OpenAI open problem #2 cracked by human theorists; paper unifies and improves MRRW bounds

abeirami · x · 2026-09-13

Problem 2 from OpenAI's August 1 open problems list has been solved by human coding theorists: Omar Alrabiah and Venkatesan Guruswami's arXiv paper derives the Plotkin, Elias–Bassalygo and both MRRW bounds from a single "pretty good criterion" via classical–quantum channels, and introduces mixed-qubit channels (MQC/2MQC) that strictly improve both MRRW bounds for all relative distances. A similar result was reportedly also reached using OpenAI models.

Original post →

More from Models

Models channel →