With Great Powers: A Practical Guide to Agentic AI for Social Science Research

Generative AI is reshaping the inputs, throughputs, and outputs of social-science research — what we can collect, how we code and measure, what we can validate. This book chapter takes stock of those changes for working political scientists, organizing the use cases and pitfalls into a framework researchers can carry into their next project. The chapter draws on my own AI-assisted research and on classroom experience teaching AI-as-method to graduate students at Stanford. The argument is that a disciplined integration of AI into the research workflow is possible without surrendering researcher control over substantive judgment — but it requires methodological norms the field has not yet articulated.