A new artificial intelligence weather forecasting model developed by researchers has demonstrated superior accuracy compared to conventional physics-based systems in global weather prediction. The AI model, trained on decades of historical weather data, can generate high-resolution 10-day forecasts in minutes on standard computing hardware, a task that typically requires hours on supercomputers using traditional methods. Initial testing shows the model provides more accurate predictions for key atmospheric variables like temperature, pressure, and wind speed across most regions. While promising, meteorologists note that AI models may struggle with predicting extreme weather events that deviate significantly from historical patterns. The technology represents a significant shift toward data-driven approaches in meteorology, potentially making advanced forecasting more accessible. Read the full article at https://technologyreview.com/ai-weather-model-forecast-accuracy.