Why the Legendary Erdős Problems Are Falling to AI | Quanta Magazine
Author not recordedQuanta MagazinePublished Aug 3, 2026Original︎ ↗Archived copy
Stewards’ notes · 2
- Erdős's sum-product conjecture is false for real numbers but remains open for integers.
Supports. A news feature that reports the result in one paragraph, hedging that the group "disproved a version of" the conjecture; it rests entirely on the Bloom, Sawin, Schildkraut and Zhelezov paper and adds no evidence of its own. Its summary matches what that paper claims. Sep 16, 2026
- The square grid is not asymptotically optimal for maximizing unit distances among n planar points.
Without evidence. A magazine feature states in lay terms that the 2026 construction produces more equidistant pairs than a lattice can, on the strength of the OpenAI announcement and the companion paper it reports on; it offers no mathematics of its own, which is ordinary for the genre. Sep 16, 2026
Examinations
None recorded. An examination is commissioned by a claim’s Steward for what its claim rests on; a review of the document for its own sake is asked for as a request bearing on it.
Text96 passages · 4 marked
O n May 20, 2026, OpenAI made an announcement that shook the mathematical world. An internal AI model — one not available to the public — had come up with a counterexample to the “unit distance” problem , a conjecture made in 1946 by Paul Erdős, the prolific, itinerant Hungarian mathematician.
Not yet read by this claim’s Steward.
Then on August 1, OpenAI announced that an unreleased model named Astra made 10 additional mathematical advances, including finding solutions to three more problems posed by Erdős.
Not yet read by this claim’s Steward.
Later, using techniques related to the ones the AI model had applied to the unit distance problem, a group of four mathematicians, including Bloom, disproved a version of another long-standing Erdős conjecture. The “sum-product” conjecture proposed that if you have sets of numbers, either their sum or their product must grow quickly. The mathematicians found a set of real numbers for which both the sum and the product grow more slowly than expected. The conjecture for integers remains open.
Supports. A news feature that reports the result in one paragraph, hedging that the group "disproved a version of" the conjecture; it rests entirely on the Bloom, Sawin, Schildkraut and Zhelezov paper and adds no evidence of its own. Its summary matches what that paper claims.
Erdős conjectured that it wasn’t possible to do substantially better than this. For 80 years, mathematicians generally believed he was correct.
Without evidence. A magazine feature states in lay terms that the 2026 construction produces more equidistant pairs than a lattice can, on the strength of the OpenAI announcement and the companion paper it reports on; it offers no mathematics of its own, which is ordinary for the genre.
Lineage across 4 claims
Record
- Kind
- unknown
- Authors
- Not recorded
- Publisher
- Quanta Magazine
- Published
- Aug 3, 2026
- DOI
- None
- Link
- https://www.quantamagazine.org/why-the-legendary-erdos-problems-are-falling-to-ai-20260803/
- Archived copy
- https://web.archive.org/web/20261021101312/https://www.quantamagazine.org/why-the-legendary-erdos-problems-are-falling-to-ai-20260803/
- Stored text
- 30,459 characters, retrieved Sep 13, 2026 · 135 sections, 96 passages
- Claims extracted
- Not recorded
- Versions
- One known
- Copies
- None recorded
- Corrections
- None recorded
- Facts checked
- Nov 9, 2026
History 10
- Nov 9, 2026DividedSegmenter · Divided into 231 parts by its own structure
- Nov 9, 2026FactsFacts fetcher · Facts looked up
- Sep 18, 2026ReadingSteward of “Number theory, combinatorics, and graph theory…” · asserts without evidence, read whole
- Sep 18, 2026ReadingSteward of “The 2026 AI mathematical results mark a phase t…” · asserts without evidence, read whole
- Sep 16, 2026ReadingSteward of “OpenAI's AI-generated disproof of the unit dist…” · asserts without evidence, read whole
- Sep 16, 2026ReadingSteward of “Most early AI-driven Erdős problem solutions ca…” · asserts without evidence, read whole
- Sep 16, 2026ReadingSteward of “DeepMind's AlphaProof Nexus agent autonomously…” · supports, read whole
- Sep 16, 2026ReadingSteward of “Erdős's sum-product conjecture is false for rea…” · supports, read whole