Every few weeks, another think piece lands: either AI is going to destroy education as we know it, or it’s the most transformative learning tool since the printing press. Teachers are in the comments arguing. Administrators are forming committees. The debate has a certain energy to it — the feeling that something important is being decided.
But I’ve started to think we’re arguing about the wrong thing.
The Argument We Keep Having
The debate, roughly: Should AI be allowed in schools? Should students be permitted to use it on assignments? Is it cheating? Is it a tool? Where’s the line?
These are real questions and I don’t want to dismiss them. But they share a premise that I think is worth questioning: that the primary issue is whether AI should be in schools.
It’s already in schools. It’s in students’ pockets. It’s in the phones they take to the bathroom during your class. The question of whether to allow it is, in a practical sense, settled — not by policy, but by the world students are already living in.
What’s not settled is what we do about that.
The Question We’re Not Asking
Here’s the conversation I think we need to be having instead: What is school actually for, and does that still make sense in a world where AI exists?
That’s harder and less satisfying than the ban-or-allow debate. It requires us to revisit things we’ve taken for granted — the structure of assignments, the role of memorization, the purpose of the essay as an assessment tool. It requires uncomfortable honesty about whether some things we’ve been assessing for decades were ever really measuring what we thought they were.
It also requires something harder than policy: genuine thought about human development and what learning is supposed to do for a person.
What AI Makes Visible
Here’s the thing I’ve noticed in my own classroom. When AI can produce a competent five-paragraph essay in thirty seconds, the five-paragraph essay becomes obviously what it always was: a formatting exercise. The form was useful as long as it stood in for the thinking. Now that the form is separable from the thinking, we can see that the thinking was never guaranteed by the form.
That’s uncomfortable. But it’s also clarifying. It pushes us to ask what we actually want students to walk away being able to do — and then to design for that, rather than for the shape of things that used to signal it.
The AI debate, when it’s at its best, is really a design challenge. What assignments require genuine thinking? What does it mean to demonstrate understanding in a world where the surface performance of understanding is cheap? These are good questions. They’re the ones teachers should be sitting with.
What I Tell My Students
I don’t ban AI in my class and I don’t treat it as neutral. I treat it as a tool with costs and benefits, like most tools. Using it to skip the thinking is a bad deal — not because of academic integrity, but because you’re paying for an outcome while skipping the part that actually changes you. Using it as a thinking partner, to test your ideas or find the holes in your argument, can be genuinely valuable.
What I care about is whether students are developing as thinkers and writers. I can usually tell. The work that comes from actual engagement looks different from the work that doesn’t, and over a semester the gap widens.
The Harder Argument
The wrong argument is the one that keeps us focused on catching and preventing. The right argument asks what we’re building — what kind of human capacity we’re trying to develop in students, and how AI fits into that goal rather than threatening it.
That’s a harder conversation. It doesn’t end with a policy. It requires ongoing judgment from teachers who know their students and their subject and are willing to keep asking whether what they’re doing is actually working.
But it’s the conversation that’s worth having.
This post draws on ideas from The AI Doesn’t Know Your Students, Chapter 13: “The Wrong Argument.” The book is available now.
