A new artificial intelligence system developed by researchers has demonstrated superior accuracy in short-term weather forecasting compared to conventional physics-based models. The AI model, trained on decades of historical weather data, can predict precipitation, temperature, and severe weather events up to 10 days in advance with greater precision. Early operational tests show it reduces error rates by approximately 20% for key metrics like hurricane track prediction. While promising, meteorologists note that integrating AI forecasts with established methods will be crucial for reliability. The development signals a significant shift in how weather predictions may be generated in the future. Read the full article at https://technologyreview.com/2024/05/15/ai-weather-model-forecast.