This course is an introduction to the wide, fast-moving range of ways in which AI methods are being used in social science, and how social science is used to understand AI impacts (e.g., in labor markets). The focus will be mostly on LLMs and machine learning, with one example from CLIP. Topics include: AI agents as substitutes for people in political and economic surveys and experiments; alignment: sycophancy, possible AI harm; machine learning for prediction in small (experiments) and large data sets (e.g., bail-setting), and AI used in analysis of financial data. Students will be graded on (1) in-class contributions, (2) two homework sets, (3) a class project either extending (or reproducing) a recent paper or doing an original analysis, and (4) an in-class final. Some funds will be available for computing power.