by Anna Muller
Why I Didn’t Assign a Final Essay?
This semester (Fall 2025), I decided not to assign a final essay. Over the past year, I’ve read many student papers whose authorship I quietly questioned. I let it go—what else could I do? Maybe they used AI, maybe they didn’t. Maybe it only helped them polish their grammar, or maybe it wrote the piece for them. I couldn’t know, and I didn’t want my suspicions to seep into my relationship with the class. AI has brought with it a new layer of scrutiny, a quiet erosion of trust, and fresh doubts about what education is—and what it should be. To be honest, it has taken some of the joy out of our work. Grading has never been the most delightful part of teaching. Rather than that it has always been a moment of reckoning: a test not only for students, but for us. Did we manage to teach them anything? What did we truly accomplish? Where did we fall short? Are we still doing what we believe education is meant to do?
AI complicated it even further.
So I didn’t assign a final paper this semester. Partially out of cowardice, partially out of fear. I understand that AI is here to stay, and that it will only get better. I even understand the arguments that AI can help bring more equity to our classes. I’m not a native speaker of English, and writing in this language is a challenge. Proofreading is expensive, time-consuming, and sometimes discouraging. AI can make my life easier and I sometimes use it. For example, it helps me with articles (which don’t exist in my native language), with syntax, with word choice. It eases anxiety, builds confidence, and saves time. I understand its value. But I also understand that by shortcutting the process, I do myself a disservice. Even small practices, like figuring out how grammar works or struggling to choose the right article, are part of that process; they teach us how language shapes thought. The effort to find the right word, to shape a sentence, to pause and revise—is not just linguistic labor; it’s intellectual formation. When I let AI fill in the gaps too quickly, I risk losing that space where ideas clarify, where uncertainty pushes thought forward. The discomfort of writing is part of what teaches me to think.
Freire, AI, and the Loss of Dialogue
Paulo Freire, a Brazilian educator, argued years ago that education works best when teachers and students engage as equals in a shared process of learning—one grounded in dialogue, conversation, and mutual inquiry. Freire critiqued what he called the “banking model of education,” in which teachers deposit knowledge into students’ minds, and students receive it passively. In this model, learning becomes mechanical and hierarchical — students memorize and reproduce information rather than engage critically or transformatively with it. Freire argued that such a system reinforces oppression by discouraging agency, curiosity, and creativity. Now, when AI “does homework” for students, something similar happens. Instead of engaging in a dialogical process of discovery — the kind of questioning, reflection, and meaning-making Freire valued — students risk returning to the very passivity Freire resisted. AI can, in a sense, become a new “banker,” depositing ready-made answers, polished prose, or summarized ideas into students’ hands. The result might be smoother and more efficient, but it circumvents the struggle, error, and reflection.
And here I think I should make a disclaimer. I’m not saying that most students will cheat — but many will, and many of those will do so unwillingly. Pressed for time, stressed, and anxious, some will reach for an easier solution because life gets intense, because they are overwhelmed. I don’t think this is surprising, or even entirely avoidable. Life often pushes us to do things we’re not proud of, and with time, those things can become normalized. Reaching for AI becomes second nature — but in doing so, students bypass the essence of learning, which should be about figuring things out slowly, about the process itself.
Close Readings and Slow Learning
So, going back to the final essay—I simply couldn’t bring myself to assign it this semester. Instead, I gave more quizzes and exams. I even considered oral exams but haven’t yet taken that leap. My Polish education relied heavily on oral examinations, and I turned out fine. And I’ll probably turn to them one day. But I also wanted my students to write—to practice the act of putting thoughts into words, even if writing is no longer the rite of passage in education it once was. To do that, I introduced written, in-class exercises that I call “close readings” in my longer 2-hour-and-45-minute class. I’ve done three of these this semester so far. For the first one, I handed out an excerpt from Crime and Punishment by Fyodor Dostoyevsky (yes, I asked them to read the entire book) and asked the students to pull out their pens and pencils. On the margins, they responded to a guiding question I posed—about motive, reason, causation, morality, and life choices. Then they summarized their reflections in a short paragraph. The exercise took about an hour. Was I pleased with the results? Not completely. The novel is dense and difficult, and we only scratched the surface in class prior to this exercise. But I was pleased by something else—the intensity of their engagement, the visible effort of thinking, the work of both brain and hand. It was tactile—paper, pens, presence—the kind of learning experience AI can’t replicate.
Recently, I took them to the Detroit Institute of Arts to see Diego Rivera’s mural. We had read about Rivera and Frida Kahlo in class, so the visit offered a perfect opportunity to reflect on their intertwined lives and artistic visions. I imagined the students sitting quietly before the mural, letting its scale and detail wash over them, choosing one fragment to contemplate and respond to in writing. And write they did. It wasn’t quite the scene I had pictured—there’s no sitting in the middle of that Rivera’s court. And they didn’t want to sit on the floor, so they found corners to perch in, hiding a little from view. Still, there was something beautiful in their quiet focus.
I realize that what I’ve written so far doesn’t really address the assigned topic for the final essay. My in-class writing exercises aren’t proper final essays or research papers—and I’m not sure if I’ll ever figure out what to do with those.
Why I Don’t Want to “Play With” AI in the Classroom?
But there are a few more things I want to add. First of all, I am not sure I need to experiment with AI in my classes. In most cases, students know it better than I/we do. A few times I brought to class tables or short surveys created by AI, and they immediately called me on it—and even told me how to make it better. Young people are smart, often remarkably thoughtful, and they adapt to digital developments much faster than I ever could. Instead of inviting more AI into the classroom, I would like to slow down its use. And if that means sacrificing the traditional at-home final essay—at least until we prepare ourselves better to deal with the present moment—I’m ready for that.
Second, I want to acknowledge that there are students who consciously refuse to use AI. Some do so because they believe in the importance of figuring things out on their own; others worry about the broader effects of AI, such as its impact on the environment. We need to create spaces that recognize and honor this kind of thinking as well.
I am embracing slow learning.
Funny thing—I was talking about this with friends over dinner. I called it slow learning, only to discover later that night that it’s actually a thing, something more and more people are embracing. One of the manifestos of slow learning notes that recent years have fostered the belief that “learning can be increasingly optimized for efficiency.”[1]
Looking Ahead
So let me summarize and return to the idea of a final essay. In the future, I would like to structure the course so that close reading exercises build toward a single, more developed paper, maybe the one that students write with a guided use of AI. If they walk via the various steps on their own, such as reading and reflecting on primary sources, and engaging a bit with secondary ones. I am not sure. What I am sure of though is that we cannot afford embracing AI fully. We should resist wherever possible, because we do not fully understand its potential impact on the student and us. And their young brains are too precious to submit to this experiment wholeheartedly. There is not much we can do to help them in the process, but slow down, to think through their ideas without the mediation of a screen. I believe in handwriting, in going back to the roots. The world will teach them to move faster and more digitally; I would like to nurture their appetite for something slower, more deliberate, more manual.
I am not entirely sure what this will look like yet. What I am sure of, however, is that we cannot afford to embrace AI uncritically. We should resist, whenever possible, because we do not yet understand its full impact—on students, on us, and on the learning process itself. Their young minds are too precious to surrender to this experiment wholesale. The world will teach them to move faster and more digitally; I would like to nurture their appetite for something slower, more deliberate, and more tactile.
[1] Thank you, Coquis Galvan Santibañez, for sharing Slow Learning with me. And thank you, Carla Vecchiola, for sharing this article (The Case for Slow Learning) with me.
[2] Diego M. Rivera, Sketch Related to North Wall, Automotive Panel, Detroit Industry Murals, 1932, graphite pencil, pen and black ink on off-white wove paper. Detroit Institute of Arts, Gift of A. Alfred Taubman, 2010.212.
Dr. Muller is a Professor of History at the University of Michigan-Dearborn. From 2010-13, she worked for the Museum of the Second World War (in Gdansk) as a curator responsible for the sections on concentration camps, the Holocaust, and eugenics. Her book on the life of women in prison cells in post-war Poland, titled If the Walls Could Speak, was published in 2017 with Oxford University Press.
