Generative AI and Large Language Models (LLMs) are disruptive technologies to all
academic disciplines that rely on writing as an artifact to demonstrate student understanding.
Writing code is not immune from these seismic changes. Learning about Computer Science
has always focused on expressing ideas in a language that is understood by the machine. As
the discipline has matured, our computer languages have evolved to become closer to
human languages. In this sense, using LLMs to create software is a natural next step.
However, this is a large step – most of the problems given in a conventional introductory
course can now be easily and correctly solved by an AI agent. It is apparent that the
traditional model of evaluating student understanding of computing by having them write
simple programs will no longer be meaningful. It is not clear what replaces this model.
This project aims to provide a collaborative environment to explore pedagogical techniques
to best engage students in a course that explicitly utilizes Generative AI and Large
Language Models. Computer Science faculty from multiple ACM schools will adapt and
deliver an introductory programming course using these techniques. This experience will be
reviewed with an eye toward generalizing the practices to a wider collection of disciplines.