Student success
AI and student success
What if students learned
to teach themselves?
Fred Gibbs
University of New Mexico
Critical Thinking with AI
Remixing Plato
Students use Phaedrus to create a dialogue
about AI, then refine its competing perspectives.
Students define the characters' positions
and decide which perspectives to develop.
Student feedback
They had not seen AI's potential to help them
explore new perspectives.
Diet, Health, Expertise
Complicate the obvious
“What is a healthy diet?”
Students use course readings to question
the answer's assumptions and omissions.
Student feedback
A better understanding of generic responses
and silences: what an answer leaves out.
History of American Food
Discussing fast food with AI
Students compare AI summaries with the article.
What did AI do with the text?
What did it miss?
Classroom feedback
Very lively engagement. My sense was that
students learned more than in our usual discussion.
It also helped me get out of my own discussion rut.
Reading: Steven Penfold, “Fast food”
Across my courses
Students valued guidance
- Many had little practice learning with AI.
- Many were skeptical and wanted a say in its use.
- Feedback was overwhelmingly positive
about learning to use AI productively.
History of American Food (16 weeks)
Critical Thinking with AI (8 weeks)
Diet, Health, Expertise (4 weeks)
More examples
Other assignments
Teaching approach
Purpose and agency
Transparency
Every assignment needs an explicit purpose
and support for doing the work.
Agency
Students need practice making decisions
about how AI helps them learn.
AI can make students more critical and more capable.
AI and student success
Learning with AI
needs teaching
What would help your students
learn when they get stuck?
Syllabifredgibbs.net/courses
Amaranthamaranth.unm.edu
Sketchbookamaranth.unm.edu/ai-sketchbook