Rhombic tiling, permutation plot, subset gamegraph, permutahedron, hyperplane arrangement, structure diagram, three-dimensional heap, Cayley diagram, braid graph, and L-tromino tiling.

AI and mathematics

A weekly digest of AI-related developments in mathematical research, teaching, and learning.

Weekly digest, 8–14 November 2026

Research and formal verification

AI’s ability to ‘autoformalize’ proofs is changing math

Konstantin Kakaes · Quanta Magazine, Transformation · 11 November 2026

Quanta reports advances in translating mathematical arguments into computer-checkable proofs, including efforts to formalize large bodies of existing mathematics. The dispatch contrasts earlier flawed attempts with recent projects that have received favorable assessments from independent experts. It explains why this matters for checking lengthy arguments and reports both company announcements and expert responses. The article’s prediction that AI will eventually formalize all mathematical knowledge remains a forecast, not an established result.

Research and interpretation

Aaronson-Ambainis conjecture proved! (By AI, of course)

Scott Aaronson · Shtetl-Optimized · 8 November 2026

Aaronson discusses an announced AI-assisted proof of the conjecture he formulated with Andris Ambainis and explains its implications for quantum computation. He emphasizes that the result rules out a particular class of quantum speedups for decision problems, rather than eliminating quantum advantage generally. His account also reflects on the experience of seeing a longstanding problem resolved with AI. This is a specialist’s explanation of the announced result and its scope, not an independent verification of the proof.

Teaching and undergraduate formation

What should we teach our (undergraduate) students?

Jana Gevertz and Steffen Marcus, guest post on Terence Tao’s blog · 8 November 2026

Gevertz and Marcus argue for curricula that teach students to evaluate AI outputs while preserving the productive struggle needed to develop mathematical understanding. They propose discussion of conceptual reasoning, procedural fluency, formalization, modeling, and communication, with coordination across disciplines that depend on mathematics. They also identify the substantial faculty time needed to redesign courses. These are questions and proposals for curricular development; the authors acknowledge that they have not yet found the right approach in their own classes.

Student formation and professional life

Our survey of early career mathematicians

Anna Brandenberger, Hanna Mularczyk, and Dora Woodruff, guest post on Terence Tao’s blog · 9 November 2026

The authors invite attention to their survey of more than 260 mathematicians and urge senior colleagues to include early-career perspectives in decisions about AI, hiring, and research. They highlight concerns about career outlook and motivation. The survey was intended as an informal snapshot, not a representative statistical study, and nearly all responses preceded 6 October. This entry covers the November discussion of results posted in October, rather than presenting the responses as a measure of current opinion.

Research review and community infrastructure

Introducing MathHelm: A place to discuss AI-generated papers

MathHelm scientific committee, guest post on Terence Tao’s blog · 13 November 2026

The committee introduces a nonprofit forum for discussing AI-generated mathematical manuscripts. Researchers can contribute observations about correctness, readability, originality, ideas, and references without undertaking a complete review. The proposal gives partial checks and explanatory comments a shared home, potentially helping readers navigate a growing volume of generated work. This is a platform announcement; its effectiveness as a review resource remains to be established.

Editorial note. This issue includes five items published from 8 through 14 November 2026. The early-career survey entry discusses an older survey through a new November post. Quanta, including its Transformation dispatch, and Numberphile were reviewed as standing sources; the 3Blue1Brown channel check was limited by access, with search results showing no new eligible upload. No video or advocacy addition was selected this week. The content is generated by ChatGPT from linked sources and may contain errors or omissions; it is a starting point for review and does not independently verify reported mathematical claims.
Coverage and access. Only freely accessible material is summarized here. Some entries cover a freely available abstract rather than a full article; these are explicitly labeled, and the full article may require a subscription. This collection is selective, not exhaustive. Paywalls and other access restrictions can limit coverage, and sources that cannot be adequately verified are omitted pending review.