Computing Beyond Code

Abstract

Abstract

Most conversations about GenAI in the intro course still assume students write code. This session takes the other fork: a pure elective for non-technical students in which programming happens entirely through conversation with an AI agent, with no syntax, no error messages, and no stack traces. Co-led with Kevin Lin and Cory Gwin, it is deliberately a discussion rather than a pitch, since none of us has taught the course yet. We set the tools aside and work through the three things that do not change: what the course is for (goals), what students must understand (concepts), and how to give it shape so learning is felt (structure).

Summary

Summary

The closing hour of the consortium’s summer workshop, following an introduction to an AI-integrated CS1 that keeps students writing code. Where that hour keeps code at the center, this one removes it entirely and asks what is left to teach when the machine can build the thing: the durable judgment that survives the tool, which is specifying, decomposing, and telling good work from bad. The hour runs as a facilitated conversation across three breakout rooms, moving through goals, concepts, and structure, with a candidate throughline that the real skill is an accurate mental model of the agent, and a nod to the architecture studio as a model for critique and felt progress. It closes, for fun, with a vote on what to even call the thing.