News for hackers. Stuff that matters.
Re:Claude Fable 5.2 solves two open problems in the physics ... by bananaflag on Thursday November 19, 2026 @01:08PM (#50563684)
Copying my comment from that discussion:

Those theoretical physics papers look like the real thing (i.e. not mathematical physics like the OpenAI Yang-Mills result). It looks like AI is finally branching out of math.

https://news.ycombinator.com/item?id=50430239

Re:Claude Fable 5.2 solves two open problems in the physics ... by pizza234 on Thursday November 19, 2026 @01:20PM (#50563889)
> It looks like AI is finally branching out of math.

Is it really, though? It's still a heavily mathematical domain, rather than one involving empirical research, where AI still struggles.

Re:Claude Fable 5.2 solves two open problems in the physics ... by ModernMech on Thursday November 19, 2026 @01:44PM (#50564320)
We used to joke in my physics undergrad that they didn’t need a math department because we were doing all the stuff they were doing but with actual applications. Then the mechanical engineers said the same thing about us!
Re:Claude Fable 5.2 solves two open problems in the physics ... by Sharlin on Thursday November 19, 2026 @01:50PM (#50564401)
One of the most efficient ways to torture a pure mathematician is to force him to watch how physicists treat math.
Re:Claude Fable 5.2 solves two open problems in the physics ... by bananaflag on Thursday November 19, 2026 @01:46PM (#50564358)
I agree, but it is the kind of math done the physicists' way (which I found recently asserted that AI cannot do) and not the kind which is formalizable in Lean.

"It's not much, but it's honest work."

Re:Claude Fable 5.2 solves two open problems in the physics ... by suddenlybananas on Thursday November 19, 2026 @01:55PM (#50564505)
Why would it not be formalizable in Lean? People are formalizing parts of linguistics in Lean.
Re:Claude Fable 5.2 solves two open problems in the physics ... by boothby on Thursday November 19, 2026 @02:22PM (#50564916)
Most famously: to a physicist, the sum of all natural numbers is -1/12 https://en.wikipedia.org/wiki/1_%2B_2_%2B_3_%2B_4_%2B_%E2%8B...

If you attempt to formalize math involving such assumptions, one expects to encounter a proliferation of contradictions. So I’m curious to see how the contributions of AI authors is handled by the physics community. With pure math, you can ostensibly write a compact statement of a theorem, and if so, trust the process.

Re:Claude Fable 5.2 solves two open problems in the physics ... by jakzurr on Thursday November 19, 2026 @02:59PM (#50565455)
OK, that goes right past me. However, the reference, in the article, to https://en.wikipedia.org/wiki/Monstrous_moonshine seems appropriate.
Re:Claude Fable 5.2 solves two open problems in the physics ... by coderenegade on Thursday November 19, 2026 @04:17PM (#50566511)
Yeah the use of divergent series in modern physics is certainly something. There is an equality that includes the sum of the natural numbers in it, and that evaluates to -1/12, but it's not equal to the sum of the natural numbers on its own.
Re:Claude Fable 5.2 solves two open problems in the physics ... by dnautics on Friday November 20, 2026 @01:03AM (#50570748)
-1/12 is a value of the zeta function so if your infinite sum is actually manipulations of the zeta function you're good.
Re:Claude Fable 5.2 solves two open problems in the physics ... by layer8 on Thursday November 19, 2026 @02:02PM (#50564623)
From skimming the papers, it’s not clear to me what it is actually doing outside of proving what can be considered mathematical theorems (even if that’s done the physicist way). It doesn’t seem to be making testable predictions (but maybe I missed those).
Re:Claude Fable 5.2 solves two open problems in the physics ... by BeetleB on Thursday November 19, 2026 @02:04PM (#50564639)
> and not the kind which is formalizable in Lean.

For people not in the know, it's common - both now and historically - for physicists to use "mathematical techniques" that have not been proven valid.

I took grad level physics courses, and this drove me bonkers. I was accustomed to the comfort mathematics provided me. If something was unclear, I could always dig deeper and find a proof. But a lot of the stuff the physics professors would do - I couldn't, and nor could they. They just responded with "intuition". Some of them had dug deep earlier in their careers and stated that the technique was likely invalid mathematically, "but still useful in getting answers experiments can verify".

I won't even talk about relying on diverging series...

Re:Claude Fable 5.2 solves two open problems in the physics ... by ape4 on Thursday November 19, 2026 @02:44PM (#50565230)
Can you declare theorems in Lean for those techniques. And list them in the paper.
Re:Claude Fable 5.2 solves two open problems in the physics ... by throwaway81523 on Thursday November 19, 2026 @06:44PM (#50568219)
Trouble is you can declare things to be true that are strictly speaking mathematically false, and from there get to proving any falsehood.

The Dirac delta function is a famous example of something mathematically non-rigorous at its inception, and mathematicians objected to it, but a few decades later Laurent Schwartz developed generalized function theory, that made it work.

Or going even further back, calculus was done in terms of ratios of infinitesimals (like dy/dx) or infinite sums of them (integrals). A century or two later the formulas of calculus were proven rigorously through classical analysis (delta-epsilon proofs) but that lost sight of the original motivations. A century or two after THAT, nonstandard analysis made a rigorous theory of infinitesimals, but reliant on relatively high powered mathematical logic that didn't exist til the 1950s or so.

Right now I think quantum field theory (renormalization, path integrals etc.) have been put on a sound mathematical basis but physicists mostly don't care. And very recently, an AI proved something about Yang-Mills theory that helped rigorize part of the Standard Model of particle physics. I don't understand that last part at all.

Besides physics, a lot of the procedures of machine learning and AI are mathematically suspicious but people use it anyway because it empirically works. People in the field have told me that yeah, they expect mathematicians will someday get around to backing it up, but for now, practitioners are moving forward with what they have.

Re:Claude Fable 5.2 solves two open problems in the physics ... by mr_mitm on Thursday November 19, 2026 @05:28PM (#50567405)
When the lecturer wrote 1+2+3+...=-1/12 on the black board, I dropped out of the string theory class.
Re:Claude Fable 5.2 solves two open problems in the physics ... by ndsipa_pomu on Friday November 20, 2026 @03:24AM (#50571738)
Why? Are you not a fan of analytic continuation?
Re:Claude Fable 5.2 solves two open problems in the physics ... by mr_mitm on Friday November 20, 2026 @04:33AM (#50572283)
Let's just say he didn't motivate it very well, not to speak of deriving it. And it was just the final drop in the bucket, I never understood much of it in the first place.
Re:Claude Fable 5.2 solves two open problems in the physics ... by teamonkey on Thursday November 19, 2026 @07:37PM (#50568638)
Can you give examples? Sometimes a simplification is enough to illustrate a theorem, especially to an undergraduate class, without the full proof being needed. The mathematics is a tool to understand what’s happening.

More mathematical rigour would be required for formal papers, though it wouldn’t need to be derived from first principles.

Re:Claude Fable 5.2 solves two open problems in the physics ... by skeledrew on Thursday November 19, 2026 @05:20PM (#50567289)
> asserted that AI cannot do

I just love to see this, particularly the response when AI then goes on to do it.

Re:Claude Fable 5.2 solves two open problems in the physics ... by keeda on Thursday November 19, 2026 @02:04PM (#50564642)
In my (limited) experience it does not struggle with empirical research — At least in the procedural aspects thereof — as long as it has access to the right data.

It is very good at understanding research goals, identifying appropriate datasets to work with, doing data wrangling and data science, determining appropriate metrics, running simulations, analyzing results, writing up reports with charts and tables, and proposing follow ups or “high potential next steps” which show a surprising amount of judgment, and if instructed to, executing on those too.

I use Astra and Fable (who delegate to cheaper models for coding) and have them review each others work, and only spot check things when they are done calling each other out. I rarely find issues, and when I do, it’s usually a misunderstanding of intentions.

I’ve only explored a few areas though, including computer vision, vector search, and wireless signal processing in applications, as far as I can tell, are genuinely novel (though of varying levels of novelty and economic value, as the models are fond of pointing out to me!)

I prefer being in the loop and setting the direction — because I do not have unlimited tokens, hah! —- and I regularly identify new directions they had not considered or had prematurely dismissed… so I am a bit more useful than just a financier of tokens! But I can totally see how they could grinding ahead by themselves and brute force useful results, especially if given very concrete goals like a theorem to prove.

I don’t think we’re at Recursive Self Improvement yet, but I don’t doubt for a minute that this is behind the insane pace of development at Anthropic and OpenAI.

Re:Claude Fable 5.2 solves two open problems in the physics ... by glitchc on Thursday November 19, 2026 @02:50PM (#50565310)
In empirical research, getting the right data represents the bulk of the intellectual effort. Designing the right experiment and collecting good data is painstaking, backbreaking work.
Re:Claude Fable 5.2 solves two open problems in the physics ... by keeda on Thursday November 19, 2026 @03:22PM (#50565814)
I agree, and this may be nitpicking, but would it not be more accurate to say “bulk of the effort” rather than “intellectual effort”?

My perception is that most data collection, once designed (which does take considerable thought, as you point out), is a slog of executing rather mechanical processes over a long time. (Which is what RA’s and lab assistants are for!) For sure, regular analysis is needed to ensure the right data is being collected and no confounding factors are sneaking in, but seems to me the majority of time goes into the relatively rote work of setting up equipment, running experiments and recording the results.

I do think that with access to MCP-enabled equipment, AI would be ideal for this too, given its infinite patience. The problem, as with all other fields, is this would rob novices the opportunity of building the necessary intuition that comes from spending time in the weeds.

Re:Claude Fable 5.2 solves two open problems in the physics ... by glitchc on Thursday November 19, 2026 @05:22PM (#50567307)
> bulk of the effort” rather than “intellectual effort”?

No, it is intellectual. Designing and implementing an experiment that eliminates confounding variables requires a great deal of creativity and ingenuity. Reality is messy and complicated.

Some notable examples include: Michelson-Morley, Millikan, Wrights, Cas-9 for CRISPR, CERN (one giant experiment).

Re:Claude Fable 5.2 solves two open problems in the physics ... by keeda on Thursday November 19, 2026 @11:16PM (#50570092)
I guess in my mind I'm separating out the cycle into experiment design -> experimental setup + data collection -> data analysis phases and feeling like the middle phase appears to be the longest, most painstaking and most mechanical. But as I mentioned in a sibling comment, my perceptions are based on limited second-hand experience, so yeah I can believe they're off-base.
Re:Claude Fable 5.2 solves two open problems in the physics ... by thaumasiotes on Thursday November 19, 2026 @05:56PM (#50567734)
The majority of time goes into collecting nonsense data in the service of doing computations that aren't useful, because the data was nonsense.

https://desystemize.substack.com/p/desystemize-1

Re:Claude Fable 5.2 solves two open problems in the physics ... by keeda on Friday November 20, 2026 @01:04AM (#50570753)
Fascinating read, thanks for sharing!
Re:Claude Fable 5.2 solves two open problems in the physics ... by epihelix on Thursday November 19, 2026 @09:58PM (#50569611)
> but seems to me the majority of time goes into the relatively rote work of setting up equipment, running experiments and recording the results.

Sadly, no. There are surely exceptions, but most data collection for experimental research is highly skilled and intellectually demanding.

The good RAs are the ones that could have aced a PhD if they'd wanted to (there's one in my lab, and they're worth their weight in gold) -- but they are rare. There's a reason that the majority of research effort in the sciences comes from PhD students and postdocs, not from RAs. This stuff is not simple, and definitely not rote.

Re:Claude Fable 5.2 solves two open problems in the physics ... by keeda on Thursday November 19, 2026 @10:49PM (#50569909)
Fair enough. My earliest exposure to scientific experimental work was working (on an unrelated engineering project) alongside some scientists and PhD students in an NMR lab. I got the impression that they spent the bulk of their time tinkering with the mechanical aspects of the NMR machines to set up experiments, then waiting hours while the experiments ran, and then staring at the results to figure out what to do next.

Then I had friends studying biotech who seemed to spend hours in labs slicing up flies or preparing cultures for experimental assays.

They were all definitely very smart and hard-working, and the science was extremely deep, but setting up and carrying out the experiments seemed to be the most time-consuming aspect, which may have skewed my perceptions.

Re:Claude Fable 5.2 solves two open problems in the physics ... by epihelix on Friday November 20, 2026 @09:41AM (#50575291)
Experiments are definitely repetitive and time consuming, sure. But they are also highly skilled, and need intelligent operators. Both things can be true.

To take one of your examples -- dissecting Drosophila is an expert task and requires extremely fine motor skills (you are pulling apart soft tissues 100-200 microns in diameter), and also requires the ability to recognise and isolate different tissues visually. You're not going to ever get a robot to do this any time soon.

And that's just a tiny part of the experiment -- before that, there was thinking through the crosses; the fly husbandry; the sorting ... and after dissection, there's the fixation and antibody labelling; mounting and imaging (an art in itself).

We don't really give these skills enough credit, I agree. But the only reason why a high-performing biomed lab can do what it does, is because it's filled with extremely smart and capable individuals.

(Yes, not everything is as highly skilled as tissue dissection, and some repetitive molecular biology can now be handled with robotics and microfluidics quite well. But there's a lot that still can't be automated. I also wouldn't dream of letting an LLM near any of this -- a hallucination that causes a bug in code is one thing; a hallucination in a biomed lab facility could kill people.)

Re:Claude Fable 5.2 solves two open problems in the physics ... by tomrod on Thursday November 19, 2026 @03:09PM (#50565614)
I'd argue that "the right data" should really be called "the right information in context."

Context is king.

Re:Claude Fable 5.2 solves two open problems in the physics ... by vvilliamperez on Thursday November 19, 2026 @02:04PM (#50564647)