Willingham's rule shifts the teacher's task from detecting AI to checking the skill was already there. Fluent output doesn't prove that, because the model predicts tokens rather than learning.
I think Willingham is saying that students should just never use AI for schoolwork. He doesn't really address, at least in that post, how schools should go about enforcing that.
How does a school verify the skill was already there before the student opened the tool? Mostly it can’t, not from the finished work. That’s the enforcement gap, and it’s wider than detection: you’d have to watch the student perform without the tool, which is a classroom practice, not a policy.
As a cautionary tale about avarice, theft, concentration of power, self-destruction on a hyper scale, and the dangers of human hubris, addiction to novelty and marketing hype masquerading as “innovation” and “intelligence.”
I actually believe there is a third possibility. AI can sometimes function as a scaffold that makes productive cognitive effort possible when a teacher is not immediately available. When students are working independently at night and become stuck, AI can provide a hint, clarification, or guiding question without doing the thinking for them.
This matters because cognitive load and critical thinking remain at the heart of learning. We want students to experience productive struggle, but not unnecessary cognitive overload that prevents them from engaging with the task. Used intentionally, AI can reduce barriers while preserving the thinking that matters: questioning, analyzing, evaluating, making meaning, and solving problems. The goal is not for AI to remove the struggle, but to provide just enough support for students to remain in the thinking (scaffold). Thoughts?
"many students don’t see why their schoolwork will benefit them once they graduate, so 'they just want to skip through it as fast as they can'—and AI is a great way to do that. At the same time, teenagers believe that the skills that matter most for their future are the things AI 'can’t or shouldn’t replace'—like 'critical thinking' and 'reading and understanding complex information.'
It might be help to inform students that by relying on AI to do their assignments, they’re actually cheating themselves—and depriving themselves of the very abilities they say matter most for their future. Without acquiring the knowledge and skills covered in the curriculum..."
I think this is an excellent point, though I'm not sure it quite goes far enough. The best way to make kids do work themselves (not with AI) is by making the work stuff they want to do themselves. Academic skills are critical to develop, likewise cognitive mindsets, intellectual virtues and habits of mind. But the specific knowledge base - the CONTENT? There's a lot of wiggle room there. In fact, there may ONLY be wiggle room there.
When it comes to developing skills and cognitive habits, I've never seen a study that insists "students MUST have working knowledge of dinosaurs" or "absent a concrete memory of the specific dates and emperors of the Chinese dynasties, subjects saw a notable decline in results." As far back as the 1930s, national education groups were pushing for curricula that didn't mandate specific content, but rather gave students more freedom within subjects to find their own knowledge sets that they would use in pursuit of broader learning skills and mindsets. Models of this sort of Dewey-inspired learning are laid out nicely in "In Search of Deeper Learning" by Mehta and Fine.
Of course, leaning away from mandated course content (other than skills) ultimately means leaning away from high-stakes testing intended to assess mastery of said content.
Willingham's rule shifts the teacher's task from detecting AI to checking the skill was already there. Fluent output doesn't prove that, because the model predicts tokens rather than learning.
I think Willingham is saying that students should just never use AI for schoolwork. He doesn't really address, at least in that post, how schools should go about enforcing that.
How does a school verify the skill was already there before the student opened the tool? Mostly it can’t, not from the finished work. That’s the enforcement gap, and it’s wider than detection: you’d have to watch the student perform without the tool, which is a classroom practice, not a policy.
As a cautionary tale about avarice, theft, concentration of power, self-destruction on a hyper scale, and the dangers of human hubris, addiction to novelty and marketing hype masquerading as “innovation” and “intelligence.”
I actually believe there is a third possibility. AI can sometimes function as a scaffold that makes productive cognitive effort possible when a teacher is not immediately available. When students are working independently at night and become stuck, AI can provide a hint, clarification, or guiding question without doing the thinking for them.
This matters because cognitive load and critical thinking remain at the heart of learning. We want students to experience productive struggle, but not unnecessary cognitive overload that prevents them from engaging with the task. Used intentionally, AI can reduce barriers while preserving the thinking that matters: questioning, analyzing, evaluating, making meaning, and solving problems. The goal is not for AI to remove the struggle, but to provide just enough support for students to remain in the thinking (scaffold). Thoughts?
"many students don’t see why their schoolwork will benefit them once they graduate, so 'they just want to skip through it as fast as they can'—and AI is a great way to do that. At the same time, teenagers believe that the skills that matter most for their future are the things AI 'can’t or shouldn’t replace'—like 'critical thinking' and 'reading and understanding complex information.'
It might be help to inform students that by relying on AI to do their assignments, they’re actually cheating themselves—and depriving themselves of the very abilities they say matter most for their future. Without acquiring the knowledge and skills covered in the curriculum..."
I think this is an excellent point, though I'm not sure it quite goes far enough. The best way to make kids do work themselves (not with AI) is by making the work stuff they want to do themselves. Academic skills are critical to develop, likewise cognitive mindsets, intellectual virtues and habits of mind. But the specific knowledge base - the CONTENT? There's a lot of wiggle room there. In fact, there may ONLY be wiggle room there.
When it comes to developing skills and cognitive habits, I've never seen a study that insists "students MUST have working knowledge of dinosaurs" or "absent a concrete memory of the specific dates and emperors of the Chinese dynasties, subjects saw a notable decline in results." As far back as the 1930s, national education groups were pushing for curricula that didn't mandate specific content, but rather gave students more freedom within subjects to find their own knowledge sets that they would use in pursuit of broader learning skills and mindsets. Models of this sort of Dewey-inspired learning are laid out nicely in "In Search of Deeper Learning" by Mehta and Fine.
Of course, leaning away from mandated course content (other than skills) ultimately means leaning away from high-stakes testing intended to assess mastery of said content.