Recent advancements in artificial intelligence are transforming weather prediction, with new AI models demonstrating the potential to surpass conventional physics-based forecasting systems. These machine learning models, trained on decades of historical weather data, can generate highly accurate forecasts in seconds, a process that typically takes hours for traditional supercomputer-run models. Key players like Google’s GraphCast and Huawei’s Pangu-Weather have shown competitive skill in predicting global weather patterns up to 10 days ahead. While these AI systems excel at identifying broad patterns quickly and efficiently, experts note they currently lack the granular detail for highly localized, severe weather events like thunderstorms. The technology represents a significant shift towards data-driven prediction, though operational integration with established meteorological centers will require rigorous testing for reliability. For the full analysis, read the complete article at https://technologyreview.com/2024/05/15/ai-weather-forecasting-models.