The mathematicians' revolt against AI: "Privatize knowledge"

Twenty-five of the world's most prestigious mathematicians, including Terence Tao, Alessio Figalli, and Cédric Villani, have sounded an alarm through an open letter titled "A Serious Misalignment of Artificial Intelligence in Mathematics." They denounce that the goals of AI labs and multinational corporations are profoundly divergent from those of the scientific community.

The mathematicians highlight how the extraordinary progress of language models in solving problems unresolved for decades is being exploited by companies as mere benchmark tests to justify billion-dollar investments and bolster stock market valuations. This transforms fundamental research into a massive corporate public relations operation.

The heart of the protest lies in the difference between providing an answer and understanding a phenomenon. Digital models produce impenetrable chains of automatic steps, churning out responses without offering any overarching vision or explanations. This mechanism sidelines human training and risks obliterating critical intelligence, turning science into dogma.

Mathematics has historically been a discipline where solving difficult problems was never an end in itself, but rather a means to develop new conceptual tools, create connections between different fields, and train new generations of researchers. This approach, based on deep understanding and critical thinking, is now threatened by the advent of artificial intelligence.

The conflict arises from the transformation of pure scientific research into a promotional tool for Big Tech. Major Silicon Valley companies have invested enormous capital in infrastructure and need demonstrations of strength to attract further investors. Thus, we are witnessing the privatization of fundamental knowledge, historically produced with public funds but now channeled and exploited in the servers of a few private giants.

The consequences could be severe: strategic fields such as engineering, energy, and finance risk becoming dependent on the decisions and algorithms of a foreign private oligopoly. Mathematicians are not asking to turn off computers, but they demand that AI remain a tool in service of human understanding and not a substitute for critical thinking.