Liam Fedus, formerly of OpenAI, offered his perspective in response to the insights shared by Terence Tao and 25 other Fields Medalists regarding the dynamic interplay between AI and mathematics. Both parties concur that the pursuit of mathematics transcends mere answer-seeking. Nevertheless, mathematicians express apprehension that AI's current emphasis on problem-solving efficiency could potentially erode the foundation and transmission of mathematical knowledge. Liam, on the other hand, maintains an optimistic outlook, envisioning a future where AI not only contributes to knowledge creation but also serves as a collaborative partner in mathematical exploration. He draws a parallel with AlphaGo, suggesting that mathematics may undergo a transformation akin to that experienced by the game of Go.

In their collective statement, Terence Tao and his fellow mathematicians highlighted a disparity between the present objectives of AI and those upheld by the mathematical community. They underscored that solving challenging problems constitutes only a facet of mathematical inquiry, with the imperative to comprehend and perpetuate the findings. AI accomplishments frequently give rise to questions concerning attribution, and the nurturing of students cannot be supplanted by mere answers. Nonetheless, they also recognized AI's untapped potential. In recent times, AI has achieved remarkable strides within the mathematical domain, transitioning from competitive mathematics to tackling open mathematical conundrums, such as disproving the Erdős unit distance conjecture and discovering proofs for singularities in three-dimensional Navier–Stokes equations. Yet, these feats have ignited debates surrounding attribution and authorship.

The American Mathematical Society, the London Mathematical Society, and the Clay Mathematics Institute have all released statements underscoring that the essence of mathematics lies in human comprehension, and AI serves as a tool whose outcomes must undergo rigorous peer review. The true value resides more in the novel ideas and methodologies generated throughout the process. Presently, the human mathematical community is tasked with exploring avenues to ensure that the newfound knowledge generated by AI is comprehended, validated, and seamlessly integrated into the existing mathematical framework.