Teaching Intro AI When the Tools Can Do the Homework: A Course Redesign and a Student Bill of Rights

Abstract

Large language models can complete most of the assignments in an introductory artificial intelligence course. This paper is an experience report on redesigning one such course, CSS~382 at the University of Washington Bothell, in response. Rather than freeze the curriculum, the redesign retained the course's classical core (search, adversarial search, Markov decision processes, reinforcement learning) and added a strand in which students build a large language model from scratch, so that a tool they are required to use is also one they are required to understand. Assessment was rebuilt around tasks that resist unattributed automation: in-class exercises, reflective writing, and a defended team project, with examinations removed entirely. The policy on AI was inverted, from unmentioned in 2023 to required in 2026. The center of the paper is a participatory ethics sequence in which a cohort of students deliberated on and endorsed a "Student Bill of AI Rights" governing their instructor's own use of AI, including a requirement that the instructor personally complete any AI-generated assignment before issuing it. The provisions were scaffolded by an AI-generated prompt and ratified by the students, and that provenance is part of what the account examines. The design, the student-authored artifacts, and the tensions that followed are reported, including student objections to AI-generated course materials, with explicit attention to the limits of what a single-cohort design narrative can claim.