A new artificial intelligence-based weather forecasting model developed by researchers at Stanford University has demonstrated superior accuracy compared to conventional physics-based models in global weather prediction tests. The AI system, named ClimaNet, was trained on decades of historical weather data and can generate a 10-day global forecast in under two minutes on a single high-performance computer node, whereas traditional supercomputer-based models require hours of computation. In head-to-head comparisons over a six-month period, ClimaNet showed a 15% improvement in predicting atmospheric pressure patterns and a 12% improvement in temperature forecasts beyond five days. The researchers emphasize that the AI model complements rather than replaces existing systems, as it currently lacks the granularity for highly localized severe weather events. The technology could significantly reduce computational costs for meteorological agencies and improve long-range planning for agriculture and disaster preparedness. Read the full article at https://technologyreview.com/2024/05/15/ai-weather-model-outperforms-traditional-systems.