There’s an old joke that certain education reformers enjoy sharing about today’s classrooms to illustrate how resistant teachers and schools have been to innovation and technology: After a 100-year slumber, Rip Van Winkle awakens and is baffled by what he sees in modern society: from cars to smartphones to operating rooms, nothing is recognizable. But a school is the one place he recognizes.
There are multiple problems with the joke (one I know I’ve made myself—mea culpa). But put aside the minutiae like the length of slumber and that age-graded schools with multiple classrooms and chalkboards weren’t around when Rip took his fictional nap. The biggest problem is that today it gets part of the narrative backward.
For the first time in my memory, schools and teachers are aggressively adopting the latest technology of artificial intelligence.
K–12 schools are just about saturated with AI. Just a few examples: Some estimates suggest Google Classroom is used by nearly 70% of US schools, and Google Classroom integrates a variety of Gemini-powered AI tools. Roughly 700,000 U.S. teachers use MagicSchool, an AI platform that automates tasks such as lesson planning, creating materials, writing and grading assessments. And if you ask teachers and students directly, the polls tell you that over 50% of both groups say they use AI for school. The older the grade level, the more teachers and students say they use AI.
Schools may even be leading in adopting AI compared to many other industries outside of tech. That’s a mind-blowing statement, yet there are reasons to believe it’s true, based on various reports on AI adoption.
That it’s an even plausible statement is actually a big problem in its own right.
In a rush to use AI, educators are unwittingly realizing efficiency gains in a schooling model that was never built to optimize learning.
Today’s dominant schooling model that was built for its own reasons of efficiency—as the most economical way to universally educate students. Sorting students into grades by age, tracking students into fixed groups based on point-in-time checks of ability, and assuming that each group would progress at the same rate were reasonable sacrifices to make in the industrial era, when teacher-delivered content and whole-class instruction were the dominant methods.
This model created a zero-sum mindset in which resources are assumed to be scarce, and that for every winner, there must be a loser. Educators preach about growth mindset and grit to children, but the system itself fails to encourage perseverance and curiosity. Instead, it does the opposite by affixing labels to students, sorting them into relatively static groups, and signaling to the students that their effort doesn’t matter.
The result is that by age 18, with most of their lives still ahead of them, the vast majority of students have already been told that they rank “below” the top, and they’ve already been labeled as “not good enough” for certain pathways.
Using AI in this model has had the effect of merely reinforcing the current system’s flaws—from its zero-sum nature to its incoherence. Lesson planning tools and the like don’t solve these problems. They perpetuate them. It’s also dangerous for students operating in a zero-sum education system, who have many incentives to use AI to automate cognition and avoid the hard, effortful work of learning.
In today’s world, which prizes intellectual capital and works best when all individuals cultivate their passions and develop their full human potential, this sorting system no longer suffices.
And here’s the central flaw of the Rip Van Winkle story.
Despite all the AI adoption, the fundamental zero-sum classroom model still persists in schools. It’s a flawed model that needs to go—and one that this fictional figure would surely still recognize.
