Done! Finished!
One might expect to hear such exclamations from exultant college students, relieved or ready to rejoice upon polishing off their latest essay assignment. Instead, these are the words I hear with increasing frequency from fellow professors who have come to think that the out-of-class essay itself is now done. It’s an antiquated assignment, some say. An outmoded form of pedagogy. A forlorn fossil of the Writing Age, a new coinage that seems all too ready to consign writing instruction to extinction.
As a new director of my college’s faculty development office, I’m privy to ongoing conversations about the teaching of writing, many of which are marked by frustration, perplexity and pessimism. “I don’t want to read a machine’s writing,” one professor laments. “I don’t want to police student essay writing for AI use,” another asserts.
Kevin Roose, a tech writer for The New York Times, who recently visited my campus, has suggested that the take-home essay is obsolete, asking, “Why would you assign a take-home exam, or an essay on Jane Eyre, if everyone in class—except, perhaps, the most strait-laced rule followers—will use A.I. to finish it?”
Whether this situation is entirely new is arguable. For decades, we’ve had online resources that might make independent student reading unnecessary, yet we haven’t stopped assigning out-of-class reading. If I assign a rigorous novel like Charles Dickens’s Bleak House, I’ve long known that students can access an assortment of chapter summaries online—CliffsNotes, SparkNotes, LitCharts and others, all of which might make unnecessary the intellectual work of deciphering Dickens’s 19th-century sentences or wading into the deep waters of his sometimes murky prose. Maybe, as a recent New York Times piece about Harvard University students not doing their reading suggests, students aren’t doing that kind of homework, either.
Still, being able to create sentences, paragraphs, essays and research papers with a single prompt—or now, having “agentic AI” engineer an entire research process in a matter of minutes—seems different from googling the plot summary for the first chapter of Bleak House.
Maybe writing via LLMs is different because it’s not just about summarizing someone’s else’s idea; it’s about asking a machine to take the glimmer of one’s own half-hatched idea and turn it into a flawless, finished product. Somehow that process seems a little more magical, like being able to create a novel or a dissertation with a Bewitched-like twitch of the nose.
Further, the problems with out-of-class writing are different from those linked to out-of-class reading because of how embedded AI has become within the most basic writing tools—from Microsoft’s Copilot to Grammarly. With tools that blur the boundaries between the student and their “copilot,” students will increasingly have difficulty discerning what’s them and what’s the machine—to the chagrin of those who do want to develop autonomous intellectual skills. As high school senior Ashanty Rosario complained in an essay in The Atlantic about how AI is “demolishing my education,” AI tools have become “inescapable” and inescapably seductive, with shortcuts to learning becoming “normalized.”
In this world of ubiquitous AI shortcuts, how do we encourage students to take the scenic route? How do we help them see, as John Warner reminds us in More Than Words: How To Think About Writing in the Age of AI (Basic Books, 2025), that writing is an act of embodied thinking and a tool for forging human community, linking one human being to another? How do we encourage them, to use the language of Chad Hanson, to see their written assignments as “investments, not just in the creation of something to turn in on a deadline, but rather, investments in your humanity”? In an Inside Higher Ed essay, Hanson describes how he tells students, “When you give yourself time to use your faculties, you end up changing the dimensions of your mind.”
But there’s the rub. Writing takes time. Teaching writing takes time. The practice of writing takes even more time. If there is still value in the time invested in developing human writing skills, where is the time to be found within the constraints of traditional writing courses? Writing practice used to take place primarily at home, on student PCs and notepads, over hours, days and weeks. Now that student writing is being chronically offloaded to a magical deus ex machina, Roose asks why teachers wouldn’t simply “switch to proctored exams, blue-book essays, and in-class group work”?
As a writing professor, my answer is: There isn’t time.
Shifting writing practice from a largely out-of-class endeavor to an in-class one doesn’t provide students with the time needed to develop writerly skills or to use writing as a mode of deep thinking. Nor does it allow for both instruction and sufficient hands-on practice. At my college, courses typically run either three days per week for a short 50 minutes per class or two days per week for 80 minutes. Even in a “pure” writing course, such time periods don’t allow for students to have the sustained practice they would need to develop skill as writers. The problem is even worse in writing-intensive courses for which a significant amount of class time is needed for discussing literary history, philosophy, political theory, religion, art history or sundry other topics.
The solution I propose is to invest more rather than less in writing instruction: Just as we require labs for science lecture courses, we should provide required “writing labs” as adjuncts to writing classes. Here I don’t mean a writing lab in the sense of a writing center where students can opt to go for peer assistance. By writing lab, I mean a multihour, credit-bearing, required time during which students practice writing on a weekly basis under the supervision of the course’s instructor or another experienced writing teacher. Such labs would be time in which students develop their autonomous critical thinking skills, tackling assignments from conception to completion, “cloister[ed]” away, as Niall Ferguson puts it, from dependency on AI machines. And if writing “lab” sounds unduly scientific for the teaching of a human art, call it a weekly workshop or practicum. (Yet, even the word “laboratory” derives, via medieval Latin, from laborare, which simply means “to work or labor.”) Whatever the name, the need is real: Writing cannot be taught without student labor.
The problem I am addressing is a critical one, with too few alarms being sounded in higher education circles, despite the plethora of articles about education and AI. Even as colleges tout writing skill as a major outcome of college education, I fear that writing education may quickly fall between the cracks, with out-of-class writing being abandoned out of frustration or despair and insufficient in-class time available for the deep learning writing requires. Quiet quitting, let’s call it, of a long-standing writing pedagogy.
If colleges still wish to claim writing skill as an important learning outcome, they need to become more deliberate about what it means to educate student writers in the age of AI. Toward that end, colleges must first reassert the importance of learning to write and articulate its abiding value as a human endeavor. Second, colleges must devote professional development resources to prepare faculty to teach writing in the age of AI. And finally—here’s the pith of my argument—colleges need to restructure traditional models of writing instruction so that students have ample time to practice writing in the classroom, with a community of human peers and under the supervision of a writing guide. Only in, with and under those circumstances will students be able to rediscover writing as a true labor of love.
