AI Solves Quantum Computing Conjecture

Physical Superintelligence Inc. (PSI) has made a significant breakthrough in quantum computing. The company announced it used an unreleased AI model from Anthropic to prove the Aaronson-Ambainis conjecture, a problem unresolved since 2008. This conjecture addresses the scenarios where quantum computers can surpass the speed of traditional computers.

The PSI team integrated the Anthropic model into their research system, achieving the proof in just three days. They described this success as resolving one of quantum computing’s core questions. The lab acknowledged that while quantum computing still has mysteries, one less remains now.

Quantum computing still has plenty of mysteries, but now there is one fewer,

In recent trends, AI’s role in solving mathematical challenges is growing. Last month, OpenAI tackled the Navier-Stokes equations, a profound mathematical problem. Such achievements are prompting reflection within the field. Terence Tao of UCLA noted that mathematics is transitioning from a ‘romantic era’ to a new phase filled with potential yet marked by loss.

Students in mathematics are reevaluating their paths. MIT researchers found that 76% of surveyed graduate students and postdocs are less inclined to pursue academic math careers due to AI’s advancements. They also reported a decrease in motivation for research.

David Turturean from PSI was instrumental in developing the research system leveraging the Anthropic model. The specifics of the AI model used remain undisclosed, and Anthropic has not commented on the matter. The proof by PSI suggests that quantum computing’s exponential advantages necessitate specific patterns, confirming previously held suspicions about their limitations. Despite these boundaries, quantum computers still represent a significant technological leap for certain applications.

The proof has been hailed as a substantial achievement by researchers who previously tackled the problem. A collective statement from Stanford’s Guy Blanc, Jordan Docter, Carmen Strassle, and Li-Yang Tan praised the integrated solution.