VibeMathedMath problems solved by AI

The Planar Steklov Analogue of Kac's Question

Can one hear the shape of a drum, in the Steklov setting and in the plane? No: there exist pairs of noncongruent bounded plane domains with identical Steklov spectra including multiplicities, simply connected, strictly convex, with real-analytic boundaries, and arbitrarily close to a disk in the C-infinity topology.

Result
Disproved(see note)
Status
Resolved
AI contribution
AI co-developed
Method
Construction
Field
Spectral geometry
Posed by
The Steklov analogue of Kac's question, raised in the Girouard-Polterovich problem literature
Year posed
2017
Years open
9y
Solved
2026-08-11
Model
ChatGPT
Vendor
OpenAI
Collaborators
Tao Hu, Jiachen Shi, Quanyu Tang
Verification
Unreviewed
Publication
Preprint
Significance
25 / 100
Disclosed cost
Wikipedia
No dedicated article

What was actually shown

Strict convexity and real-analytic boundaries are what make this sharp: the classical Gordon-Webb-Wolpert drums are non-convex polygons, so the obvious escape routes are closed off.

What the AI did

The division of labour is set out in the AI statement: the idea of adapting the Sunada construction of Gordon, Webb and Wolpert to the planar Steklov problem "was proposed by the authors". From there, "ChatGPT provided substantial assistance in developing the concrete counterexample construction, including the passage from the orbifold construction to weighted Steklov problems on the disk and their subsequent realization by Euclidean plane domains. It also assisted with several technical arguments." The authors checked, revised and rewrote the arguments.

Verification

A preprint days old, with no independent review.

Source

Changelog1 change
  • Rasmus Lindahladded this entry

Discussion