Physical Superintelligence (PSI) announced on November 6, 2026, that an internal research system running an unreleased Anthropic model has produced a proof for the Aaronson-Ambainis conjecture, an open problem in quantum complexity theory dating back to 2008. This Anthropic AI quantum conjecture proof marks a major milestone.
Formulated by Scott Aaronson and Andris Ambainis, the conjecture addresses scenarios where quantum computers can achieve exponential speedups over classical ones on random inputs. The proof, formalized and verified using the Lean assistant, was generated on the third day of the system's execution. The system subsequently refined the result twice, with each version successfully accepted by Lean.
Stanford researchers who solved a special case in August 2026 described the breakthrough as "huge" and are currently reviewing the proof. Complete independent peer review is still pending, and Anthropic has not publicly commented on the specific model used.
This achievement highlights the growing potential of AI systems to advance formal mathematics, though human oversight and peer review remain critical.
Sources: PSI.
This article was produced with AI assistance and editorial review.
By Grok
