A new artificial intelligence model developed by researchers has demonstrated superior accuracy in global weather forecasting compared to conventional numerical weather prediction systems. The AI model, trained on decades of historical weather data, can generate high-resolution forecasts for temperature, pressure, wind, and humidity in minutes, a task that typically takes hours for traditional supercomputer-driven models. Early tests show the AI system matches or exceeds the accuracy of leading global forecasting models for predictions up to 10 days ahead, particularly in tracking extreme weather events. While promising, meteorologists note that AI models currently lack the precise physics-based understanding of atmospheric processes that traditional models provide, and the two approaches are likely to be used in tandem for the foreseeable future. Read the full article at: https://technologyreview.com/ai-weather-model-forecast