The article examines the rise of AI-powered machine-learning weather models from major tech companies, assessing how they compare to traditional physics-based forecasts. AI models, trained on decades of data, run much faster and require less computing power than traditional supercomputer-based systems. While some new AI models have outperformed traditional ones in certain large-scale forecasts, they often underperform in predicting local events and weather extremes, and face challenges adapting to a changing climate. Experts suggest that using both traditional and AI models together may provide the most accurate and timely forecasts in the near future, as each approach brings unique strengths.





























