Session Presentation: "What liberal arts and sciences students should know about AI: The Case of Alien Anatomy," Mark Guzdial (11.13.23)

Abstract: The University of Michigan has started a new Program in Computing for the Arts and Sciences which offers computing education to arts, humanities, and science students – separate from programs in Computer Science, Information, or Informatics. The learning objectives and student activities were created in a participatory design process with faculty in the liberal arts and sciences. I've been teaching a course for humanities students (for the last year) on "Computing's Impact on Justice: from Text to the Web". When we designed the course in summer 2022, ChatGPT wasn't a thing yet. My advisors wanted me to teach about sentence recognition and generation, building chatbots, and training an ML system (gesture recognizer). Students program in all of those activities in the course, in either teaspoon or Snap languages. While not explicitly designed for ChatGPT, we were in a good place for relating students' activities to how LLM systems work when they emerged on the scene. We're now building a separate course (“Alien Anatomy: How ChatGPT works”) with those elements in a tighter, explicit framing on what humanities students should know about generative AI. This talk is an exploration of what arts and humanities students should know about AI and how our teaching methods should change for a student population who are not aiming to become software professionals.

Mark Guzdial. 2023. What liberal arts and sciences students should know about AI: The Case of Alien Anatomy. In Uppsala University Conference: Generative AI: Implications for Teaching and Learnings (November 13, 2023). Uppsala, Sweden