When I saw the headline title for the online version of Hua Hsu’s New Yorker essay “What Happens After A.I. Destroys College Writing?” I was ready to hate. Destroys college writing?!? Have these people not read my blog posts about this?? What the hell does this guy think he’s talking about?
Interestingly enough, Hsu’s essay, as published in the July 7 & 14, 2025 print version of the magazine, has the nondescriptive title “The End of the Essay.” I think that says a lot about the differences between the two publishing formats– online and clickable versus on paper, just like it was 100 years ago. Neither headline is right because Hsu is not writing about college writing assignments “ending,” let alone being destroyed. Rather, this is more about how AI is challenging the college experience, including the essay assignment, and about the anxieties of both teachers and students around these changes. I think a more accurate title might be “AI is Changing and Complicating College Writing and Learning Itself,” but that’s not exactly a clickable link, is it?
I think it’s a really good read because it taps into the anxieties that both teachers and students have about AI and its role in college (especially in writing classes) without characterizing students as constant cheaters and teachers as all hopelessly out of touch and unwilling to change. And I think it also highlights the problem of an overemphasis on the technology of education at the expense of actually trying to learn something.
Hsu is an English professor at Bard College and the author of Stay True, which won the Pulitzer for Memoir in 2022, and he’s been a writer for The New Yorker since 2017. He’s a talented and accessible writer, plus he has also “been there” as someone who has had to deal with AI in his own teaching. Though it is worth mentioning that he realizes that the teaching situation he has at Bard, where “a student is more likely to hand in a big paper a year late (as recently happened) than to take a dishonorable shortcut,” might be a little different from those of us teaching at less elite institutions. He interviews several students on their experiences in using AI in classes, some of which are pretty straight-up cheating, but a lot of which is not– or probably not. One of the students he interviewed used AI more or less like Google as a search tool, and others talked about how they used AI more as a study tool in ways I try to teach in my classes.
Hsu also talks to faculty, some who are returning to handwritten blue books for exams, including the University of Virginia media studies scholar Siva Vaidhyanathan: “Maybe we go all the way back to 450 B.C.” (For what it’s worth: blue books are a bad idea, and I think Elizabeth Wardle articulates why in an excellent op-ed in The Cincinnati Enquirer called Students aren’t cheating because they have AI, but because colleges are broken,” a commentary that is similar to Hsu’s, but with a more accurate title). But he also talks to Dan Melzer, the director of the first-year writing program at the University of California, Davis, about what pretty much everyone in my field has advocated for the last 50 years, teach writing as a process. He talks about the many problems of higher education nowadays– the costs, the constant assessments, the shifting perceived values of different majors and higher education itself.
The only thing I wish is that there was citation for the various studies Hsu notes. Though perhaps that will be part of what I will ask students to do when I assign this in the fall, to track down Hsu’s evidence. I’m (probably) going to be assigning this in the fall in my classes, and not just because my classes this fall are going to be about AI.1 Actually, I think everyone who teaches in college– and that is especially true in fields like mine– needs to have “the AI talk” with their students.
I’ve informally polled my students over the last couple of years and asked, “How many of your professors in your other classes have said anything about AI?” A few students told me they were actively using AI in their courses, and there was more of that last year than the year before. A few students told me that their professors have forbidden them to use AI. When I asked those students if they thought their professors could actually tell if they were using AI or not, they generally shrugged. But in most cases, certainly more than half of my students said their professor didn’t say anything about AI at all.
By “the talk,” I’m thinking about sex rather than the conversation Black American parents have with their children about racism. But really, I mean talking with someone about something potentially embarrassing and uncomfortable for everyone, not unlike telling a student that they failed.
I can understand why a lot (most?) professors do not want to have the AI talk. It makes us vunerable. Most professors don’t feel like they know enough about AI, they don’t want to look like idiots or as hopelessly out of touch, and also because AI is scary. I also think that a lot of professors think that refusing and ignoring is enough: that is, just tell your students don’t use AI because it’s bad, m’kay?
Again, that’s what I like so much about Hsu’s essay. It’s a good starter for “the AI talk.”
- I’m teaching an advanced special topics class called “Writing, Rhetoric, and AI,” and the two sections of first year writing I’m teaching this fall will have “your career goals and AI” as the research topic. ↩︎