A research team from the Massachusetts Institute of Technology (MIT) published a study in MIT News on October 8, 2025, in which machine learning was used to analyze the behavior of metals under extreme conditions. The scientists used AI to process vast amounts of data from simulations of metal alloys at ultra-high temperatures and pressures. It was previously believed that under such conditions, the materials atomic structure becomes completely chaotic just before failure. However, the machine learning algorithm was able to identify hidden patterns and discovered that the atoms briefly self-organize into previously unknown, ultra-stable nanocrystalline structures. This discovery fundamentally changes the understanding of metal deformation and failure processes. It paves the way for the targeted creation of new alloys designed to form these stable structures, which could lead to much stronger materials for the aerospace and nuclear industries.
MIT Researchers Use Machine Learning to Uncover New Physics in Metals
