With the rise of generative AI tools, the approach to programming projects in large computer science classes may need changes. Traditionally, these courses, which often cater to 500-1000+ students, manage coursework through tightly specified projects for automated grading. However, generative AI tools can write code that fits these specifications, allowing students to depend too much on these tools.
Using a new course that will be part of the Computer Science and Data Science degree programs, this project aims to create innovative classroom projects that both (1) feel valuable to students because they cannot be solved using a generative AI tool and (2) are gradable at a scale of 500 students. Building on insights collected from a CRLT-Engin survey of EECS 485 students regarding their use of generative AI, we will explore several models for open-ended projects where students exercise more design skills. The materials developed in this new course will serve as a model for how student projects in other large courses can also be adapted.
Project Team
John Kloosterman
Computer Science and Engineering
Melina O’Dell
Computer Science and Engineering
Funding
This team was awarded $5,496 in funding in Summer 2024.