A new artificial intelligence weather forecasting model developed by researchers has demonstrated superior accuracy compared to conventional physics-based models in global medium-range weather predictions. The AI system, trained on decades of historical weather data, can generate 10-day forecasts in minutes on a single computer, a task that typically requires hours of computation on supercomputers using traditional methods. Initial testing shows the AI model matches or exceeds the accuracy of leading global forecasting systems for predictions up to seven days ahead, particularly in tracking extreme weather events like hurricanes and atmospheric rivers. While promising, researchers caution that AI models may struggle with predicting unprecedented weather phenomena not represented in their training data. The development signals a potential shift toward hybrid forecasting systems that combine AI speed with traditional physics-based reliability. For the full details, read the complete article at the provided source.