Three Quick Examples of Teaching with and about Generative AI

This morning I met with a small group of UVA faculty to talk about generative AI. We’re reading Teaching with AI: A Practical Guide to a New Era of Human Learning by Bowen & Watson, but haven’t actually started the book yet. I wanted to share three examples of AI use mentioned during our meeting today, since I thought they were all pretty interesting.

Text-to-Podcast. You may have seen that Google’s NotebookLM app (an AI-enhanced notetaking app) rolled out an “Audio Overview” feature which will turn your document into a “lively” podcast episode. It’s a little goofy, with the AI podcast hosts engaging in more banter than you’ll find in my podcast, certainly. But one of the faculty members this morning used it on a journal article she was having an independent study student read. NotebookLM created an 8-minute overview of the article that was easy to listen to and gave the student an accurate overview of the article, setting the student up well to dive into the article itself for specifics. I thought that was a clever use of a new tool! You can read more about the tool in this story on the Verge. And you can listen to this podcast episode generated by NotebookLM from this recent blog post of mine.

 

Assigning Students to Groups. Another instructor, Charlie Gleek, talked about the challenge of putting students into groups for out-of-class discussions and projects, given all the conflicting schedules students have that make meeting times hard. This semester, he used a Microsoft survey to get information about student availability, dumped that information into an Excel file, and had Copilot figure out groups so that each group would have a common meeting time. He said it worked like a charm, and just took a couple of minutes of his time. This would be hard to do with a from-scratch interaction with ChatGPT, but since Copilot already had the structured data from the form, it was quick and easy. Please don’t ask me for details since I haven’t replicated this feat! I just appreciated knowing this was possible for someone interested in figuring it out.

AI Acceptable Use Scale. We talked a bit about ways to help students be transparent in their AI use on assignments and other course activities. One faculty shared an AI “acceptable use scale” designed by Vera Cubero based on the work of Leon Furze and others that maps ways students might use AI in their work on a five-point scale.

This wasn’t the first time I had heard of this scale, but it was the first time I had actually seen it! You can read more about it on Vera Cubero’s LinkedIn post, and see other versions of this scale in her Canva post.