GPT-5.2 Milestone: AI Proves Novel Gluon Interaction

OpenAI’s GPT-5.2 has achieved a landmark in theoretical physics by identifying a novel formula for gluon interactions previously thought to be impossible ^[1][2]. The discovery, detailed in a research preprint titled "Single-minus gluon tree amplitudes are nonzero," challenges a long-standing textbook assumption that specific scattering amplitudes in quantum field theory must always equate to zero ^[2][3]. A collaborative research team including experts from Harvard University, the University of Cambridge, and the Institute for Advanced Study utilized the AI model to simplify complex mathematical expressions ^[1][2]. GPT-5.2 identified a hidden pattern in these equations and conjectured a general formula valid for all cases ^[2][3]. Following the conjecture, an internal version of the model spent approximately 12 hours reasoning through the problem to provide a formal mathematical proof ^[2]. The resulting discovery was subsequently verified using standard analytical methods, including the Berends-Giele recursion relation ^[2]. Leading physicists, such as Professor Nima Arkani-Hamed, described the findings as a significant demonstration of how modern AI can excel at deep pattern recognition in fundamental science ^[1][2]. This breakthrough marks a transition for large language models from simple assistance to generating original, journal-level scientific knowledge ^[3].