Monday, April 27, 2026 - 4:00pm
Event Calendar Category
LIDS Seminar Series
Speaker Name
Avery Cohn
Affiliation
Tufts University
Building and Room Number
32-155
"Making AI Earth-Aware: An Open, Public-Interest Agenda"
Earth observation (EO) should be one of the most valuable data categories on the planet. Satellites monitoring Earth's surface with increasing scope and frequency fill cloud buckets with hundreds of petabytes of imagery including a great deal that is freely available. The collection is rich enough to support much-needed action on agriculture, disasters, climate, public health, and much more. Yet the field underdelivers today. Challenges with costs and access result in disappointing levels of EO use, users, and value. Our non-profit initiative, Clay, is part of a community working to combat this problem. We see a path to help make widely-used AI aware of Earth (observation) in ways that support a vast array of public-interest applications. I'll discuss accomplishments and lessons from our first chapter–during which we released Clay, an open foundation model of EO and explored its applications with partners. I'll also discuss our ambitions for Clay's next chapter, which rests on three linked bets: creating, storing, and serving Earth embeddings as data for downstream workflows enables new potential; practitioner-driven benchmarking guides both model development and the workflows to help realize this potential; and a deeply collaborative approach increases speed and scale.
Avery Cohn co-leads Clay, an applied AI initiative putting Earth data to work on public-interest problems, incubated at Renaissance Philanthropy. He is on the faculty at Tufts University, a Senior Partner at Ode Partners, and an advisor to LGND AI and the Climate Curve Methane Prize. Previously, he founded and led the food and agriculture program at ClimateWorks Foundation. He holds a Ph.D. in Environmental Science, Policy, and Management from UC Berkeley.