
Today I published my first lengthy, public reflection on what I would call the AI crisis in higher education. I cover the extent of the problem, the difficulties professors face, and the real risks of proliferation of AI cheating (namely to the moral and intellectual development of our students). I would appreciate it if you gave that piece a read over at Fusion, and let me know what you think in the comments.
I want to briefly touch here on something I think that I left out of the piece. In the broader conversation around AI use in schools, one consistent theme has emerged from AI-positive folks critical of the panic I and others express. That theme is: if your assignment can be performed by AI, it was a useless assignment to begin with. If a machine can write a perfectly serviceable B-minus or C-plus level paper, and the student would have done the same or worse unaided, then the assignment itself was broken. The problem is not with the new machines, these people argue, but instead with outdated and outmoded forms of education that ask students to do silly things like write sub-par essays on topics already exhausted in academia.
But this critique misses both the point of assigning essays and the point of human life more generally. Yes, it truly is that far off the mark.
First, on the purpose of essays. When I ask my students to write an essay assessing a text, or summarizing a thinker’s ideas, or comparing and contrasting two political ideologies, it is not because I think the student will (at least immediately) produce some never-before-conceived insight into these texts, thinkers, and ideologies. It is not because I even expect them to have a well-developed understanding of the same. The process of writing itself is developmental, the process of making an argument and supporting it with evidence itself a type of education.
I occasionally tell my students that I am going to let them in on a secret, give them a peek behind the curtain: in how I design my coursework, my goal is for them to always read the text at least three times. They should read the assigned work in preparation for the course. Then, in class, in the course of our discussion, we approach the text together, reading relevant sections and interpreting them in group discussion. Then, when they begin their writing assignments, I hope that they will again pick up their texts as they seek to explain to the reader why they have interpreted the text as they have.
This entire process is educative. Students do not learn completely or well simply through one exposure to an idea, a single presentation of a single fact on an isolated PowerPoint slide (this is empirically provable as well as intuitive). Frankly, I don’t care if students understand a text badly the first time they read it. I don’t care if they make mistakes in class discussion. I don’t care if they write a below-average paper. I care that at each of these stages, they are putting in the intellectual effort in educational community to come to a better understanding of the ideas under discussion.
This brings me to my second point: those who say “if an AI can do the assignment, it’s a bad assignment” miss the entire purpose of human life. I recently read a wonderful collection of lectures and essays by James V. Schall entitled On the Unseriousness of Human Affairs: Teaching, Writing, Playing, Believing, Lecturing, Philosophizing, Singing, Dancing. In one of his essays, entitled “On the Mystery of Teachers I Never Met,” Schall discusses Chesterton’s quip that “If a thing is worth doing, it’s worth doing badly.” Chesterton goes on, Schall says, to say that things worth doing badly are dancing and writing “one’s own love-letters.”
I myself have taken dance lessons in the past. I do not, however, consider myself a particularly good dancer. Nevertheless, when we attend weddings or parties or events with time for such things, my wife and I take great joy in being mediocre dancers together. In this unserious life we lead, I would never dream of replacing my role in the dance with my wife with a robot who could more capably lead her through the dance. Similarly, I would never dream of asking ChatGPT (or, heaven help us, DeepSeek) to write my wife a love letter, and I would expect her to slap me if she found out I had done so. The point of writing a love letter is not to complete a perfunctory task, and it is not made better if someone else does it for me, nor is it made obsolete if something else could do it just as well as I could.
Of course, it’s equally true that if something is worth doing, it’s worth doing well, and this is part of Schall’s (and Chesterton’s) point. To do something lovely and worth doing badly is the first step on the educational path to doing it well. I would not even begin to dance without lessons, and I could not improve in my dancing if I did not begin to dance badly.
Similarly, I expect most of my students would never voluntarily read Plato’s Republic or the ethics of Aristotle if I did not assign it to them in class. I expect, too, that they would not re-read those works if I didn’t compel them to bring their texts to class. I expect, too, that after bringing their texts to class, they would remain forgotten on their shelves if I did not compel them to provide me with accurate citations to textual evidence in their writing. But if a teacher does not make them do those things, even to read, discuss, and write perhaps badly at first, then they will never become good readers, good conversation partners, or good writers.
This, then, is what my critics miss: That some one or some thing exists out in the world that can do an assignment I give my students better than they can do it themselves does not, in any way, make the assignment less worth doing. You might as well say that Shakespeare’s existence ought to keep me from writing poetry, or the existence of virtuosos keep me from learning an instrument. Education is not exclusively about content and information, which can be had in greater quantity and quality from the library than in anything a student will produce in my class. Education is instead a shaping of the moral and intellectual character of the student, and for that, there can be no cheap machine substitute.
"I care that at each of these stages, they are putting in the intellectual effort in educational community to come to a better understanding of the ideas under discussion."
Creating opportunities for intellectual effort is the whole point of almost any task I create for my classrooms. The extent to which any tool (calculator, LLM, word processor, piece of paper) is useful in a classroom is the extent to which it empowers or enables the intended focus of the intellectual effort. Using AI in the classroom is positive if it supports the intended intellectual effort. It it takes the place of the intended intellectual effort then it only cheapens the education and narrows the learning taking place.
It's the same with art, as I wrote elsewhere... I think in general, AI-mongers and devotees are only focusing on the results. They are ignorant of, or choose to ignore, the other, far more important part of the creative process: the journey that results in an improved human being.