This summer, 38 million farmers across India benefited from AI-powered forecasts predicting the start of the monsoon season, enabling them to make better-informed decisions about planting their crops. These advanced predictions were made possible by NeuralGCM, a cutting-edge weather modeling system developed by Google Research that combines traditional physics-based forecasting with machine learning to deliver more accurate and efficient climate simulations. For decades, weather and climate models have been resource-intensive, typically requiring access to supercomputers to run. Google Research aimed to change that by creating a more efficient and scalable approach. NeuralGCM was designed to learn from decades of historical weather data, identifying complex patterns and relationships that traditional models might miss, while still incorporating fundamental physical laws of the atmosphere. What sets it apart is its remarkable efficiency—it can run on a single laptop, significantly lowering the barrier to entry for researchers and institutions worldwide. When Google Research open-sourced NeuralGCM, it hoped the scientific community would build upon the tool to drive new innovations. That vision became a reality through a collaboration with the University of Chicago’s Human-Centered Weather Forecasts Initiative. The team recognized that one of the most critical challenges for smallholder farmers in tropical regions is knowing when the monsoon will begin. The timing of the rainy season directly impacts crop yields, livelihoods, and food security for hundreds of millions of people. Despite decades of research, predicting the exact onset of the monsoon—especially weeks or even a month in advance—has remained a persistent challenge. The University of Chicago team tested multiple AI-driven weather models and found that NeuralGCM, when combined with other advanced systems like the European Centre for Medium-Range Weather Forecasts’ AIFS model and historical climate data, delivered the most reliable results. The hybrid approach not only predicted the monsoon’s start with high accuracy but also captured unusual weather patterns, such as a sudden dry spell during the monsoon progression, which could have gone undetected by traditional models. The success of this project demonstrates how AI can bridge the gap between complex climate science and real-world impact. By making high-quality, localized forecasts accessible and actionable, AI is empowering farmers with the tools they need to adapt to climate variability, reduce risk, and improve agricultural outcomes—proving that advanced technology can serve the most vulnerable communities with tangible results.