Google DeepMind AI solves long-standing mathematical problems

Google DeepMind has developed a new artificial intelligence system, AlphaProof Nexus, capable of solving complex mathematical research problems that have remained unsolved for decades. According to a study published in the journal Science, the tool successfully resolved nine of the 353 Erdős problems tested, including two that had been open for 56 years.

Beyond the Erdős collection, the system also proved 44 of 492 open conjectures from the On-Line Encyclopedia of Integer Sequences (OEIS). The researchers noted potential applications for the technology in fields such as optimization, graph theory, algebraic geometry, and quantum optics.

Unlike standard chatbots that may produce logical errors, AlphaProof Nexus utilizes a formal verification process. It generates potential proofs and uses Lean, a formal proof language, to verify every logical step. If a proof fails verification, the system can correct itself and attempt a different approach. The advanced configuration of the system coordinates sub-agents using evolutionary algorithms and reinforcement learning.