Millions of farmers around the world rely on rain for their crops, but don’t have access to accurate weather forecasts that could help them decide when to plant, when to harvest, what crops to plant, and when to use inputs like fertilizer, pesticides, and irrigation.
A new project funded by the Laude Institute's inaugural Moonshots program—a research competition that asks how AI should be used to solve humanity's hardest problems—aims to change that.
The interdisciplinary group, led by researchers with the University of Chicago, is working to develop AI-based forecasting technology to support farmers and citizens around the world to plan agricultural decisions, public health decisions, and avoid extreme heat.
The hope is to deliver forecasts that combine AI weather and climate models, data from developing countries, and metrics that are explicitly designed to reflect farmer and public health concerns.
“AI is redefining what’s possible in modeling the Earth system. For example, it allows for the development of faster and cheaper weather forecasts that can be tailored to local needs, making them a game changer for developing countries on the frontlines of climate change,” said project member Pedram Hassanzadeh, UChicago Associate Professor of Geophysical Sciences.
“By bridging disciplines and moving the rapidly advancing theoretical work of AI and climate scientists beyond the lab—just two years after our team helped co-develop FourCastNet, the first pioneering global AI weather model—this project will deliver transformative tools to harness this AI-driven revolution and better prepare vulnerable communities for the climate realities they face today,” said Hassanzadeh, who directs the AICE: AI for Climate program, a collaborative initiative led by UChicago's Institute for Climate and Sustainable Growth and the Data Science Institute.
Hassanzadeh is one of four UChicago scientists on the project, along with Michael Kremer, the University Professor in Economics and the Harris School of Public Policy and Nobel laureate and Ian Foster, the Arthur Holly Compton Distinguished Service Professor of Computer Science and senior scientist at Argonne National Laboratory; the project is led by Rebecca Willett, the Worah Family Professor of Statistics and Computer Science in the Wallman Society of Fellows and Faculty Director of AI in the Data Science Institute.
Providing ‘human-centered’ AI weather forecasts
The team is actively working on partnerships to reach hundreds of millions of additional beneficiaries. Through partnerships with the Asian Development Bank, the researchers have already initiated conversations with officials from four other countries in Asia.
Hassanzadeh’s team also runs a training program on AI-based weather forecasting for officials from Meteorological Offices, with five countries (Bangladesh, Chile, Ethiopia, Kenya, and Nigeria) participating in 2025 and a further five (Colombia, Indonesia, the Philippines, Rwanda, and Senegal) scheduled to participate in 2026.
For the Laude seed grant, Foster, Hassanzadeh, Kremer, and Willett said they are developing software systems to make the forecasting and circulation of forecasts simple and user-friendly so that other countries can incorporate their own data.
The interdisciplinary team will develop new benchmarks that enable easy assessments of different AI forecasts across countries using their own local data, as well as AI tools that will yield more accurate forecasts and uncertainty estimates.
“Human-centered AI weather forecasts have the potential to save lives and boost prosperity by providing accurate, actionable information to millions in developing economies,” said Willett. “We are delighted by the Laude Foundation’s support of this effort to translate state-of-the-art AI methods to massive societal benefits for the people most exposed to climate risk.”
The Laude Institute backs computer science researchers turning breakthroughs into real-world impact from early-stage ideas to open-source projects and multi-year labs. Anchored by a $100 million pledge from founder Andy Konwinski, the co-creator of Apache Spark and co-founder of Databricks and Perplexity AI. Its Moonshots program selected eight teams to receive $250,000 seed grants, with each team competing for $10 million to fund a three-to-five-year lab to scale the open-source system globally.
—Adapted from an article first published by the Data Science Institute.
