VibeMathedMath problems solved with AI

The Erdos-Borwein Constant is 2-Dense

Is the Erdos-Borwein Constant 2-Dense?

Result
Proved(see note)
Status
Candidate (review pending)
AI contribution
AI-discovered
Method
Argument
Field
Analytic number theory; digit distribution of constants
Posed by
Richard E. Crandall
Year posed
2002
Years open
24y
Solved
2026-09-07
Model
ChatGPT-6 Astra
Vendor
OpenAI
Collaborators
Verification
Lean-checked, statement unaudited
Publication
Preprint
Significance
25 / 100
Disclosed cost
Wikipedia
No dedicated article

What was actually shown

This work proves that the Erdős–Borwein constant's binary expansion contains every finite binary sequence as a consecutive block, each occurring infinitely often. This implies that the Erdős–Borwein constant is 2-dense.

Verification

The Lean development proves disjunctivity of the binary expansion conditional on two hypotheses supplied as theorem arguments, not as axioms: AGP, the Alford-Granville-Pomerance estimate in the form of Vandehey's Proposition 2.1, and PrimeIntervalSupply, a standard prime number theorem consequence bounding the primes in (L, 2L) below by L/(3 log L). Both are published theorems rather than conjectures, so the mathematics is conditional only on known results; neither is formalized here. Read directly from lean/ErdosBorwein/PrimeInputs.lean on 8 September 2026. The audited endpoints depend on propext, Classical.choice and Quot.sound only. Because the statement is the author's own rather than anchored to a canonical tracker, and because what the kernel certifies is the implication, this takes the statement-unaudited tier. No specialist in analytic number theory has read the argument.

Sources

Submitted by nufrogcaca on

Changelog2 changes

Discussion