Beyond the Panic
Educators, rightfully so, are concerned about the impact of GenAI on their profession, myself included. The fear is that it will make it too easy for students to outsource the work of learning, leading to a decline in critical thinking, creativity, and intellectual engagement.
I understand the fear. I also think and hope it may be mostly misplaced or at least will be a springboard for us to rethink and improve our practices.
GenAI is genuinely disruptive. But the disruption is also revealing something we should have seen all along: many of our standard practices were never well-aligned with how people actually learn (looking at you 250 student lecture hall). GenAI did not create that gap. It made it impossible to ignore.
The gap
Consider the book summary. For decades, we assigned it as a proxy for reading comprehension, critical thinking, and synthesis. But what we were actually assessing was a product — a piece of text. The learning was supposed to happen in the process of writing it: the struggle to clarify your own thinking, to hold competing ideas in tension, to find the thread. Writing is thinking.
The problem is that a student can now generate that product in seconds. And many do, not because they are lazy, but because (for the most part and contexts) the assignment never made the process visible or valuable in the first place. The task was always framed around the deliverable, not the doing.
GenAI did not break the book summary. It exposed that the book summary was already broken — or at least, that we were grading the wrong thing.
What AI cannot replicate
Here is what I keep coming back to. Learning is an active, constructive, and deeply human process. You cannot learn to solve problems without solving problems. You cannot learn to write without writing. The struggle is not an obstacle to learning — it is the learning. As Andrew Heiss put it recently, “the conditions that produce the most errors during acquisition are often the very conditions that produce the most learning.” LLMs, by design, eliminate that friction.
Learning is also social. It is students in a room, challenging each other’s assumptions, building on each other’s ideas, working through confusion together. And it requires a why — not just grades, but the intrinsic experience of mastery and the personal relevance of what you are creating.
AI can produce a final product. It cannot replicate the cognitive or emotional experience of getting there. That distinction is everything.
Thinking with, thinking of, thinking through
Gavriel Salomon and David Perkins offered a framework decades ago that I find myself returning to constantly. Technology, they argued, has three cognitive effects: you can think with it (partnering with a tool to accomplish a task), think of it (the cognitive residue left behind after the tool is gone), or think through it (the tool fundamentally reorganizes how you understand something).
For a veteran teacher who has designed thousands of lessons, using GenAI to brainstorm is a gift. That is thinking with — the expertise is already there, and the tool amplifies it. But for a novice teacher or a student still developing those skills, the risk is different. If the tool does the cognitive work before you have internalized it, you end up with what researchers have called cognitive debt: the illusion of understanding without the substance. That is the dark side of thinking of.
This is why the lower-order versus higher-order distinction matters so much. AI is excellent at offloading lower-order production tasks. The danger is when we let it offload the higher-order thinking that was supposed to be the whole point.
What changes, what stays
Our practices have to change. We can no longer design assignments where the product is the point. We have to assume AI is being used and design around it — not by detecting it, but by making the process matter more than the deliverable.
Our theories do not need to change. Active learning, constructivism, situated cognition, transparent pedagogy — these are not outdated. If anything, GenAI makes them more urgent. We need to explain to students why we are asking them to struggle. We need to design tasks where the doing is the learning, where the process is authentic enough that delegating it to a machine would feel like giving away the interesting part.
I have seen this work in my own game design classes. When students are building their own games, they put in the effort because they are intrinsically motivated to play and improve their own creations. The concept of “cheating” becomes irrelevant — not because we solved it, but because the task made it beside the point.
The role that remains
GenAI is a mirror. It forces us to ask whether our educational practices actually reflect what we know about how people learn. In most cases, the honest answer is: not as well as we thought.
The foundation of education has never been the technology. It is the relationship between a good teacher and a curious student. That relationship is now more important, not less. Our job is to design learning experiences so engaging and authentic that students would not dream of outsourcing the thinking to a machine.
A teacher benefitting from offloading lower-order tasks will hopefully have more time and energy to focus on that relationship. A student who has internalized the process of learning will be more motivated to engage with it, not less. Maybe there is blessing in the disruption after all.