The Riemann Hypothesis and the Industrialization of Mathematics

It’s time for an emergency update. Levent Alpöge, working with Anthropic’s Claude, has announced a proof of the Riemann hypothesis—arguably the most important problem in mathematics (although I would argue that resolving P vs. NP has far more important implications). He also claims to have proved the generalized Riemann hypothesis. The operative word is working with, because this detail matters enormously for the future of mathematics.

The default response, as we’re already seeing online, is skepticism, despite Alpöge’s claim that the proof has been formalized in Lean. Independent verification is still pending, but I strongly suspect that, for all practical purposes, it’s done. Alpöge is an expert. He wouldn’t casually release a half-baked proof of the Riemann hypothesis. Presumably, he has been working behind the scenes for some time, checking everything, and waited until he was confident it would hold up before going public. He also had access to Anthropic’s enormous computing resources. This isn’t some random crank claiming to have solved a famous problem. If the Lean formalization checks out, with no unproved assumptions and the correct theorem statement, there is little room left for the usual skepticism. My expectation is that it will.

The rise of AI-assisted mathematics has produced a range of emotions from elation to anomie or even existential dread, along with endless think pieces and hand-wringing about the future of the field. With this discovery, assuming it holds up, the debate is effectively over in one sense: mathematics has been forever changed. At this point, it’s reasonable to assume that, given enough computing power, resources, and expert guidance, AI will eventually solve most, if not all, of the remaining important open problems. Mathematicians now have to decide whether it’s worth devoting years of their lives to a problem when they risk being scooped by AI. Or whether they can ever hope to make a meaningful contribution to mathematics when overshadowed by enormous AI-assisted discoveries by people and firms with vastly more computing power.

The other problem is that almost any individual contribution looks quaint by comparison. Compared to proving the Riemann hypothesis, what is left? Some say mathematics is more than solving problems or proving or disproving conjectures, but those are also what get all the attention. So how are we supposed to reconcile these two competing narratives—that mathematics is about some higher pursuit of understanding or truth, when solving hard problems is really what matters as far as careers or recognition are concerned?

I discussed this dilemma with ChatGPT weeks ago, below:


Discussion with ChatGPT about AI and mathematical research

(I’m glad I was not working on the Riemann Hypothesis—I spent maybe an hour total on it. Then I would be pissed.)

You don’t get credit for an unsuccessful approach, nor can you recover the lost time. It’s all or nothing, unlike, say, patents, which can still be sold after a business fails. An approach that almost proves the Riemann hypothesis, but ultimately fails, gets little or no credit unless it produces some independently publishable result. (Note: This is distinct from the proof that more than two-thirds of the nontrivial zeros lie on the critical line, which established a complete mathematical result toward the original conjecture.) The distinction matters: a partial result can be valuable, but an approach that almost works and ultimately leads nowhere has no value.

Now back to the beginning, the operative word being “computation.” It’s not as if any individual could have done this. He credits Claude, but he makes it sound as if he used an ordinary AI subscription, like anyone else. He glosses over the enormous computing resources presumably involved (17.4 million lines of Lean across 175,187 files, plus 250,000 pages of notes). How much computation was involved? How much of the work was done in parallel? 250,000 pages cannot fit into the context window, and would have to be divided into thousands of chunks, with some mechanism for managing dependencies and retrieving relevant earlier results. What specialized infrastructure and tools were available? These are hardly minor details. In his usual fashion, he glosses over this. The computation is not a small detail relegated to a footnote; it’s central to understanding how the proof was possible at all.

To use myself as an example, and to give an idea of what is involved for a much smaller (albeit still big) problem, my recent result took nine days of computation for the second part and another week for an earlier part. The Lean formalization will take a month, minimum. The bottleneck isn’t simply how long it takes Lean to check a finished proof. Instead, it’s how long it takes to produce a finished proof that Lean accepts, including all the failed attempts, revisions, and compilation cycles.

For his vastly larger project , to accomplish all this this in 6 weeks, from producing a draft to a correct Lean proof, is indicative of extensive parallel computing and resources, well beyond a single computer and subscription. For my paper, I was using a single subscription each to Claude and ChatGPT on the same laptop ($200/month between them), for a total of $1,300 spent on subscriptions and extra “boosts” to use the most powerful models available. This pushed the limits of what a retail or consumer user could reasonably hope to achieve with these models at home. I’m guessing the amount of computation involved in his proof was at least 1,000 times what I used, if not considerably more. So yes, Claude helped, as he says, but the comparison is meaningless to how people actually use it.