Anya Martin is an interdisciplinary scholar working at the intersection of social science, artificial intelligence, and meteorology. She received her master’s degree in machine learning from UCLA, and is currently pursuing a doctorate in human-centered computing at Georgia Institute of Technology. Her work focuses on infrastructural and human factors in AI/ML applications at a time when AI/ML methods are moving outside of their traditional homes in tech companies. But in this shift and expansion, AI/ML methods are forced to reckon with “alien” infrastructures of data, compute, and trust. This is perhaps most clearly seen with meteorology, where data interoperability is founded on physical law, and where model “trust” is hotly contested given the high-stakes nature of disaster prediction. Anya is studying how these epistemologies of large-scale data interact as it is crucial for understanding where AI/ML methods succeed and fail.
In her Fulbright-Nehru project, Anya is examining how AI is being used to predict the weather in India. Meteorology is a physics-based “big data” science reliant on massive data and compute infrastructures to effectively model the atmosphere. This project aims to understand how these infrastructures work with the new AI/ML methods developed over the last few years, most notably the 2022–2023 AI weather models developed using (almost exclusively) tech company data and compute infrastructure. Anya’s ethnographic work to capture this crucial phase will not only inform AI transition but also provide a critical case study for the effective use of AIs.