This chapter explores a study on using AI language models to analyze speech patterns and diagnose schizophrenia. The researchers found that the speech of people with schizophrenia had subtle differences compared to those without the condition, potentially allowing for early detection and treatment assessment.
Madeleine Finlay meets neuroscientist and psychiatrist Matthew Nour, whose research looks at how artificial intelligence could help doctors and scientists bring precision to diagnosis of psychiatric conditions. He describes his latest study looking at patients with schizophrenia, and explains how he thinks large language models such as ChatGPT could one day be used in the clinic. Help support our independent journalism at
theguardian.com/sciencepod