Toby Ord argues mathematics is more than proof

Toby Ord argues mathematics is more than proof

On August 23, 2026, Toby Ord argues that automating mathematical proof would not end mathematics. He distinguishes proof from the work of choosing fruitful questions and creating new fields that give mathematicians new concepts and ways to investigate problems. Ord says current AI systems are beginning to automate proof, but there is little evidence that they can ask powerful questions or find profound new fields of mathematical inquiry. He notes that good questions help direct attention toward important problems, while new fields can make previously unaskable questions possible. Ord expects proof automation may change mathematicians’ work, but he does not claim that human mathematicians will always retain an advantage. He says AI could also automate question-finding and field creation, perhaps swiftly. How to assess AI’s ability to ask valuable mathematical questions remains open: Ord says there are no good benchmarks for measuring that skill.

Publication: f278e473-769f-6f99-270c-70bba7d25b5e

attribution — pass: The piece is a single essay by Toby Ord on his own site, and the reading attributes its arguments to him throughout. The claims about AI's current abilities are presented as Ord's assessments, not as established facts.

August 23, 2026Toby Ord

summary — pass: The reading conveys the central thesis that proof is one part of mathematics, while question-asking and field creation remain. It keeps Ord's caveat that AI may come for these areas swiftly and identifies the missing benchmarks. Benchmark is defined beside the reading. It omits Ord's historical arc of calculation and symbolic manipulation being automated, and his view that a shift to setting direction would be a promotion rather than a demotion. These omissions reduce color but do not misrepresent the position.

the automation of proof alone would not be the end of mathematics.

date — pass: originalPublished is 2026-08-23T13:48:52+01:00, which is August 23, 2026 in UTC, and it matches the dateline. The modification timestamp equals the publication instant. The page is a publisher page, so no archive disclosure is needed. The reading frames nothing as happening after publication and reports AI's future prospects as Ord's conditional possibility.

August 23, 2026Toby Ord

support — pass: Each claim is matched by a verbatim excerpt. The summary's statements about questions directing attention and new fields enabling previously unaskable questions are grounded in the text. The question about benchmarks follows from Ord's stated gap and his doubt about the quality of mass-generated conjectures. No benefits or costs are invented.

I’m sure people can already make language models that spit out large numbers of mathematical questions and conjectures, but I’m doubtful about their quality.

media — pass: No media was selected, and no verified media candidates were supplied, so having no media is appropriate.