Google DeepMind launches GenCast AI to enhance weather forecasting accuracy

Google DeepMind has introduced GenCast, a new AI model that improves weather forecasting accuracy and extends prediction timelines beyond traditional methods. Launched on December 4, GenCast outperforms existing models, especially in predicting extreme weather and long-term forecasts. GenCast uses ensemble forecasting based on AI rather than numerical weather prediction. Trained on 40 years of weather data, it generates multiple forecasts simultaneously, achieving results in eight minutes compared to several hours for supercomputers. While GenCast shows promise, experts note that traditional models remain crucial for providing foundational weather data. Google plans to collaborate with weather agencies to enhance forecasting methods while continuing to rely on established models.