“There’s a broad and increasing sense from students that something is being stolen from them,” observed Elmira College English professor Matt Seybold, in a recent article in The Guardian on AI’s impact on student learning. That article showcases an emerging consensus in many humanities fields that a host of negative consequences follow from the uncritical embrace on college campuses of AI (which for our purposes here means Large Language Models). Such a critical perspective on AI is sorely needed at OU, where AI’s adoption is often taken to be both inevitable and beneficial. We reject both assumptions. In what follows, we briefly review the body of AI-skeptical research and writing and then share how we’ve created AI-free spaces for our students.

The Limits of AI

Too often, discussions of AI within higher education set aside, or ignore, the broader planetary and social harms of this new technology. Yet those harms are well-documented. Widespread AI adoption is a disaster for the environment, driving new investment in fossil fuel extraction. AI data centers consume massive amounts of water and gobble up enormous swaths of earth. AI models are built on stolen data and stolen labor, perpetuating “AI colonialism.” AI deepens systemic racism and threatens to destroy democratic civic institutions. AI intensifies income inequality, enriching a handful of (often far-right-wing) tech oligarchs, who profit from the extraction of a generation of young people’s personal information and attention. AI places students on college campuses under nearly constant surveillance. 

AI, moreover, brings a particular set of problems into the college classroom. Recent studies have shown that AI use has detrimental effects on human cognition and erodes brain function. On college campuses, students are using AI to avoid the cognitive strain and hard thought associated with reading and writing. This avoidance stifles the development of students’ linguistic capacities–their abilities to read and to write, to process and manipulate language, and thus to think independently and creatively. Many students already understand this. Such concern with preserving students’ rights to intellectual growth is what has led us to develop classroom practices to help them learn to think without AI.

Analog Classrooms

A starting point for us has been to cultivate analog classroom communities. Our classes have no screens: no phones, tablets, or laptops (notwithstanding, of course, appropriate accommodations for students who need them). Students read printed books and take notes by hand. We share with them our reasons for this choice, and the research supporting it (on the advantages of print over digital reading, and on the learning benefits of taking notes by hand). Our students have responded enthusiastically, feeling unburdened by their devices, and we’ve been excited to see how they talk more freely and happily with each other before, during, and after class. Such an analog classroom, moreover, makes it easier for students to learn to think without AI.

Reclaiming Attention

Thinking requires focus and attention, but AI threatens to further deplete our attentional capacities. One step towards creating AI-free classrooms, then, is helping students reclaim their attention—refusing to let technology have (and monetize) it. Sustained attention is necessary for a functioning democracy, of course, as well as for thoughtful deliberation, problem-solving, listening, understanding, cultivating social relations and friendships, and simply for reading well. In class, we make attention an explicit topic of discussion (sometimes by assigning portions of Johann Hari’s Stolen Focus).

Just as importantly, we devote class time to exercises that cultivate habits of paying attention; we practice sustained focus. We might begin class by reading silently for 5 minutes—then 7, then 10 as the semester proceeds. At other times, students might pair with a classmate and engage in casual conversation for a designated period of time. Weather permitting, we might ask students to take a walk on campus, phone-free, and notice specific features of the built environment, or observe a particular tree or plant, then reflect upon these experiences. 

To let students grapple with their ideas and prepare for class discussions, free from the temptations of AI-generated summaries, our students do ungraded in-class writing. These exercises encourage students to reconnect with texts, the world, and one another without distraction. They also equip students to engage in the work of the class, whether that means watching a 30-minute documentary film without interruption, lingering over a line of poetry, or exploring the richness of a metaphor. Rather than hurrying towards “answers,” students learn the pleasures of discovery, the rewards of working through their own perplexity, and the power of trusting their own minds.

Seeing Literacy as Power

When we help students reclaim their attention, they’re well-positioned to think directly and critically about AI. We’re particularly concerned with helping students understand the links between AI usage and literacy, and between literacy and power. As professors of American literature, we find African-American abolitionist writing to offer an especially powerful archive of reflections on literacy and power. Writers like Frederick Douglass and Frances Ellen Watkins Harper spotlight how slavery and racial inequality could only be sustained by circumscribing the literacy rights of enslaved and formerly enslaved people, and how acts of reading and writing allowed Black Americans to reclaim their agency. Douglass, for instance, cast literacy as the “pathway from slavery to freedom,” and Harper recounted how learning to read led a once-enslaved woman to feel “as independent / As a queen upon her throne.” Such work brings into focus just how intimately related linguistic facility and worldly power are.

While taking care to avoid collapsing the distinctions between the situations of these writers and our students, we ask our classes what this literature helps us notice about the world we live in. What pathways to freedom get foreclosed via the linguistic and cognitive atrophies that attend generative AI use? Can a regular user of, say, ChatGPT feel the independence of a queen? To be sure, U.S. abolitionist literature need not be a singular reference point for thinking about AI, language, and power; other bodies of writing could occasion no less meaningful questions. What we’re suggesting here is not that students must read one particular text or another to think about AI’s liabilities but rather that writing about literacy and power from the past like Douglass’ and Harper’s can defamiliarize the present and aid our students in thinking critically about it.

You Can Say No

The examples of Douglass and Harper also remind us of what is at stake in offloading cognition: freedom. The integration of AI across digital platforms and devices has largely taken place without users’ consent. Universities have participated in this coercive project, whether by contracting with tech companies to embed AI into learning management systems, or by encouraging its use in the name of “AI literacy,” or simply by capitulating to the claims of inevitability advanced by those who stand to profit from AI’s integration into every sphere of life.

But for us it’s important to remember that students and faculty alike are still free to say no. Rather than acquiescing to technofeudalism, universities can (and should) be the places where students learn how not to use AI–-where they learn, rather, how to exercise agency by insisting upon their own humanity. Like so many faculty, we remain dedicated to this ideal of college, in which education is not just preparation for the many careers our students will hold but also, even more fundamentally, preparation to lead the rich and flourishing lives they deserve.

About the Authors

Timothy Donahue is an Associate Professor of English at OU. Jeffrey Insko is a Professor of English at OU. 

Others may share and adapt under Creative Commons License CC BY-NC.

Receive Weekly Tips view all teaching tips Submit a Teaching Tip