We have to talk about what it’s going to mean, in an increasingly AI-saturated landscape, to teach without AI. Because one thing it won’t be is what we’re used to.
As we know, many people have serious concerns about the rapid spread of this new technology through our educational spaces. Some cite the wide array of dangers (environmental, economic, cognitive, democratic, etc.) it poses; others are skeptical of the wealthy and unregulated corporations designing and aggressively disseminating AI; still others simply note that there’s a lot we don’t know about this technology, and that we need to approach it with caution rather than a rush to implement. Altogether, many of us are saying—at the very least—slow down.
This is an uphill battle, perhaps, and one that might take a while. In the meantime, though,
I’m arguing that we need to do something that sounds simple, but isn’t: we need to learn how to teach without AI. And the reason that this isn’t simple is because teaching without AI means—has to mean—something different from what it used to mean back when this tech was unavailable.
Teaching without AI can’t, for example, mean ignoring these developments and pretending like we live in the past. No, we all have a responsibility to know what AI is and what it might be doing to the world—and, in particular, to our students.
After all, these new tools claim to offer a lot of appealing stuff: ready answers, speedy information, a lightened workload, and even therapy, medical advice, spiritual counsel, and companionship. Those offerings may be shaping many students’ experiences and expectations, all of which they’ll bring into the classroom. Against this backdrop, we’ll need to be particularly alert to signs of student distress. And it’ll be important for each of us to find out exactly what our individual students are bringing—surveying them, having full-class and individual office-hours conversations—so that we can respond with the most effective and transformative learning opportunities possible.
We also need to know what AI promises in the educational sphere, because some of those benefits are worth pursuing—by humans, without AI. In particular, advocates claim that AI has the advantage of being able to personalize learning, tailoring the journey to each student’s needs. That’s a great idea, actually, so we need to make changes in our institutions and classrooms to shape learning around our students’ diversity and individuality. How much agency, for example, can we give students to identify their own learning goals in our courses and to design course-relevant projects of personal importance? How responsive can we be to their individual challenges and progress, creating somewhat different arcs for each student?
Teaching without AI also can’t mean just returning to old practices. In fact, we have to confront the likelihood that those practices apparently did not make a sufficiently convincing case for the importance of human struggle, discovery, and growth. If they had, we might not now be dealing with ubiquitous arguments about the supremacy of efficiency, a flood of misinformation and disinformation, and persistent offers to take meaningful labor off our hands.
If a process of struggle, discovery, and growth does matter to us, we need to make space for all of it in our schools. Some of this is out of an individual teacher’s control: every student is dealing with a pile of obligations, social, professional, and academic, only some of which are coming from that one teacher’s course. One broader question for higher education, then, is: what are we asking of our students, and do our demands leave room and time for expansive learning?
But there are things we can tackle in the classroom. In what new ways can we offer our students low-stakes opportunities to experiment, make mistakes, hit dead-ends, try again, try again, and try again? This is a good argument for implementing multiple smaller assignments rather than a small number of big ones, for breaking major projects into discrete parts, and potentially allowing students to redo assignments in order to improve.
How can we better help students experience and explicitly reflect on the pleasure of reaching understanding after a messy, winding path? First of all, foregrounding our own reflective practices—both as we do our work and as we observe students doing theirs—creates a model for them to emulate. If informal reflection prompts aren’t part of your assignments, consider including them. Meetings with individual students, when possible, would allow that reflection to happen in conversation. We can also guide them to focus their reflection on key issues: the connection between discomfort and breakthrough, the value of doing the work oneself—even difficult or mundane work—at each stage of the process, and so on.
Of course, we can add extrinsic motivation, too, by rewarding students for throwing themselves into the work via process and completion grading. We all ought to be asking ourselves: to what extent does my grading system privilege outcomes, and to what extent does it value the process of learning?
Altogether, teaching without AI can’t just mean teachers slapping tools out of students’ hands. There’s plenty of reason to eschew this technology, but our approach can’t be, in other words, just about the exclusion of a technology. Teaching without AI in this moment has to be about thoughtfully promoting teaching and learning that is powerful, personal, meaningful, and lasting. It’s got to mean teaching with, and for, humanity.
David Ebenbach is the author of twelve books of fiction, non-fiction, and poetry, including his new collection The AI Suspects It Might Be a Hungry Ghost. He teaches creative writing and literature at Georgetown University’s Center for Jewish Civilization and creativity through GU’s Learning, Design and Technology Program, and he’s the Assistant Director for Graduate Student and Faculty Programming at Georgetown’s Center for New Designs in Learning and Scholarship.
References
Alpha School, “Welcome to Alpha: Where Learning Transforms Lives.” Alpha School, accessed July 17 2026. https://alpha.school/the-program/
Boehm, Jessica, Russell Contreras, and Isaac Avilucea, “AI arrives at church: How chatbots are shaking up religion.” Axios Phoenix, November 17 2025. https://www.axios.com/local/phoenix/2025/11/17/churches-ai-sermons-prayer-apps
Corrigan, Paul T., “A Taxonomy of Harm for the AI Age.” Inside Higher Ed, June 10 2026. https://www.insidehighered.com/opinion/views/2026/06/10/taxonomy-harm-ai-age-opinion#
Cox, Jay, “The Rise of AI Companionship: Exploring love, loss and ethics in the age of artificial intelligence.” Syracuse University, January 28 2025. https://www.syracuse.edu/stories/ai-at-ischool/
Frances, Allen and Luciana Ramos, “Preliminary Report on Chatbot Iatrogenic Dangers.” Psychiatric Times, August 15 2025. https://www.psychiatrictimes.com/view/preliminary-report-on-chatbot-iatrogenic-dangers
Georgetown University Institute of Politics and Public Service, “AI: The Promise and Peril.” YouTube, November 18, 2025. Video, 1:13:33. https://www.youtube.com/watch?v=oz3gvM7Oj3g
Klein, Ezra, “How Afraid of the A.I. Apocalypse Should We Be?” New York Times, Oct 15 2025. https://www.nytimes.com/2025/10/15/opinion/ezra-klein-podcast-eliezer-yudkowsky.html
Rosenbluth, Teddy and Maggie Astor, “Empathetic, Available, Cheap: When A.I. Offers What Doctors Don’t.” New York Times, November 16 2025. https://www.nytimes.com/2025/11/16/well/ai-chatbot-doctors-health-care-advice.html