A new artificial intelligence weather forecasting model, developed by researchers at a leading technology institute, has demonstrated superior accuracy compared to conventional physics-based models in a comprehensive global evaluation. The AI system, trained on decades of historical weather data, can generate high-resolution, 10-day forecasts in under a minute on a single desktop computer, a task that typically requires hours of computation on supercomputers using traditional methods. The study, published in a peer-reviewed scientific journal, compared the AI model’s predictions for key atmospheric variables against the current gold-standard global forecasting system. Results showed the AI model matched or exceeded accuracy in over 90% of the 1,300 test cases, particularly in tracking the intensity and path of major weather systems. Researchers emphasize the model is a complementary tool for meteorologists, not a replacement, and its speed could enable more frequent updates and ensemble forecasting to improve severe weather warnings. The full details of the methodology and results are available in the published study at https://sciencedaily.com/releases/2024/05/240507123456.htm.