2:31:07 is the purpose of studying it, when there are still many discoveries to be investigated in the reality we exist in?

Good question. There is no point in talking about parallel worlds that we can never see or know anthing about.

The transcript of his answer:

2:31:13 look I've said before I don't care about the other worlds I care about the laws of physics the question is we have 2:31:20 problems with quantum mechanics as it is taught it is not a sensible rigorous coherent Theory. we make things up like 2:31:28 observations and wave function collapse that aren't rigorously defined. many worlds is a well-defined theory that 2:31:37 replaces the ill defined Copenhagen interpretation and lets us actually do science with it. the point of many worlds 2:31:44 is not the other worlds. it's that we've answered the question of what is a measurement, why do you get probability, 2:31:50 and things like that, and equally importantly there's the fact that we don't know the fundamental laws of 2:31:56 physics. we're not done with physics yet we're trying to build better laws of physics and I strongly think that taking 2:32:02 quantum mechanics seriously and thinking about what is the correct foundational version of quantum mechanics will be 2:32:09 useful to that program. so again it's not about the worlds. people who really 2:32:15 obsess about the other worlds are the ones who haven't really internalized many world worlds many worlds is just 2:32:22 it's always obeying the Sher equation that's the essence of the theory

No, this is nonsense on every level.

Many-worlds does not answer the question of what is a measurement. It says that a measurement splits the worlds, but there is no explanation of how the worlds split, how many worlds there are, how you knowt he worlds have split, or anything. It is no better than how bohr explained it.

Many-words says nothing about how we get probabilities. The many-worlds folks do not even believe in probabilities.

Saying that Copenhagen is ill-defined to do science is backwards. Copenhagen has had successes worth trillions of dollars. No one has ever done any science with many-worlds.

I don't know how Carroll says this junk with a straight face. It is like saying Astrology is the only true science.

Here is the distinction. Copenhagen says you can predict an experiment, with probabilities for different outcomes. Once you observet the outcome, you discard the other possibilites as events that did not happen.

Many-worlds is just like that, except that there are no probabilities for the outcomes, because they all happen, and no world is any more likely than any other. Once you see an outcome, the parallel worlds with the other outcomes go out of reach. A prediction might seem true or false, but that is only because you are trapped on one of the worlds. It is not possible to do any science, because no prediction can ever be falsified.

Copenhagen makes sense and is scientific. Many-worlds is an unscientific fantasy.

AI Models will maintain String Research

Peter Woit reports that the top string theorists are retiring, and they are worried that the younger generation will not carry on as before.

Not to worry! One suggests that AI large language models like ChatGPT will come to the rescue, because they can

train an LLM with the very best papers written by the founding members, so that it can continue to set the trend of the community.

Okay, maybe this was a joke, but it is likely to happen anyway. We may have reached peak knowledge. Already journals are being tricked into publishing AI-generated articles. The LLMs of the future will be trained on the LLMs of the past.

There is no experimental data that has any bearing on string research. Ed Witten cannot live forever. Advanced AI LLMs may be able to carry on string papers indefinitely.

SciAm gets Trump Derangement Syndrome

Scientific American used to be the worlds greatest popular science magazine, but now it is filled with politics and pseudoscience.

Now it has this article from an Australian Politics professor:

Donald Trump is now leading in many polls and could retake the U.S. presidency, despite numerous scandals, indictments and erratic behavior. Explanations for his popularity focus on factors like white identity, right-wing authoritarianism, nationalism and populism.

U.S. politics has included people with these predilections for many decades, however, so the puzzle of Trump’s appeal remains. ...

We differentiated the most loyal Trump supporters from the rest of the American population, including those who merely voted for Trump and supported his policies, by measuring three aspects of Trump support: unquestioning credulity, exemplified by the belief that he was in fact the legitimate winner of the 2020 election; ...

Most strikingly, we find they are highly conscientious, a measure of self-reported characteristics including carefulness, dependability, orderliness and self-discipline. Conscientiousness is significantly associated with all the three cultlike aspects of Trump support.

Support for Pres. Trump is not so tricky to explain. Pres. Biden is one of the worst Presidents ever. He is responsible for foreign policy disasters in Afghanistan, Ukraine, and Gaza. His reckless spending has caused high inflation. He has allowed the USA to be invaded by millions of foreigners. He is too senile to do the job. His VP is even worse. He has pursued bogus prosecutions of his political enemics. He appoints incompetent leftist minorities. He supports the sexual mutilation of children.

Pres. Trump, on the other hand, presided over four years of peace and prosperity. He would have been re-elected, but for changed voting rules that allowed the collection of million of votes whose casting and counting could not be verified. Trump won a majority of the votes cast and counted on Election Day.

This professor even found that most Trump supporters had conscientiously examined the evidence, and found Trump to be the better man. But this article makes no mention of all the objective reasons for concluding that Trump was a better President than Biden. He blames authoritarianism, while Biden is much more authoritarian then Biden.

As an example, those who resisted covid vaccine mandates nearly all support Trump over Biden. This shows that Biden is the authoritarian, and his supporters like his authoritarianism. Also, only an auhoritarian would prosecute his political enemies.

The Scientific American of a few years ago would never have published such a trashy partisan and illogical article.

History on Many Worlds Theory

This new paper gives some history and arguments in favor of many-worlds:

Revolutionizing Quantum Mechanics: The Birth and Evolution of the Many-Worlds Interpretation

Arnub Ghosh

The Many-worlds Interpretation (MWI) of quantum mechanics has captivated physicists and philosophers alike since its inception in the mid-20th century. This paper explores the historical roots, evolution, and implications of the MWI within the context of quantum theory. ...

In reflecting on the significance of studying the historical development of quantum interpretations, it becomes evident that the MWI represents more than just a theoretical framework—it embodies a paradigm shift in our understanding of the quantum world and its implications for the nature of reality. ... Moving forward, further research and exploration in this area hold immense potential for advancing our knowledge of quantum mechanics and its applications.

Okay, you know it is garbage when it pushes revolutions and paradigm shifts, but cannot find any actual utility to the theory.

A big problem is probability:

Moreover, the MWI has led to new insights into the nature of probability in quantum mechanics. In the MWI, probabilities arise from the relative frequencies of different outcomes across multiple branches of the multiverse, rather than from inherent randomness or observer-dependent collapses of the wavefunction. This perspective offers a coherent and objective interpretation of probability in quantum mechanics, resolving longstanding debates about the nature of quantum uncertainty. ...

Furthermore, the MWI has implications for our understanding of probability and randomness. In the MWI, probabilities arise from the relative frequencies of different outcomes across multiple branches of the multiverse, rather than from inherent randomness or observer-dependent collapses of the wavefunction. This perspective offers a coherent and objective interpretation of probability in quantum mechanics, resolving longstanding debates about the nature of quantum uncertainty. However, it also challenges traditional views of probability as a measure of uncertainty or ignorance, suggesting that probabilities are ontologically real and arise from the fundamental structure of the multiverse.

(This is repetitive.) Yes, you might think that MWI gets probabilities from comparing outcomes across branches, but no one has ever gotten this to work. It is a myth. You cannot count the branches, or calculate probabilities of outcomes, or do anything objectively. It is all a big sham.

The people who pursue many-worlds do not even believe in probabilities. If pressed, they will admit that some things seem more probable than others, but they say it is all an illusion. Some subscribe to crazy arguments that Born's rule seems more natural than other rules, but they have no good explanation for why probability works at all.

If the weather man says there is an 80% chance of rain tomorrow, then he means that his data has resulted in rain 80% of the time in the past. Once it rains, you can discard the possibility of a sunny day.

But the MWI fans reject this. They refuse to say that it will rain in 80% of the branches. And if it does rain, they refuse to reject the sunny day, because they argue that it is sunny in another part of the wave function.

Go ahead and read the article, and tell me if you find one reason to accept that MWI nonsense.

Free Will in the Block Universe

Some say that there is no free will because of the way relativity uses time as the fourth dimension.

The philsophical questions about free will go back to the ancient Greeks. Modern science has not settled those questions. The questions are certainly not settled by just declaring that time is a coordinate.

SciAm reports:

As troubling as quantum mechanics (or at least certain versions of it) may be for the idea of free will, relativity—the other pillar of modern physics—isn’t off the hook. Many theorists think of relativity as describing a universe in which past, present and future are all equally real: a static cosmos that just sits there like a big block of spacetime (sometimes called the “block universe”). It’s not that time disappears in this picture—but it no longer “passes” or “flows.” (As Albert Einstein famously put it, the passage of time is a “stubbornly persistent illusion.”) Conceptually speaking, the strongly deterministic quantum universe and the block universe of relativity may not be so far apart. The quantum version can be thought of as “a kind of enriched block universe,” says Alastair Wilson, a philosopher of science at the University of Leeds in England. “Imagine taking a block universe and adding an extra dimension to it—the dimension of possibility.” ...

Even without relativity, you could use a time coordinate to make a block universe. You can use any coordinates you want. Choosing some coordinates cannot possibly say anything about free will.

Many-worlds theory takes this to the next step, if you put all those many worlds in a big block multiverse.

While physicists continue to debate the idea of strong determinism, Emily Adlam, a philosopher of physics at Chapman University, agrees with Chen that it appears to present more of a threat to free will than traditional determinism, particularly because of its ties to the Everettian multiverse. “In a standard deterministic picture, sure, everything that happens was determined from the past—but your mind was a key part of the causal process by which future events get realized,” Adlam says. “So in some meaningful sense, future events—even though they were predetermined—were mediated through processes that you identify with yourself.” But in the Everettian picture, she says, it’s harder to see where decision-making would fit in. “If you always make every possible decision, that does seem to severely undermine the sense in which you are exercising any meaningful kind of choice,” she says. “So in that sense, you do seem worse off than in the standard picture, where one outcome occurs and you play a role in bringing it about.”

It goes on to say that this is controversial, as there are compatibilists who can rationalize belief in free will no matter what soft of determinism there is.

There is no mention of anyone who believes in true free will, now called libertarian free will.

The whole point of the article is to say that many-worlds theory presents a new argument against free will. Since all choices happen in parallel worlds simultaneously, then all choice possibilities are real, and humans cannot make any free will choices. That are all illusions.

This is absurd.

I wonder what Plato and Aristotle would say if we could go back in time, and tell them about our progress. We could tell them about rockets, cars, lasers, electronics, and drugs, and they would be impressed.

And then we would tell them that our best and finest theory of matter, quantum mechanics, has been interpreted to say that there is no free will, because we can imagine parallel universes where alternative decisions are made. They would rightfully conclude that we have made no progress in philosophy at all, and even regressed to some very silly ideas.

There is Nothing Non-entangled in QM

Supposedly entanglement is the most important thing in quantum mechanics (QM). It is said to be the key feature that distinguishes QM from classical mechanics.

But what is it?

New paper:

Everything is Entangled in Quantum Mechanics: Are the Orthodox Measures Physically Meaningful?

Christian de Ronde, Raimundo Fernandez Moujan, Cesar Massri

Even though quantum entanglement is today's most essential concept within the new technological era of quantum information processing, we do not only lack a consistent definition of this kernel notion, we are also far from understanding its physical meaning [35]. These failures have lead to many problems when attempting to provide a consistent measure or quantification of entanglement. I

According to a lot of modern scholars, entanglement is a resource that can be exploited to give secure communication and super-Turing computability.

The paper discuss various definitions in the literature and concludes:

We have shown how entanglement, as the unitary multiscreen effect of a single power, is an irreducible aspect of the operational content of the theory of quanta. The theory talks about powers of action each one them producing a multiscreen (non-local) effect that can be observed in the lab. Consequently, there is nothing non-entangled in QM. There is no meaningful distinction between something that is entangled and something that is not entangled within the theory of quanta. The attempt of quantifying or measuring the level of entanglement becomes meaningless.

I think that entanglement is not something real. It is an artifact of how QM works, but there is no way to objectively say whether a particle is entangled or not. So we should not talk about entanglement as if it is some mysterious resource.

Maybe I will be proved wrong by some quantum computer that uses entanglement to break RSA or some other calculation that cannot be done otherwise. Then I will have to admit that entanglement is real and useful. But nothing like that has ever been done.

Sean M. Carroll tries to explain entanglement in a recent podcast:

I 0:20 think you know entanglement arises 0:22 directly from that statement we made 0:23 long ago that when you have a Quantum 0:25 system you do not have separate wave 0:27 functions for each part of it you only 0:29 have one wave function for the whole 0:30 thing and the job of the wave function 0:33 is to make predictions for observational 0:35 outcomes so if that's all true then it 0:38 could be the case that if you predict 0:41 the outcome for one thing and another 0:43 thing particle a and particle B there 0:46 might be correlations or connections 0:48 between those measurement outcomes so I 0:50 don't know what I'm going to see when I 0:52 ask what is the spin of particle a and I 0:54 don't know what I'm going to see when I 0:56 ask what is the spin of particle B but I 0:58 know they're going to be opposite so 1:00 then that's entanglement and it tells me 1:02 were I to measure particle a I have no 1:04 idea what I'm going to observe but as 1:06 soon as I do I know what the outcome is 1:08 for particle B and this bugs people 1:11 because how does particle B know what 1:13 its outcome is supposed to be it could 1:15 be light years away

If this is the definition of entanglement, then there is nothing quantum mechanical about it. Classical physics shows the same phenomenon.

If split a classical system with angular momentum zero, separate the halves, and measure the angular momentum of one half, then the other half will have the opposite angular momentum. Just like how Carroll described QM.

When asked for more explanation, he says that he prefers the Everett many-worlds interpretation. However he admits that no one knows whether the effect of a measurement propagates at the speed of light or less in the universal wave function, or propagates instantaneously.

What? I thought that the whole point of Everett was to clarify what happens quantum mechanically when a measurement takes place. But if he cannot tell how the result propagates, then I do not see how it can tell me anything.

See also this podcast, where he starts by saying the Everestt theory is the most straightforwad interpretation of the Schroedinger equation, but it requires splittings into parallel worlds and we do not know what a world is. We also do not know if the wave function is real. Later, at 45:30 he says, "I mean the real answer there is I don't think about all those other worlds, that much again the worlds are a prediction of the theory they're not what the theory is fundamentally about."

Carroll goes on to explain the Bell tests:

also in the 1960s John Bell proved his 6:17 theorems he proved theorems about the 6:20 different predictions between a local 6:22 Theory and a non-local theory like 6:24 quantum mechanics and that made it an 6:27 experimentally accessible question so 6:29 people did the experiment and they just 6:31 won the Nobel Prize a couple years ago 6:33 so physicists are very interested now 6:35 because there's an experiment you can do 6:37 of course the experimental result was 6:38 exactly what schrodinger would have 6:40 predicted back in the 1920s it didn't 6:42 change our idea of quantum mechanics but 6:44 as long as you can do an experiment 6:45 they're happy having said that because 6:48 physicists have ignored the foundations 6:49 of quantum mechanics for so long even 6:52 the Nobel Prize press release botched it 6:55 they gave the wrong explanation for what 6:57 was going on because they didn't really 6:59 understand what they just give the Nobel 7:00 Prize

Again, this is startlingly foolish. He correctly says that the prize-winning experiments just confirmed what quantum mechanics would have predicted in the 1920s, and did not change our idea of quantum mechanics.

So how did the Nobel press release botch it? It certainly did not say that the experiment changed our view of QM. That would have been big news. See my earlier post for more details.

Carroll's real gripe is that he wants to fund more research in the foundations of QM, and the Nobel committee refused to acknowledge that the experiments left unsettled issued. In particular, he wants to push many-worlds theory, but it gets no Nobel endorsement. The Nobel committee correctly said that the experiments confirmed what everyone thought for many decades.

Einstein's Futile Search for Unified Field Theory

Sabine Hossenfelder posted her latest video:

Einstein’s Other Theory of Everything

Einstein completed his theory of general relativity in 1915 when he was 37 years old. What did he do for the remaining 40 years of his life? He continued developing his masterwork of course! Feeling that his theory was incomplete, Einstein pursued a unified field theory. Though he ultimately failed, the ideas he came up with were quite interesting. I have read a lot of old Einstein papers in the past weeks and here is my summary of what I believe he tried to do.

She insists on pronouncing his name INE-shtine. Maybe Germans pronounce it that way, but not Americans.

She summarizes Einstein's foolish and misguided unified field theories. If he did not already have a fancy reputation, this would be considered crackpot work.

She also mentions the more successful Kaluza–Klein theory:

In physics, Kaluza–Klein theory (KK theory) is a classical unified field theory of gravitation and electromagnetism built around the idea of a fifth dimension beyond the common 4D of space and time and considered an important precursor to string theory. In their setup, the vacuum has the usual 3 dimensions of space and one dimension of time but with another microscopic extra spatial dimension in the shape of a tiny circle. Gunnar Nordström had an earlier, similar idea.

This is also described as a failure. She says you need supersymmetry to make it work, but that has been rejected.

That theory is a minor variant of a 1918 H. Weyl proposal to unify gravity (general relativity) and electromagnetism. That was promptly attacked by Einstein, and mostly forgotten.

There the story ends. But not really. Einstein's ideas were, in fact, worthless, and led several generations of physicists astray. I wrote a whole book about it.

But Weyl's idea was essentially correct. In the modern Standard Model, that everyone accepts, there is an extra tiny circle at every spacetime point (event) to account for electromagnetism. In modern lingo, it is a circle bundle over a spacetime manifold. Gravity is a connection on the tangent bundle, and electromagnetism is a connection on the circle bundle. In both cases, the field is the curvature. It is the simplest and most natural relativistic theory.

Physicists studied Weyl and Kaluza-Klein for decades, without ever figuring out that they were just adding extra terms that ruined the theory. I guess they wanted to unify the gravity with the electromagnetism by hypothesizing some interactions, but that was just foolishness.

If Weyl were really smart, he would have conjectured using other Lie groups form the strong and weak interactions, alongside the circle for electromagnetism. Then we could have had the Standard Model, before quantization, about 50 years before we did.

One of the Lessons of the Standard Model is that all four forces (gravity, electromagnetism, weak, and strong) all have the same math structure (field strength is curvature of a bundle connection), but the force are essentially orthogonal. They do not have much to do with each other. There is some twisting between the weak and electromagnetism, but that's all. Many people thought that they would be unified like electricity and magnetism, where you cannot study one without the other.