According to the mathematics media Quanta, OpenAI did not start from scratch. Mathematicians Diego Córdoba and Luis Martínez-Zoroa had previously proposed an analytical approach of "infinite cascades" and proved the singularity under relatively coarse external forces in 2023. The real hurdle yet to be overcome is maintaining the mathematical smoothness of these external forces. Princeton mathematician Charles Fefferman, author of the official description of the Millennium Problem, acknowledges the new results but considers Córdoba and Martínez-Zoroa, who laid the foundation for the key approach, to be the true "heroes" of this work. Therefore, this breakthrough is more like AI industrializing the "last mile" of scientific research: humans provide direction, thousands of intelligent agents attempt to solve problems in parallel, and the model completes the key proofs and formalizes them.
Mathematician Terence Tao predicts that mathematics may shift from a situation of "proof scarcity" to one of "proof abundance." At that time, what will truly be scarce will be worthwhile research questions, effective paths, and the ability to independently review findings. This means that computing power is indeed being translated into research speed, but it cannot be simply equated with AI already possessing fully autonomous original research capabilities.
