Researchers have developed an AI model named 'SleepFM' that can predict the risk of over 130 health conditions using sleep data. Trained on nearly six lakh hours of data from 65,000 participants, the model analyzes multiple physiological signals recorded during sleep studies. Initially tested for sleep analysis tasks, it was then applied to predict future disease onset by correlating sleep patterns with health records, demonstrating reasonable accuracy for a significant number of diseases.