Tao Zhixuan, a laureate of the Fields Medal, and Deng Yu engaged in a profound discussion, expressing their perspectives on the ongoing debate concerning the widespread dissemination of mathematical proofs generated by commercial AI. They emphasized that mathematical proofs produced by AI do not inherently guarantee that humans have truly grasped the essence of the problem at hand. This is because the act of solving a problem and the comprehension of its proof are two fundamentally different notions. Instead of fixating on the sheer volume of problems resolved by AI, greater emphasis should be placed on the clarity of the proofs, the verifiability of the logical steps involved, and whether mathematicians can build upon these proofs for further exploration.

They contend that while AI is adept at managing monotonous derivations and can serve as a source of inspiration for mathematicians, the importance of slow-paced scientific research should not be underestimated. Such research plays a pivotal role in deepening understanding, fostering academic communities, and ensuring the sustained mastery of the discipline by humans. Simultaneously, it is imperative to trace the intellectual lineage of AI-generated content, elucidate the attribution of contributions, and preserve the academic heritage chain.

To this end, both scholars actively participated in the inception of the Open Math Model initiative and spearheaded the development of the Open Math OS work platform. This platform seamlessly integrates existing open models, research tools, and collaborative functionalities, with the aim of involving the mathematical community in the development of tools that enable AI to genuinely support mathematical research. They underscored the necessity of continuously nurturing core human research capacities, such as formulating valuable questions, devising proof strategies, and cultivating innovative ideas. They firmly believe that high-quality AI tools should augment research capabilities rather than supplant human endeavors.