One reason why reinforcement learning—a technique for improving AI models—has become so popular is because researchers can ...

Physicists used a machine-learning method to identify surprising new twists on the non-reciprocal forces governing a many-body system.