Two theoretical physicists, including Nobel laureate Giorgio Parisi, have solved a decade-old mathematical problem with help from Claude, the large language model (LLM) developed by Anthropic. The problem, which stumped Francesco Zamponi and Parisi for 10 years, originates in the field of complex systems and relates to jamming, a sudden transition where a fluid system becomes completely rigid yet remains disordered. Zamponi, a physicist at the Sapienza Università di Roma, Italy, explains that jamming is an example of a general mathematical problem known as constraint satisfaction, where spatial positions of spheres act as variables and the rule that no two spheres can overlap serves as the constraint. These concepts are widely applied in neuroscience and artificial intelligence, as an artificial neural network undergoes a jamming transition similar to physical spheres. The result is significant for fundamental physics and offers further evidence that artificial intelligence is reshaping the nature of research in some fields.