Friday, November 20, 2026
Home / Science / An Anthropic mathematician claims to have proven t...
Science

An Anthropic mathematician claims to have proven the Riemann hypothesis. We have to rethink how we reward expertise.

CN
CitrixNews Staff
·
An Anthropic mathematician claims to have proven the Riemann hypothesis. We have to rethink how we reward expertise.

An Anthropic mathematician using Claude claims to have proven the most famous problem in math — the Riemann hypothesis — in the latest, but potentially most consequential, example of artificial intelligence (AI) producing stunning math results.

AI has steamrolled its way into the math community. Anthropic's proof will certainly and rightly be scrutinized, but rather than picking apart every single result these models produce, we should respond by changing how we evaluate and reward expertise — placing less weight on the volume of results and more on the human judgment required to decide which questions and answers matter.

Without this shift, we risk drowning genuine insight in a flood of plausible but poorly understood mathematics.

A gift from Live Science Pro

This is a Live Science Pro article that we're putting out for free. If you liked this story, please consider subscribing.

Mathematical judgment is the ability to distinguish a result that is correct, but boring, from one that might fundamentally change how we think about the problem or even the whole area of mathematics surrounding the problem.

AI could generate hundreds of thousands of complex variations on a theorem, but these variations might add very little to mathematical knowledge, each one ultimately proving a dead end. A truly consequential result might settle a long-standing unsolved problem or connect to apparently unrelated areas. It is mathematical judgment that allows us to tell the difference between the two.

Earlier this year, an OpenAI system produced a counterexample to the long-standing planar unit distance problem, overturning a conjecture of Paul Erdős that had resisted mathematicians for decades. More recently, OpenAI announced an apparent solution to the Navier-Stokes problem, a celebrated question about whether equations describing fluid flow can develop singularities, or infinite values. It is one of the Clay Mathematics Institute's Millennium Prize Problems — a set of seven well-known and complex math problems, each with a $1 million prize attached.

Whether OpenAI's result ultimately stands up to scrutiny remains to be seen. But what is clear is that AI systems are now operating at a level that many mathematicians did not expect to witness for years, if not decades. OpenAI has since released hundreds of further mathematical results, leaving mathematicians scrambling to assess their correctness and significance.

This does not mean that mathematicians are about to become obsolete. But it does mean we need to carefully reconsider what their precise purpose is.

Traditionally, expertise has been closely associated with production: Experts solve difficult problems, prove theorems, construct models and write papers. If AI can generate proofs, conjectures and probe research areas at unprecedented speed and scale, mathematical production is no longer necessarily the bottleneck. Instead, many pure mathematicians would argue that it is mathematical judgment that will become the scarcer resource.

Human judgment remains necessary because proofs generated by AI are not necessarily illuminating, comprehensible or even reliable. Some proofs may be so long or opaque that checking them requires substantial work by specialists and even more work to reinterpret them in a way that is meaningful to human mathematicians.

A proof can be valid without offering insight into why it is true or how its ideas might be employed elsewhere. The job of human mathematicians will be to become arbiters of the correctness of proofs, but also to scrutinize the reasoning and to extract the underlying ideas, determining whether the result says anything that meaningfully advances the field.

Significance vs. output

Mathematics, as a research community, needs to react rapidly to this change. Mathematicians and journals must prioritize significance over output. Mathematics involves deciding which results are important as well as generating them. A proof can be correct without being insightful. To give a simplified example, AI might "prove" that the square of a number divisible by two is divisible by four and then prove that the square of a number divisible by three is divisible by nine, and so on with ever larger integers. However, each new "proof" isn't an example of insightful mathematics, just a variation on the same elementary idea.

AI can create arguments faster than the community can verify, interpret or absorb them. Existing safeguards were not designed for that imbalance.

In contrast, a significant proof might settle a long-standing question, reveal an unexpected connection between disparate fields or develop a technique which is useful in the solution of many other problems.

Journals should therefore require authors to explain clearly what is genuinely new, why it matters — and what role AI played. Reviewers should judge whether a result deepens understanding, rather than merely verifying that the argument works. That does not mean rejecting work whose eventual importance is uncertain. The history of mathematics is full of ideas whose significance became apparent only decades later, so journals must continue to make room for foundational, exploratory and curiosity-driven research.

But there is a difference between a result with as-yet unknown implications and one that is unlikely ever to be insightful. Mathematical judgement cannot guarantee that nothing valuable is ever overlooked. It can, however, distinguish work that contains genuinely new ideas or possibilities from the automated production of uninteresting dead ends.

Universities and funders should also stop treating publication counts as a convenient proxy for quality. Careers are built around papers, citations, grants and prizes, but these measures were developed when producing sophisticated mathematics demanded substantial human time and effort. In the age of AI, work that might previously have taken years of careful thought could now be produced in weeks or even days.

Consequently, as the cost of generating mathematical results falls dramatically, those measures of quality become less useful. A decreased reliance on traditional metrics may be no bad thing, but institutions need credible alternatives: assessments of a researcher's most important contributions, evidence of intellectual leadership and explanations of how their work has changed understanding. This is urgent because mathematicians are already struggling to evaluate the sheer volume of material being produced.

Finally, mathematical training must continue to reward understanding. Written work alone will become a less reliable guide to individual capability when papers, reports and applications can be written mostly using AI. Seminars, discussions and collaborations can reveal whether someone understands an argument, can defend it and can connect it to other ideas. Demonstrating understanding has always been a more important quality in university students than simply presenting the answer; education and recruitment should now reflect that principle more explicitly.

Exclusive to Live Science Pro

Critics might argue that this is an overreaction. Mathematical claims still require proof, peer review already evaluates significance, and AI is simply another tool, like a calculator or computer algebra system. There is a degree of truth to each objection. Yet the explosion in the number of published papers changes the nature of the game. A calculator does not generate hundreds of research claims that specialists must spend precious time checking. AI can create arguments faster than the community can verify, interpret or absorb them. Existing safeguards were not designed for that imbalance.

Recent advances have understandably generated both excitement and anxiety. The right response is neither to reject AI nor to unguardedly celebrate every finding it produces. Journals, universities and funders should adapt their incentives now, before quantity overwhelms judgment completely.

In an environment increasingly capable of automatically generating research-level mathematics, the most valuable experts will be those who know which questions are worth asking, which results deserve trust and which ideas are worth our attention.

Opinion on Live Science gives you insight on the most important issues in science that affect you and the world around you today, written by experts and leading scientists in their field.

Help us improve Live Science Pro: We're always trying to make our content better. Leave us feedback about Pro here.

Originally reported by Live Science. Read the full story at the original source.