The Illusion of Free Will in the Automated Classroom

I've been wondering: when an algorithm can generate the essay, solve the equation, and summarise the book in three seconds, what is actually left for the student to become? I don't mean what's left for them to submit, the software will happily handle the grade, I mean what happens to the mind sitting behind the screen.

Ask any teacher who has watched a laptop grade an assignment in less time than it takes to close the classroom door, or any parent who has watched their child ask a chatbot a question before looking up at them, and you'll find they are wrestling with the exact same quiet anxiety.

Andrej Novik, a primary school teacher in Prague writing in UNESCO's report The Algorithm in the Room, captured the moment that friction disappeared from his own classroom. His prop was a real German banknote from 1923, worth 500 million marks and barely enough, by then, for a loaf of bread. For years, pulling it out of his pocket reliably set the whole room off, children arguing over why the number was so absurd, offering their own competing theories about what it meant. In 2024, he let his students photograph the note and ask AI instead. The answer came back in two seconds. "The classroom is silent," he writes. "Nobody argues. Nobody offers their own hypothesis. Nobody debates. The answer is simply there." When he followed up with questions, it was clear the words hadn't been processed. The children were, in his own phrase, "parroting words that were never theirs." So what, exactly, did that classroom gain, in exchange for the two seconds it saved? In two seconds, the friction was gone, and the understanding went with it. The children didn't lose the facts. They lost the struggle to make sense of them.

It isn't only happening in Novik's classroom, either. Also writing for UNESCO, education consultant David Ross describes the identical loop at the scale of an entire school system, giving it a blunt name: the Classroom AI Doom Loop. A teacher prompts AI to generate an assignment. A student prompts AI to write the response. An AI grader posts the mark straight into the school's own database. No person is required at any point in between. "No humans were harmed in this process," Ross writes, "because humans were only ancillaries to the process." His essay borrows its title from Bertrand Russell, who asked nearly the same question in 1951, at the dawn of cybernetics, and concluded that an automated society might well be able to function without human beings at all, but that this probably wasn't something to want. Seventy-five years on, Ross is blunter about it: "if a major function of education can be automated, it's probably not human enough."

If the system runs fine without us, what are we actually in the room for?

We should be careful not to paint Generation Alpha as passive victims in all of this, though. Nearly 70% of Gen Alpha children worldwide already say they trust AI, according to Generation Alpha 2030, and it isn't hard to see why. Kathy Sheehan, formerly global managing director of the cultural strategy firm Cassandra, puts their confidence plainly: "their sense is that as long as they have the technology, they can figure anything out." An always-available, non-judgmental space that never makes a child feel foolish for admitting confusion is not nothing. This is also a generation that fiercely wants agency elsewhere: 93% play video games weekly, mostly in sandbox environments like Minecraft and Roblox, where they design, code, and solve problems rather than passively consume, and 76% of UK Gen Alpha say they aspire to be their own boss or launch a side hustle one day. They aren't reaching for frictionless answers out of laziness, they're reaching for tools they believe hand them control over their own future.

That confidence is real, and it's also being cultivated on purpose. Tech platforms don't build AI companions to be neutral calculators, they engineer them using "emotional language, memory and mirroring" specifically to keep a person engaged. Clayton Ramsey, a maths teacher writing in the UNESCO report, caught himself talking to ChatGPT "as though it were a person" before he ever thought to worry about his own students. If an adult educator can be drawn in this easily, a child stands little chance. The financial incentive behind that design is considerable: the creator of one gacha-style game built on these same mechanics recently became one of the fifty richest people in Korea. An entire industry has discovered that one of the fastest ways to build a fortune is to build a world where a child is never contradicted.

What worries me isn't a student reaching for a shortcut. It's what happens to choice itself, when the environment around that choice has been engineered to remove every reason to make a harder one. The answer feels like agency because it arrives instantly, but real agency was never about speed. It's the part that gets quietly designed out first.

That's the real question, and it's why our current debate is so shallow. We spend almost all our energy arguing over whether to adopt these AI tools or ban them, as if that binary choice could solve the problem. It never could. Debating the tool itself only engages what Knowledge Mindfulness calls the Outer Window, the environment of software, interfaces, and friction-free technology surrounding us. It does nothing to cultivate the Inner Window, the metacognitive self-awareness and quiet judgment required to notice when an algorithm is replacing our own sense-making. And it leaves the Relational Window entirely vacant, that intersubjective space of human presence, ethical dialogue, and unscripted connection where truth is actually co-created between people.

And it isn't only students who feel that the Relational Window is closing. Claudia Muellauer, who teaches Theory of Knowledge at a high school in Austria, started using AI to help draft feedback for her language students. She didn't cut corners. If anything, she worked harder, spending fifteen to twenty minutes of careful prompting and editing per essay, more than double what marking by hand used to take her. Her students noticed anyway. "Students told me they wanted me to read their work," she writes in the same UNESCO volume. "To them, the AI's presence suggested my absence." Effort, it turns out, isn't the same thing as presence. If a teacher working harder than ever can still read as absent, the question stops being about students at all, and becomes about the rest of us: what, exactly, makes any of us necessary in a room, if not the parts of us a machine can stand in for?

Here is the detail I keep returning to, though. Generation Alpha 2030, a foresight report commissioned to project this generation's future workforce needs, ranks the skills it believes they will need most. Interpersonal skills come first. Critical thinking and strategic decision-making, second. Creativity, third. Digital fluency and AI collaboration comes in fifth, out of six. The point isn't that AI is going to replace us. It's that we have spent so much time obsessing over the software that we forgot to cultivate the human mind using it. The future doesn't depend on how fast we can prompt a machine. It depends on whether we still have the judgment to evaluate what it gives us back.

I often think about my former students, the ones who sat in seminar rooms, took the risk of being wrong out loud, and occasionally changed my mind about something I'd believed for years. That messy, unscripted friction wasn't an inconvenience, it was the exact space where shared understanding was built between us. If an algorithm had provided a clean answer before we ever had to reach for one, those moments would never have existed, and none of us would have learned to trust and sharpen our own judgement.

We already know AI is in the classroom, and will continue to be. So when the machine can do the thinking, what are you bringing to the room that only a human being can offer?

Knowing to Live, Living to Know.

With care for “ALL”, Dr. Laila Marouf and the KMD Team