Every technological revolution is, underneath the noise, a single move: it floods one thing into abundance, and by doing so it makes something else scarce. The value of a skill, a job, an education then quietly migrates — away from whatever the machine just made cheap, and toward whatever it accidentally made rare.
The printing press made the written word abundant. The scarce thing became literacy; the ability to read turned into the dividing line of a society. Centuries later the internet made information itself abundant, and the scarce thing became attention — so completely that the largest companies in history were built to mine it. The pattern is reliable. Find what a new technology makes free, and you can predict what it is about to make valuable.
Artificial intelligence is running the same move on the faculty we assumed was permanently ours. It is making competent thought — recall, calculation, summary, the serviceable first draft of almost anything — abundant.
So ask the question the frame forces on us. If competent thought is becoming abundant, what just became scarce? Not intelligence — intelligence is exactly the thing being mass-produced. The scarce faculty, the one whose value is about to climb the way literacy and attention once did, is the will to think when you no longer have to. The desire to move toward a hard question instead of away from it. I am going to call that volition, and I am going to argue it is the only thing left worth building a school around.
We panicked about the wrong thing
When AI arrived in classrooms, nearly every institution asked the same first question: how do we stop them cheating? The frame above tells us why that question is a dead end. You cannot protect the scarcity of something that has already gone abundant. Competent answers now cost roughly nothing — guarding the exam is guarding a vault after the gold has quietly walked out the back.
And yes — the machines are imperfect. They hallucinate, they miss context, they produce confident and shallow work. Those flaws are real, and they matter day to day. But they no more change the trajectory than early printing errors saved the profession of the scribe. Even imperfect abundance reprices what is scarce. The question was never whether students will use these tools. They will, the way we all now use a calculator without a second thought. The question is what remains worth learning once the answer is the cheap part.
It is not memorization, and it is not speed. What remains is the willingness to care about a question whose answer no rubric can fully specify — to keep going after the obvious answer has already arrived. Schools have always rewarded completion. The age of abundant cognition asks them to start rewarding continuation.
You cannot teach a child to out-calculate the machine. You can only teach them to want something the machine will never want.
What volition actually is
Volition is not motivation, even though we use the words as if they were interchangeable. Psychologists have kept them apart for decades. Motivation is the choosing of a goal — the spark, the wish, the resolution made on the first of January. Volition is what happens after the spark: the sustained will to keep crossing toward the goal once the glow is gone. The researchers Heinz Heckhausen and Peter Gollwitzer called the moment of commitment “crossing the Rubicon,” and drew a hard line between the motivational phase of wanting a thing and the volitional phase of actually doing it. Motivation gets you to the start line. Volition is what is left at mile twenty, in the cold, with nobody watching.
The sustained will to move toward cognitive difficulty in pursuit of a self-chosen future.
Intelligence is what you can do. Volition is what you choose to become. It is the bridge between ability and becoming — the force that keeps a person moving after talent, instruction, and reward have all run out.
Hold that definition against intelligence and the two pull cleanly apart. Intelligence answers the question on the page; volition asks whether the page holds the right question. Intelligence solves the assigned problem; volition notices the problems no one assigned. Intelligence stops when it is correct; volition keeps walking, because correctness was only ever the trailhead. For two hundred years we could afford to blur the line between them, because intelligence was the bottleneck and volition rode along for free inside it. Abundance has pried them apart. From here, they are two different resources — and only one of them is still scarce.
The evidence has been hiding in plain sight
This is not only a philosopher’s intuition. The science of motivation has been circling it for half a century. Edward Deci and Richard Ryan’s decades of work on intrinsic motivation found, across a large body of experiments, that people sustain genuinely difficult work not when they are paid or pressured into it, but when the pursuit feels chosen and their own — and that piling on external rewards can actually crowd out the internal drive that does the durable work. Volition is what that internal drive looks like when it survives contact with real difficulty.
Then look at what abundant cognition has already accomplished. In 2020, DeepMind’s AlphaFold cracked in hours a problem that had resisted biology for roughly fifty years: predicting how a protein folds from its raw sequence. It was a staggering feat of machine cognition, and in 2024 the work behind it shared a Nobel Prize. But notice precisely what the machine did not do. It did not wake one morning wanting to understand proteins. Generations of human biologists wanted that — wanted it through decades of dead ends — and aimed the tool at a question worth answering. The cognition was the abundant part. The wanting was the scarce, human part. That is the shape of every good future with this technology: the machine supplies the horsepower; the human supplies the direction.
And the shape of every bad future is already sitting in classrooms. A student who can finish a week of homework in five minutes now faces a choice no earlier generation met so young: spend the time saved going further, or spend it disappearing. Almost everything about how school is scored — the grade, the deadline, the rubric — rewards the disappearance, because the finished work looks identical either way. We are, at this very moment, optimizing for the wrong student.
I argued in an earlier essay that we must raise AI natives, not AI tourists — that hands-on fluency with these tools is the new literacy. I still believe every word of it. But fluency is the floor of this building, not the roof. You can be flawlessly fluent and intellectually inert: a very fast tourist who never actually wanted to go anywhere. Fluency without volition just reaches the emptiness sooner.
Two students, one machine
Picture two students handed the same tool on the same morning.
The first uses AI to make school disappear. It drafts the essays, solves the problem sets, summarizes the books she was meant to read. Her grades hold steady. And year by year, invisibly, every delegated struggle is a withdrawal from an account she does not know she has — judgment, stamina, the specific confidence that comes only from having once been badly stuck and clawed her way out. She grows more efficient and less able at the same time, and because the output looks the same, the two trends hide each other until the day they can no longer be hidden.
The second student uses the very same AI to make school bigger. She delegates the routine work for the identical reason — but the time it frees, she spends somewhere she could never have reached alone: harder questions, longer projects, the territory that used to be walled off behind years of prerequisite drudgery. For her the machine is a tutor, a research partner, a simulator, an amplifier. She is not avoiding the climb. She is using a faster vehicle to attempt a taller mountain.
The technology is identical. The entire fork runs through the student — through whether anyone ever taught her to want the climb in the first place. That is why volition cannot be an afterthought tacked on to a “real” curriculum. It is the variable that decides which of these two futures the same machine produces.
The machine is the same on both roads. The only difference is whether the human still wants to walk.
What a school for volition would do
You cannot lecture a child into wanting. Volition is caught, not taught — it grows in conditions, not lessons. But the conditions are buildable, and most of them are simple inversions of what school currently rewards.
Grade the question, not the answer
When answers are free, they are worthless as a signal. The rarer, more honest thing to measure is the quality of what a student went looking for. The child who arrives with a sharper question has shown you the scarce faculty directly.
Make struggle the high-status thing
In a room where the AI hands out finished work, visible effort on hard problems becomes the only currency that still means anything. Celebrate the student wrestling at mile twenty — not the one who breezed through mile one and called it talent.
Protect difficulty and boredom
Volition is born in the gap between what a student can already do and what they wish they could. Frictionless tools close that gap the instant it opens. Sometimes the most generative thing a teacher can do is decline to hand over the shortcut.
Assign pilgrimages, not problems
Give children long, hard, self-chosen work — measured in months, not minutes — that AI can accelerate but cannot choose on their behalf. The point was never the deliverable. It is the lived experience of wanting something enough to cross difficult ground to reach it.
And none of this stays inside the classroom. The same law — abundance makes cognition cheap and volition dear — is about to reprice hiring, leadership, creativity, and parenting alike. The valuable colleague will no longer be the one who can produce the work; the machine can produce the work. It will be the one who still knows what work is worth producing, and wants to badly enough to see it through. We are all, shortly, going to be graded on our volition. School is simply where that repricing ought to begin.
My own children will grow up beside machines that know more than they ever will, calculate faster than they ever could, and produce a competent first draft of nearly anything before they have finished reading the prompt. I have no wish for them to compete on those axes. That race is already lost, and it was never the race that mattered.
What I want for them is the scarce thing. To stay restless in a world engineered, year by year, to make them comfortable. To keep asking after the machine has finished answering. To choose hard problems on purpose — because those are the only problems that change the person who solves them. Intelligence, they can now rent by the token. Volition is the part that has to be theirs.
Teach them to want. The machines can help with almost everything else.
For every child who deserves to stay hungry in a world built to make them full.
Keep wanting. Keep climbing. The view was never the point — the climb is.
— VJ
A parent, wondering where to start on Monday? Read the companion essay: You can’t wait for the school. →
A note on how I write
I am not a writer. I am a person with strong opinions and scattered notes. Every essay on this site started as a messy brain-dump — half-formed arguments, bullet points, and “you know what I mean” — that I hand to an LLM. Another LLM handles the background research needed to find the facts that support an argument. And then it all gets translated into writing far too good for me to pretend is mine. The ideas are mine. The craft is not. They say blogs are dead — but I am falling in love with this. It gives me an outlet for expression that would otherwise have stayed buried in my head. I believe you deserve to know all of that.