Ashlyn Campbell
PhD Student, University of Michigan
ashlynnn (with 3 n's ♡) [at] umich [dot] edu
Thinking Critically About Large Language Models
In 2026 the University of Michigan’s School of Information hosted a summer camp in collaboration with Wolverine Pathways for high school students. I worked alongside Xingjian (Lance) Gu to develop a workshop for students to critically consider the ramificiation of large language models (LLMs). The workshop began with one question: Can we trust LLMs? Large language models (LLMs) are a type of AI technology behind tools like ChatGPT, and given their growing role in information platforms such as Google, Bing, and Twitter/X, this is the central question we focused on. The workshop centered on critical AI literacy and began by exploring the basic ideas behind the algorithms that power LLMs through Python conditional statements. Afterwards, we placed the students into teams to address one misconception about LLMs for non-experts and create a one-slide presentation. Finally, each team presented solutions to the entire camp. It’s important to note that this workshop was created for students with both prior programming experience and little to no experience and provided an opportunity to develop beginner programming skills while exploring the ethical implications of AI.
To support educators, students, and others interested in critical AI literacy, I have included the workshop resources developed for this session. These materials include the workshop agenda, instructional slides, and an example of a student presentation addressing a common misconception about large language models.