Joseph Austerweil

Joseph Austerweil is a cognitive psychologist and machine learning researcher. He investigates how people retrieve, learn, create, and teach knowledge, formalizing these processes as computational models and testing them against empirical results from psychology and neuroscience. He has multiple publications in top-tier psychology journals (Psychological Review, Trends in Cognitive Sciences) and machine learning conferences (NeurIPS).

He held faculty positions at Brown University (2013-2015) and the University of Wisconsin-Madison (2015-2025, Associate Professor of Psychology and Computer Science), where he directed the Austerweil Lab studying memory, concept learning, decision-making, computational social cognition, and the development of computational tools for understanding mental representation. His lab trained graduate students who now hold faculty positions (e.g., Mark Ho, NYU; Jeffrey Zemla, Syracuse) and postdocs who have moved into academia and industry (e.g., Yoed Kenett, Technion; Nolan Conaway, industry).

In June 2025 he joined the Henkaku Center as a senior researcher and Chiba Institute of Technology as a co-founding faculty member for the new School of Design and Science (SDS), where he serves as Professor and Academic Director.

His current research centers on human-machine collaboration in higher education: how people learn to work with AI tools, how those collaboration skills develop, and how curricula can be designed to build durable judgment under rapidly-changing AI capabilities. He leads a longitudinal study at Chiba Tech measuring how students’ task delegation, calibration, and prompt sophistication evolve over a semester, and how individual differences (cognitive reflection, need for cognition, intellectual humility) moderate effective AI collaboration. He is also developing “Teaching AI and Teaching with AI” as a framework for AI-native higher education: teaching students to work effectively with AI tools (workflow design, project management, security awareness, human-AI-human collaboration) and using AI as a medium for developing problem-posing skills, assessment innovation, and durable judgment. He retains broader interests in computational social cognition (formal models of how people reason about other minds, pedagogical reasoning, and the emergence of norms), semantic memory and fluency, concept learning, and Bayesian and reinforcement-learning models of human cognition.