AI-Aware Computer Science Education

Earlier this year on the podcast, I shared a panel conversation on the evolving impacts of generative AI on mathematics education. I really liked that format and I thought I might try it with an adjacent discipline—computer science. There were AI coding tools in wide use before ChatGPT launched, so CS as a discipline has had a little more time to adapt to AI than other disciplines. Plus, CS is a field used to change--new coding languages, new programming heuristics, now technologies. How has it adapted to the arrival of generative AI?

Not as well as I thought.

This week on the Intentional Teaching podcast, I welcome a panel of CS educators to discuss the challenges AI poses to the field and the strategies they're using to respond. On the show are Briana Morrison, professor of computer science at the University of Virginia; Greg Nelson, assistant professor of computer science at the University of Maine; and Elijah Rivera, assistant professor of computer science at Brandeis University

We talk about the need to align CS program learning outcomes with the needs of industry, strategies for building self-reflection and community building into CS courses, and lessons from the history of automation in software development that we can use to make sense of AI's impact. To be clear, my panelists are responding in creative, intentional ways to AI! I was just surprised to learn how little the CS curriculum has changed since I was a CS major in the 90s. Why on earth are they still teaching compiler courses?

Warning: This episode is a long one! My panelists had a lot to say, and I circled back around to Elijah after the recording to ask a few more questions about automation. But it's worth listening through the end. CS education hasn't found some magic wand to wave to "solve" the AI problem, but my panelists have practical advice on AI-aware teaching that cuts across disciplines.

​You can listen to my conversation with Brianna, Greg, and Elijah here, or you can search "Intentional Teaching" in your podcast app.

On the Importance of Unlearning

Back at the start of the semester, I ran a workshop at UVA called "Opening Lines: Starting (and Continuing) the Conversation with Students about AI." We focused mostly on AI syllabus statements, not because a good AI syllabus statement is all it takes to "solve" the AI problem, but because a good AI syllabus statement can crystalize for the instructor how they want AI to show up or not in the student learning process. And a good AI syllabus statement is an excellent starting point for ongoing discussions with students about how they think about and use AI. Thus the workshop title, "Opening Lines."

I shared an AI syllabus statement for a first-year writing course on attention and distraction taught by my UVA colleague Ethan King. Given the course topic, Ethan takes a red-light approach to AI in this course. Here's part of how he talks about AI on his syllabus:

​You can read the rest of Ethan's AI syllabus statement here. It's a quality statement. He's clear about what uses of AI are allowed (none) and which are disallowed (all of them), and he provides a rationale for that particular boundary line, which might seem arbitrary otherwise. And the whole statement conveys an attitude of care for students. Ethan wants them to learn and he wants them to understand his role in helping them learn.

I had already decided to share Ethan's statement before I knew that Ethan would be a participant in the workshop. Awkward? Not really, more like good fortune. Ethan was able to provide more context for his statement and field some questions from other faculty about his approach to AI in this course. During the discussion, Ethan said something that I keep thinking about. It was in response to this part of his AI statement:

This is at the heart of his rationale for banning AI, but what Ethan shared during the workshop is that it's a way of thinking about writing that's new to some of his students. Given their academic writing experiences, many of his students over-index on the product of the writing process instead of the process. If the product is the thing, then using AI to make a better product is awfully tempting. But if writing is about experiencing the world, making sense of those experiences, and finding ways to share that sensemaking with others, well, as Ethan says, there's not a good reason to outsource that to AI.

Ethan's comment reminded me of the notion of an enduring understanding that Grant Wiggins and Jay McTighe write about in their classic Understanding by Design. It's a way to think about our high-level course goals: What understanding of the course topic or your field do you want students to internalize and use to shape their way of being in the world? The thing about an enduring understanding is that often involves unlearning existing and naive understandings. College writing isn't just about the final paper you turn in, it's the thinking and sensemaking you do as you write. For some students, maybe many students, that requires a kind of unlearning.

I thought about this unlearning when I read a recent post by Elissa Yukiko Weichbrodt, a professor of art and art history at Covenant College in Georgia. Here are a couple of paragraphs from her essay:

There's a vision of grading in her essay that is quite beautiful and likely very different from how many of her students think of grading. She's pointing to a kind of unlearning, helping her students see grading as "opportunities for connection, not just correction," as she writes. What kinds of unlearning might benefit your students and open up opportunities for connection?
​
Elissa's entire essay, "When I Say I'm Grading Your Paper, What I Mean Is...", is worth reading. Her Instagram version of the essay is quite charming, too!

The Norton Guide to AI-Aware Teaching

My new book, The Norton Guide to AI-Aware Teaching, co-authored with Annette Vee and Marc Watkins, is now available! It's available through major online retailers like Amazon and Barnes & Noble as well as from smaller retailers like Bookshop and from Norton. The ebook version is free for faculty who have adopted a W.W. Norton textbook, so if that's you, ask your Norton rep. And if you'd like to order multiple copies for a campus reading group, Norton offers a bulk order discount. Email me for details.