Teaching Programming When Students Have AI

Abstract

Traditional programming assignments have long been used to evaluate students’ ability to write code, a task that can now easily be completed with generative AI assistance. Students can produce functional code and meet assignment requirements without necessarily developing a conceptual understanding of the programs they submit. When students submit AI-generated solutions they cannot explain, learning becomes shallow. AI-generated code may also create a false sense of confidence, giving students the illusion of understanding without developing the deeper problem-solving skills necessary for future learning and adaptation. These challenges highlight that programming education is about far more than producing correct code.

This shift requires reconsideration of the broader purpose and design of programming education. Student-centered programming education in the age of AI requires both intentional course redesign and explicit guidance for how students engage with AI tools. This presentation will explore how instructors can redesign programming courses to preserve meaningful student learning and how students can use AI in ways that support rather than replace learning. Discussion will focus on examples of where AI can support novice programmers and where it may undermine learning.

Presenters