Jacob Barandes and Me
Please enjoy Harvard’s Jacob Barandes and yours truly duking it out for 2.5 hours on YouTube about the interpretation of quantum mechanics, and specifically Jacob’s recent proposal involving “indivisible stochastic dynamics,” with Curt Jaimungal as moderator. As always, I strongly recommend watching with captions turned on and at 2X speed.
To summarize what I learned in one paragraph: just like in Bohmian mechanics, Jacob wants classical trajectories for particles, which are so constructed to reproduce the predictions of QM perfectly. But unlike the Bohmians, Jacob doesn’t want to commit to any particular rule for the evolution of those particle trajectories. He merely asserts, metaphysically, that the trajectories exist. My response was basically, “OK fine, you can do that if you want, but what does it buy me?” We basically went around in circles on that question the entire time, though hopefully with many edutaining disgressions.
Despite the lack of resolution, I felt pretty good about the conversation afterward: Jacob got an extensive opportunity to explain his ideas to listeners, along with his detailed beefs against both the Many-Worlds and Copenhagen interpretations. Meanwhile, even though I spoke less than Jacob, I did get some opportunities to do my job, pushing back and asking the kinds of questions I imagined most physicists would ask (even though I’m not a physicist, I felt compelled to represent them!). Jacob and I ended the conversation much as we began: disagreeing on extremely friendly terms.
Then, alas, I read the comments on YouTube and got depressed. Apparently, I’m a hidebound academic elitist who’s failed to grasp Jacob’s revolutionary, paradigm-smashing theory, and who kept arrogantly interrupting with snide, impertinent questions (“OK, but what can I do with this theory that I couldn’t do before?”). And, I learned, the ultimate proof of my smug, ivory-tower malice was to be found in my body language, the way I constantly smiled nervously and rocked back and forth. I couldn’t help but wonder: have these people watched any other YouTube videos that I’m in? I don’t get to pick how I look and sound. I came out of the factory this way.
One commenter opined that I must hate Jacob’s theory only because I’ve poured my life into quantum computing, which depends on superposition, the confusing concept that Jacob has now unmasked as a farce. Presumably it’s beyond this person’s comprehension that Jacob makes exactly the same predictions as I make for what a quantum computer will do when built; Jacob just prefers a different way of talking about it.
I was reminded that optimizing for one’s scientific colleagues is wildly different from optimizing for YouTube engagement. In science, it’s obvious to everyone that the burden of proof is on whoever is presenting the new idea—and that this burden is high, especially with anything as well-trodden and skull-strewn as the foundations of quantum mechanics, albeit not infinitely high. The way the game works is: other people try as hard as they can to shoot the new idea down, so we see how it fares under duress. This is not a sign of contempt for new ideas, but of respect for them.
On YouTube, the situation is precisely reversed. There, anyone perceived as the “mainstream establishment” faces a near-insurmountable burden of proof, while anyone perceived as “renegade” wins by default if they identify any hole whatsoever in mainstream understanding. Crucially, the renegade’s own alternative theories are under no particular burden; indeed, the details of their theories are not even that important or relevant. I don’t want to list science YouTubers who’ve learned to exploit that dynamic masterfully, though I’m told one rhymes with “Frabine Schlossenfelder.” Of course this mirrors what’s happened in the wider world, where RFK Jr. now runs American health policy, Tulsi Gabbard runs the intelligence establishment, and other conspiracy theorists have at last fired all the experts and taken control of our civilization, and are eagerly mashing the buttons to see what happens. I’d take Jacob Barandes, or even Sabine, a billion times over the lunatics in power. But I do hope Jacob turns out to be wrong about Many-Worlds, because it would give my solace to know that there are other branches of the wavefunction where things are a little more sane.
Follow
Comment #1 March 4th, 2025 at 5:48 pm
Scott, one day I will finally convince you that you should not care what random idiots are saying online. I promise you that Jacob (who I know well) was not thinking any of those things, and if these random idiots can’t stand watching actual scientific discussion they can go back to watching videos about how the earth is flat, Trump won in 2020, and Putin is really a decent guy protecting his people from NATO.
Comment #2 March 4th, 2025 at 5:57 pm
If all of the experts have been fired, I presume that means that once the next democrat president gets brought in ( if we survive that long), I should just continue to ignore what the government is saying? Well, unless they go out of their way to rehire most of the people who have been let go in this administration, I suppose.
Expertise is a precious thing. Hard to cultivate, easy to undermine. For every true expert, there will be many pretenders who will present themselves as experts to try to claim a job. And if layman struggle to tell experts from non experts, that creates a difficulty in finding them. And that is before you add any corrupting influence from the political process.
If the government has lost all grasp on expertise, should we expect that it cannot be restored in a timely fashion?
Comment #3 March 4th, 2025 at 6:29 pm
Day by day I feel suffocated by idiots. Its easy to get the impression that they vastly outnumber the reasonable people. But I suspect they are just louder, in part due to the fact that they don’t understand enough to appreciate how little they understand. In part because the idiots are winning politically at the moment. As per Daniel Harlow’s comment, please ignore these people. They demonstrate themselves unworthy of your attention.
Comment #4 March 4th, 2025 at 6:42 pm
To be fair, you could have asked this about Bohmian mechanics for years, until John Bell came along and was so inspired by it that he derived Bell’s theorem. Never underestimate the explanatory power of a real ontology!
Comment #5 March 4th, 2025 at 6:48 pm
Daniel Harlow #1: Twenty years ago, it would’ve been reasonable to say “don’t worry what some random online idiots are saying.” But now, as I pointed out in the post, the random online idiots rule the world! For those of us who still care about getting things right, I don’t know what the best response is, but I’m pretty sure it’s not “more of what we were doing before.”
(And yes, Jacob is my friend too, and I don’t even need him to tell me that he wasn’t thinking those things.)
Comment #6 March 4th, 2025 at 6:50 pm
Sandro #4: But as I say in the video, Jacob’s view is sort of just “Bohm minus minus”: Bohm without a commitment to any particular distribution over the trajectories. So I think the question stands: if I already knew well that many other distributions over trajectories were possible, then what could such a view give me that Bohmian mechanics didn’t give me already?
Comment #7 March 4th, 2025 at 7:32 pm
I’m in complete agreement with you. Modern social media and the podcast circuit are turning popular physics into a grotesque series of WWE matches, where actual physicists are always the villain, never given a speaking role, and always scripted to lose.
I’ve seen it happen over and over again in particle physics. Consa and co. claimed that all modern QED calculations are pure fraud, on the basis of the fact that several wrong two-loop g-2 calculations were published in the 1950s, so we must have gotten stuck then and just made up all the results since. Ben Recht claimed that the Higgs discovery probably never happened because it was so complicated to be beyond human comprehension (only true if you regard graduate QFT as “beyond human comprehension”). Recently we had Sabine waving around a perfectly real anonymous letter saying, “all of us particle physicists gleefully waste taxpayer money on pure fraud, and don’t tell anybody I told you this.” That’s millions of views in total.
Even when the discussion has physicists on both sides, it gets distorted and politicized. Just the other day, I was talking with a guy who was convinced that we weren’t listening to Susskind’s take on the black hole information paradox, because we were still in the thrall of Hawking’s “woke progressivism”.
The problem isn’t just with podcasts; even thoughtful people who only read long-form content get reliably misled. If you’re a literary person interested in physics, you will go to outlets like Aeon and Noema, which are dominated by philosophers like Tim Maudlin, who will tell you that anybody who isn’t a strict Bohmian has given up on rationality and intellectual standards. The shortcomings of pilot wave theories are never mentioned, because the authors of the articles usually don’t even know what they are.
If you’re an engineer interested in physics, you will regularly hear on Ars Technica and Phys.org that dark matter is dead. For instance, every few months, somebody messes up a basic calculation and posts a paper claiming that _unmodified_ GR explains galaxy rotation curves and thus removes the need for dark matter. It doesn’t (a one-line estimate shows the effect is 10^6 times too small), but each time the claim leads to dozens of news articles. The fact that we get a “there’s no DM after all!!” article every week makes the majority of engineers fairly sure that it doesn’t exist. Heck, even Elon is on record saying DM is probably fake.
As a result, the reputation of physics has fallen precipitously in the past decades. At one point, your average well-read person would at least be vaguely aware that physicists use mathematics to make quantitative predictions using well-motivated principles. Nowadays, the majority of smart non-physicists I talk to will shortly confide their mistrust in the whole enterprise. Why do we ignore theory X, which they’ve heard solves everything? Are we afraid of how good it is? Why do we search for particle Y, when they’ve heard there’s no way it can exist? Are we just siphoning off the grant money to buy Ferraris? Why do we believe in principle Z, when they’ve heard so many decrying it in the most emphatic possible terms? Are we just too dogmatic to change? Of course, most people phrase these questions in a less combative way, but the mistrust is always there. It’s hard to trust an enterprise if you only see, year after year, tweets, podcasts, and articles unanimously saying it’s all wrong.
Most of my colleagues in particle physics laugh when they hear about this, but I think it’s a serious issue. If you only care about quantum gravity thought experiments, then maybe you can hide in the ivory tower, or live off Simons’ money. But if you want our society to build real experiments and gather real data, you need to make a case. It is very hard to fund ambitious science even when the public is excited about it, so it certainly isn’t possible if the dominant opinion is that it’s waste and fraud.
For years, I’ve been typing up rebuttals to wrong claims, in various corners of the internet. Not a single one has reached 1% of the views of the original claim. Some friends of mine like to talk about science to schoolkids. Not one of them has achieved even 0.1% of the impact of a mediocre podcaster. These old fashioned approaches are not working, because they can’t match the reach of algorithmically-spread falsehoods. We need an entire generation of physicists to master these algorithms and start fighting back. NASA spends a huge amount of effort on making pretty pictures; why don’t we? Crackpots dominate the podcast circuit, but we actually outnumber them, so why aren’t we showing up? If you’re willing to trade zinger for zinger, then it’s not actually hard to win an argument, because we actually know what we’re talking about! And to not show up is to automatically lose.
Comment #8 March 4th, 2025 at 8:12 pm
I don’t like the YouTube comment interface, no chance to reproducably find specific comments again. So I rarely write comments below YouTube videos myself, and don’t put much weight on such comments either.
Not saying that comments in other parts of the internet would be more substantive or correct, but at least … I don’t know. Let me quote two comments from YouTube and two comments from elsewhere. My guess is that it is easy to identify the ones from YouTube.
Comment #9 March 4th, 2025 at 8:36 pm
Daniel Harlow #1 and Scott Aaronson #5 – I know neither of you need to hear me say this, but I’m happy to confirm that I wasn’t thinking any of those things!
Comment #10 March 4th, 2025 at 8:36 pm
Haha surely there must be one sane branch, if it forks as often as is claimed.
Comment #11 March 4th, 2025 at 9:06 pm
Kevin Zhou: “For years, I’ve been typing up rebuttals to wrong claims, in various corners of the internet. Not a single one has reached 1% of the views of the original claim.” This is precisely why it is not a good use of your time. Due to the second law of thermodynamics it is far easier to spew BS about any nontrivial topic than it is to be right, so you can never keep up. I certainly agree that as scientists we should do outreach, both for self-serving reasons and also because it is valuable to society (and in fact there will be a long form interview with me and Brian Greene showing up soon on youtube so I’m not just saying that). My point is just that arguing with (or worrying about) ignorant trolls online is not a cost-effective way of doing outreach. I think Brian has the right approach, as does Lenny Susskind. What we do is fun, and there are quite a lot of non-conspiracy-theorist people who like to hear about it without any of the dumb stuff. And maybe, just maybe, if we do it well we can suck in a few of the conspiracy-adjacent as well.
Comment #12 March 4th, 2025 at 9:16 pm
Kevin Zhou #7: Thanks, that’s beautifully put.
I worry that anyone using YouTube, phys.org, etc to learn science could end up with a wildly distorted picture. They’ll of course hear many crackpot ideas of loons, but more subtly, they’ll also hear heterodox ideas of serious people wildly out of proportion to the mindshare that those ideas earned in the gauntlet of expert criticism. Some of that is unavoidable and fine: people are naturally drawn to the heterodox, and many heterodox thinkers like to make their case to broad audiences, and lots of mainstream work is too hard to explain anyway. But it becomes a problem when people come away believing important falsehoods — for example, that quantum mechanics is embattled, in crisis, etc., and about to be toppled any day in favor of the latest classical alternative hawked on a podcast.
Comment #13 March 4th, 2025 at 10:54 pm
Daniel Harlow #11: you’re probably right. For early career people, getting nontrivial reach takes too much time. I know some people who “succeeded”, but they did so by essentially living on Twitter, which killed their research output. It’s easier if you’re an established professor, so that there will already be an audience of people to listen.
But the thing is, I haven’t managed to convince a single professor to give what you’re doing a try! And weirdly, hep-ph seems even more reluctant to do outreach than hep-th, even though we depend more on public funding for progress. I personally should post less, but I think if everybody else shifted a standard deviation upward we’d be in a much better place.
Scott #12: and even the mainstream works are described on these sites in an unnecessarily, intentionally confusing way. How often have you seen an AMO physicist describing their latest submission to Nature or Science as proving that “quantum mechanics is even weirder than we thought” or “we have no idea how quantum mechanics works”?
For instance, there’s a whole genre of articles claiming that some experiment has shown that quantum processes can rewrite the past, foresee the future, or take a negative amount of time. The AMO physicists hopefully all know the real story is “our experiment checked that textbook QM works exactly as expected, but if QM _wasn’t_ true you’d need some crazy retrocausality to get the same results”, but they choose not to communicate that subtlety. It seems to be a game they play to get into top journals. But then the public just gets more confused, and convinced that physicists don’t know what’s going on.
Comment #14 March 4th, 2025 at 11:43 pm
Hi Scott, off-topic, my apologies:
Gil Kalai seems to still insist that QC at a scale that will break (say) RSA is not likely to be feasible in the near future, possibly ever, in a few recent and quite detailed blog posts.
I fully understand if you don’t want to go there, but does he have reasonable arguments? Is there a realistic possibility that he is right. He also claims that Leonid Levin and some others have similar positions vis a vis QC.
Comment #15 March 4th, 2025 at 11:54 pm
kodlu #14: From my perspective, Gil’s comments have gotten more and more wildly disconnected from the reality of what’s happening experimentally—e.g., Google’s demonstration this past year of a surface code qubit that gets better with increasing scale, which worked exactly like theory said it would in the 1990s. To be honest, though, I’m happy to have Gil and other QC skeptics around; I no longer feel the burning urge to debate them that I once did. I’m satisfied that Nature Herself will get more and more chances to weigh in on these questions over the next few years. Obviously I have my own guess about what the result will be, but the outside possibility of the QC skeptics being vindicated will make things all the more exciting! 🙂
Comment #16 March 5th, 2025 at 12:50 am
Jacob’s claim that there are always definite configurations and trajectories seems to clash with his notion of an “indivisible stochastic process”, in which you can ask about the probability of how things end up, but you’re not allowed to ask about the probability of being in a particular state e.g. at the halfway point.
If we’re being realists about configurations and trajectories, then any concrete instance of an indivisible stochastic process consists of one particular member of an ensemble of possible trajectories. Do we have a measure on that ensemble? Several possible measures? If we do have a measure, then couldn’t we calculate the forbidden intermediate probabilities? But if we do have an ensemble, how can we not have a measure on it?
Comment #17 March 5th, 2025 at 12:56 am
I was about to say I’m glad it was you and not me, but then there came an unexpected turn in the last paragraph. I’m not sure I understand what you mean there. Did you just compare me to RFK Jr? Wtf, Scott.
I can be blamed for many things, but pushing alternative theories, mine or someone else’s, isn’t one of them. Yes, I talk about new physics ideas. But I almost always end up saying I think it’s nonsense. So much so that the most frequent complaint I get is that I am too “negative”. But what can I do. Most of it is nonsense.
Fwiw, I read Barandes’ first paper when it came out and had pretty much the same reaction. A lot of people asked me to comment on it, and I put it on the list to re-read but. Well. Evidently I wasn’t enthusiastic enough to actually make time for it.
Comment #18 March 5th, 2025 at 1:21 am
I think you’re being unfair to Sabine. From what I’ve seen, she makes strident but principled and mostly-correct criticisms of academic science, very similar to those by Scott Alexander.
Comment #19 March 5th, 2025 at 2:28 am
Sabine #17: I very often agree with your acerbic takes, or at least enjoy them. Not always. To me, superdeterminism is a candidate for the most insane idea in the history of physics—certainly 1000x more insane than anything the string theorists have ever come up with. But even your superdeterminism advocacy wouldn’t merit a comparison to RFK Jr.
No, what merits the comparison to RFK Jr. is this recent video of yours. There, with comically unconvincing caveats (“I’m not necessarily saying this should happen, just that it will“), you speak approvingly about the imminent destruction of publicly-funded academic science, in favor of just letting Musk, Bezos, et al. fund whatever they feel like. As Daniel Litt tried to explain to you (see here), there’s a massive public goods problem here: again and again, basic research has upgraded the “power level” of entire nations and of civilization as a whole, but it’s hard for any one company or investor to capture the gains. That’s why so much of the world-changing research of the 20th century was done either at universities, at companies that subsisted on defense contracts (like the ones that built the early Internet), or at Bell Labs back when it was a government-enforced monopoly. And indeed, after the monopoly ended, Bell Labs dwindled to a shadow of what it was, precisely because AT&T could no longer capture the gains from curiosity-driven basic research. I was astonished that not once in your video did you grapple with this obvious point.
RFK Jr. knows that he despises the mainstream medical establishment, and he can even point to many genuine failures of the mainstream medical establishment. The only problem is that his recommended alternatives—e.g., give kids vitamin A instead of the measles vaccine—are horrifying and will return humanity to the dark ages.
You, likewise, can endlessly point to genuine failures of mainstream academic science. But the wholesale destruction of academic science is no longer hypothetical. And this recruiting season, we’re already getting a glimpse of what’s going to replace it: not better science, not more efficient science, but nothing. No more pipeline for basic research in math, physics, or computer science, unless maybe someone convinces Elon that algebraic number theory could lead to larger, more phallic rockets. Just about anyone with the option will go to finance or Silicon Valley rather than science PhD programs. Traditions of knowledge that thrived for centuries will die, to be continued, if at all, only by the AIs.
At this life-or-death moment for science, you, just like RFK Jr., have chosen to cast your lot with the witch doctors. You’re of course free to do that, but you can’t expect the rest of us not to notice.
Comment #20 March 5th, 2025 at 2:34 am
Isaac King #18: I find it ironic that you should compare Sabine to Scott Alexander, given that less than a month ago, Scott wrote a stirring defense of publicly-funded research designed to appeal to Ted Cruz and other anti-woke Republicans.
(Or, as the MAGA forces gleefully march the whole of academic science to the gallows, good and bad and everything in between, are we not supposed to notice who’s advocating for our pardon and who’s volunteering additional arguments for our execution?)
Comment #21 March 5th, 2025 at 3:59 am
Scott #19: The video from February 27th which you mention here
should be the reason why you publicly attached Sabine on February 10th with similar words like you used in the past to attack Jill Stein and RFK Jr?
Maybe your attack was one of the causes why she made that video in the first place? Yes, it is well known that she opposes building the next bigger particle collider. You are in favor, without even knowing the details of that collider. Maybe Sabine’s opinion could change, if the proposal were really good and made sense. I don’t see how your opinion could change, because your support for it is based on support for the group behind it. But that group has not treated Sabine well, so it is little surprise that this won’t cut it for her.
It is a different story that Sabine has not yet realized that the destruction which the new administration will bring is unrelated to any rationalizations you might find.
Relatedly, Barandes mentioned your “The Ghost in the Quantum Turing Machine” in the video. I asked the following AMA question on March 1st:
Comment #22 March 5th, 2025 at 6:14 am
I still cannot fathom why you would disregard the only way to formulate quantum mechanics in a way that is consistent with the locality of Einstein’s theory. I can understand that you may not think this is an important point, fine. But calling it insane is somewhat… bold.
Regarding the public good. I did address that point in the video, but maybe you didn’t notice, so let me try it again. You have known me for a while, so hopefully you can believe me when I say I come at this from a theoretical, not from a political perspective.
In my eyes, the problem with the current organization of basic research is that much of the supposed public good isn’t good for anything. It may be a public good in the economic sense, but it’s not a good that anyone asked for. The public neither needs nor wants it, yet they are forced to pay for it.
In my video I am pointing out that the arguments that people in academia make for why this is a good system will no longer work. It’s too wasteful. You need to come up with better arguments than the “public good” argument.
The question is then, how do you decrease the part of basic research that no one has use for?
Daniel Litt’s answer to this was literally “we agree on what we mean by good science”. Well, your previous reply to me proves that we do not. Clearly that doesn’t solve the problem. So how do you solve the problem?
I am pointing out that for one thing companies do invest into basic research (in fact just statistically most basic research in the USA is done by private business). But more importantly, if there are supposedly so many people who understand the value of basic research, they will support it. I am sure that if someone got on the task they could find a way to leverage that agreement. If Trump can raise support with crypto, why can’t scientists?
In any case, this is where I see things going. Massive defunding of academia, a shift to crowdfunding in some sort. I don’t think crypto is going away, rather the opposite. As I said, I don’t know if this will be a net good or bad, but someone’s going to try.
I don’t understand why you are blaming me for the US billionaire problem.
In any case, let me assure you I understand that you want to hate on me because I am a nasty European who is curious to see how the defunding of US academia will work out. I am neither surprised nor particularly offended because I know you feel strongly about these matters.
But the idea that I am pushing “alternative theories” on YouTube is offensive. I barely even ever talk about superdeterminism, despite the fact that people constantly ask me about it. I have in the past 5 years made maybe 2 videos about it. I do for the most part talk about other people’s ideas and explain what the problems are. I just yesterday had an update on the Majorana-1 story. Today’s video explains why I don’t believe post-quantum gravity will survive for long. Tomorrows video is a criticism of a climate science paper. You get the idea: This is what I mostly do and I do think that your dismissive comment about my efforts in science communication are off the mark.
Comment #23 March 5th, 2025 at 6:59 am
This is just getting increasingly bizarre. I have literally burned papers with ‘fringe ideas’ on camera, yet Scott cannot find a fringe idea that I dislike?
I also made quite a few videos on what you could call ‘mainstream’ in the foundations.
Eg, I recently did a video about a proof on the uniqueness of the Veneziano amplitude that I found quite interesting. I stumbled across this because Peter Woit seemed to not like it, but when I read the paper I didn’t share his opinion. I’ve also talked twice about the amplituhedron stuff, in both cases to say that it seems to be genuinely useful but it remains to be seen whether it’s any good for quantum gravity. And last week I talked about a rather technical proof about the GHZ theorem. Again, I did my best to convey to people why I think this is interesting and/or useful.
So really I think that the only way you can make such a criticism is to simply not follow what I do in the first place.
Comment #24 March 5th, 2025 at 7:52 am
Re the academic model of scientific research, two words: string theory.
It could be however that the problem is that humanity is reaching its level of incompetence, or less disparagingly, that the far frontiers of science are too difficult for the best of us–which probably had to happen sooner or later. This universe is not a simple system.
My worthless guess on super-determinism is that it is the other way around from Dr. Hossenfelder’s opinion: that General Relativity’s determinism has to be altered, not QM.
On Dark Matter, on Dr. H’s recommendation I started reading Dr. McGaugh’s website (Triton Station) and in my also worthless opinion, he seems to be correct that MOND’s record of successful prior-predictions vs. DM’s zero (although DM has several free parameters for post-predictions and MOND has just one) is trying to tell us something fundamental. (No, MOND is not a complete theory which always works. Neither is Dark Matter.)
Comment #25 March 5th, 2025 at 8:24 am
Scott #20: I haven’t watched her most recent video, I didn’t realize your complaint was about that one in particular. If so, I think your framing in the original post is a little dishonest, or at least misleading. You don’t frame it as a disagreement with her video about defunding academia, you pretty explicitly present it as her promoting “alternative” (i.e. pseudoscientific) theories. It is this criticism that I claim is unfair. I believe this is also the criticism that she’s primarily responding to in her comments here.
I know very little QM, so I can’t speak meaningfully about the details of superdeterminism, but from an outsider’s perspective, it seems obviously more mainstream and more plausible than something like free energy devices or other stereotypical crackpottery. Its Wikipedia page, for example, presents it as a perfectly respectable theory.
Regarding Sabine’s most recent video, having seen some of the discussion on Twitter, I’m inclined to think that you and Daniel are correct; scientific progress seems like a pretty standard public good, and Sabine is refusing to engage with her critics there and her own sources on what “public good” means seem to contradict her. If your original post had said that her understanding of economics was poor, or that her idea would be harmful to scientific progress, I wouldn’t have objected. But you made a much stronger and more reputationally-damaging claim: that she’s a physics crackpot in the same way that RFK is a medical crackpot.
Comment #26 March 5th, 2025 at 9:06 am
Hello! Long time reader / first time commenter. (I think!)
First, thanks to you Scott, Jacob, and Curt for an intelligent, intellectually scrupulous, and inspiring discussion. These days, putting these sorts of things out into the world feels like a warm ray of sunshine. I really appreciate it!
Second, RE comment #16 from Mitchell Porter: Mitchell, I’ve been a fan of your comments many times on this blog. Your concern about a possible inconsistency between the existence of trajectories and the workings of indivisible processes is understandable, but there isn’t an actual issue. The picture is that the particle has a position at any instant in time and if you had a way to know the position then you could draw the trajectory. However, the indivisible process is governed by a law that tells you the conditional probability that the particle ends up at point \(x\) at time \(t\) *given* that it started at \(x_0\) at time 0. Using this law, you can compute the standalone probability \(p(t)\) that the particle ends up at \(x\) at time \(t\) if you are given some distribution over possible starting positions. However, you won’t be able to pin down a particular trajectory, or even a measure on the space of trajectories.
To sum up: there are trajectories in this picture, but there is no guarantee of a well-defined measures on the space of trajectories. However, one can show that in the usual sorts of circumstances—for instance, when there are frequently occurring division events, such a measure naturally emerges on the effective space of trajectories.
Jacob and I make these points in a paper on a different interpretation (https://arxiv.org/abs/1807.07136). I also encourage anyone to really read Jacob’s stochastic-quantum paper (https://arxiv.org/abs/2302.10778) and subsequent work. If you have questions, Jacob and I are very friendly and would be delighted to chat via email.
Comment #27 March 5th, 2025 at 10:14 am
Scott #12
I would not waste too much time worrying about folks getting corrupted by YouTube. Speaking as a merely highschool-educated curious 40-something, it’s usually pretty obvious whether someone is trying to communicate in good faith. So I think you have the causality wrong. While there are cesspools of garbage on YouTube, the people drawn to them were probably not interested in facts in the first place, and would not consume good content if the bad content didn’t exist.
I’ve enjoyed Sabine a lot over the years. I’ve even read her book Lost in Math and valued that perspective. But her recent shift towards ever more fringe contrarian opinions leaves me near the point of deliberately avoiding her work. That exchange with Daniel Litt crossed my feed and made me lose a lot of respect for her. It’s crazy that she can’t appreciate the scale difference between giving mathematicians a bit of coffee and building a supercollider megaproject. It’s clear she’s more interested in playing the victim (for the clicks) than having a good faith conversation. Except !! apparently she feels the need to play the same character on this humble blog, so who can say what’s really going on. I miss the old Sabine.
Anyhow, in this moment it’s perhaps too easy to lose faith in humanity. So hopefully a random nobody demonstrating reasonable judgement can preserve at least a little.
Comment #28 March 5th, 2025 at 10:19 am
Scott,
1. Try not to get upset with with the attacks of idiots, save your energy for those that at least have a clue.
2. As to comment #15, “Gil’s comments have gotten more and more wildly disconnected from the reality of what’s happening experimentally—e.g. Google’s demonstration this past year of a surface code qubit that gets better with increasing scale, which worked exactly like theory said it would in the 1990s.”
Maybe so, but it is also possible that you lose perspective by being an enthusiast and also too close to the action. I mean, is this a breakthrough that is going enable UQC (“useful QC”, doing something that someone cares about) in the next few years, or is this one of a million steps that might lead to UQC in the next few centuries, if ever?
Comment #29 March 5th, 2025 at 11:29 am
Scott, I don’t get the harsh words on superdeterminism. Shouldn’t we, fans of Peter Shor, let everyone adopt whatever interpretation they feel useful?
It feels a bit like different shade of leftists arguing what’s the true truth while antivaxers win the election.
Comment #30 March 5th, 2025 at 11:41 am
Sabine #22: It’s disappointing (but not surprising) to see these types of takes from someone with a substantial audience;
>In my eyes, the problem with the current organization of basic research is that much of the supposed public good isn’t good for anything. It may be a public good in the economic sense, but it’s not a good that anyone asked for. The public neither needs nor wants it, yet they are forced to pay for it.
You must surely know that R&D spending by the federal government is down 25% as a percentage of GDP versus the 1950s (and down ~66% versus the beginning of the 1960s. (https://ncses.nsf.gov/pubs/nsf23339) And frankly, the take doesn’t even make sense — the public didn’t know about, want, or need the internet until the Department of Defense created it. The same goes for nuclear weapons, spacecraft, microprocessors, largescale aviation, radar, GPS, etc. In fact, the public actively delegates the knowing and wanting away from themselves (hint: it’s called representative democracy).
But I’m not surprised you don’t mention that — after all, you’re financially incentivized not to. And I think largely that’s a major problem that people who get their “information” from YT face — is that the people they consume content from are not incentivized to be objective, or tell the truth, or present “both sides”, or dispassionately discuss scientific fact/consensus…they’re incentivized to drive comments, to drive clicks, to increase viewing/dwell time. You do this by way of creating discord, driving anger, increasing confusion, creating conspiracies where there are none, and obfuscating scientific consensus in favor of “This 1 Fact The String Theorists Don’t Want You To Know.”
You can’t hide your conflict of interest from everyone — we know that you gain financially from the racy fake news that you generate at the expense of people doing real work; they’re now paying the price as the federal government hacks and burns down the system that provided them with actual contributing employment (rather than just contributing to the outrage machine that you’re a part of).
Comment #31 March 5th, 2025 at 11:45 am
JimV #24
What are “prior-predictions” and “post-predictions”?
Kevin Zhou #7
Dark matter is the perfect theory to hate for conspiracy-minded non-crank engineering types. Its historical motivation readily lends itself to a negative, evil interpretation.
Astronomers do some calculations, realize that the amount of matter expected from the amount of light measured doesn’t match the amount of matter expected from the rotational speeds.
The basic scientific method interpretation of this situation is: the theory makes an unexpected prediction. Cool!
The emotionally satisfying conspiracy explanation is: the scientists realized their cherished theory was totally wrong, so they made up some dumb shit to patch the hole. The theory of dark matter is obviously a kludge to cover up their incompetence.
The fact that working astronomers are nearly unanimous in accepting the basic premise of dark matter is not evidence that the theory is good — it is evidence that the big-money scientific establishment has, in the interest of self-preservation, systematically shut out dissenting voices.
Critically, this is a non-crank position. Quantum mechanics and relativity are contrary to “common sense”, but from an engineering perspective the proof is ultimately in the pudding: lasers, satellites, etc. Dark matter on the other hand, is not connected to anything tangible, leaving space for skepticism and bad-but-not-crazy philosophy of science.
Comment #32 March 5th, 2025 at 11:46 am
Raoul Ohio #28: The countless quantum computing hypesters I’ve skewered here over the past 20 years would surely be amused to hear me called an “enthusiast.”
Mostly, I just look at the fidelity of 2-qubit gates as a function of time. I see that after 30 years of continual progress, it’s now mostly to what you would need for scalable fault-tolerance, but not quite all the way there, with zero surprises along the way that would change the basic theoretical picture from the 1990s. My best guesses about the next decade all flow from that.
The Kalai/Goldreich/Levin/Alicki position, however, is not about the next decade: it’s that scalable QC can never work. My confidence that they’re wrong comes from the rock-solidness of the underlying theory, combined with “never” being a long time.
Mostly, though, I’ve decided that this debate is no longer worth my rhetorical energies, since the engineers now have billions of dollars with which to make the argument for me, much more convincingly than I could. 🙂
Comment #33 March 5th, 2025 at 11:58 am
Sabine, I would just like to leave two empirical observations.
First, although you often say that your message is nuanced or even-handed, none of this nuance ever seems to reach your fans. I’ve met them everywhere, at dinner parties, university events, and all around the internet. They all sound extremely similar. They tell me, for instance, that you showed them that superdeterminism is obviously the single correct interpretation of quantum mechanics — but not a single one of them knows how it differs from regular determinism, even though that’s a crucial distinction! They also tell me that particle physicists don’t do anything besides SUSY model building, even though papers on that subject have fallen over 90% in the past decade. They tell me that we all want a 100 km collider, with no regard for public expense. But the consensus of America particle physicists in the 2024 P5 report was that a 100 km collider is indeed too expensive, and we should look into more efficient alternatives like muon colliders or even plasma wakefield acceleration.
If you intend to communicate nuance, does it not alarm you that none of it is apparently received by your audience?
Second, most physicists of my age and younger are afraid to be seen criticizing you. On Twitter, they are afraid that you’ll Tweet something like, “this whiny little physicist is afraid of the truth”, and thereby direct your enormous fan club to attack them. On YouTube, it doesn’t matter, as any comment critical of your videos will get piled on and downvoted to invisibility. However, this produces an echo chamber effect, whereby your average fan believes that there are _no_ legitimate counterarguments to your claims, because they never see them. Perhaps it also prevents you from seeing those counterarguments. Could this ever become a problem?
Comment #34 March 5th, 2025 at 12:29 pm
Nick @ 31, please take a look at Triton Station for definite examples. From my understanding, both the second peak of the CMB spectrum and the current JWST data were predicted in advance by published papers which used Newtonian and Mond calculations without any dark matter. Contemporaneous papers assuming dark matter gave a range of values for the former and contradicted the latter, although efforts are now being made to fine-tune DM properties further and/or reject the data. (Prior means before the data was available and post means after the data was available.) (I suppose post-dicted would be a better word.) (DM however can be made to fit the first several peaks well, whereas MOND with its single parameter consistently under-predicts the peaks after the second. Looking at the CMB alone, DM would be reasonable.)
In a data-vacuum it is natural to find DM more sensible than MOND, as was also my first reaction. The issue is, why does MOND work at all, for thousands of spiral galaxies (basically all of them for which there is good data) and in several different situations? What is it telling us? Meanwhile, after extensive testing, no viable DM particle has been found.
Comment #35 March 5th, 2025 at 12:44 pm
JimV #24 and Nick Drozd #31:
This is precisely the kind of one-sided narrative I was complaining about. Yes, if you don’t have anything besides galaxy rotation curve measurements, you can’t decisively favor DM over a modification of gravity. The fact that many more people work on the former today is a consequence of a lot of data collected over the past 50 years, which is always left out of the story told on blogs and podcasts. For example:
– 50 years ago, the null hypothesis was that there was just normal matter we weren’t seeing, and astronomers worked for decades to rule that out.
– 50 years ago, we didn’t actually know if general relativity worked that well. After many precision experiments, we learned that it works extremely accurately within our solar system, with some relativistic effects predicted at the part-per-million level. We even saw gravitational waves behaving exactly as predicted. These facts killed a lot of modified gravity theories.
– 50 years ago, cosmology wasn’t an actual field of science. Today, it is one of the most rapidly developing fields of physics, with a huge amount of quantitative data. Standard general relativity plus a few parameters (including dark matter) fits thousands of data points from diverse sources exceptionally well. As usual in science, some anomalies exist at the edges of our sensitivity. But modified gravity theories are in a much worse situation. Either they don’t fit any of the data at all, or they sneak in a new field that behaves _exactly_ like dark matter but call it something else.
– And of course, we discovered all sorts of weird systems that are hard to make sense of with MOND, like the Bullet cluster and the weird galaxies that seem to have “no DM”.
If you’re not told any of this, then it’s indeed inexplicable that so many people think DM is more likely, so you start conjecturing fraud or peer pressure. That doesn’t make sense for two reasons. First, why don’t you try to guess the total amount of funding ever given for (1) all dark matter searches combined, (2) the Deep Underground Neutrino Experiment, (3) the James Webb Space Telescope, and (4) the Space Shuttle? You’re pretty well-read people, so it shouldn’t be that hard to at least get the order of magnitude right.
The answers are (1) $0.1 billion, (2) $3 billion, (3) $10 billion, and (4) $100 billion.
There has simply never been that much money in searching for dark matter, and the vast majority of the money that does go into it goes straight to buying equipment, not paying salaries. People work on it _despite_ its low funding because they think it’s such a compelling idea!
Second, the peer pressure angle doesn’t make sense, because scientific glory comes from being the first to realize something, which encourages pursuing directions that others _aren’t_ following. I myself have tried to get MOND to emerge from a weird theory I was working on, though it didn’t work out. And I can recall quite a few well-known professors who have thought about testing modifications of gravity, but most don’t stay on the subject because, again, it doesn’t work out.
You will never hear any of this from Stacy McGaugh. And I know, because when I was younger, I was deeply curious about MOND and made a point of reading Triton Station regularly for a few years. But after a while, I started seeing how one-sided and repetitive his blog posts were. Seriously, have you noticed that he rehashes the same 5,000 word rant every week? He literally has 3 old figures that he posts again and again. He has written dozens of times about a supposed 1980s conference hallway conversation where he was right and a random other scientist was wrong. Another favorite topic is to imagine what a DM proponent would say, and then rant about how closed-minded and dumb it sounds to him.
If you read Triton Station for years, you won’t learn much, but you’ll just get angrier and angrier. As an alternative, why not try reading McGaugh’s actual MOND papers? I have. For example, let’s take this one, which McGaugh often touts as his greatest triumph: https://arxiv.org/pdf/2009.11525
The introduction and conclusion make bold claims about how this work shows MOND works perfectly. But the raw data in figure 5 shows a different story. The prediction of vanilla MOND is on the line “e = 0”. That clearly isn’t right; many data points are many standard deviations away. The “improved” prediction of MOND with external field fudge factors is the line “e = e_env” where e_env is a positive galaxy-dependent parameter. Does that look right to you? It barely improves the fit; many data points are many standard deviations away. Also, many data point strongly favor negative e, which is impossible within MOND.
This is the true state-of-the-art in describing galaxy rotation curves with MOND. It’s not particularly accurate, even when you do add fudge factors. And the blogs will never tell you that.
Like that bell curve meme, you have three choices. First, you can simply not bother to have an opinion. There’s no shame in that! Second, you can believe whatever narratives you hear from full-time bloggers and podcasters. You can live in a comforting, simple world where every answer is obvious and anybody who disagrees is a liar and fraud. Or, third, you can think for yourself, by reading actual papers and looking at real data. Which will it be?
Comment #36 March 5th, 2025 at 12:54 pm
Kodlu and Scott (#14,#15), overall, Scott’s comment reflects well (without being hostile or offensive) the nature of our current disagreements. Scott (as other members of the quantum computing community) thinks that the recent experimental claims, and mainly the claims by Google, come close to refuting the skeptical views. My view is different and I am fairly critical of the Google supremacy claims (in a similar way to criticism made by various researchers against Microsoft topological qubit claims).
I haven’t studied Google’s surface code paper and I agree that it is in some tension with my skeptical theory. (But unlike supremacy claims it does not refute my argument or come close.) One technical question I am curious to know is what can be expected from distance-5 surface code in terms of the quality of logical qubits. 13 years ago both Scott and I listened to a lecture in Jerusalem given by John Martinis in the inaugural conference of our quantum science center. John described a plan for logical qubits based on distance-5 surface code that gives 1-2 orders of magnitude improvement compared to the physical qubits. This assessment was based on optimistic (but not unrealistic) expectations for the fidelities of the individual components (in 2012) and on computations from a paper by Fowler, Mariantoni, Martinis, and Clelandand. I wonder what the theory precisely gives on the quality of logical qubits based on distance-5,-7 surface code as a function of the noise parameters. (I am sure some readers here can answer from the top of their heads.)
Of course, I welcome technical discussions both on matters related to the large debate and on matters related to the evaluation of Google’s effort (and Microsoft’s and other groups). I had some very useful public and private discussions with many colleagues (and SO was quite instrumental for discussions). It goes without saying that nobody is obliged to debate with me and I share Scott’s belief that matters will become clearer in the coming decade. There are a lot of exciting questions (small and large) related to quantum computing, to the debate about quantum computing, and to related questions on quantum physics. As Kodlu commented, I wrote a few posts devoted to this (presenting views of people on both sides of the discussion). As I often said, I am curious not only in Nature’s ground truth but also in the many points of view of different people, and I am also curious about connections to mathematics.
Comment #37 March 5th, 2025 at 12:54 pm
YouTube Content Creator #27 and Kevin Zhou #33: I’ll post more replies later. For now, though, a huge THANKS to both of you for helping to reassure me of my sanity, by affirming that you saw the same thing I did in Sabine’s evolution over the past few years.
With one side of her mouth, she rallies her hundreds of thousands of fans to burn down the corrupt, elitist, discredited, moribund scientific establishment, with all its terrible ideas like standard quantum mechanics and dark matter (!), and she provides ideological cover to the empowered ignoramuses who are indeed right now setting our little world ablaze. With the other side of her mouth, she comes to this blog and says “who, me? no, this is all a misunderstanding!” Again, just like RFK Jr.: “DON’T VACCINATE YOUR KIDS, IT CAUSES AUTISM!” / “who, me? No, I’ve never opposed vaccines, I just want more research!”
I could accept Sabine doing one or the other, but not both.
Comment #38 March 5th, 2025 at 1:01 pm
Kevin Zhou #7:
Ironically, this is not solved among scientists either. It’s well established now that surprising study results are more likely to get published in high impact journals despite questionable methodology, get cited more, and citations continue to grow even after the original study was refuted or retracted. This is worse in some fields than others, typically those fields with worse replication rates.
I think this is a mistake. The reputation of physics hasn’t fallen because lots of lay people are talking nonsense, it’s because most of the core problems from 50-100 years ago remain unsolved. Do you think we would be in this same situation if every new problem physicists encountered was conquered within 10 years? I doubt it. Doesn’t it then seem reasonable for an outsider to conclude that physicists might be stuck in some local minimum and some out the box thinking might be needed?
To be fair, I also think that the replication crisis and science journalism has been so bad over the past few decades and really contributed to general distrust of science, some of which has bled over to physics.
To steelman this, I think when most people say DM is wrong or doesn’t exist, they mean “particle DM” and not the set of anomolous gravitational observations that physicists sometimes mean when they say DM. Too many people are imprecise when using this term. The observations for DM are obviously real and meaningful, but the skepticism around attributing them specifically to particles has good justifications (and bad ones too of course, both of which also apply to attributing it to particles).
Comment #39 March 5th, 2025 at 1:12 pm
Incidentally, Kevin Zhou #13:
For instance, there’s a whole genre of articles claiming that some experiment has shown that quantum processes can rewrite the past, foresee the future, or take a negative amount of time. The AMO physicists hopefully all know the real story is “our experiment checked that textbook QM works exactly as expected, but if QM _wasn’t_ true you’d need some crazy retrocausality to get the same results”, but they choose not to communicate that subtlety. It seems to be a game they play to get into top journals. But then the public just gets more confused, and convinced that physicists don’t know what’s going on.
On the basis of this one paragraph, I hereby award you the Shtetl-Optimized Order of Merit.
You are welcome to collect your award in person, over tacos and beer, if and when you’re ever in Austin.
You are also entitled to write a guest post on this blog whenever you wish.
Congratulations! 😀
Comment #40 March 5th, 2025 at 1:27 pm
If parallel universes is a thing, everyone of us will eventually see its reality by noting that they’re surviving against all odds (survive multiple diseases, multiple close call accidents, …) into very old age while everyone else around them dies according to the average life time.
I.e. if I believe there are multiple versions of myself in separated branches, all claiming and feeling that there are “me” (rightly so, e.g. many of them will all have the same memory about reading this very post), if we wait long enough, one of these versions of me will be the very last one that survives the longest, and, for most people, it means living into the very end tale of life time distribution (around 130 years old and possibly longer).
The downside is that this apparent “luck” is most likely a curse (surviving as nothing to do with quality of life).
Comment #41 March 5th, 2025 at 1:38 pm
Scott #19:
de Broglie came up with the first superdeterministic theory in this “double solution”. Its simplicity is brilliant and there’s nothing insane about it. I find it ironic that physicists so readily accept the 4D manifold of GR, but seemingly have a conniption when some of the same concepts are applied at the quantum level.
Isaac King #25:
Sabine has acknowledged that some research is a public good (non-rivalrous and non-excludable). I didn’t see any evasion or refusal on this point. The contention has always been whether all such research is necessarily good for the public.
drift_agent #30:
You’re not making the point you think you’re making. Sabine explicitly said that governments would and should still fund research that has implications for public goods that are explicitly within their mandate, like national defense. Guess what all of the funding for the research you just pointed out came from?
Maybe considering a counterfactual would be helpful: how different would our world look right now if government hadn’t funded things outside of its clear mandate, eg. national defense, law enforcement, justice system, etc.? What innovations of the past 50 years would definitely not have happened because the first mover advantage was insufficient to recapture the costs?
Comment #42 March 5th, 2025 at 1:39 pm
Sandro #38:
> This is worse in some fields than others, typically those fields with worse replication rates.
I agree it’s a problem. Theoretical particle physics has its own solutions, but they’re unfortunately opaque. When I was a beginning grad student, I would often stumble upon some older paper in a top journal like PRL that sounded interesting, but inexplicably had very few citations. When I’d ask a senior person they’d say “everybody knows that paper is wrong, because of X, Y, and Z”. Now that I’m older, I often deploy similar warnings to students. The system kind of works, but it makes it very hard for the public to know what’s going on.
> The reputation of physics hasn’t fallen because lots of lay people are talking nonsense, it’s because most of the core problems from 50-100 years ago remain unsolved.
First, I think that sometimes the history of science gets oversimplified to erase progress. People like to say we already had the Standard Model in the 70s, but at that point literally less than half of its particle content had been discovered, and it was just one of many alternatives. The missing pieces were put together decade by decade through the 2010s.
Second, there’s a simple reason that we’re having trouble probing higher energy physics: in the past ~40 years, we have not gotten any better at building physical infrastructure. Measures of Western industrial production, such as energy consumption and steel production, have stayed flat or declined since 1980, and the Large Hadron Collider sits in a tunnel dug in 1983. That’s why the reach of particle colliders, which rose rapidly from the 50s to 70s, reached an inflection point in the 80s and has since barely budged.
Now, I’m not saying that the solution is to break the bank by digging an even larger tunnel. I’m just saying that in the long view, it’s not surprising that once you stop increasing inputs, you don’t get great outputs. You don’t need some deep reason to explain that.
> The observations for DM are obviously real and meaningful, but the skepticism around attributing them specifically to particles has good justifications (and bad ones too of course, both of which also apply to attributing it to particles).
Indeed, I’ve spent a lot of time thinking about how to search for exotic fields which might be responsible for the apparent effect of DM, using new precision experiments. However, the real obstacle to progress is that there’s essentially zero funding for any of these experiments. In my previous comment, I pointed out that the total amount spent on _all_ particle DM experiments was only 0.01% of the Space Shuttle’s total funding. The situation for alternative “wave-like” experiments is even worse, totaling to perhaps only 0.001%.
It’s a circular problem, due in part to the massive public interest in the subject. We spend, in relative terms, almost nothing on these experiments. Then, every day we don’t find anything, dozens of news articles are written, and thousands of people say the experiments must be a waste of time and clamor to defund them. It’s no surprise that progress is slow!
Comment #43 March 5th, 2025 at 1:40 pm
fred #40
that’s also probably why, whenever a 120 year old is asked what they did to live that old, they always say something along the lines of “nothing special, except maybe having a glass of brandy every night before going to bed, and I smoke a cigar once a while”.
Rather, we’re really just living in the same branch as a very “lucky” version of themselves… i.e. every branch has its own longevity winners, and we’re the winner in one of them out there 🙂
Comment #44 March 5th, 2025 at 1:41 pm
Kevin Zhou #35
My comment #31 was maybe unclear. I am NOT arguing against dark matter. I was just trying to lay out what seems to me to be the motivation behind anti-DM skepticism, which I consider to be mostly bad. Bad, but not crazy, which contributes to why you find it so frustrating to argue against. Contributing to your frustration is the last thing I would want! (Please re-read with the assumption that I agree with you.)
Personally, I don’t understand why people get so worked up about the possible existence of dark matter. The proposal is that there is stuff out there that interacts with gravity but not with light. People hear this and they go nuts because it is different from human experience. Okay, but why should we expect that physics across the universe comports exactly with human experience? What is this, Kant?
Further, if the existence of dark matter is obviously false, then that means the non-existence of dark matter is obviously true. This is equivalent to saying: it is obviously true that all stuff that interacts with gravity also interacts with light. Nobody ever says that out loud, because it sounds stupid, but that’s what the position amounts to.
Comment #45 March 5th, 2025 at 1:47 pm
Scott #39: Thanks! I’ll take you up on that if I ever visit. Fingers crossed that in the next few years, I write a paper interesting enough to UT Austin’s particle theory group to merit an invite. 🙂
Comment #46 March 5th, 2025 at 1:53 pm
It can be interesting to ask – if quantum computing is a huge win, why hasn’t nature taken advantage of it when evolving brains?
Maybe it has and brains aren’t “classical” systems but quantum computers?
Maybe it’s just too hard to achieve using normal biochemistry, but nature is doing it in two steps – by first evolving the most complex classical brains (humans), then those brains are building artificial quantum brains (QCs)?
Or nature hasn’t used it because the usefulness of QCs is too limited for the vast majority of biological organisms.
Comment #47 March 5th, 2025 at 2:23 pm
Regarding random idiotic online commenters, the silver lining is that it should be relatively easy to use an AI to engage/argue with them until they’re physically/mentally exhausted (unless of course an AI is also used to write their own idiotic posts).
Comment #48 March 5th, 2025 at 3:01 pm
Sandro #41: No, I disagree. The idea that something funded by DoD is definitionally related to the national defense is just inaccurate. But sure, have it your way, let’s just move all science funding to DoD and then your requirement is met.
In reality, the internet is an adjacent technology that arose from a non-defense need within DARPA. There are many other examples of government funded technologies that arose out of non-specific funding; Doppler, MRIs, flu vaccinations, closed captioning, barcodes. Perhaps you think that the development of Goodyear tires vis-a-vis Nasa was out of scope.
The problem I have with this whole ‘efficiency’ theory that comes out of Silicon Valley tech-bro-speak is that it completely ignores the other side of the same ethos with regard to the federal government. For some reason, in SV, only tech startups should be allowed to fail often and fast (and to learn from those mistakes). The government should be perfectly efficient, constantly validating it’s existence, never making mistakes.
Comment #49 March 5th, 2025 at 3:22 pm
On this issue regarding Sabine’s video, I wholeheartedly agree with Scott. I believe Sabine has consciously or unconsciously fallen into the influencer pipeline and made a “deal with the devil.” However, I don’t want to be too harsh on her because, based on her comments, it seems she genuinely doesn’t understand why so many scientists are criticizing her.
One issue that, to me, needs to be clarified and that Kevin Zhou mentioned in #33 is how come she has this large follower base, that has developed such deep resentment toward science and scientists. Over the past centuries, incredible discoveries have profoundly transformed our understanding of nature, so it seems illogical that people have already forgotten this. I am sure that many still appreciate science, yet there is an undeniable hostility.
One possible explanation is envy or resentment toward those fortunate and intelligent enough to dedicate their lives to researching fascinating topics. But scientists have existed for centuries, and it doesn’t seem that people have always held this attitude (correct me if I’m wrong). Figures like the founding fathers of quantum mechanics, along with many lesser-known physicists who contributed to the field, were respected, or at the very least, not actively hated. Perhaps this varies by culture (for instance, in Germany, being a professor still commands deep respect among the general public).
I also don’t believe this resentment stems from financial concerns. Fundamental research doesn’t receive an exorbitant amount of funding, and the public is likely aware of this to some extent. Similarly, I doubt that lengthy discussions about dark matter are the real trigger.
It would be unfair to claim that scientists have actively promoted “wokeness” in universities, at least not in the sciences, but they have allowed (self?)-indoctrination to take root among students. Instead, they should have taken on a more parental role, showing genuine concern for the psychological and moral development of the younger generation. Turning academia into an industry certainly hasn’t helped.
This backlash, this wake-up call, did not come from scientists, but from billionaires and corporations, which are now falling in line with these cultural shifts faster than universities. The public’s frustration isn’t driven by a belief that academic discussions and research topics are useless, but rather by a sense of betrayal. Instead of standing against DEI initiatives and ideological conformity, many academics (virtually all) have bowed their heads and played along. And for what? Unlike those in truly oppressive regimes, they weren’t forced into compliance at gunpoint. They went along merely out of fear of losing their jobs, while others suffered real injustices under DEI policies.
What the public sees is: “These scientists are no better than me. They don’t deserve to spend their lives doing what they love any more than I do, especially not with money that could have gone toward something I value for myself instead.”
Intellectuals, the very people who should have resisted and guided society away from ideological capture, are now seen as having broken the social contract. “We fund you with public money so that you don’t have to fear speaking the truth, not just in your research, but in your broader intellectual life”. And now, like someone discovering they’ve been cheated on, the public is retroactively reevaluating their past admiration for them.
They merely use Sabine as a post-facto justification for the backlash they already wanted. At the same time, they still seek validation from experts, so when they find one who tells them their resentment is scientifically justified, it only strengthens their conviction. Sabine is providing them with a rationalization for what they already wanted to do.
@Scott: I only recently tuned into this discussion. If what I’ve said is redundant or outdated, feel free not to post it.
Comment #50 March 5th, 2025 at 3:24 pm
By some puzzling coincidence the (failed) scientists who are relentlessly attacking the science institutions (which treated them unfairly by ignoring/stealing their breakthrough theories when they were Phd students or refused them the Nobel when they were ‘professional’ scientists) are always the ones trying to make a living from the monetization of their Youtube channel…
Comment #51 March 5th, 2025 at 4:05 pm
According to what I have read at Triton Station, over the last year or so, while many of the experimental searches for DM particles (such as observation in deep mines) are not very expensive, what has been done so far comes close to ruling out all proposed candidates. The data Dr. McGaugh shows on galaxies looked like a good fit to me, due to inherent difficulty in getting precise astronomical data. That is, there is scatter about a trend curve, but that curve is the one predicted by MOND calculations. He shows that presumably better data from closer galaxies or from better equipment is a subset with a much better fit. (Metallurgical fatigue and creep data on GE proprietary alloys taken over 50 years–prior to Jack Welch–has more scatter, while costing a lot for 100,000 hour tests.) Particle-physics standards for data are rarely applicable elsewhere, I think.
Recent posts have been about the JWST data, and his recently published paper on it. I don’t get the impression that other commenters here have engaged with that data yet. He also has done some recent work on gravity lensing calculations. He has reviewed and discussed fairly recent wide-binary studies, and his own calculations on out-of-plane motion in the Milky Way galaxy (they seem to fit basic Newtonian calculations, on stars whose in-plane motion fits MOND.) Maybe once a year or so he reviews the MOND story from the 1990’s on, and occasionally puts an old anecdote in a footnote.
Comment #52 March 5th, 2025 at 4:06 pm
DM, MOND, and all that:
The only thing we can say for sure is that the simplest situation (no DM, no MOND, no nothing) doesn’t work.
Thus we look at the next simplest solution first: DM. A key fact is that you can’t see DM. The good news is that you can thus assume it is wherever it needs to be to make observations work. DM is the best candidate and my guess for the probability of it being correct is 75%. The bad news is that in 50 odd years, zero evidence has emerged that DM actually exists, which is why my guess has gone from 90% down to 75%.
So, if not DM, then what? MOND is pretty interesting and at least plausible, and my guess for it turning out to be right is 10%.
I am leaving a 15% chance of “something else”, or that we have a total misunderstanding of what is going on.
Comment #53 March 5th, 2025 at 4:29 pm
Sabine #22
You claim that the public doesn’t care to fund science, which I honestly doubt but sure let us grant this. Do you not agree with the public? If not, why do you not use your video to make the case against them, rather than signal-boosting their general misunderstanding and also saying that most science is bullshit. If you do agree with the public, then stop hiding behind this whole “oh I’m just pointing out that other people don’t think funding science is worthwhile” shtick and just say what you believe. If you think that the government should defund higher education and basic research then simply say this. But now you are indeed just pushing anti-science rhetoric, saying that science is mostly useless and we shouldn’t fund it. Understand that this is what you are doing and own it, or stop doing it.
Comment #54 March 5th, 2025 at 6:16 pm
JimV #51: These are examples of the one-sided narratives I was warning about. According to Triton Station, if there’s a single galaxy that seems unusual under the DM hypothesis, then DM is dead. If there’s a single fine feature of galaxy structure that isn’t nailed by simulations, like the number of satellites, then DM is dead.
But when you turn to MOND, the standards suddenly get much lower. MOND does not even remotely fit the vast majority of galaxies — it’s not just that it doesn’t match in precise detail, the basic features are order-one off and far outside observational error bars. (This is precisely what I was trying to get at when I suggested you look at the actual data.) The only observational triumph of MOND is that, when you average over many galaxies, sampled in the right way, blow up the error bars, put everything on a log scale, and zoom way way out, it looks like they satisfy the radial acceleration relation. If you go back to a normal zoom level, you’ll find that what it actually says is “we get the average right to within ~200% error”. By contrast, standard cosmology nails down the DM fraction to better than 1%. And forget about predicting fine features; MOND doesn’t do those at all, so it can’t be criticized for getting them wrong.
I think it’s good that some people work on MOND, and I give something like it a nonzero probability of being correct. But the way it is promoted is systematically misleading and basically designed to encourage distrust in science at large. If a bunch of careful and dedicated people (with no financial stake either way!) believe something you don’t, the default question to ask is what evidence convinced them. Instead, MOND popularizers encourage their followers to forget about such questions and just assume the other side’s all liars and frauds.
Comment #55 March 5th, 2025 at 7:00 pm
Ever since I started out on the tenure track in 1971, the subject of classicalization has been at the top of my topics. This came through studying intramolecular vibration relaxation of small isolated molecules (e.g. methyl formate, benzene, toluene, vinyl bromide.) Classicaliation is the heart of “the measurement problem” which really is no problem. A collegue, a NAS member, also does this, and we agree.
This discussion misses Barandes’s main point, which is that one can come up with the Schrodinger equation and Copenhagen, starting with randomness with some odd preconditions … which oddities lead to the oddness of QM and presumably quantum field theory too. This is not a big deal because it matters. It does not … as mentioned , Barandes predicts what QM does, including entanglement. Its a big deal because it offers a mathematically satisfying statement.
On a related subject, Strassler’s current thread, toy that it is, is a very good and accurate statement of what measurement really is.
This whole discussion is not anywhere remotely related to Woit and McGaugh.
That its not is relevant.
Including Sabine is not relevant.
Comment #56 March 5th, 2025 at 7:48 pm
The ToE discussion was enjoyable but it never got to anything substantial despite the absolute journey that it was.
For example, what’s the key intuition behind why indivisible stochastic dynamics are able to model quantum interference? At what point do conditional probabilities overlap with amplitudes?
And what would a simple quantum circuit look like, if given as an indivisible stochastic process?
I guess I differ from you in that i’m specifically not interested in a comprehensive reinterpretation/reformulation of QM. It’s kind of how i’m not interested in a proposed proof of P!=NP or P=NP, beyond the supposed secret ingredient.
What’s the point of wrapping a key insight in a contrived, over engineered and ultimately insufficient package?
Any potential overlap between classical probability and quantum interference has my immediate attention, but you didn’t seem too enthusiastic about it.
Comment #57 March 5th, 2025 at 9:05 pm
Kevin Zhou @7: Your typical physicist has essentially no incentive to care about the opinion of anyone that’s not a physicist. Will doing outreach to the public help them get tenure or otherwise improve their professional reputation? No, and in fact it’s probably more likely to hurt than help.
https://www.astralcodexten.com/p/on-priesthoods
Comment #58 March 5th, 2025 at 9:08 pm
drift_agent #48:
That’s not what I said, I said that every example you had cited very clearly fell under the purview of national defense. That’s not an argument to “move all research under defense”.
Doppler clearly falls under defense. Vaccines could partially too, eg. don’t want your army taken out by the flu. All of the others, even vaccines, have clear market value that people would pay for. That they happened to have been publicly funded is not an argument that they would not have existed otherwise, which was the point of my question.
This is not really about efficiency or mistakes but about feedback loops. Markets are efficient resource allocators because there are direct feedback loops that enforce accountability, eg. if I waste time and resources and have nothing to show for it, I get reprimanded or fired. Are there direct, effective feedback loops in a sprawling government bureaucracy? Seems like they’re mostly diffuse and indirect, and even more so for public funding of research. Elected representatives by and large have no idea what’s being done with funds, so the democratic feedback loops that are supposed to keep government accountable don’t work here.
Even if public funding of basic research were kept, I think there are good arguments that it should take a different form. Sabine is correct to point out that it’s mostly acting as a planned economy, but in many other spheres, we try to create an artificial market to solve our resource allocation problems (like carbon taxes). What would an artificial market solution look like for public research funding?
Kevin Zhou #54:
The reason why this is not as convincing as you portray is because the distribution of particle DM needed to fit observations cannot be predicted a priori, in general.
I also think your characterization of Triton station is a little unfair. McGaugh has always seemed pretty clear that he didn’t believe MOND was the final answer, but that it has made some surprisingly successful predictions that it had no reason to get right, predictions that particle DM didn’t and couldn’t make, and that particle DM had to be fitted a posteriori to match many observations, all of which suggest that particle DM’s superstar status is kinda sketchy.
I also don’t think the notion that some institutions could get captured by some ideas is necessarily wildly off the mark either. It’s not like this never happens in science. The fraudulent amyloid Alzheimer’s research is the most recent example of 20+ years of wasted work. I don’t think skepticism of standard narratives carries an intent to engender distrust in science writ large. Skepticism is generally good, so I wish it were clear how do this better, eg. avoiding tribalism and seeing people as “opponents” to defeat.
Comment #59 March 5th, 2025 at 10:11 pm
“Instead, MOND popularizers encourage their followers to forget about such questions and just assume the other side’s all liars and frauds.”
My impression from Triton Station and from some Youtube videos arguing for DM is about 180 degrees from that. I have seen Dr. McGaugh vilified on Dr. Woit’s Not Even Wrong Blog (by an anonymous commenter). He has been called a liar in a conference in which he mentioned his paper that calculated the second peak of the CMB spectrum. (One person was good enough to read the paper later and apologize.)
As for being one-sided, I have stated that DM was the most reasonable a priori hypothesis, and does a much better job (after fine-tuning) on the CMB spectrum. I feel that you have been somewhat one-sided in basing your critique on a single paper, ignoring the JWST and gravity-lensing data, and in not allowing for the difficulty of precise astronomical data, when, for example, the exact inclination of distant galaxies with respect to our point of view can only be estimated. I guess people on different sides will always think the other side is one-sided.
I am in pretty much agreement with Raoul Ohio’s comment at #51, with some differences in the numbers. I don’t believe and don’t think Dr. McGaugh believes that MOND is the final answer, but rather an effect of some deeper issue.
Comment #60 March 5th, 2025 at 10:14 pm
drift_agent #30
You, like many other ill-informed people, confidently spread false opinions about me.
First you claim that I didn’t mention the US R&D investment and its breakdown. In fact I did, it’s here at 9:29
https://www.youtube.com/watch?v=QtxjatbVb7M
Second, yes, of course I make money with my videos. I do not get paid by taxes you know. But my main business is in science news as you can easily find out by having a look at my channel.
It has as a matter of fact been difficult to find sponsors to talk about academia. Why do you think I talk about it nevertheless, if I know that a big part of my audience — which has a large fraction of left-wing US academia — will hate it. Do you really think I cannot read the comments? Do you think I do not hear what people say about me? Why do you think I do this?
I have been talking and writing about the problems with academia for 15 plus years. (Maybe Scott remembers.) No one cared. I feel that now the time is right that people might actually listen, so of course I talk on YouTube about it. People still don’t want to actually hear it. But if they now get pissed off about it, I consider this great progress.
Let me repeat this again: I talk about this despite the very evident fact that my audience does NOT want to hear it. I talk about it because it matters to me. I talk about it because I am greatly worried by the lack of progress in science that I think is caused by the way we fund academia. I talk about it because scientific progress is more important than me.
You can disagree with me on that however your heart desires. But the idea that I am doing this to make money despite knowing how many people will get upset about it, and despite the rather obvious fact that I could instead talk — as Scott suggests — about yet another alternative theory of something, is nonsense. It’s a classic ad hominem argument.
Comment #61 March 5th, 2025 at 10:18 pm
Scott: You need to specify whether your criticisms of Dr. Hossenfelder are based on her physics videos or her non-physics videos. The physics ones are honestly mostly about *rejecting* fringe theories (after engaging with them on their own terms, first, which I really appreciate!) The others… I don’t watch many of them. I haven’t seen the single video you seem to be basing all of this on. But it’s entirely rude and inappropriate to compare any of her videos to RFK — he operates on a purity-based moral framework, with no grounding in science. Even if you don’t like what Dr. Hossenfelder considers in her videos, she’s at least approaching it and picking it apart from a scientific perspective.
Kevin Zhou: You wanted someone who watches her videos *and* gets the nuance? Here I am! I am such a person! I like some of her videos, others I find problematic. I’ve appreciated how she approaches issues from multiple sides and is often left with uncertainty. I don’t like the clickbait-y ones. The physics explainers are great. The comments below the videos? Mostly terrible! But you can’t really pick your fanbase. That’s just an unfortunate part of being a public figure.
Comment #62 March 5th, 2025 at 10:29 pm
Kevin Zhou #33:
I did talk about the P5 report and I did mention that Americans seem to prefer a muon collider so I’m not sure what your criticism is there. Though I grant you the possibility that you didn’t notice, which is fine because we all know I talk too much and make too many videos.
Let me make a general empirical observation: The vast majority of people do not understand the nuances of any argument they hear anywhere. This has very little to do with me. I am sure you would find the same thing in the audience of literally any other website or news outlet or YouTuber. In short, you are blaming me for human nature.
I try to do the best I can. Of course I don’t succeed. Every single video that I have ever made I have misgivings about. After they have run, I know how I could have done them better. The next time I talk about the topic, I try to do it better. This is how I got to where I am. I learn by doing.
I have to my best knowledge never personally attacked or insulted anyone. It’s just not what I do.
If someone posts nonsense on twitter they will have to live with the fact that I (or someone else) points out it’s nonsense. If they’re not prepared for that, they shouldn’t be on twitter in the first place.
And yes, of course superdeterminism is obviously the right explanation. It is obvious because per Bell’s theorem, it’s the only way to locally explain our observations. I cannot fathom how physicists cannot see this.
Sure I am aware of the “counterarguments”. Alas, I have addressed them all, multiple times. It’s just that people like Scott ignore them. Has it occurred to you that maybe the problem is that rather than actually engaging with my arguments, a lot of physicists just declare that I am insane and go straight to ad-hominem attacks. And then they complain that I am not being nice to them.
Comment #63 March 5th, 2025 at 10:45 pm
Scott #37
It is true what you say that I get a lot of comments of the sorts “dark matter is a fraud” etc. Needless to say, this is not specific to me, you find this in the comment sections on every article on dark matter etc.
I have in my videos explained numerous times that dark matter in astrophysics terms was a *good* hypothesis. I have noticed that people do not understand why I say this but I think it’s forgivable seeing that physicists, too, don’t understand it.
That said, I think you are giving me too much credit by claiming that I am the origin of these misgivings. The origin is all the crap that physicists publish. Which returns us to the problems with academia etc.
Basically, I see the same problem as you but have drawn very different conclusions.
Comment #64 March 5th, 2025 at 10:53 pm
Kevin Zhou #42
“When I was a beginning grad student, I would often stumble upon some older paper in a top journal like PRL that sounded interesting, but inexplicably had very few citations. When I’d ask a senior person they’d say “everybody knows that paper is wrong, because of X, Y, and Z”. Now that I’m older, I often deploy similar warnings to students. The system kind of works, but it makes it very hard for the public to know what’s going on.”
I know this wasn’t addressed to me, I just wanted to say that I have made the exact same experience. It’s why I think that training an AI on scientific literature without guidance by people who work in the field is going to cause a lot of trouble. I wish that sociologists would chime in here, because there is a lot of knowledge in science which is not to be found in the published literature. It’s in people’s brains. (Which are not public goods…)
Let me add though that there is a flipside to this, which is that any paper that hasn’t received at least 5 or so citations within a bunch of years will look fishy and will remain ignored because people suspect there is something wrong with it.
Comment #65 March 5th, 2025 at 11:22 pm
Drizzle McGloom #49
“how come she has this large follower base, that has developed such deep resentment toward science and scientists”
I don’t know why you would say something like this. My follower base is extremely pro-science. Why do you think someone who is anti-science would even watch a science news channel?
You people really need to stop inventing things about me and my audience just because you don’t like what I say.
Look, I personally engage with my followers almost every single day. They are mostly people who are interested in science on a level that is technically somewhat higher than what much of the popular science media aims at. That’s my niche — I do the same thing, but with a little bit more detail, and I offer my own opinions to go along with it. I always try to make clear where my reporting ends and the opinions start, and encourage people to disagree with me.
Several years ago I very consciously made a decision that I would not shy away from topics that I know are going to be controversial. It is primarily a problem with the left-leaning US-American audience, that I estimate to be about 1/5 of my viewership. Why this estimate? because I see in the stats that about 1/3 of my audience is based in the US and I expect that on a science channel, the majority of those leans left, but not all who lean left are of the crazy sort. So 1/5 seems roughly plausible.
I made this decision because I think we need a culture in which people can talk about what they think without being afraid of being personally attacked and I want to set an example for this.
As a European I also think something is very wrong with American debate culture, but that’s another story.
“One possible explanation is envy or resentment toward those fortunate and intelligent enough to dedicate their lives to researching fascinating topics. But scientists have existed for centuries, and it doesn’t seem that people have always held this attitude (correct me if I’m wrong).”
You should take into account that for most of this time, scientists were not paid by taxes. They were mostly rich people themselves or directly paid by rich people.
Does it really surprise you that especially in a country like the US that is very against social welfare in general, and that has “do it yourself” as its unofficial motto, people do not want to pay for academics writing useless papers.
You say they “research fascinating topics”. But for most research that isn’t so. Most people in academia do what they do because they get paid for doing it. It’s yet another job. And it’s a nice one because it’s reasonably well-paid and safe, as opposed to many jobs that the average American has to settle for.
And that’s the very problem: By paying this research by taxes on governmental programs, we have created a class of people that produce output that no one wants and needs — and the public pays for it. Why are you surprised that people don’t like that?
“They merely use Sabine as a post-facto justification for the backlash…”
The backlash hasn’t even yet fully begun. In any case, I am a small fish. I have next to zero influence of anything, so don’t worry. I would suggest that you just continue to ignore me.
Comment #66 March 5th, 2025 at 11:48 pm
Sandro #58
“Even if public funding of basic research were kept, I think there are good arguments that it should take a different form. Sabine is correct to point out that it’s mostly acting as a planned economy, but in many other spheres, we try to create an artificial market to solve our resource allocation problems (like carbon taxes). What would an artificial market solution look like for public research funding?”
Yes, indeed, this would be a possibility and is exactly the sort of conversation that we need.
I would like to use the opportunity to explain how I arrived at this rather unfortunate opinion as many think. I have spend a lot of time looking into measures for scientific success or impact. Indeed if you look at my publication list you will find that I wrote a few papers about this.
First I thought that the problem with academic incentives can be fixed by using better measures. Then I thought it can be fixed by not using any particular measure, but letting scientists create their own measures and combine them. (And I worked on a software solution for that.) Eventually I realized that none of that could ever solve the underlying problem, namely how do we, collectively, identify what “good” science even is?
Eventually it was my personal experience that made me realize that this is a non-problem. Because I see every day that a lot of people value science very highly. Indeed, this is also what the polls say. People want to learn how nature works and how the universe works etc. They do understand that in many cases it takes a very long time for such research to become profitable, and maybe it never does.
But if they understand that, then why do we need to force them to pay basic research by taxes? That makes no sense.
Now economists say you can’t monetize the result of basic research if it’s been published because it’s already out there anyway. But (a) scientific publications aren’t useful for most people and (b) as I mentioned above, a significant part of scientific knowledge never gets published. The real value of scientific research is locked away in the brains of people who have spent their life studying these problems. Like Scott, like me, like Daniel Litt.
This is literally how I make money: I sell the knowledge I have gathered.
Of course private business does that already. They’re buying in knowledge locked away in researchers like soccer teams buy players. But this also means there are multiple ways to use market forces to self-regulate science.
That said, I acknowledge that there are sectors where this will not work or not work as we want to for one reason or another. It could also to just not turn out to work, period.
But look at the world. Look at where it’s going. It’s seeing a strong trend to decentralization, to crowd-funding, to crypto. This is why I think academia is going to see a shift to a market-based self-regulation.
Comment #67 March 6th, 2025 at 12:03 am
> The fraudulent amyloid Alzheimer’s research is the most recent example of 20+ years of wasted work.
I completely agree, but we have much, much higher epistemic standards than biology and social science. In your average biology paper, nobody ever sees the raw data besides a single overworked grad student, and you just have to trust they haven’t falsified or manipulated the few images they display.
Meanwhile, experimental results in particle physics and cosmology are vetted by collaborations with dozens to thousands of people, which each contain many people qualified to understand the whole thing. The analysis is usually blinded, so that they don’t even know what result they’ve found until just before they’re done. Most important quantities are measured by multiple experiments which operate independently and cross-check each other. The standards for statistical significance are very high. The data analysis procedure is described in detail, the raw data is often released, and sometimes even the data analysis code is released, so you can reproduce it at home.
It is literally the most transparent and reproducible stuff in all of science. There is no room for faking data or p-hacking, unless you posit massive conspiracies.
> [MOND] has made some surprisingly successful predictions that it had no reason to get right, predictions that particle DM didn’t and couldn’t make, and that particle DM had to be fitted a posteriori to match many observations.
I mean, that’s definitely the picture McGaugh paints, but I repeat again that it’s a framing. For him, the failures of DM are cases where it doesn’t nail its prediction of every complicated quantity. The successes of MOND are cases where it gets a simple quantity (averaged over many galaxies) right within a very generous error bar.
For instance, DM is often criticized for not getting the number of satellite galaxies right. That’s a subtle quantity that’s hard to calculate and hard to measure (we keep finding new ones, so the gap keeps closing). But MOND can’t be criticized on the same grounds, because it makes no prediction at all!
With JWST early galaxy formation, it was a failure for DM because it didn’t get all features of galaxy formation _quantitatively_ right, and a triumph for MOND because you can _qualitatively_ argue that adding MOND effects pushes galaxy formation earlier. But we ignore the fact that the large-scale structure formation has never worked in MOND even approximately.
There’s a more subtle problem here, which is the implication that the existence of any anomalies at all in a theory that predicts many quantities is evidence that it’s about to fall apart, and resolving the anomalies means we’re just covering up the theory’s flaws. Actually, there has never been a scientific theory that wasn’t afflicted by tons of anomalies. For instance, people often say that we shouldn’t look for DM because standard cosmology is full of anomalies, and we shouldn’t look for particles beyond the Standard Model because the Standard Model explains everything perfectly. But it’s actually the other way around: there are many more existing anomalies in Standard Model measurements, with higher statistical significance! For instance, see the list here: https://hepcomm.github.io/hepmist/
Anomalies are a fact of life. Based on what’s happened in the past, it’s most likely that none of the ones listed above will overturn the Standard Model; instead, they probably result from bad luck, subtle differences in what experiments are measuring, or subtle physical effects not accounted for. Finding and explaining anomalies is the bread and butter of science. It’s not a new thing either. Scientists debated countless anomalies in special relativity, quantum mechanics, and general relativity for decades after their founding. Since the theories held up, we don’t bother teaching what the anomalies were, but then people get the mistaken impression that any good scientific theory must be perfect at conception.
Anyway, my point is, I think it’s good that people work on MOND. I even think it’s a good thing that some MOND proponents use motivated reasoning to trumpet how good it is! That’s all part of the scientific process; it’s fine if individuals are biased, because the best arguments will win after vigorous debate. In an ideal world, you would have Triton Station, but also 5 other blogs from other scientists arguing just as emphatically in the other direction.
The problem is that we don’t live in that world. In our world, almost every well-known podcaster or blogger strongly promotes the notion that most physicists are totally mistaken, in dozens of contradictory ways. (Partly because these folks tend to be less busy than most physicists, and partly because that message is inherently more viral.) Some of these folks are fair and well-intentioned. Others are less scrupulous and like to whip their audience into a frenzy. Regardless, the net effect is that every smart non-physicist who engages with popular science ends up thinking us physicists are idiots!
Comment #68 March 6th, 2025 at 2:48 am
Great discussion! Would love to listen to a debate between Jacob and a committed many world-er as well. Scott seemed to concede many of Jacob’s critiques of MWs, but a card carrying MWer pushing back would have been helpful in clarifying where the crux of the disagreement is.
Comment #69 March 6th, 2025 at 5:45 am
Comment #35 Comment #54 Kevin Zhou. From someone who has spent so much time criticizing erroneous opinions on the Internet I’m surprised to see you reserve those directed at Stacy McGaugh, his blog and even one of his paper for another blog; why do it here and not on Triton Station ?
Comment #70 March 6th, 2025 at 6:51 am
Comment on Gil Kalai’s skepticism (comments #14 and #15): I am commenting here from the perspective of a philosopher of science. A central question in scientific skepticism is whether the skeptic is genuinely open to being proven wrong or whether their position is, in effect, unfalsifiable. Is Kalai’s skepticism a rigorous scientific position, or does it function more as a logically structured belief system resistant to empirical refutation?
Kalai’s argument against scalable quantum computing rests on conjectures rather than formal no-go theorems. In physics and computer science, we distinguish between: 1. *Proven no-go theorems*—absolute mathematical constraints that explicitly forbid something, such as the no-cloning theorem in quantum mechanics. 2. *Conjectures*—theoretical expectations that something should fail based on classical analogies, complexity-theoretic reasoning, and extrapolations from current challenges but without a definitive mathematical proof.
Kalai’s skepticism falls into the second category—a structured and intricate set of conjectures, but still conjectures. If scalable quantum computing were fundamentally impossible, one would expect a physical law prohibiting it, akin to the second law of thermodynamics or the impossibility of superluminal signaling. Yet no such law has been identified, nor has Kalai produced a mathematical proof of an insurmountable constraint preventing fault-tolerant quantum computation.
Of course, shifting the burden of proof onto experimentalists is a reasonable stance—experimental failure would certainly validate Kalai’s argument. However, experimental progress has not yet provided a definitive failure of scalable quantum computing. If Kalai’s conjectures are correct, they must eventually manifest as an absolute roadblock in experiments. However, without a universal mathematical proof of failure, his argument remains an open hypothesis rather than a definitive theorem.
Kalai suggests that all NISQ devices are confined to the LDP complexity class, making error correction infeasible at scale. If this claim were rigorously proven, it would be of extraordinary significance—yet no peer-reviewed proof has been produced. Instead, Kalai’s reasoning, while logically structured, remains empirically unverified. The assertion that no future quantum device can surpass LDP is not a theorem but a testable hypothesis. If an experiment demonstrates computational power beyond LDP, the entire framework collapses. Given that the ultimate judge of computational feasibility is Nature Herself, we may soon see whether Kalai’s position withstands empirical scrutiny or remains an elegantly formulated, yet ultimately refutable, conjecture.
Comment #71 March 6th, 2025 at 6:53 am
Here’s my 2 cents to this discussion. The internet has enabled a phenomenal sense of “participation”, even “contribution” to a vast majority of people on any topic of their interest (or the topics the algorithms goad them into). In this scenario, it is the fringe that always wins over the mainstream due to the fact that mainstream is boring, needs rigor, incremental and time consuming. The fringe gives all the thrills of adventure and participation in that gives people a sense of purpose (who wouldn’t want to be part of the Theory of Everything and AGI ?). With even Nobel winners, especially physics ones, jumping on to this bandwagon (Roger Penrose, Gerad’t Hooft, even Frank Wilckzek), of pitching their fringe theories to unsuspecting public, gives it a lot of credibility in the minds of even the most educated lot.
In summary, whether in physics or everything else, human life seems to be already in control of algorithms, which try to find the local minima of the capitalism’s evolutionary instinct of greed
The solution to this train wreck seems to be way beyond the ability of humans as the deep learning algorithms ever increase the chasm, with the only hope that some sentient AGI finds a graceful way out of this mess….
Comment #72 March 6th, 2025 at 8:01 am
Scott #37,
“YouTube Content Creator #27 and Kevin Zhou #33: I’ll post more replies later. For now, though, a huge THANKS to both of you for helping to reassure me of my sanity, by affirming that you saw the same thing I did in Sabine’s evolution over the past few years.”
FWIW (it might be nothing) as I’m not sure how much you value or care about my opinion, but I think Sabine and Isaac’s responses are far more accurate and fair here. I don’t want you to question your sanity, but I think you should question your post and whether it was fair.
I too don’t think your framing in the original post is honest or fair and find it somewhat misleading. While you disagree with her about superdeterminism, it would seem your real beef here is with her last video which you deem not sufficiently denunciative of the disinvestment of taxpayer funded basic research.
Other than your strident opposition to her position on superdeterminism can you state any other *scientific* disagreements where you believe she has engaged in promoting pseudoscience?
Comment #73 March 6th, 2025 at 8:05 am
To compound the issue of contributing to the current anti-science sentiment in the name of criticizing the institutions and being an “open-minded” contrarian (this gets more attention than going with the mainstream), and why we should be all careful/responsible about it:
the new administration claims that all the cuts/cancellations of publicly funded science is targeting woke stuff, with people like Scott saying something like “cut all the woke stuff if you want, but leave me alone”, but they miss that the ‘woke stuff’ is of course just the excuse to target across the board…
in reality the move is about controlling science in general, to shut down all things that are not fitting their right-wing narrative, e.g. blocking flu/covid seasonal data, censoring/defunding climate science and drastic cuts to the NOAA.
Also we seem to have forgotten that the war on science from the right already started during the W.Bush terms, with the very aggressive pro-creationism and anti-evolution movements in red states, trying to take science out of schools in the name of “skepticism”.
And you never know what the right-wing-bibble-thumper nuts are going to target next… cosmology?
Comment #74 March 6th, 2025 at 8:20 am
We should also question the idea of professionals washing their dirty laundry with the general public as a witness… sure, it’s about the tax money of the public, but the allocation of that money is best decided by the relevant experts, and if they disagree (which is normal), done behind close doors until an decision is met and then explained. Doing it in the public eye on youtube just creates confusion in the general public, there’s only a few mainstream narratives, but an almost infinite number of fringe theories, most of them having no values… but such a flood of supposed expert opinions just gives the public the impression that the field is a shit-show of lies and driven by corruption and personal interests.
Imagine if WW2 had been conducted with Patton, Einsenhower, Montgomery, … all settling their personal extreme disagreements on how to pursue the war or pushing their ego-driven agendas by publishing youtube videos. Noone would have any trust in the role of the army anymore.
Comment #75 March 6th, 2025 at 8:58 am
Scientists are truth seekers.
The current administration and their supporters want to cancel science because the concept of truth gets in their way, i.e. unlike gender, the rest of ‘facts’ ought to be fluid… as a result they also need to erase the idea that there are ways for arriving at the truth independently of one’s opinions.
Comment #76 March 6th, 2025 at 10:53 am
Having finished watching the entire youtube video discussion between Scott and Jacob, it seems that quantum mechanics is (as Donald Rumsfeld put it) all about agreeing on the known knowables, the known unknownables and the unknown unknownables.
Comment #77 March 6th, 2025 at 11:02 am
Kevin Zhou #13:
Maybe it would be better advertise these results under the headline: Einstein way off, as usual, confirm scientists.
Comment #78 March 6th, 2025 at 11:11 am
@Sabine:
“People want to learn how nature works and how the universe works etc. They do understand that in many cases it takes a very long time for such research to become profitable, and maybe it never does.
But if they understand that, then why do we need to force them to pay basic research by taxes? That makes no sense. ”
It does make sense because of the time scale involved. An example:
35 years ago a spanish PhD student investigated archaebacteria in order to find out how they survive in their hostile environment of salt marshes. He did some DNA sequencing and found interesting palindromic structures. You know, how that went.
I can’t imagine a business funding research, which will pay off, if at all, after a quarter of a century. Nations (I’m excluding the current US administration) think in other dimensions, so they are the natural funders in this case.
Unless you have some crystal ball that tells you which lines of endeavour lead to fruition, you will just have to keep digging in all sorts of places to find the gold nugget.
The part which makes no sense to the public is possibly, that inevitably there is some digging where there’s no gold.
Comment #79 March 6th, 2025 at 11:39 am
RE #67 “In an ideal world, you would have Triton Station, but also 5 other blogs from other scientists arguing just as emphatically in the other direction.”
I haven’t been looking for them, but know of at least two anti-MOND blogs (and one additional pro/neutral MOND blog). I didn’t bookmark them and would have to search for links, but one featured a young woman physicist who was playing some video game at the same time as making her argument (impressive but a little confusing); the other concluded, from a paper last year stating that no evidence of DM had been found from data analysis of a particular galaxy that, okay, DM might not be universal, but MOND was supposed to be as far as rotation curves were concerned, so it was thereby disproved. The paper itself stated that its data was not from the low-acceleration MOND region and no conclusions about MOND could be drawn from it.
As far as I know there have been no scientific conferences on MOND and several on DM. Which might be fair enough on the merits, but does not support the above premise.
Comment #80 March 6th, 2025 at 12:31 pm
@JimV #79
The woman is Angela Collier. She is quite clear about the intent of her videos — “edutainment.” But, when speaking about physics, she can sometimes be worth watching.
https://m.youtube.com/@acollierastro
As for nonphysics, I could appreciate her videos on Feynman bros and oligarchs. I don’t expect any Feynman admiring physicists to like her very much.
My appreciation arises from my first day of college. I found myself surrounded by “Uncle Milty” acolytes who would never comprehend the subtleties of how differing academic perspectives cannot be decided by equations on blackboards.
They bragged about credentials and were impressed by credentials. I expect that those who completed studies have all enjoyed success.
Comment #81 March 6th, 2025 at 1:24 pm
@77: Your link is broken. I think you meant this one:
https://www.smbc-comics.com/comic/einstein
Comment #82 March 6th, 2025 at 2:06 pm
Sabine #62: thanks for the response. I think we’re all falling into the trap of social media: everybody prefers negativity. Your video on all of particle physics being a bubble and fraud has over 20x more views than your video on the P5 report. And my feelings on the P5 report are more positive than yours, but there is no way my opinion could ever get even 1/200 of your viewcount. Of course there exists reasoned disagreement to your opinions, but it predominantly exists in papers with 2-digit views.
I mean, this actually strikes me as an argument for public funding of basic science. If we all had to make money privately, then I suspect most people would quit, a small number would get sponsored by their rich buddies, and a much smaller number would get supported by ad revenue from the crowd. The ratio of theoretical speculation to experimental progress would greatly increase, because theorists are cheaper. Most funding wouldn’t even be market-driven, it’d be determined by who you know. People would make even larger promises, and make videos with even less nuance, and drag even more politics into science to boost their audience.
Jean #69: I’m talking about Triton Station here because people specifically asked me about it here. Also, it’s never a good idea for a nameless person like me to disagree with somebody like that on their home platform. What always happens is that you get dogpiled by commenters calling you a psychopath or a religious fanatic, and if you’re lucky, the blog owner will make fun of some out-of-context snippets of your comment in the next post. I’ve been through that too many times to bother again. I am always happy to disagree with people in person, but the dynamics of the internet are different.
Adam Treat #72: It’s not about the latest videos. I’ve personally talked to dozens of fans of Sabine over the past 5 years, and they usually come right out the gate pronouncing physicists wrong about everything. Maybe they’ve heard nuanced arguments at some point, but in practice they don’t even seen to know that counterarguments exist. Something in the communication process is breaking down.
suomynona #77: Haha, that’s funny! Yeah, it’s weird how science communication relies on a small number of templates, endlessly repeated. There are so many ways to blow people’s minds by saying stuff that’s unquestionably true but doesn’t fit in a template.
Comment #83 March 6th, 2025 at 2:20 pm
Sabine #65
““They merely use Sabine as a post-facto justification for the backlash…”
The backlash hasn’t even yet fully begun. In any case, I am a small fish. I have next to zero influence of anything, so don’t worry. I would suggest that you just continue to ignore me.”
I don’t understand why you persist in using this strange defense that you’re just a herald for the will of the people. No, you are actively trying to push public opinion in a certain direction, and if you don’t realize that then you’re irresponsible with your platform. You are telling people that a large percentage of science is bullshit and that it is useless and no one should want or need it and that we shouldn’t pay for it; you might think this is true, but you have to understand that when you say this you *are* going to attract the ire of the people who do science and find you calling them useless bullshit artists incorrect and also deeply insulting. I’m not sure why you act so affronted by people being offended when you say this, because it is deeply offensive and I presume you said it on purpose.
Sabine #66
“But if they understand that, then why do we need to force them to pay basic research by taxes? That makes no sense.”
This betrays a pretty shocking ignorance of the very basic economics involved here. Would you say this about the mail service? Imagine a parallel argument “The USPS is extremely bloated and wasteful, we pay people to have cushy jobs no one needs, and the public increasingly wants to cut the USPS and save their tax bill. If there really is a need for the USPS then people will fund it privately.”
The problem, of course, is
A) The public does not want to cut the USPS, just like the public doesn’t want to cut NSF funding. Right wing politicians want this, and you just assume this is popular amongst laypeople. It isn’t.
B) It isn’t true that just because the public agrees something should be funded that automatically means that it would get funded under a pure free market. That’s literally why we pay for things with taxes, unless you’re some kind of uber-libertarian who thinks that we shouldn’t have a public sector at all but I don’t think you are.
I’m just not sure what this argument is supposed to be doing; you haven’t made any sort of economic case that science research would just get more private funding if public funding is cut (this doesn’t not seem to be true, to me) and so in lieu of that you are just saying that anyone who currently has a publicly funded research job is a useless bullshit artist who shouldn’t have a research job at all. I hope you understand why people who think that research is valuable and worthwhile and have publicly funded research jobs (read: the vast majority of researchers and scientists including many you’ve been feuding the past few days) see you saying “I think what you do is useless and your job should not exist” as a pretty vicious attack, aside from being something they disagree with.
Like you say that you just want a rational discourse about this, but you are actively peddling the idea that the majority of researchers are useless and the work they do is meaningless during a time when science is actively being cut. I don’t know why you launch such a vicious and directed attack against the vast majority of scientists and expect people to not take disagreement to that.
P.S. I think that the latter half of Sabine #66, where you quibble about how “there might be sectors where this doesn’t work out” is pretty strange. If you have a reasonably high certainty that cutting public funding for research won’t decrease the amount of progress then you should make that case better, because the vast majority of scientists, policy experts, and economists disagree with you. It is an unorthodox economic standpoint and needs to be argued for. But I suspect that this economics argument is not really what this discussion is about, because even if not a cent more was spent on science after the NSF gets gutted you would still be happy because you think publicly funded research is almost entirely vapid bullshit. That is what the disagreement is about, and it would be nice if you would keep it there instead of talking about your highly unorthodox layperson economic stances.
Comment #84 March 6th, 2025 at 2:55 pm
@Gali Weinstein #70
Dr. Weinstein, the failure of falsifiability as an effective demarcation is a central issue for some of this discussion. Relative to string theory, there is the popularity of Dr. Woit and Dr. Hossenfelder. Relative to dark matter, there is Dr. McGaugh. A fundamental appeal for these blogs lies with how paper after paper manages to justify “falsifiability” with rhetoric. When these same authors then adjust parameters to fit new evidence repeatedly, it strongly suggests that the community will not hold their peers to this standard.
This is why Dr. McGaugh incessantly insists that confirmation of a priori prediction is “the gold standard” in science (I cannot judge Mr. Zhou’s criticisms here).
To the best of my knowledge, some portion of the philosophy community has abandoned falsifiability as a demarcation. Even Popper had to adjust his “alignments” by accepting some disciplines as scientific after formerly describing them as pseudoscientific.
Again, to the best of my knowledge, Popperian falsifiability had been a reaction to Reichenbach’s verificationism. Reichenbach, in contrast to modern authors like Sokal and Bricmont, fully understood that verificationism cannot be a correspondence theory of truth. This very distinction lies at the heart of Goedel’s incompleteness theorems — they lie at the heart of distinguishing between truth and provability. Researchers who simply choose to ignore this study varieties of semantics not based upon correspondence with objectual ontologies. One can also avoid the incompleteness of arithmetic with self-verifying theories in which addition and multiplication are not total.
Once again, to the best of my knowledge, falsifiability arises by analogy with logic studies. I seem to recall a paper attributing this to Popper. To the best that I can discern, this refers to the passages from Aristotle’s Topics. This book focuses on “argumentation from common belief,” and, the context is how any (dogmatically) stipulated definition is undermined by a counterexample. I ought not attribute that particular knowledge to Popper, though. It comes from sorting out description theory from among logicism, first-order logic, Herbrand logic, Skolem’s descriptive functions, and usage for the sign of equality in the semantics of positive free logic and negative free logic.
You also suggest that Gil Kalai carrys a burden of proof.
I always find this curious. Aristotle referred to rhetorical logic (Topics) in legal contexts as forensic logic. There is, indeed, an established protocol for who has the burden of proof in courts of law.
To the best that I can tell, the common approach to burden of proof in “science” is based upon the logical fallacy of argumentum ad populum. Dr. Hossenfelder recently rejected the idea that “science” ought to be a consenus opinion decided by polls. She is certainly correct in this, although it is difficult to agree with her bare claim that scientific “truth” is not consensus-driven.
In the end, science as a human endeavor constructs narratives. Those narratives are grounded in observation and data collection. Only a fool would knowingly act contrary to essential components of a scientific narrative — it is intended to conform with our objective experiences. The problem with believing that it is “logical” in the sense of slinging “zingers to win arguments” is that this only provides tu quoque retorts from anti-science cohorts. I do not arrive at this conclusion naively. Declaring mathematics to be reducible to algebraic language signatures leads to the (postmodernist) view that knowledge reduces to linguistic interpretation. W. W. Bartley III documents this problem in his book “Retreat to Commitment.” It is particularly poignant because it does so in the context of science decoupling from liberal Protestantism.
I can assure all of you that capture of scientific language by anti-theists like Dawkins, Kraus, Coyne, etc. does far more to undermine science than pessimism about falsifiabilty from Dr. Hossenfelder.
Anyway, there are severe problems with falsifiability as a meaningful demarcation.
Comment #85 March 6th, 2025 at 3:07 pm
Doug S. #81
Oops, not sure what happened. Yes, that one exactly. Thanks!
Comment #86 March 6th, 2025 at 3:28 pm
Sabine #65 asks “Why do you think someone who is anti-science would even watch a science news channel?”
There’s a very clear answer to this question. People like news that reinforces their own worldview. If a science news channel is mostly news about bad/nonsense science and criticisms of science academia, then it mostly reinforces their “science bad” worldview.
Of course, Sabine’s channel is not intended to have a “science bad” message. So it comes down to a fundamental question: If someone says something factually correct publicly, and it’s misinterpreted by a large number of people, how responsible is the speaker for the misinterpretation?
Obviously the answer can’t be “not at all”, or we’d have to accept all the propaganda and fake news that presents out-of-context and cherrypicked facts to construct a false narrative. But the answer probably shouldn’t be “always 100%” either, else we’d have to condemn every researcher that was just trying to publish a cool result and had a misleading article written about it.
It seems like Scott’s answer is “very responsible” in most cases. Here’s a video of a popular science YouTuber supporting his opinion; what do you guys think of her?
Comment #87 March 6th, 2025 at 3:53 pm
Comment #82 Kevin Zhou. Yes, I know what you mean: who would want to face the pack you’re describing? It’s also the surest way of avoiding a reply from the person you’re criticizing about a specific paper and, more generally, about your remarks in Comment #54. It would be sterile to continue this exchange here. Good luck to you.
Comment #88 March 6th, 2025 at 4:05 pm
Kevin Zhou #82,
“It’s not about the latest videos. I’ve personally talked to dozens of fans of Sabine over the past 5 years, and they usually come right out the gate pronouncing physicists wrong about everything.”
I’ve followed Sabine and Scott via this blog for longer than that and I have not noticed any change in the general tenor of their scientific views over the course of that time. Sabine is saying more or less the same things now as she was before Trump was a blip on our worldwide subconscious. Same with Scott. I don’t detect anything much at all has changed with their overall scientific worldviews.
When it comes to the science I can’t think of many things where they out-and-out disagree that doesn’t come packed full of nuance and careful attention to the details. Superdeterminism being the biggest exception.
Comment #89 March 7th, 2025 at 4:00 am
To return the discussion to a technical question, one thing that immediately caught my eye is that Scott defines stochastic matrices via the absolute value of unitary matrix elements (https://www.scottaaronson.com/papers/qchvpra.pdf), while Jacob uses the absolute *square* (https://arxiv.org/abs/2302.10778). I wonder if there is an obvious contradiction here? What am I missing?
Comment #90 March 7th, 2025 at 4:54 am
@mls #84
I appreciate your historical and philosophical context, though it largely sidesteps the empirical issue I raised about Kalai’s skepticism of quantum computing.
To clarify, my comment did not rely on Popperian falsifiability as a strict demarcation criterion. Instead, I pointed out that Kalai’s argument is structured as a set of conjectures rather than a formal no-go theorem. The issue is not whether his claims fit within a rigid Popperian framework but whether they are empirically testable meaningfully.
You mention that science constructs narratives, which is undoubtedly true, but that does not mean all narratives are equally viable. A scientific hypothesis—whether supportive or critical of quantum computing—must eventually be confronted with empirical results. Kalai’s skepticism implies constraints that should manifest experimentally. If those constraints do not materialize, his framework is not merely an alternative “narrative” but an empirically refuted position.
The real question is not whether Popperian falsifiability is a perfect criterion for science but whether Kalai’s claims can be tested and challenged by Nature itself. That is the standard by which scientific claims—whether mainstream or contrarian—must ultimately be judged.
Comment #91 March 7th, 2025 at 5:55 am
@Sabine #66
Your critique of academia as a planned economy with distorted incentives resonates, especially regarding how institutional gatekeeping prevents unconventional thinkers from contributing. But while the current system is undeniably flawed, a fully market-driven approach risks introducing its own set of problems—particularly for foundational theoretical research, which often lacks immediate commercial applications.
Your concerns about academic exclusion and public funding inefficiencies are absolutely valid. Gatekeeping is real, and many brilliant researchers are sidelined for reasons unrelated to merit. Public funding, while essential, often incentivizes careerism over intellectual risk-taking, pushing researchers to chase grants rather than focus on fundamental questions. There’s an inherent paradox: academia provides the stability needed for long-term breakthroughs, yet it excludes many of the very thinkers most capable of producing them.
Would a decentralized, market-driven model fix this? Perhaps in some cases. But there’s also a real danger that it would create a new form of exclusion—where funding prioritizes what is commercially appealing or immediately practical, sidelining deep theoretical work that only reveals its value decades later. The problem isn’t public funding itself, but how it is distributed and controlled. The real challenge is reforming academia so that it supports unconventional thinkers while maintaining the long-term stability that fundamental discoveries require.
I’d be curious to hear your thoughts on how you see these trade-offs playing out.
Comment #92 March 7th, 2025 at 7:50 am
David Kagan #26: Thank you for your support! Unfortunately, what I conclude from the obstructions is not that ontic trajectories must obey a “non-probabilistic uncertainty” from which probabilistic uncertainty occasionally emerges, but just that there aren’t ontic trajectories in such a theory.
At some point in my study of quantum interpretations, the idea arose of a particle having a position, but not having a definite position. I think this was not how anyone expressed themselves, but rather me trying to put into words the reification of uncertainty that goes on.
I put it that way to make the problem more explicit. One would normally suppose that you can’t “have a position” without having a definite position. This may be an instance of a broader ontological principle, that you can’t have a “determinable” without having a “determinate”. You can’t have a color without it being a particular color, you can’t have a shape without it being a particular shape, and so on.
In this case, I guess the apparently contradictory concept is that of “non-probabilistic stochastic trajectories” – randomness that is not even probabilistic. At the very least, if you are going to make such a radical concept fundamental, I think it deserves a lot more attention! How could such a thing even be?
By the way, if anyone thinks of Knightian uncertainty at this point – that is an *epistemological* non-probabilistic uncertainty – you don’t know something and can’t put odds on the possibilities. Here we are talking about *ontological* non-probabilistic uncertainty.
Ontological uncertainty here boils down to not having a cause. The premise is that fundamental physical properties are changing and there’s no cause for why they take one value rather than another. In quantum theory we say, but we can at least associate a probability to the different outcomes. What exactly a fundamental probability is, is a good question – is it a frequency over counterfactuals? a frequency over possible worlds?
Anyway, what Jacob and David discovered in their attempt to think about quantum mechanics as arising from a particular scenario of fundamental properties changing randomly, is that at a certain point, the probabilities became inconsistent. There was no probability distribution that stayed within their scenario, and still gave the same predictions as quantum mechanics.
One reaction to this would be to say that their scenario, their premises, had been falsified. You can’t get quantum mechanics from the kind of random trajectories they were investigating. But instead they say, it’s a special new kind of randomness that isn’t even describable by probability, except when it’s quantumly safe to do so.
I’ve already made it plain how I would prefer to interpret the obstruction: locally random continuous trajectories are not the answer. Maybe you need instead some kind of discontinuous quantum jumping. I’ve always been intrigued by the idea that there might be some kind of maximal coarse-graining of decoherent histories that provides a natural subquantum ontology…
But if we *are* going to propose that the world runs on non-probabilistic randomness (aka “indivisible stochastic processes”), I’d like to see more effort devoted to explaining *that*. What could it be? The other day on Max Tegmark’s homepage, I saw a critic saying that you can’t have a uniform prior on the set of all mathematical structures. I suppose it’s like the claim that you can’t have a uniform probability distribution on the integers. (Though could you, if you allow your probabilities to be actual infinitesimals? But I digress.)
Maybe this non-probabilistic uncertainty would be a kind of “indexical uncertainty” about which possible world you’re in. Then you’d have to explain how probabilistic uncertainty emerges within that context. It sounds like a long shot to me. But in my opinion this is what it would take for this to be an ontological hypothesis, rather than just a formalism.
Comment #93 March 7th, 2025 at 8:55 am
Mathematicians seem pretty safe in that sense.
Seemingly irrelevant audience just can’t have any opinion and so doesnt engage in such YouTube trolling.
Chemists also seem safe.
Somehow trollers have lots of opinions about biology and physics.
Comment #94 March 7th, 2025 at 1:34 pm
Scott, what you’re seeing here is not Youtube culture in general, it is specifically the audience that Curt has cultivated by platforming outright cranks uncritically, promoting fringe views in the most sensationalist/clickbait ways and going on his own diatribes against e.g. string theory.
While he does also get serious scientists on (case in point), the channel in general is not responsible science communication! The comment sections of many other long-running science channels (sixty symbols or PBS space time, say) tend to not be a fraction as bad.
Comment #95 March 7th, 2025 at 1:46 pm
Jean #87: I think you’re missing the underlying dynamic here. It has been going on for decades, even before I joined physics! I’m not the first to criticize how MOND is popularized; I’m not even one of the first hundred. If you ask around at any physics department you will hear similar arguments from anybody who’s looked into the issue.
However, internet popularity allows one person with a platform and an axe to grind to get more views than the combined efforts of hundreds of independent people with carefully thought out disagreements. There are many people like me who will explain their points of view in a private conversation to one, or in a referee report to five, or at a coffee break to ten, or deep in a comment section to a few dozen. Stacy McGaugh ignores all of them, or just beats an (occasionally entirely made up) strawman, then laments that nobody has any real counterarguments. Then, when you see my counterargument, you suspect it’s some secret information I’m too lazy to divulge. But the reality is that you’ve briefly, accidentally peeked out of a carefully constructed bubble, centered around Triton Station, which nothing is ever allowed to pierce. That is the playbook run by most bloggers and podcasters promoting alternative theories.
Comment #96 March 7th, 2025 at 4:14 pm
@Sabine, science funding has historically come from patrons, whether noble lords in the past or, in modern times, the state in a democratic republic. The key difference is that the power structure has changed, but the underlying model remains the same: a centralized authority with resources supports scientific endeavors. This stands in stark contrast to a crowdfunding model, which turns researchers into content creators competing for public attention. The “OnlyFans-ification” of science would be disastrous. Take yourself as an example. Your prominence online comes not from your scientific theories but from your ability to generate engagement through rants and critiques. That alone illustrates what an “OnlyScientist” future would look like.
Comment #97 March 7th, 2025 at 4:50 pm
Did you think the response on YouTube was predictable? (Sorry if this is a repeated question. There are many of them and I only read the first few.)
Also, an am applied mathematician. I found the debate very hard to follow. I kept googling terms I didn’t understand. I watched about the first 20 minutes.
Comment #98 March 7th, 2025 at 10:06 pm
> I don’t get to pick how I look and sound. I came out of the factory this way.
I think you’re wrong about this. Those are just the default settings your body came with. You can absolutely change them. There are a ton of settings you can tweak to make yourself look and sound exactly the way you want to (even tho the current administration is trying to forbid some of them). But of course if you like the way you came out of the factory, there’s nothing wrong with that. You shouldn’t change yourself just to please other people, especially random youtube commenters.
Comment #99 March 8th, 2025 at 1:48 am
scott: you write: “I’d take Jacob Barandes, or even Sabine, a billion times over the lunatics in power.” why the seemingly gratuitous ‘even Sabine’ dig?
Comment #100 March 8th, 2025 at 2:52 am
mike comment #3.
“Day by day I feel suffocated by idiots. Its easy to get the impression that they vastly outnumber the reasonable people. But I suspect they are just louder, in part due to the fact that they don’t understand enough to appreciate how little they understand.”
your initial impression was correct – the idiots do outnumber the reasonable people.
Comment #101 March 8th, 2025 at 3:22 am
I originally wanted to ask Scott (or Jacob, as he lurks here as well, as I see) a physics related question, but the conversation took a wild turn, and now people are knee deep into something else entirely. But of course even though the video was about physics only, the post itself was not, so I shouldn’t be surprised.
I really enjoy Sabine’s videos usually, even though her range of interests does not entirely covers mine, so… let’s just say I won’t miss her science news segments…
I’m sad to see her being more emotinal in her clips though, because more emotional really means more angry – she curses a lot, and sometimes even attacks character. She is always clear about why she does that, and I don’t even say it’s not well deserved to be attacked by her, still… For me personally it’s not something I look forward to in her future videos, and frankly, it seems more like real stress in her life materializing this way, than anything else. She could do so much more with clever wit, than giving tantrums. Or even without that.
As for attacking science, it seems to me that she creates these videos for people who understand the – not particularly small – difference between attacking certain aspects of academia and attacking all academia or science itself. But I agree that she should recognize that her clips resonate way too well with certain people. The same folks who are just as happy with Sean Caroll on Joe Rogan as with Bob Lazar. And this goes back to my previous point. Had she been less emotional but more objective, her video would have generated smaller waves. For some people strong words and passion is much more crucial than nuances.
And just to try to get tracks back on physics, I’m gonna post my initial question anyways: you guys talked about Bell and von Neumann at the end, I would like to know what you think of recent works revistiting the „vN proof”, by Acuna, Golub and others? Not sure if you’re familiar with those, and I know Scott said no asking about papers, but as far as I know most people involved in the discussion are respected scientists, so what the heck? It’s more like a history of science question anyways, as far as I can tell, and it seems to me that it’s a case of „everyone was right and eveyone was wrong” in some sense.
Comment #102 March 8th, 2025 at 5:56 am
>Then, alas, I read the comments on YouTube and got depressed.
Dear Scott, I watched the whole discussion and in my view, the criticism you provided was exactly the kind that someone who is submitting a potentially significant novel idea for consideration by the scientific community needs. You checked practically all points that in my opinion make for excellent constructive criticism:
1. You argued in good faith and with an open mind
2. You were precise in your criticism
3. You attempted to state your criticism in several different ways for the benefit of being better understood
4. You asked questions which were effective in helping you (and the audience) understand his work and its implications better
5. You were firm but fair, not backtracking on issues you identified as being in need of resolution without a good reason to do so
6. You articulated the conditions necessary for you to change your mind
7. You gave your discussion partner room to consider differences in what each of you deem important.
These are all aspects of criticism one cannot take for granted, sadly.
I was also impressed by Jacob’s defense, he did in my view make the strongest case one could make with what he has and our current state of understanding of quantum mechanics. I happen to think that he has found something potentially important, but not necessarily in the way that he seems to think is important, and that is why I actually agree at least partially with both of you, even though you did not agree with each other (but that is another topic).
So the debate between you and him was in my view a model of how scientific debates on new ideas should be.
The comments you refer to reflect in my view two contexts which bias them toward belligerence:
first, within the context of youtube, debates are often advertised in personal terms (e.g. “x destroys y”) for the sake of clicks which deflects the purpose from coming closer to the truth to a kind of verbal battle. Now, Curt does not do that, but enough other people on youtube do it that it becomes a kind of general expectation, very unfortunately.
The second and broader context is that we live in an age in which tribalism has become a more important part of our identity (generally speaking) and so we often see exchanges pretty much everywhere that are more adversarial than they need to be.
At any rate, you are not there to please the youtube commenters who frankly couldn’t care less about whether you are a great scientist or not because that is irrelevant to their tribal identity. Therefore, there is no reason to let this affect you.
Comment #103 March 8th, 2025 at 10:13 am
I believe higher education and academic research are highly beneficial to society and reflect its values. Therefore, both should be publicly funded. This has always been my stance in the intense debates on these issues in Israel over the past four decades.
Overall, public funding for university education and academic research in Israel is fairly adequate. Tuition at public universities is subsidized, making higher education reasonably affordable.
Here are some additional thoughts:
1. Public funding for research should not be taken for granted. Scientists must continuously advocate for it and make a strong case for its importance.
2. The position that higher education and research should not receive public funding is a legitimate one.
3. In general, I oppose the practice of delegitimizing opposing views or excessively mocking them. That’s why I commended (#36) Scott for avoiding this approach in our debate (#15) this time, in contrast to times in the past when he did use it. However, I acknowledge that such tactics can be effective in debates and in drawing attention.
4. Sabine’s views on public funding seem to align with an economic right-wing perspective, as do the actions of the new U.S. administration. That said, there are also valid left-wing arguments that public funding for university research and education primarily benefits the privileged and wealthier segments of society.
5. Like many debates, this is not just a clash of ideas but also a clash of interests.
6. About a decade ago, a commentator here suggested that my remarks on the blog were overly contrarian due to my disagreement with Scott on the feasibility of quantum computing. At the time, I disagree with the comment—I saw myself as contrarian to everyone—but I still chose to significantly reduce my non-technical comments on SO. That said, I suppose a small deviation from my policy once or twice per decade is acceptable. 🙂
Comment #104 March 8th, 2025 at 12:05 pm
OK, enough complaining about how people engaging with the issues in a shallow way can easily be directed by demagogues to support whatever they want them to support, a fact that has troubled human society for thousands of years in innumerable historical examples… can someone please suggest what we could do about it? I’d be glad even to see an unworkable idea, just to change the target of the discussion.
Here is what I think we have all agreed on about the problem:
1. Anyone can play the role of the duped rioter / “YouTube commenter” if they don’t think about it for long enough – innate qualities have nothing to do with it. The variance of thinking speeds between people is nothing compared to the difference between thinking for ten seconds and thinking for ten years, or having access to a thousand years of previous thought through study.
2. People will learn not to think about things the same way they can learn not to do anything else: if it’s not rewarded. In many cases, people who are careful and attentive to things that affect them directly learn not to spend any mental effort on “volunteering.”
3. Education is difficult when it isn’t wanted.
Comment #105 March 8th, 2025 at 1:29 pm
I guess the reason there will always be a cyclic argument is that your take is:
comes from your position as a computer scientist and tangential physicist, where my assumption is that your focus is usually on understanding for utility. Maybe I’m wrong. In the philosophy of physics, I see time and time again that they just toss out the utility component, which I guess is fair game in philosophy. Then, it’s just about trying to get a new interpretation for clarity, simplicity, and possibly unification.
Comment #106 March 8th, 2025 at 1:44 pm
Xirtam Esrevni #105: In case it needs saying, for me the answer to “what does it buy me?” could have nothing whatsoever to do with engineering utility. It could be: a new angle on quantum gravity, a new way to understand Wigner’s friend, a better way to explain QM to undergrads, a new meaty open question. Indeed I’m personally more interested in all of those things than I am (for example) in the practical problems of building quantum computers. But whatever the answer, it needs to be explained to me in language I understand.
Comment #107 March 8th, 2025 at 2:14 pm
Re: #99
As Jackie Gleason used to say (on “The Honey Mooners”, when his character was flustered), ad hominem ad hominem ad hominem. Dr. McGaugh has many published papers and thousands of citations. He has worked in astronomy and cosmology for over 20 years. He has opinions based on his work. He has received a lot of criticism for those opinions, which annoys him. What is his crime that has inspired all this vitriol by “hundreds”? In a better world, those people would go to his blog and debate the facts. They could try, “MOND can’t do large structure formation” for example, and see how that works. [Obligatory Warning: Dr. McGaugh knows what MOND can and can’t do better than any of those hundreds. The issue is why does it work well enough to predict phenomena before they are measured, and why is everything about galaxies directly correlated with the estimated number of baryons.]
Comment #108 March 8th, 2025 at 2:18 pm
Xirtam #105 Scott #106
I am not sure where the “throwing away utility” angle is coming from, because engineers typically go farther on intuition alone than anyone else does, and have a much greater use even for conditionally approximate intuitive ideas that are usually phrased in terms of a nonfundamental noun like “vortex street,” because that is how people think. Something like post-hoc trajectories is actually very likely to succeed as a useful concept in environments where people have a reason to think about what it does imply and no reason to think about what it doesn’t, and to construct those environments you’d have to get away from the general case and find a lot of value in a specific one, which is what engineering is all about.
Comment #109 March 8th, 2025 at 3:42 pm
I should probably watch the video and also study more carefully what Jacob has done. Superficially I thought he was trying to extend the Nelson idea so that it is non-Markovian. Hasn’t this already been shown to fail long ago (e.g. https://journals.aps.org/pra/abstract/10.1103/PhysRevA.19.2440). What am I missing?
Comment #110 March 8th, 2025 at 6:31 pm
Scott, you and Nathan briefly touch on a framing of MWI that I’ve never quite agreed with, and I’d be interested to get your take.
People often loosely describe the MWI as implying that “Everything (that’s compatible with the basic rules of QM) happens in some branch of the universal wave function.” In popular accounts, they’ll often throw in some wild examples like “There’s a branch in which you’re the President of the United States, and another branch in which the Earth is populated entirely by intelligent lizards.”
But this claim does seem to me to follow at all from the MWI. As you point out in the video, just like Newtonian physics, classical E&M, and GR, QM specifies a particular mathematical form of state space and gives a dynamical differential equation for time-evolution, but it is completely silent on the initial conditions. So a priori, it seems fully compatible with the MWI that there are in fact literally zero branches of the wave function in which I’m the President, or in which intelligent lizards populate the Earth.
I can imagine three possible responses to this point, and I’m curious which one best describes your view (if we take the MWI as given):
1. It’s true – we have no particular reason to believe that there are any branches of the current universal wave function in which I’m the President, or in which intelligent lizards populate the earth.
2. The vast majority of possible initial conditions for the universe that lead to a universe anything like our observed current one would probably time-evolve to a universal wave function that contains branches with lizard worlds or me-as-President. So by the principle of maximum entropy, there are almost certainly such branches in the actual universal wave function.
3. The time-evolution of the universe probably displays quantum chaos, so every reasonably choice of initial condition for the universe probably leads to a current universal wavefunction that contains lizard-world branches and me-as-President branches. More generally, the universal wave function almost certainly “densely fills the space of all physically possible classical macrostates of the universe.” (The notion of “classical macrostate” is implicitly implying some kind of preferred basis of pointer states.)
(Options #2 and #3 are closely related but (I think) slightly distinct: I think that option #3 is stronger, because it asserts that the initial condition of the universe is essentially irrelevant for determining which universe “macrostates” can be found in the universal wave function.)
Comment #111 March 8th, 2025 at 7:19 pm
Ugh, sigh. In my comment #110, “This claim does seem to me to follow” should have been “This claim does NOT seem to me to follow”. Sorry for any confusion.
Comment #112 March 9th, 2025 at 4:36 am
There are these other guys who doesn’t believe MWI, who I’ve been trying to understand. They acknowledge that the Bell inequalities rule out hidden variables, and their idea instead is that the outcome of measurement somehow gets baked into a superposed state during its preparation. So you can have two identical states that willl measure differently, because of information cooked into them during preparation. But that sounds likea hidden variable to me….. they seem loike legit phsycists, maybe you know of them. Anyone here have an idea if this is bogus?
https://arxiv.org/abs/2404.18621
Comment #113 March 9th, 2025 at 6:36 am
@Gil #103.
Your concern about public mockery in debates is understandable. Still, it raises a more profound question: Why is public ridicule a problem, while quieter forms of exclusion and professional diminishment within academia receive far less scrutiny? Public debates may shape reputations, but institutional gatekeeping—determining who is included, who is heard, and who is given opportunities—is far more consequential.
Skepticism is often tolerated, even respected, when it comes from those who already hold institutional power. You, for instance, have built a career questioning quantum computing, and despite the pushback, your skepticism is still engaged with at the highest levels. But skepticism alone is not originality. It is one thing to challenge prevailing ideas; it is another to contribute something truly new.
What happens when unconventional ideas or independent thinkers emerge from outside the academic hierarchy? When originality does not come in the form of sanctioned skepticism but in actual breakthroughs that disrupt entrenched narratives? The system tends to respond not with debate but with exclusion. It ensures that those thinkers remain on the margins, unrecognized and unsupported, while others benefit from their insights without ever acknowledging them.
If you oppose delegitimization, then shouldn’t that concern extend beyond public debates? What about professional exclusion? If institutions quietly sideline those who think differently—keeping them in marginal positions, downplaying their contributions, or ensuring that their careers never progress—isn’t that a far more effective form of delegitimization than any blog post mocking skepticism?
If we are serious about fostering intellectual diversity, then the issue is not simply about who gets mocked in public. The real question is: Why does academia allow some forms of questioning while shutting down others? Why is skepticism embraced when it reinforces existing power structures but rejected when it comes from those who challenge them in ways that institutions find uncomfortable?
Would you agree that skepticism should not be a privilege reserved for those with secure positions but a space where true originality—whether it aligns with institutional expectations or not—is allowed to flourish?
Comment #114 March 9th, 2025 at 12:36 pm
Sabine #22, #23 (also a response to others): I rewatched your video, and you don’t respond specifically to the public goods argument—i.e., the argument that companies are underincentivized to fund basic research because no one company can adequately capture the gains. Instead, you just argue in general that basic research doesn’t produce enough gains, despite the contrary evidence of the entire 19th and 20th centuries (and, e.g., the studies estimating that public research funding produces $5 for every $1 spent).
The special problem with superdeterminism is that, once you’ve gone there, you’ve basically thrown out scientific explanation as your goal. The reason is that, pretty much regardless of what science found to be true—for example, if there were correlations even beyond Bell/CHSH, or even superluminal communication—you could easily cook up a superdeterministic “explanation” for that as well.
I hereby retract my comment about it being hard to find fringe ideas that you dislike. A more careful formulation would be: you promote multiple fringe ideas that mainstream physics has considered and rejected (e.g. MOND, superdeterminism), as part of what’s now a completely explicit broader narrative that the mainstream physics community is so dysfunctional that we might as well end its existence by taking away its funding.
Comment #115 March 9th, 2025 at 12:41 pm
Pissed Canadian #29:
Scott, I don’t get the harsh words on superdeterminism. Shouldn’t we, fans of Peter Shor, let everyone adopt whatever interpretation they feel useful?
As I explained in #114, superdeterminism is less an interpretation than an “anti-interpretation”—basically, the quantum analogue of the young-earth creationist saying “God must’ve planted the fossils in the ground 6000 years ago, with exactly the patterns you’d expect to see if evolution were true.” I’m willing to change this stance if anyone ever proposes a superdeterministic theory with any explanatory power. Until then, I find the idea that anyone could find superdeterminism “useful” for some other problem to be ludicrous.
Comment #116 March 9th, 2025 at 12:48 pm
Sandro #58: I’m actually extremely sympathetic to the idea that the government should fund research in a better way—if nothing else, one that’s less onerous for scientists (some of whom spend more than half of their time writing and servicing grant proposals). Crucially, though, if that was the goal, you’d first start your new funding mechanism as a pilot program, while keeping the existing mechanisms! You wouldn’t gleefully take an ax to funding streams that are the only thing keeping the current crop of PhD students employed, while having nothing to replace them with.
Comment #117 March 9th, 2025 at 12:51 pm
Gali Weinstein #70: I agree with your summary of Gil’s stance and how it differs from mine!
Comment #118 March 9th, 2025 at 2:59 pm
Gali Weinstein (#70), I also largely agree with your summary of my stance and how it differs from Scott’s.
Of course, when we delve into the details, there are many additional intriguing physical, mathematical, and computational aspects to consider.
I also agree that understanding experimental reality is crucial, and I have devoted significant effort over the past five years to examining the evidence.
Gali (#113), you have raised important issues, and I will take time to reflect on them. (I hope others will as well.)
Comment #119 March 9th, 2025 at 7:15 pm
I’m certainly not anti-Sabine but I have generally thought of her as the superdeterminism lady and because of that, not watched her vids much recently. Maybe I should look more closely. I’m not into videos anyway though.
For what it’s worth, Gerard ‘t Hooft is or was himself a superdetermism guy and has a big textbook (pdf online*) where he argues this point, so it’s not like the idea has only fringe adherents. The book explicitly says that a quantum computer would falsify his theory though. It is from 10+ years ago so idk what he considers the current status to be.
I still would like to know if anyone has paid attention to Collins/Popescu and understand what they are saying. Popescu has been on this track for a very long time.
* https://arxiv.org/abs/1405.1548 Hmm, “cellular automaton”, yes I guess it does sound fringe, oh well. I had forgotten that part. I looked at the book some years back and don’t remember if he account for complex-valued probability amplitudes in any plausible way.
Comment #120 March 10th, 2025 at 5:42 am
On the topic of private research funding, the recent controversy surrounding Microsoft’s Majorana claims shows the kind of problems which can occur when research is performed by private companies. A lot of the information on which Microsoft bases its claims on are hidden from the community due to IP concerns, which makes verification of their claims problematic.
https://www.nature.com/articles/d41586-025-00683-2
Comment #121 March 10th, 2025 at 8:15 am
Scott #115,
`As I explained in #114, superdeterminism is less an interpretation than an “anti-interpretation”—basically, the quantum analogue of the young-earth creationist saying “God must’ve planted the fossils in the ground 6000 years ago, with exactly the patterns you’d expect to see if evolution were true.”`
This is the most common criticism of superdeterminism. That it *necessarily* involves finetuning and or wild conspiracies in the initial conditions of the universe. If the criticism were accurate I’d agree that it’d be a pretty big nail in the coffin of superdeterminism.
Now, I don’t have the expertise to adjudicate whether the nail is real, but I do want to point out that Sabine *has* addressed this nail and argued it is not real.
I haven’t seen you rebut or respond to this paper: https://arxiv.org/abs/2010.01327
“A local, deterministic toy model for quantum mechanics is introduced and discussed. … We show that it is neither fine-tuned nor allows for superluminal signalling”
If you’re going to continue to insist that Sabine’s arguing for a super finetuned theory of QM that invokes wild conspiracies in the initial conditions, then I think you should respond to this paper and show why it doesn’t work when she says that her toy model, “demonstrates just exactly why these finetuning arguments are misleading.”
I humbly ask that you read this paper and respond in detail so that I might learn something.
Here is her conclusion:
“We have shown that fine-tuning, which has frequently been used as an argument against superdetermism, it[sic] unnecessary: To make predictions with this model one does not need to introduce a huge number of delicately chosen parameters. This toy model shows that violating
statistical independence is a scientifically sound option for resolving the quantum measurement
problem.”
Comment #122 March 10th, 2025 at 5:29 pm
asdf #112
That paper does not claim that “the outcome of measurement somehow gets baked into a superposed state during its preparation.”
On the contrary, they write
> It’s tempting to use this line of thinking in the quantum case also, and imagine that the angular momentum always had a specific value, even in the initial superposition state.
but reject this because of Bell inequality:
> Hence, when we measure the angular momentum of one of the spins and the state collapses to one of the possible values, it is not the case that we uncover a value that was already there.
What worries Collins and Popescu is the potential nonconservation of classically conserved quantities in the face of quantum randomness. As they put it:
> However, in quantum mechanics a state can be in a superposition of different values of a conserved quantity. If we then measure the conserved quantity, we get one of the values it had in the superposition: It seems that the conserved quantity, which is supposed to be unchanging, jumps to the new value upon measurement.
Standard quantum mechanics says that conservation laws hold only on average, not in every individual interaction. MWI says that the disappearing quantity goes into an alternative universe where the measurement outcome was different. Collins and Popescu do not accept either of these answers.
Instead, they prove that every measurement-induced random jump in angular momentum corresponds to an instantaneous transfer of angular momentum to an entangled “preparer” state so that total angular momentum is always conserved exactly. They consider this conservation law acceptably localised because it involves only a small number of nearby particles. “Nearby” could still mean outside the lightcone (i.e. faster-than-light transfer) so not what I’d call “local” but it’s much better than the totally inaccessible universes of MWI.
It’s all standard quantum mechanics but there’s something there that might not be obvious: conservation laws hold exactly in the classical limit; not just “up to quantum noise.” For example, an isolated spaceship will not gain macroscopically-observable rotation just from doing an immense amount of angular momentum measurements on spin-1/2 particles inside it.
asdf #119
‘t Hooft continues to promote these ideas in recent papers. Not so much cellular automata, though; his main idea is a hypothesized “ontological basis” such that time evolution operator acts as a permutation on the basis states (when time is discretized at Planck scale). This amounts to classical mechanics dressed in quantum formalism.
The math in https://arxiv.org/abs/2005.06374 is simple enough that I think I can see where he goes wrong.
The interacting Hamiltonian in Eq 8.3 is does not generate ontological evolution (i.e. permutation). Both exp(i H0) and exp(i H^int) are ontological but exp(i H0 + i H^int) cannot be when the matrices don’t commute. The closest ontological operators are exp(i H0)exp(i H^int) and exp(i H^int)exp(i H0), these are matrix products whose relation to Hamiltonians is complicated. Astonishingly, a footnote acknowledges that noncommutativity is a problem but simply advises the reader to work in continuous-time limit — a limit that is unphysical according to the model.
If the dynamics truly conserved ontology, ‘t Hooft could ditch the Hilbert space formalism in favor of ordinary probability distributions. But that doesn’t get you any quantum phenomena.
Adam Treat #121
While Scott didn’t respond [the last time you posted that paper](https://scottaaronson.blog/?p=6871#comment-1944781), other people did, including one superdeterminism proponent who said the paper was bad. Why do you keep pestering Scott?
Comment #123 March 10th, 2025 at 7:22 pm
N #122, Because he keeps repeating his criticism of Sabine’s version of superdeterminism being a big conspiracy and/or finetuned and has never responded to the fact that she wrote a paper specifically refuting that criticism. As for why I ask Scott, it is because I have some faith in his judgement and honesty whereas a lot of the other people in the linked responses… well, I don’t have such faith.
The only comment in that old thread that seemed pertinent was Dimitris #174 which faulted the toy model for being retrocausal and then implied that the finetuning was being shoved under the rug and would re-appear inevitably if it was made compatible with relativity. I don’t buy it as necessarily so.
Comment #124 March 10th, 2025 at 11:23 pm
Not that you need to hear it from a random blog commenter, but ignore those YouTube commenters! You did a great job expressing my thoughts as a physicist: nothing in what Jacob says seems wrong, but I also don’t see what it buys me that I don’t get from the quantum mechanics I already know and love.
One thing that I did get out of the video was that Jacob’s ideas sound related to the decoherence functional picture proposed by Gell-Mann and Hartle. I wonder if the “indivisibility” in Jacob’s picture is related to the coarse graining in the decoherence functional approach. At least at a words level they seem related; Jacob can only assign probabilities at certain times, the decoherence functional only can assign probabilities after “enough” coarse graining has been done. I haven’t done any work to check if there is a deeper connection beyond the words level similarity though.
Comment #125 March 11th, 2025 at 1:03 am
n #122, thanks much for the responses to my two posts. The clarification about Collins and Popescu’s paper is very helpful. It’s harder for me to understand the one about ‘t Hooft’s current theories but I’m appreciative either way.
A followup question if that is ok: I had thought that laws like energy conservation were connected with symmetries like time translation symmetry (Noether’s theorem). If repeating a quantum measurement can give differing (random) outcomes, then time translation symmetry doesn’t hold at that level, so is anything wrong if the corresponding conservation law goes away too? I remember hgearing that the symmetry (and conservation law) were also absent at the cosmological scale, since the universe is expanding. So those laws are local, classical approximations. Is it even worrisome if they don’t hold at every scale?
Comment #126 March 11th, 2025 at 10:09 am
Adam Treat #121, #123: The basic problem here is that Bell’s Theorem is a theorem. And it shows that you can’t have a secretly classical theory that
(1) reproduces the prediction of QM for two entangled particles and
(2) is free of “fine-tuning”—i.e., a conspiracy of initial conditions sufficient to force random number generators to produce particular outputs, human brains to make particular choices, etc., in order to cause the entangled particles to be measured in certain bases and not others.
Furthermore, over the decades, the “Bell deniers” (superdeterminists and others) have shown that they’re willing to generate an unlimited amount of verbiage, and even “math,” in an attempt to evade these simple points—thereby wasting unlimited amounts of everyone else’s time, in a sort of intellectual denial-of-service attack.
Whereas, if there were actually an error in the simple argument above, it ought to be possible to explain that error in a few sentences in such a way that mainstream physicists and quantum information experts would understand the error. And I’ve never seen that happen.
Maybe I’ll someday spent the days to work through Sabine’s (and Tim Palmer’s, and Gerard ‘t Hooft’s) papers on this if sufficiently provoked, but this is the crucial context you seem to be missing.
Comment #127 March 11th, 2025 at 10:22 am
Hi Scott #126,
“this is the crucial context you seem to be missing”
FWIW, I get the context and understand the criticism you level at superdeterminism. I believe Sabine understands it as well. She has written a paper rebutting that criticism as misguided or as a misunderstanding. I’ll let you determine the best use of your own time, but until you do get around to reading the paper I’ll say the rebuttal hasn’t been answered :=)
Comment #128 March 11th, 2025 at 2:45 pm
asdf #125
I don’t know why Collins and Popescu hold conservation laws sacrosanct. All I can say is that before reading their paper I wasn’t sure if quantum noise modified classical conservation laws and it never bothered me.
I agree with your point about cosmological scale, though I’ve also heard it said that if you could somehow quantify nonlocal gravitational potential energy, you’d find that the total energy of the universe is conserved after all.
I have one comment on Noether’s theorem. You seem to think that the “symmetries” refer to constantness of laws of physics when written in the form “if cause X, then effect Y.” That’s not quite correct. Noether talks about symmetries of the Lagrangian function, which is a bit more subtle construct.
In classical physics, a symmetry in the Lagrangian really does imply unchanging equations of motion (though the converse need not be the case).
But the way quantum physics uses its Lagrangian means a symmetry only implies unchanging probabilities, not outcomes. Which is why everyone thinks energy is conserved in the usual quantum mechanics. (But this energy-from-time-symmetry is the many-worlds-averaged energy and not necessarily equal to the energy of classical mechanics in the realized outcome, which is what Collins and Popescu are after.)
Comment #129 March 11th, 2025 at 11:07 pm
Scott #126, I don’t think that’s a fair statement of Bell’s theorem. The extra assumption it relies on is Statistical Independence (SI). Critically you can only really get to your “anything goes” / conspiracy universe (as in your (2)) by dropping SI and replacing it with absolutely nothing at all.
OTOH it may be entirely viable to drop SI and replace it with a weaker (or different) assumption SI’ that evades the conclusion of Bell yet is constraining enough to still give you a sensible theory. I think that is (at least implicitly) what the superdeterminists are claiming to have done, or at least outlined a plausible programme for doing so.
(Some form of retrocausality would be one way of violating SI. And indeed retrocausality might look like a contentless conspiracy theory *if* you insist on viewing the theory as an *initial* value problem, but presumably that simply wouldn’t be the right way of formulating the theory.)
I don’t believe in superdeterminism btw. I think it’s a speculative long shot at best, and that the motivation for even looking in this direction is weak (but not non-existent). But I don’t think it’s necessarily as crazy as you make out, at least in its best form.
Comment #130 March 12th, 2025 at 7:33 am
flergalwit #129,
Thank you! That is a very succinct and informative version of why I think the usual criticism of superdeterminism is overblown.
Comment #131 March 12th, 2025 at 9:18 am
Adam Treat #121,
for whatever my opinion is worth, the discussion here provoked me to skim through the paper that you linked (Donadi and Hossenfelder) and what I got from it is the following: To me, the key feature of the model presented there is what they call “future input dependence” (quoting directly from the paper, “the evolution of the prepared state depends on what the measurement setting will be at the time of measurement”). So, in short, they avoid the non-locality implied by Bell theorem for models with hidden variables at the expense of introducing a form of retro-causality. I can tell you that, personally, I do not feel this trade-off (non-locality for retro-causality) buys anything good for me and I am somewhat surprised to find out that there are people ready to go for it, but whatever, as long as it makes anyone happy…
Anyway, I think the paper helped me pinpoint the source of misunderstanding when the critics of superdeterminism claim that it requires conspiratorial fine-tuning (I would tend to claim this myself before) while the proponents vehemently deny this. (1) I had no idea before reading the paper that superdeterminism allowed for retrocausality. I imagined that it is deterministic in the sense of the usual past-determines-future causation. Without the retro-causality, I believe, one would still need fine-tuned initial conditions, essentially conspiring to simulate the effects that otherwise the retro-causality takes care of. (2) Even with the retro-causality (future input dependence), the model can boast to avoid fine-tuning mostly because it brings the information about future detector settings into the dynamics simply via a postulate. If the model were to be derived from some more fundamental principles, such that they should not be a priori aware of which process constitutes a measurement and which process constitutes a preparation of something that is going to be measured later, well, I suspect these detailed principles or laws would still have to be rather fine-tuned in order to reproduce the observable effects of quantum mechanics. And yes, I concede that the standard Copenhagen interpretation also fails to provide a description independent from what we deem to be state preparation and what we deem to be a measurement. This is actually what I do not like about Copenhagen, but than again, if the superdeterministic interpretations have the same problem, I do not feel like they can buy me much.
Finally, I feel compelled to note that I understand what Scott #126 meant with his somewhat derogatory mention of `even “math”` generated by “Bell deniers”. Not that the math in the paper we discuss deserved the scare quotes for being wrong, it’s perfectly sound as far as I can tell, it’s just that…you know, it rather misses any points that I do care about when I talk about interpretations of quantum physics. What clicked for me in the paper was the one sentence about future input dependence I quoted above; that was essentially all I needed to understand what is the beef with this interpretation from my perspective and all the math I was trying to go through then felt kind of as a distraction.
Edit: I did not see flergalwit’s comment before submitting, there may be some duplicit insight regarding the role of retrocausality in superdetrministic models in our comments.
Comment #132 March 12th, 2025 at 9:37 am
Martin #131,
I’m not personally turned off by the retrocausality, but by the ill-motivation for the retrocausality. The paper itself acknowledges this ill-motivation. The point of the paper was to try and clarify with a concrete example what flergalwit #129 said so well:
> “Critically you can only really get to your “anything goes” / conspiracy universe (as in your (2)) by dropping SI and replacing it with absolutely nothing at all.”
BTW, the retrocausality is only *one* possible:
> “assumption SI’ that evades the conclusion of Bell yet is constraining enough to still give you a sensible theory.”
Another possible one is the fractal model put forward by Tim Palmer. It is conjectured that every single possible assumption SI’ that evades the conclusion of Bell will necessarily hide the fine-tuning of the initial conditions under the rug to be revealed later with a full accounting. However, I do not see why this conjecture is necessarily true.
There might very well be a limiting version of SI’ that evades the conclusion of Bell without any fine-tuning of initial conditions. I see no a priori reason why this is ruled out. If people want to dissuade me of superdeterminism they should tackle this head on and show it. The toy model of Sabine’s for me is enough to rebut this criticism if itself is not rebutted by showing a math error or so on.
Comment #133 March 12th, 2025 at 5:21 pm
I watched this video on Curt’s channel before I read your blog post and if it’s any consolation at all, none of the things those YouTube commenters said occurred to me. I enjoyed listening to a polite conversation where you were still very clear about your reasons for not being taken with his idea.
Comment #134 March 12th, 2025 at 7:01 pm
Regarding retrocausality as a way to evade the Bell theorem, I think that when extrapolating this idea to the macroscopic level, you run into pretty much the same problems that you have with superdeterminism regarding the scientific validity of the idea. Imagine if people believed that your action of digging into the ground made you find dinosaurs, while if you hadn’t done that, you would have found two fossilized trees on a mountain and a snake skeleton. You would again be in the position of losing objectivity itself if people genuinely believed that the act of looking for fossils makes creationism retroactively false.
I guess one would need some dynamical theory that explains why retrocausality is possible at the quantum level and not at the macroscopic level. What happens then when we do a Bell test with macroscopic objects? One can claim that with retrocausality or superdeterminism, we can evade the Bell theorem, but no one is able to put forward a dynamical theory that doesn’t contradict the experiments or allow for too much “craziness”, in the same way that quantum mechanics doesn’t contradict classical intuition: since we have the theory of decoherence we know more or less about scale separation.
Comment #135 March 13th, 2025 at 9:23 am
Drizzle #134,
You lost me completely with, “in the same way that quantum mechanics doesn’t contradict classical intuition” LOL
More seriously, what you’re doing is taking the formalism of retrocausality for this specific case of QM experiments and saying it *must* be wrong because it violates your intuition. From an instrumentalist perspective it makes not one iota of difference if it violates your intuition. The question is can it be useful and allow a predictive framework. Sabine’s paper answers in the affirmative and says it is exactly as useful and predictive as standard QM. Whether you dislike that it involves retrocausality is immaterial.
Comment #136 March 13th, 2025 at 11:10 am
David Skelton #133: Thanks so much!
Comment #137 March 13th, 2025 at 4:22 pm
Scott #115 on superdeterminism re Peter Shor’s style
You asked for an example of usefulness. That’s a fair request. Let me take a quick detour, though. When Sir Penrose talked nonsense about the computability of the brain, most scientists simply shrugged and turned their attention elsewhere. But you didn’t, because you recognized the appeal and potential usefulness of an interpretation of the world where our minds would not be replicable, and because you justifiably thought you could provide a (much!) more coherent framework for this concept. I don’t know Sabine personally, but my understanding of one of her responses above is that she finds appeal in a form of Occam’s Razor where superdeterminism might be useful. Is that not all we should ask?
(Thank you, Gemini, for your assistance in refining this text.)
Comment #138 March 14th, 2025 at 4:11 am
Some weeks ago, when YouTube’s algorithm offered me a video of Scott Aaronson and Jacob Barandes I was very surprised. I had seen some stuff from Barandes and my spidy instinct said: “not really worth it”. I saw a very well read individual who knew a lot of stuff and that liked precision and completeness when speaking, but with a tendency to end up in square one.
I watched 2/3 of your interactions, and was very happy to see you “representing me” (yes, I’m a physicist). I really enjoyed the discussion and the tone from both of you. Really elegant stuff.
Then I also went to see the comments, and left my own, where essentially I said that you two are on different intellectual levels (which may be a bit harsh, but I was pissed at how many comments were supporting Jacob essentially because he appears nicer than you).
Then I came here because I just saw a paper on the arXiv that discusses some problems with real quantum theory (quantum theory using real numbers), which is a topic that makes me think of you. [“Gap between quantum theory based on real and complex numbers is arbitrarily large”, arXiv:2503.09724].
Anyways, I was sad to see that you got depressed from the comments picking on your body language (!), so I wrote this as emotional support 🙂
Comment #139 March 14th, 2025 at 5:36 am
This was a wonderful discussion to listen to! Thank you for both the podcast and this post. I 100% agree with every position you’ve taken in this discussion. I think you presented the sane, standard view in a polite but firm manner. I don’t understand the negative reaction — it could be that some of these folks don’t appreciate how deep/well-thought out your insights were.
I tuned in while driving for a couple of hours toward a U.S. consulate for my visa interview. By the time I arrived, I was wondering whether you pulled your punches and let Jacob off the hook. I did find your criticism crisp and to the point, but by the end he was still standing….
I was surprised you didn’t mention Bell’s inequality. I’m guessing this was a deliberate choice—perhaps owing to past encounters with Jacob or the fact that bringing up Bell’s inequality opens the door to more nuances and complications that muddy the issue.
I suspect he would answer a Bell inequality critique by saying it’s too high a bar, which even QM doesn’t clear (i.e., why permit non-local collapse in QM but not non local variables in his theory?).
Regarding his repeated claims that his theory might someday prove to be useful in the same way that Hamiltonian and Lagrangian mechanics proved to be, I would say that Lagrangian mechanics was useful from the start—e.g. the Brachistochrone problem—and Hamiltonian mechanics was developed precisely because of its utility with constrained problems. By contrast, his new theory is good for nothing so far (except ruffle feathers and make headlines).
Comment #140 March 14th, 2025 at 7:53 am
Some of the existing interpretations start to “pay rent”, if you acknowledge their week points, and try to attack those head on. This happend with Bohmian mechanics and density operators, for exampe:
For the current discussion about retrocausality and superdeterminism, interpretations with time symmetric versions feel most relevant to me. (By the way, even Bohmian mechanics can be interpreted in a time symmetric way, even so this is rarely ever done.) The consistent histories interpretation has a nice and simple time symmetric version. A “well known” weak point of consistent histories is its “robustness issue”, when it comes to “inexact zeros”. One way to try to attack this weakness head on might be to try look at the entropy of the probabilities of possible sets of histories. Not sure whether this will really “pay rent” at some point, but at least it leads to
non-trivial conjectures about bounds for that entropy:
Now proving or disproving such bounds might not pay much rent yet, but understanding why even those bounds themselves still show signs of the original robustness issues could be illuminating.
Comment #141 March 14th, 2025 at 4:54 pm
Dave #109, Looked over the paper: https://www.physik.uni-augsburg.de/theo1/hanggi/Papers/18.pdf, and I believe its definition does not include the more general class of stochastic processes that Barandes is discussing.
From my understanding, indivisible stochastic processes are indivisible because they can not be decomposed and so we don’t have the linearity condition that a “classical stochastic process” would have. So (5.2) would not apply here.
Indivisible stochastic processes are not “classical” in the sense of standard stochastic process theory (as referenced in that paper), but they are “classical” in the sense that they use standard probability theory with real-valued probabilities and normal probability axioms.
Comment #142 March 15th, 2025 at 8:56 am
Kevin Zhou #67:
Agreed, but “massive conspiracies” are not needed for bias to enter, people naturally become obsessed with fads, and there is definitely gatekeeping.
I don’t want to derail this thread with a MOND debate, so I’ll just conclude with saying that the picture I’ve always gotten when I’ve read about MOND successes from MOND researchers like McGaugh has never been “MOND is correct”, it’s mostly been “particle DM is adding too many epicycles”, and “theories of alternative gravity are justified by the evidence and should get far more attention”. A qualitative win for MOND and a quantitative failure for particle DM requiring more epicycles to correct is exactly that.
That said, your position that anomolies are common and nothing to worry about is also plausible, but it becomes less plausible over longer time frames. I think at this point, being very skeptical of particle DM is pretty reasonable.
Kevin Zhou #82
I think that’s a mischaracterization. Sabine has never said anything like “all of particle physics is a bubble and fraud”. Believing there is a bubble in particle physics does not mean it’s all fraud or all a bubble.
I don’t think Sabine is suggesting everyone go her route, there are many possible alternate routes. For instance, prediction markets are a big thing these days. What about a globally permitted science prediction market, where people could bet on the outcomes of experiments and theoretical predictions, and where X% of the proceeds from the resolution of a bet are pooled for future experiments. Predictions with the largest buy-in from the public get experiments funded from that pool. This eliminates gatekeeping, and provides a direct feedback loop in quantitatively determining who has a reputation as a good forecaster, and what teams and theories have a good track record.
Scott #115:
This analogy is ridiculous is because we have strong reasons to accept statistical independence in classical settings, and we have strong reasons NOT to accept it in quantum settings. This is a point Sabine has hammered on many times. But superdeterminism doesn’t even depend on this fine-tuning regardless, so it’s not even a well motivated analogy despite that failure.
If you’re genuinely interested in reading how superdeterminism could have more explanatory power and don’t find Sabine’s work sufficient, then I recommend Whence Nonlocality? Removing spooky action at a distance from the de Broglie Bohm pilot-wave theory using a time-symmetric version of de Broglie double solution, which furthers de Brogie’s work on superdeterminism which I mentioned earlier. At the very least you’ve found Bohmian mechanics a reasonable interpretation in the past, and this derives the only awkward feature (non-local pilot wave) from a more fundamental basis (which is also what Martin Ondráček in #131 says he is after). The main benefits are as Sabine said, providing local realism and thus compatibility with relativity.
Scott #116:
I don’t think that’s always possible. Some markets can only tolerate a single player (natural monopoly), not to mention vested interests politically sabotaging alternatives that could make them obsolete. The politics has always been and will always be the problem here, so solutions that free research from political interference are ideal.
Anyway, this has been an interesting set of discussions and I hope some people got some value from it.
Comment #143 March 17th, 2025 at 11:52 pm
Sabine #17: What’s so striking to me about your comment and others is that you’re claiming Scott said something he didn’t. You say: “I can be blamed for many things, but pushing alternative theories, mine or someone else’s, isn’t one of them.” But that just isn’t what you’re accused of.
So it begs the question as to why folks are pretending this is the criticism when instead it’s that you are misinforming your audience to believe most of science is useless, and done by ivory tower elitists “lost in math” and unconcerned with real world falsifiability, application, or understanding so that:
> …anyone perceived as the “mainstream establishment” faces a near-insurmountable burden of proof, while anyone perceived as “renegade” wins by default if they identify any hole whatsoever in mainstream understanding.
It seems to me a secondary point was that you simply don’t subject your own theories, like superdeterminism, to the same level of skepticism as every single other study you cover on your channel.
In **many** of your videos about quantum computing for example, say this recent one about Google’s Willow chip, you are extremely negative and misleading.
In this video, you discuss one sentence in a press release (there is a paper!) about the Willow Chip. You don’t discuss the genuine breakthrough in demonstrating that one can scale up the surface code size while maintaining coherence for longer. This is central to fault tolerant quantum computing and the viability of the whole enterprise, yet you focused on a claim whose legitimacy turns little on facts and largely on one’s interpretation of quantum mechanics.
We are left believing once again that Google did little else but another random circuit sampling experiment, with an arguably unscientific claim about parallel universes.
Comment #144 March 21st, 2025 at 10:28 am
I found it very brave to agree to participate in this youtube format that also has the tendency to give way over proportional weight to fringe views on science (because it lacks the depth to actually investigate details at a technical rather than conversational level). I also resist to join the discussion about Sabine’s videos and her relation to her vocal audience or reasons why the academic system is not more welcoming to outsiders (hint: if there were a shortcut to 10+ years of studying to contribute to high energy physics or cosmology where students learn about the many things that work for the “orthodox theory” but which your finge idea would throw out with the bath water, we would not make students do that) to get to what I am here to do: Shameless self promotion.
Scott, you give the Bohmians too much credit when you say their trajectories “are so constructed to reproduce the predictions of QM perfectly”. Well they might be so constructed and they will tell you they succeeded, but as soon as you make predictions at more than one instant of time (and thus invoke non-commuting position observables), they fail! I am not the first one to point this out but in a recent paper I have a particulary simple (and thus completely solvable) system to demonstrate this: https://arxiv.org/abs/2502.14449
Comment #145 March 21st, 2025 at 4:08 pm
Robert #144: I read that paper after a colleague pointed me to it, but I think it’s completely wrong—or rather, that it knocks down a straw man that not a single Bohmian believes. A key clue is the weasel sentence in the abstract, where it claims Bohmian mechanics is empirically false “unless enlarged by a microscopic theory of collapse of the wave function after observation.” By that, the author simply means you need to be willing to talk about the wave function of the entire universe (just like the Everettians do), including the humans, the measuring devices, etc. But all Bohmians are willing to talk about that. At some level, the only difference between Bohmians and Everettians is that the Bohmians add a hidden variable, which is governed by a specific differential equation that they postulate, such that it obeys the |ψ|2 distribution at every individual time.
Ah, but what about multi-time correlations? As soon as we’re talking about the wave function of the entire universe, it becomes obvious that multi-time correlations are not directly observable in QM. What’s observable is only, e.g., the correlation between the position of a particle at time t2, and your record of where the particle was when you measured it at an earlier time t1. But to measure the particle you needed to interact with it, and then we’re again just talking about the standard Born rule, and the Bohmians will make the same predictions as everyone else.
A quarter-century ago I wondered: what would happen if we changed QM so that multi-time correlations were directly observable, as they aren’t in Bohmian mechanics? The result was my paper Quantum Computing and Hidden Variables, where I show (e.g.) that this change would slightly enhance the power of quantum computers, letting them break cryptographic hash functions in polynomial time and do Grover search in N1/3 instead of √N time.
Comment #146 March 21st, 2025 at 5:12 pm
Robert #144: I have not read your entire paper, but I disagree with Scott’s evaluation of it. My point is that you tried to communicate with Tim Maudlin and other Bohmians before, and there was this point where a disconnect happened. This happened to me too, and I then even tried Scott’s suggestion to “apply Everett’s strategy” to overcome it:
It didn’t help either.
But let me also say that I am extremely grateful to Tim Maudlin and other Bohmians that always stayed friendly and tried to bridge the disconnect. And they stayed friendly even after “we established” that we probably won’t be able to bridge it.
The reason for the disconnect in my case is probably that I want to use BM in an instrumentalistic way. Then I get exactly the problem you describe at the bottom of page 10:
However, I did not succeed to convince the Bohmians that such a way to use BM could be worthwhile, even less in explaining how it could work, more or less:
Comment #147 March 22nd, 2025 at 3:57 am
Scott, regarding your question “but what does it buy me?”, what do you think about what Jacob offers in his recent paper https://arxiv.org/abs/2402.16935?
He extends Bell’s definition of causal locality in a way that aligns with a well-established body of work in Bayesian theory and what one can actually say about causation.
If this holds up, I’d say it buys us a lot, intellectually. For instance, it would resolve the measurement problem (as he argues in the paper) and could eliminate the need to invoke non-local mysteries in Nature, because it provides a causally local interpretation of them.
Comment #148 March 22nd, 2025 at 4:40 pm
#135 Adam Treat, I agree with just about everything you are saying about superdeterminism. Regarding the following, responding to Drizzle #134:
> You lost me completely with, “in the same way that quantum
> mechanics doesn’t contradict classical intuition” LOL
I think what Drizzle meant is that QM doesn’t contradict classical intuition *in the realm in which the latter applies*. Or IOW, the correspondence principle.
Similarly a good superdeterministic theory, which replaces Statistical Independence (SI) with a different constraining principle SI’, should be able to explain (ideally in a non-contrived way) why SI works so well as an assumption in so many situations – for example, in the classical realm – in spite of its failure in the Bell-type experiments.
In addition of course they need to show how their replacement principle SI’ evades the conclusion of Bell, and (perhaps most importantly) how SI’ is constraining enough to make the theory epistemically sound (i.e. “make science possible”) without SI.
This all seems like a daunting task to me, albeit not necessarily impossible afaik. That said, the onus is very much on the superdeterminist proponent to do this, rather than the skeptic to prove it can’t be done, in order for superdeterminist theories to be considered viable or even a promising direction. (Though if the skeptic is making the strong claim that *any* approach along these lines is *bound* to fail, leading to a vacuous theory devoid of scientific merit, I think the onus switches back to them to show this, which to my knowledge no-one has.)
Comment #149 March 22nd, 2025 at 5:14 pm
dennis #147: The idea of trying to get both “local causality” and the observed results of the Bell experiment, by redefining what one means by “local causality,” has been mystifyingly popular for decades yet has precisely zero appeal to me. Rather than fighting over the definitions of words with established meanings, why not just accept that the words mean what they mean and the theorems prove what they prove, and then clearly postulate whatever new thing one wants, using new words?
Comment #150 March 22nd, 2025 at 5:47 pm
#gentzen 146, the issue you quote in that paragraph starting “Another problem for a Bohmian model …” has always bothered me as well. The true positions \(q_i(t)\) in the theory are down a kind of “causal cul-de-sac”. The value \(q_i(t)\) neither influences the wavefunction nor even the positions of other particles \(q_j(t)\) for \(j\neq i\). Since measurements involve interactions of some kind, when we measure the position of a particle hitting a detector screen, we’re not really measuring the “true position” \(q_i(t)\) at all as that variable doesn’t influence anything else. What we’re actually measuring is some aspect of the wavefunction. That seems rather to undermine the base for having positions as objectively real in the first place (“ultimately all measurements are of position”), as these objective positions aren’t even really what we’re measuring!
What I do find of interest in Bohm from a theoretical point of view is comparing it to Everett. The configuration \(q_i(t)\) at time t will pick out one of the (approximate) Everettian branches of the wavefunction \(\psi(t)\). But since \(q_i(t)\) is a deterministic function of \(q_i(0)\), this means that the initial conditions \(q_i(0)\) parametrise the branches of \(\psi(t)\) at any time \(t\) in the future. This seems kind of wild – the number of branches will presumably generally be exponential in the time \(t\) and yet the entire future history is encoded in the \(q_i(0)\). Presumably the “branch function” at time t becomes exponentially (in t) sensitive to the initial conditions \(q_i(0)\). Which actually isn’t so wild after all, since something similar happens in (classical) chaos theory. Still, it would be a good exercise to map out what this relationship looks like, at least for simple toy systems. (Others have almost certainly already done this, to the extent the problem is tractable…)
Comment #151 March 22nd, 2025 at 6:04 pm
self-correction (sorry). The \(q_i(t)\) in Bohm *do* actually influence the other \(q_j(t)\) for \(j\neq i\), because the former appears in the argument of the current field guiding the latter. This is indeed the very reason for the non-locality of Bohm.
So you can call the \(q_i(t)\) as a whole a causal cul-de-sac (in that they don’t influence the wavefunction) but not each particle position individually.
(Scott, obviously feel free not to post my last comment or this one, as the former contained a significant error.)
Comment #152 March 23rd, 2025 at 6:53 pm
flergalwit #150: My impression of Robert’s paper was that he understands pretty well how BM “can work”, especially if you want to use it in an instrumentalistic way: In the Copenhagen interpretation, not only do measurements collapse the wavefunction, but you can also change the Hamiltonian depending on outcomes of measurements. (Quantum error correction is often presented in that way, even so it is not strictly required.)
BM takes care of wavefunction collapse in its own way, but BM remains silent on how the Hamiltonian is allowed to depend on the outcomes of (position) measurements. I guess Robert had this part in mind with that paragraph. Of course, if the positions could influence the Hamiltonian, then they could indirectly also influence the wavefunction. But I don’t think that it “bothers him” in the way that it bothers you. Rather, he just points out an important limitation of BM (compared to Copenhagen). But it is not total a showstopper, because this type of feedback is also not easy to realize in actual experiments.
The particle positions still provide a description of space to complement the wavefunction, which is missing from Everett MWI. Let me try to illustrate this with my attempt to explain how it (to use BM in an instrumentalistic way) could work, more or less:
The reaction of the Bohmian was that the Hamiltonian is given by nature, so it is unclear what I mean when I claim that I can change it. He also remarked that he doesn’t fully understand what I mean by “some linear canonical transformation of spatial variables”. So I tried to better explain that part:
Comment #153 March 28th, 2025 at 7:54 am
Corey Leander #143:
A bunch of headlines talked about “proof of the multiverse”, Sabine gets asked about these articles, she posts a video discussing the multiverse and quantum computing to answer those questions, and you’re upset she didn’t also talk about specifically what you’re interested in discussing?
That’s not serious evidence of being unfair, unscientific, disingenuous, misleading or any of the other things she’s been accused of in this thread.
Comment #154 March 30th, 2025 at 8:14 pm
I got lost somewhere in the above thread’s discussion of Sabine, science as a public good, fringe vs mainstream, etc., and so I will simply add some points and references that might help clarify the issues. Probably no one will read it, but the discussion so far has been far too conclusionary and superficial in common ways for me to overlook it.
1) Basic science is empirically not dependent on taxpayer funding. It MAY be dependent on the margin, i.e. there might be more of it because of government support than there would be otherwise, but that depends on whether public funding crowds out private funding of various kinds (both for-profit and charitable/consumption).
The first industrial revolution (textile mechanization and iron and steam power) and second industrial revolution (electricity and steel and radio) proceeded without any significant government support at all, and the identification of the elements and the laws of electromagnetism and relativity and the rest up to WWII were not dependent on centralized government support. Even the first giant telescopes, the beginning of Blg Science, were funded privately (Hale was good at raising money).
These are basic facts, and pretending that history occurred in some alternate universe where government funding of curiosity-driven research was crucial to modern science and technology is pointless.
2) Privately funded science, like publicly funded, is subject to fads and attraction to shiny objects, etc. (I suggest a reading of Sinclair Lewis’s Arrowsmith, where microbiologist Paul deKruif was his science expert, for a satirical but ultimately optimistic description of this in biology.) OTOH, the various foundations set up by industrialists seemed to do a pretty good job of picking things to fund (this was before their woke descendants diverted their resources to identity politics and such). Interdisciplinary fields such as geochemistry were largely started by the funding from these foundations.
3) The best idealized description of how a self-regulating set of would-be discoverers in basic science can flourish and advance the frontier of knowledge is probably the essay by Michael Polanyi “The Republic of Science.” He analyzes the feedback loops and norms and practices that he observed as a physical chemist and generalized to other fields he encountered. It is useful to compare and contrast the explicit and implicit conditions in that essay with how today’s science institutions behave.
4) A useful serious “endpoint” writer on science policy is Daniel Sarewitz. He takes and defends the position that decentralized curiosity-driven basic research on a mass scale doesn’t work either to society’s benefit or even to the advancement of the frontiers of knowledge, because the incentives are wrong. (Also worth comparing and contrasting with Polanyi’s discussion.) See his essay “?Saving Science” in The New Atlantis https://www.thenewatlantis.com/publications/saving-science
5) Another serious “endpoint” writer on public support of science is Terrence Kealey, a practitioner who has argued that basic science is paid for by the private sector when it is a tool to recruit talented people to do applied research, and that public support of R&D outside of military applications crowds out private funding and distorts and reduces scientific progress. Here is a hostile review of his book which inadvertently concedes much of what he argues. https://issues.org/nelson/
6) As to the “problem” of science-curious people being attracted to “fringe” theories and not being deferential enough to the professional specialists, this is a natural and indeed healthy thing on the whole. (That judgment only applies to the genuinely curious, of course, and not to those who are acting out other impulses by verbally attacking the motives and intelligence of professional researchers. Such people can be found on any Internet comment thread; for example, when female thirst-trap influencers show off their barely clad bodies one can usually count on someone to opine that they are really men in drag. It has nothing to do with science or the quality of public discourse about it.)
It is healthy in the sense that many scientists have remarked on how occasionally going back to teach introductory courses was helpful to them in restoring their big-picture curiosity about the body of their field, how it fits together, and how strong the evidence and logic really is for various claims commonly made in both the specialist and pop-science literatures. (Economists are obligated not just in every generation but in nearly every year to go over the grounds of why trade protectionism doesn’t work the way most people think, despite oodles of evidence and detailed reasoning behind their position.)
If one thinks of the history of a scientific field as a path through a branching tree, with the “consensus” choice at each node representing an evidentiary judgment, then experts are naturally obsessed with the latest node, taking for granted all the previous branching points. But was the field too hasty or wrong at an earlier node? This is not an unworthy question even if it vexes the experts.
Specialists not used to persuading the public (unlike economists, for example) about the basics of their field get frustrated by such queries. The leading researchers arguing at the “which branch off the dark matter node is the best one to fit the data?” get exasperated even by their own dissenting colleagues who wonder if DM is even the right way to go. And in the comment threads of the researchers asking about the DM/MOND node, you’ll run into Arpists who question the validity of redshifts as measures of distance, and on the Arpist threads you’ll run into Alfven/Lernerists who wonder if elecromagnetic plasma threads rather than gravitational forces dominate celestial mechanics, and the plasmaist threads probably have Velikovskyite “electric universe” proponents.
Having people question the earlier branching choices at least keeps the frontier specialists grounded in the evidentiary and logical basis for their field as a whole. And it may provoke serious reconsideration when a field seems “stuck.”
Comment #155 August 18th, 2025 at 4:13 pm
No offense dude you’re a computer scientists you don’t need the tools Barandes is providing to solve any of your problems. Unclear why you were the moderator here, the discussion should have steered towards barandes claims about using his formalism to make better quantum algorithms to make your experience relavent.
Comment #156 March 26th, 2026 at 4:08 am
“OK fine, you can do that if you want, but what does it buy me?”
If you are a Bohmian, it buys you clean extension to more general quantum theories like QFT.
If you are an Everettian, it buys you probabilities and a parsimonious ontology.
If you are a Copenhagenist, it buys you a microphysical ontology.
etc.