When I played hockey, the hardest part of practice came at the end. Ice time was precious, and our coach—who was also my academic advisor—saved the wind sprints for last, when we were already spent. His reasoning was simple: the only point of sprinting was to push past what we could already do. The discomfort wasn't a side effect of the training. It was the training.
I think about that a lot now, because I work at the intersection of learning and technology, and the most important question in education right now is what we do with struggle.
Here's the thing about learning: it requires struggle the way muscle requires resistance. You don't get stronger by lifting what's easy. And for a long time, school had a reliable way of making students do the hard reps anyway—the grade. The grade was the lever. If you wanted the outcome, you did the work, and somewhere in the doing, you learned.
That lever is breaking. When a student can produce a polished essay or a worked solution without the reps, the grade stops standing for what it used to. The reflexive worry is that this is a cheating problem, or a detection problem, or a policy problem. I don't think it's any of those. I think it reveals something that was always true and easy to ignore.
The lever was never really the grade. It was whether the student saw the point.
What's striking, once you start looking, is how little of this is about ability. When I tried to name the conditions under which a student will choose the harder path—the productive struggle that actually builds capacity—almost none of them turned out to be cognitive. They were relational. A student leans into difficulty when they want to understand the material, when they believe they can, when they trust that the adult in the room cares more about understanding them than grading their output, when an honest attempt feels safe even if it earns a low mark, when the work connects to who they're trying to become. Struggle, it turns out, is gated by conditions far more than by talent. And conditions are something a school can build.
This is the work I find most interesting: not asking whether AI is good or bad for education—that's like asking whether electricity is good or bad—but asking the more specific and more useful questions underneath it. Where does a tool genuinely extend a student's thinking, and where does it quietly replace it? What does it feel like to learn something, and how do we build a culture that talks about that honestly? How do we pair the discomfort that learning requires with feedback fast enough and kind enough that students seek the struggle out rather than route around it?
I don't have all of this figured out, and I'm suspicious of anyone who claims they do. The pace of change outruns the pace of curriculum, and that gap is exactly where the thinking has to happen. But I'm convinced it's the right gap to be standing in—and I'd rather work it out in conversation than alone.