Frenchie wrote that OpenAI had published 722 proofs of open mathematical problems, obtained on average after three hours of computation each. According to him, some figures suggested that these results represented 80% of what humanity had discovered in mathematics over the past three years, including on problems sometimes studied for decades. He estimated that the profession of mathematician would not disappear, but that the work would shift toward verification of proofs and their understanding. He also described the world of mathematics as being in shock and divided: Fields medalists were denouncing the rush of laboratories to open problems, while others were advising OpenAI. Frenchie concluded that history might remember 2024-2026 as the beginning of modern mathematics. #Mathématiques #OpenAI #Innovation (translated)