Researchers developed an AI model, WARN, that predicts cardiac arrhythmia 30 minutes before onset with 80% accuracy. Trained on data from 350 patients, WARN can be integrated into smartphones and wearables, offering real-time, early warnings for atrial fibrillation, enhancing preventive healthcare.