Connor T. Jerzak is an Assistant Professor of Government at the University of Texas at Austin. His research lies at the intersection of causal inference, machine learning, and epistemology. He develops frameworks for “planetary causal inference,” integrating Earth observation, satellite imagery, and AI to study cause-and-effect relationships, treatment-effect heterogeneity, and generalizability at global scales. Through Neuristemic.ai (where he serves as AI Research Lead alongside epistemologist Ethan Jerzak), he examines AI systems as instruments of knowledge, investigating the epistemic reliability of large language models and multimodal re…
Connor T. Jerzak is an Assistant Professor of Government at the University of Texas at Austin. His research lies at the intersection of causal inference, machine learning, and epistemology. He develops frameworks for “planetary causal inference,” integrating Earth observation, satellite imagery, and AI to study cause-and-effect relationships, treatment-effect heterogeneity, and generalizability at global scales. Through Neuristemic.ai (where he serves as AI Research Lead alongside epistemologist Ethan Jerzak), he examines AI systems as instruments of knowledge, investigating the epistemic reliability of large language models and multimodal representations, treatment leakage in text-based causal inference, debiasing for trustworthy scientific measurement, and the limits of automated knowledge production. His work engages core questions in the philosophy of science and social science methodology: the nature of causal identification in high-dimensional, unstructured data; the validity of machine-learned proxies and representations (including Platonic representations in vision-language models); and the epistemological foundations of measurement, confounding, and evidence when scaling empirical inquiry across complex social and planetary systems.
