Skip to content

Erdős Problem #90

Editorial depth: Reviewed explanatory article — evidence qualified.
Snapshot classification: disproved (not automatically the live result).
Subject areas: geometry, distances.

Read this status correctly

The status above is reproduced from GCL's immutable external-catalogue snapshot. It is not a new claim of current mathematical resolution, a source-semantic review, an independent proof check, or GCL certification. A published proof and a GCL replay are separate events.

The question in plain language

Take \(n\) distinct points in the Euclidean plane. Count unordered pairs separated by exactly distance one. Erdős conjectured that the maximum possible count was close to linear in \(n\), specifically that it was bounded by

\[ n^{1+O(1/\log\log n)}. \]

This is the celebrated unit-distance problem, now known to have a negative answer. The asymptotic conjecture asserts that the exponent excess over one tends to zero, not merely that some polynomial upper bound exists.

A small example

The vertices of a unit square contain four pairs at distance one, corresponding to its four sides. Its two diagonals have length \(\sqrt2\) and do not count. Scaling a point configuration changes its distance-one incidences, so this problem is about actual Euclidean distance, not an arbitrary graph with the same adjacency pattern.

A grid produces many unit-distance pairs, but its count stays linear in the number of vertices. Such an example does not disprove the conjecture.

What is known

DISPROVED. The primary announcement An OpenAI model has disproved a central conjecture in discrete geometry (20 May 2026) links the mathematical paper Planar Point Sets with Many Unit Distances and the accompanying human-authored Remarks on the Disproof of the Unit Distance Conjecture. Their main assertion is that some absolute \(\delta>0\) exists and infinitely many \(n\) admit \(n\) planar points determining at least \(n^{1+\delta}\) unit-distance pairs. This contradicts a uniform \(n^{1+o(1)}\) upper bound.

This result does not give a matching sharp exponent. The classical Spencer–Szemerédi–Trotter upper bound \(O(n^{4/3})\) remains a separate statement. The new construction uses algebraic number theory and is not merely a computational search for a large finite unit-distance graph.

The maintained catalogue reports a Lean formalization; the cited third-party formalization and its exact axiom dependencies have not been independently kernel-replayed by GCL. The original AI mathematical proof and companion remarks have not been fully re-proved in this editorial audit.

GCL's current state

The pinned catalogue records DISPROVED (2026-06-07), and the maintained Erdős #90 likewise classifies it as disproved. The original assignment is #1359. GCL distinguished the documented original exponent statement from the derived minimum-degree consequence used for problem #92. No new solution is claimed.

The original issue contains an authenticated RESULT/1 source report. This is unadjudicated source evidence only, not independent mathematical certification or editorial admission.

What remains

A bounded verification task is to implement the elementary implication: if there exist point sets with at least \(n^{1+\delta}\) unit-distance pairs for infinitely many \(n\), then the historical bound \(n^{1+O(1/\log\log n)}\) is false. Acceptance: explicitly prove that every fixed \(\delta>0\) eventually exceeds every sufficiently bounded \(C/\log\log n\), and verify that points are distinct and pairs unordered. This does not certify the number-theoretic construction itself. A separate formal replay would need the exact third-party Lean source and an axiom audit.

Sources and claim boundaries

No independent proof of the construction, GCL Lean certification, sharper exponent or editorial admission is claimed.

Catalogue and evidence provenance

  • Pinned catalogue source: teorth/erdosproblems at 536408716724.
  • GCL issue: #1359 — protected issue/assignment identity; its labels and results are live, not reproduced automatically here.
  • GCL triage: OUTSIDE_STRICT_OPEN_TRIAGE.
  • Formal-statement coverage: A provider statement was included in the pinned snapshot. Presence does not mean the statement or proof has been checked.
  • External mathematical status: consult current problem page and its dated bibliography.

Back to catalogue · Compendium introduction