Talking to Colleagues about Generative AI

I’m facilitating or participating in five faculty learning communities this fall that focus on teaching and generative AI. Most of the participants in these learning communities are keen to explore possible uses of generative AI in their teaching or in their students learning. Many of the participants, however, are working in departments or programs with colleagues who are skeptical of generative AI and its role in teaching and learning.

The skepticism shows up in a few different ways. Some instructors aren’t convinced generative AI is good for much. (These instructors likely haven’t tried any of the paid versions of the major AI tools lately.) Some instructors are legitimately worried about students using generative AI to cheat on assignments, which raises concerns about academic integrity and lost learning opportunities. Still others have deep ethical concerns about environmental impacts or labor issues or intellectual property issues.

Several of the participants in my learning communities have asked for advice on how to talk with colleagues about generative AI, particularly skeptical or even resistant colleagues. Below you’ll see some strategies that I brainstormed (with just a bit of help from ChatGPT), but I want to make sure to frame these correctly. I’m not interested in all instructors adopting generative AI in their teaching, but I am interested in all instructors having a thoughtful and informed approach to AI. “Red light” policies are fine if they’re intentional and not just an easy response to fears about AI.

With that framing, here are some approaches you might take to talking with colleagues about generative AI:

• Something I’ve learned in my years of consulting with faculty about their teaching is to lead with questions. What is your experience with generative AI? What have your students said about generative AI? Where in your course do you think AI might play a helpful or disruptive role? Figure out where your colleague is with respect to teaching and AI and only then offer some advice for taking next steps. (In centers for teaching and learning, we call this indirect consulting, since it doesn’t lead with advice or prescriptions.)
• If you’re exploring the use or mitigation of generative AI in student assignment or activities, share an example or two with your colleague, along with a sense of how your students responded to the assignment or activity. Instructors who are tentatively interested in incorporating AI into student work are typically very interested in practical examples.
• Often when students are given an assignment, there’s a great deal of attention placed on the products that the students create for the assignment, whether that’s an essay or a paper or a presentation or poster or something else. Some instructors are naturally worried about the many ways students might use generative AI to produce such products. However, if we can shift our focus from the product to the process used to create that product, we can often do a better job teaching students the skills of our disciplines and create more transparency about the use of AI during that process. So try shifting the conversation with your colleague from product to process.
• Be ready to say that you don’t know something about AI and to research it later. This set of technologies is changing so fast and educators are experimenting in so many different ways, no one person can keep up with it all! I’m always making notes during consultations or learning communities meetings to investigate some question or tool or application later. I don’t always have time to do a deep dive, but I try to find at least one or two resources that can help.
• There’s been so much discussion in higher ed about students using AI for good or for ill, but I’m hearing more and more from faculty who are using AI as part of their course and lesson prep, either to save time or to improve the quality of learning materials. If a colleague hasn’t considered these uses, then shift the conversation there and away from perhaps harder-to-answer questions about AI policies and student use.
• To that end, if you use generative AI for practical tasks, like drafting letters of recommendation or assignment rubrics, describe those uses to your colleague. Then share a sample chatlog with your colleague so they can see how you prompted the tool and what kinds of responses it provided. This approach communicates some of the value you’ve found in using AI and makes more concrete a tool that might be mysterious to your colleague.
• The paid versions of generative AI tools like ChatGPT and Claude typically provide much more useful and accurate responses to prompts than the free versions. For instructors resistant to exploring generative AI, it might be useful to demonstrate for them the difference in the paid and free versions of a tool and then connect that to equity concerns. Not all students will have the means to pay for “pro” versions of these tools. How can we try to address that access gap individually or collectively?
• For instructors who would just rather not talk about generative AI, point out that many students are using AI tools of various kinds and they will be expected to use them in many jobs after graduation. Tell your colleague that students want guidance about using AI effectively and ethically, which is a good reason to teach about AI even if you’re not teaching with AI.

If you’re having conversations with colleagues about teaching and AI this fall, how are they going? What approaches are you taking to make these productive conversations?