spekcular

My experience in academia has taught me the following: Any slowdown in productivity among mathematicians as they age is more a result of increased administrative duties, childcare, or loss of interest than it is cognitive decline. The slowdown is real, but the commonly suggested reason (biological aging) is not quite right.

You can look up cognitive decline studies and see that it really doesn't hit in full force until the 60s and 70s, with lots of heterogeneity. I've seen plenty of sharp 70-year-old academics doing great work (though I can't remember someone at 80 who I thought was still going full steam).

PNAS publishes a lot of junk. If you're connected to the Academy you can get papers through with very light review. In fact, one can even seen this happened in the Tipler paper: "This Direct Submission article had a prearranged editor." [0]

Andrew Gelman calls it a "tabloid": https://statmodeling.stat.columbia.edu/2014/06/17/hurricanes....

I will check the paper out though.

[0] https://www.science.org/content/blog-post/no-more-prearrange...

Edit: I read the paper. It is borderline crank work. There are decent arguments for the Born rule in MWI but this is not one of them. (I like David Wallace's book, for example.)

Your reference [2] is somewhat suspect. As far as I know, it is not generally accepted in the physics community that the Born rule can be derived directly from the Schrodinger equation. In fact, it seems generally accepted that such a derivation does not currently exist, and MWI has to stick it in as an additional postulate in some form (or otherwise give an auxiliary argument for it, none of which seem generally accepted either).

Further, the author of that paper, Frank Tipler, is a notorious nutjob: https://en.wikipedia.org/wiki/Frank_J._Tipler.

I don't have time at the moment to review the paper in detail to locate any potential flaws, but I wanted to flag this for others. I think there is substantial reason to be skeptical.

In your opinion, what are the best electric designs available today (that aren't crazy expensive)? Would also be interested in acoustic options!

A nasty integral! https://mathworld.wolfram.com/DiskLinePicking.html

Also, no sunglasses in "A photo of a raccoon wearing sunglasses and a red shirt riding a bike in a garden," and a few similar prompts (e.g. surfing).

I think the author makes some good points, and certainly I find this article more more compelling than the poker one. Trick-taking card games seem like a great choice for kids. However, when I looked into bridge specifically, the following things turned me off.

1) It requires a dedicated, long-term partner. If your partner isn't available, you can't play. "Pick up" bridge with strangers simply isn't feasible in the way that randomly matching teams on, e.g., an online FPS or tower defense games is.

2) The game is fundamentally broken by certain bidding strategies, which must be banned.

3) There's no way to prevent opponents from cheating by using out-of-game signals (e.g. timing of bids). This can even be done unintentionally. So fair play is always an open question.

Perhaps someone who knows more can comment about whether I'm wrong?

My problem with this article: I did not see a concrete suggestion for where children should play poker.

Clearly, they cannot play at casinos or for real money online. I suppose they could play for fake money online, but there the dynamics are much different. (Why not go all-in and risk it all if you can just create a fresh account if you lose?) Fake money tournaments, where the goal is to come out on top instead of maximize profit, are probably the best option. But the word "tournament" doesn't appear in the article (and even then, is it really good for kids to be hanging around with the people on online poker sites)?

I suppose implicitly the author is talking about playing heads-up against a parent at the dinner table, or with close family members, with fake money. But people who don't know anything about poker are going to play essentially randomly, and the game won't last long enough/will be too variable for the kid to see any reward from playing "correctly," so I'm not sure there's a point.

Also, the first thing to teach the kid is to fold most of their hands. But sitting out of hands is boring and the kid is going to get bored. It's called "grinding" for a reason...

Something like MTG that isn't pay-to-win would be a better choice.

The electron's magnetic moment is not a free parameter that must be determined experimentally. It can be computed from other parameters, as done for example here: https://arxiv.org/abs/hep-ph/0406325.

The discrepancy in the article you linked is completely unrelated to how well we conceptually understand renormalization (which is broadly applicable to many quantum field theories, not just the standard model). It could be the case that the standard model is wrong, but my claim would still stand.

This paper came up previously: https://news.ycombinator.com/item?id=31217791

I stand by my comment from before:

The person who wrote that paper doesn't understand the basics of the field that he's talking about.

For example (quote from the other thread): "Consa gives an analogy wherein Indian mathematician Srinivasa Ramanujan has claimed that the sum of all positive integers is not infinite, but is instead -1/12. It’s wrong, it’s absurd, but renormalization has now been accepted, and is even sold as a virtue."

One when performs zeta function regularization, one gets -1/12. This isn't some mystery; it's a perfectly reasonable thing to do. Analytic continuation has been understood since the 1800s.

Reference: https://en.wikipedia.org/wiki/Zeta_function_regularization

Also, the claim that Karplus and Kroll committed "fraud" is basically libel, as can be seen by reading the complete account. The worst one can say is that people didn't publish full details of calculations due to page limitations or laziness, but this is hardly a special feature of QED. For instance, Onsager famously solved the 2-d Ising model exactly in 1944 but never provided details in print, just the final solution.

This is an outdated, pre-1970s view of renomralization. Thanks to work of Wilson (1982 Nobel prize) and others on the renormalization group, we have a much better understanding.

One good article that explains this shift (in the context of a debate in the philosophy of physics) is here: http://philsci-archive.pitt.edu/8890/.

This article is representative of my experiences as well. I'm thankful to the author for writing it. My favorite parts:

> This slowness affects other software written in Julia as well. For example, my editor formats on save. For most languages, the formatting time-delay is not noticable, but for Julia it sometimes took more than 10 seconds.

> The idea of “You only need to start the interpreter once a day” failed for me.

For me, this is the biggest obstacle to using Julia. It's not designed to support the traditional C/Python/whatever workflow of editing a text file and then interpreting/compiling it; it wants you to be in the REPL all the time. If you don't use the REPL, you have horrible problems with package loading times. So if you're someone who doesn't like developing in a REPL, you're out of luck.

Some comments by Julia users on the post are available here: https://discourse.julialang.org/t/blog-post-about-my-experie...

Not Japan, for instance: http://gregmankiw.blogspot.com/2022/05/an-inflation-puzzle.h...

Thanks!

> The purpose of building a collider that big and expensive is to find new particles.

This is not true! There's much more to particle physics than simply hunting for new particles. Our knowledge of even the particles we know to exist is massively incomplete. Several next-gen collider proposals are for "Higgs factories" to study the Higgs boson better, for example.

Yes, I'd love to read it, if you have time to find the DOI. Thanks!

"There's a lot of engineering and construction involved but the new ones are really just very expensive and huge without netting necessarily much in terms of engineering science."

This is extremely false. The new engineering science required to make the incredibly strong magnets for each collider (newer colliders requiring stronger magnets) is one obvious counterexample.

Damn.

I never actually learned this stuff. Is there a good textbook account of this isotropic-to-nematic transition that includes full details? Or is there still a gaping hole in the published literature?

In other words, did your colleague do this as a kind of history project, or because the details weren't available anywhere else?

The person who wrote that paper doesn't understand the basics of the field that he's talking about.

For example (from the blog post): "Consa gives an analogy wherein Indian mathematician Srinivasa Ramanujan has claimed that the sum of all positive integers is not infinite, but is instead -1/12. It’s wrong, it’s absurd, but renormalization has now been accepted, and is even sold as a virtue."

One when performs zeta function regularization, one gets -1/12. This isn't some mystery; it's a perfectly reasonable thing to do. Analytic continuation has been understood since the 1800s.

Reference: https://en.wikipedia.org/wiki/Zeta_function_regularization

Edit: I read more of the linked paper. The claim that Karplus and Kroll committed "fraud" is basically libel, as can be seen by reading the complete account. The worst one can say is that people didn't publish full details of calculations due to page limitations or laziness, but this is hardly a special feature of QED. For instance, Onsager famously solved the 2-d Ising model exactly in 1944 but never provided details in print, just the final solution.

"Not a whole lot to show" from the LHC? We got the Higgs Boson, and piles of data providing more accurate measurements of fundamental constants in the Standard Model, along with a ton of progress in various "applied" areas related to developing the materials/tools for constructing the thing.

You're right, of course. It's just much less clear to me why local anesthesia would be implicated in cognitive impairment. But the brain is a complex thing, so I suppose my intuitions don't count for much!

IIRC there are still post-operative cognitive problems in major surgeries done without general anesthesia, suggesting that there's something else going on too. (I don't doubt that anesthesia plays some role, though.)

edit: Found the reference: https://www.frontiersin.org/articles/10.3389/fpsyt.2018.0075....

"More convincingly, a recent study demonstrated that open abdominal surgery under local anesthesia caused increases in hippocampal IL-6, TNFα, and memory impairments (71), suggesting that anesthesia per se is not necessary for the production of neuroinflammation and subsequent development of POCD."

"As seen in observational studies, a prospective randomized clinical trial comparing the use of general vs. spinal anesthesia in extracorporeal shock wave lithotripsy showed no significant difference in the incidence of POCD defined by a neurocognitive battery (142), suggesting that surgery and not anesthesia causes POCD."

> Still, SR followed from theories that were well proven empirically, and had obvious experiments that could falsify it.

Sure, my point is just that reasoning based on thought experiments, parsimony, etc., played a key role in its discovery.

> Well, I think here there is a difference between trying to work out consequences of an otherwise well-established theory (such as working out exactly how gravitational waves would look like in GR, or arguing based on Maxwell's equations) and trying to come up with a new theory that modifies existing ones (such as grand unification or some solutions to BHILP).

I agree, but this claim gets a bit far from what I take Hossenfelder's point to be (as does the UV catastrophe stuff). You and I both accept that certain theoretical activities are productive, even in the absence of empirical evidence (e.g. working out the consequences of a well-established theory). Hossenfelder's argument, as best I can tell from that blog post, implies that she does not accept these activities as worthwhile. I think there is an interesting conversation to be had about exactly where we should draw the line between worthwhile and not worthwhile theorizing when we don't have data forthcoming, and I think there's a good case to be made that the study of BHILP falls on the "not worthwhile" side of that line. But Hossenfelder doesn't make that case, or seem to admit any nuance: If there's no data, you're just "lost in math" and not doing productive work.

> By that logic, you shouldn't believe in black holes / GR at all, right? In practice, we can always replace infinities with some arbitrarily large numbers that don't grow all the way to infinity because of yet-unknown physics (probably quantum gravity, in the case of black holes).

I don't see how this follows. The ultraviolet catastrophe (or other divergences) says we have to fix something, it doesn't say that we should choose a weird ad hoc fix.

> I don't think this would have been promising if the MM experiment seemed a hundred years away, and in general I don't think SR would have been as compelling in that case.

It probably wouldn't have been as compelling, I agree. Don't get me wrong; the gold standard is empirical evidence. But the invariance arguments and procession of mercury would be grounds for taking it seriously. Einstein's argument for the theory depends only weakly on MM.

> Sure, it's a difference in degree in the end. The plausibility of having an experiment "soon" is the degree here. I don't think that Hossenfelder would argue that if an experiment is only possible 2 years from now, or maybe even 20 years from now, you shouldn't work on some theoretical subject. But, when that time horizon stretches well beyond your lifetime and the lifetime of your students, it's perhaps time to reconsider.

This seems like an unnecessarily narrow view of what constitutes worthwhile physics. First, because the experimentalists are very clever, and there's no knowing what indirect tests they might propose. But mainly because, if we can make theoretical progress on an important question, why not do that (even if empirical data is not forthcoming)? This view suggests that a physicist in 1915 (or 1925, etc.) should not try to work out properties of gravitational waves, for example, which seems obviously ridiculous to me. (The direct confirmation came about 100 years later.) If the theoretical motivation for doing so is solid, why not?

I agree that BHILP paradox is probably a different case, one where we might really have no shot of saying something useful theoretically. But this requires getting into the specific details of BHILP. These general statements about what is and isn't good physics because of near-future testability all seem clearly suspect.

If I told you "some unobservable region of spacetime" had to display a certain feature, or there would be a mathematical contradiction, would you accept that?

If not, would you endorse believing in gravitational waves on the basis on the success of the rest of GR, before they were experimentally confirmed?

> Well, that theory also predicts that the universe can't exist for very long, which is a pretty big experimental inconsistency.

Sure, but I don't need to appeal to this fact to know there's a problem that must be solved. The infinities are enough.

> Well, Maxwell's equations were relatively well experimentally verified by other experiments, so there were good reasons to at least tentatively accept their prediction of a constant speed of light* in any rest frame as a given. Even if the Michaelson-Morley experiment was not big on his mind, it was still relatively clear that this type of experiment could be performed with technology already available at the time, as the speed of light had been measured with pretty good precision already for a few decades.

But again, even if MM couldn't have been performed, Einstein still would have had good theoretical reasons (consistency with Maxwell's equations) to posit SR. And indeed, this was the path the discovery seems to have actually taken.

> So, SR was not some highly speculative theory based only on extrapolating other theories, as solutions that only address BHILP but not other problems of QM/GR are currently.

Sure, but this is a difference in degree, not in kind. You accept that theoretical arguments for new phenomena that are not directly testable (at present) are acceptable if they're convincing enough. The question about BHILP under this view (which I agree with) is then about how convincing the theoretical arguments are.

Hossenfelder, taken literally, suggests that no theoretical arguments will ever good enough in the absence of empirical evidence.

> So, why not work directly on the other inconsistencies, which might be directly testable, and see if this one disappears that way?

These are not mutually exclusive options. The community works on both.

Again, I agree that so far, practically speaking, work on BHILP has not been super compelling. I take issue with the stronger stance that Hossenfelder seems to take, that we know a priori that work on BHILP will be worthless because experimental data is not forthcoming. ("And that’s why I stopped working on the black hole information loss paradox. Not because it’s unsolvable. But because you can’t solve this problem with mathematics alone, and experiments are not possible, not now and probably not in the next 10000 years.")

Sorry, I wrote poorly. I don't mean to say that GR or gravitational waves were in some sort of indeterminate level of credence. I mean to say that our credence in gravitational waves in the absence of experimental confirmation is justified by exactly the kinds of non-empirical evidence that Hossenfelder rejects. So it's clear that she's being too restrictive in what she believes is worthwhile theoretical physics.

I don't think I misunderstand her position. She seems to take the position you give:

"However, if you come up with a theory that makes the exact same predictions as QM and GR except for the BHILP, then your theory is not very interesting, since we will never be able to test this theory, and there are other inconsistencies between QM and GR that you haven't solved, so we can't just rest on our laurels and say physics is over."

I think this is wrong. There are non-empirical reason reasons we might prefer the new theory to the old one. For example, we know QM are GR are inconsistent, so if I give you a parsimonious, consistent theory that captures both QM and GR in the appropriate limits, then that's a great reason to prefer it. In principle, we could prove theorems (and some theorems of this form have been proved) that pin down the possible resolutions of the GR/QM inconsistency. If we tighten the net so that only one theory remains, then obviously we should choose to believe that theory.

Now, will that actually happen? I don't know. But merely waving your hands and going "not testable" is not enough, because we all accept that mathematical consistency ought to have a huge influence on theory choice. You need to make some argument about BHILP specifically saying that theoretical arguments will never produce productive physics.