The scientists at the University College London's Institute of Neurology have developed a new tool using an AI language model that can characterize subtle differences in the speech of individuals with schizophrenia. This research explores how automatic language analysis can improve the diagnosis and assessment of mental illnesses. The findings indicate that the linguistic performance of schizophrenia patients, which is less predictable than that of the control group, can be predicted by AI models, possibly related to the way the brain learns and stores cognitive maps. The scientists plan to extend this technology to a larger sample of patients, bringing new possibilities to the diagnosis of mental illnesses.