A child spends roughly 16,000 hours in compulsory education between the ages of five and eighteen.

Reading, writing, arithmetic, basic scientific literacy, the core skills that everyone needs, require a fraction of that. With focused, competent instruction, you could cover the essentials in under an hour a day. Call it 2,000 hours, generously.

So what happens in the other 14,000?

Mostly crowd management. Standardized pacing. Making sure forty kids move through identical material at identical speed.

You might be thinking about starting a family soon. I am. And the closer that gets, the more a question keeps nagging at me: what exactly am I handing my kids over to?

I've spent a lot of time with this question lately. What I found unsettled me enough that I wanted to write it down. Not as a manifesto but as an invitation to rethink an imperfect system most of us have come to accept.

This essay is about what education used to mean, how we lost it, and whether technology might help us recover something ancient and valuable.

The Three Losses

The Greeks had a word for what we now awkwardly call "holistic human development": paideia. It meant the deliberate formation of a person's character, judgment, physical capability, and capacity for flourishing.

Not training. Not credentialing. Formation.

Three ancient institutions embodied this ideal, and modern education has lost something essential from each.

Plato's Academy gave us the method.

Education was dialectic - structured conversation where the teacher asks questions and the student discovers answers through their own reasoning. In the Meno, Socrates teaches a slave boy geometry without telling him a single fact. He just asks questions until the boy derives the proof himself. Education as drawing out, not pouring in.

We've reversed the direction entirely. The modern classroom treats students as receptacles. A teacher stands at the front and delivers information. Students absorb it, regurgitate it during an exam, and forget most of it within weeks. The information flows one way. The student's existing understanding, their curiosity, their particular way of making sense of the world; none of that enters the equation.

Aristotle's Lyceum gave us the purpose.

Aristotle didn't educate people for employment. He educated them for eudaimonia - human flourishing. The curriculum spanned biology, ethics, politics, rhetoric. The breadth was seen as necessary because logic and judgment requires it. You can't reason well about ethics if you don't understand human nature. You can't understand human nature if you've never studied a living system. The Lyceum produced people capable of thinking across domains, not specialists trapped inside one.

Lessons at the Lyceum were taken while walking. Schooling happened in gardens, in motion, in the physical world. Why not teach biology while walking through an ecosystem? Why not teach geometry while building something? The idea that education requires fluorescent lights and plastic chairs would have struck Aristotle as absurd.

The Hellenistic schools (the Stoics, the Epicureans) gave us the relationship.

Education wasn't lectures. It was mentorship. Students didn't just learn Stoic philosophy; they practiced it daily through exercises, journaling, and accountability to a teacher who knew them deeply. The relationship between teacher and student was closer to what we'd now call coaching or apprenticeship. The teacher wasn't an information-delivery system. They were a moral exemplar. A character-former.

These three institutions look different on the surface but they share an identical architecture. Small groups. Extended duration. Relationships deep enough that the teacher understood each student's particular way of thinking, their intellectual blocks and their natural curiosities.

Socratic dialogue requires an understanding of where each individual's reasoning breaks down so you can ask the question that unlocks it. Character mentorship requires knowing the person well enough to see what they don't see in themselves. Tailoring curriculum to curiosity requires time to discover what that curiosity actually is.

All three models shared one assumption: a low ratio of students to teachers. All three produced extraordinary results. And all three have been dismissed as impossible at scale.

The Factory

Before the 1840s, most children received no formal education at all. Some attended church-run charity schools. Some learned a trade through apprenticeship. Most remained illiterate. Education was either elite or non-existent.

The Industrial Revolution changed the equation. Factories needed literate workers who could read instructions, follow schedules, and perform specialized tasks. Armies needed soldiers who could read orders and operate increasingly complex equipment.

For the first time in history, mass literacy became an economic and political necessity, not a luxury. The question was how to produce it at scale when nobody had done it before.

Prussia was the first to solve this problem, and they solved it the way you solve any manufacturing challenge: standardization, efficiency, scalability. Standardize the curriculum so you can train teachers quickly. Group students by age so you can process them in batches. Create clear metrics so you can prove to taxpayers that the system works.

Horace Mann visited Prussia in 1843 and saw an answer to a different American problem. Massive immigration. Rapid industrialization. Dozens of languages and customs colliding in cities that hadn't existed a generation earlier. The factory school would do more than teach reading; it would manufacture Americans. Shared curriculum, shared language, shared civic values. The Pledge of Allegiance was mandated in schools in 1892 for a reason.

The mission kept expanding. Universal literacy became vocational preparation (adding shop classes, emphasizing punctuality and rule-following). Vocational preparation became college preparation (the post-WWII expansion, the GI Bill era). College preparation became credentialing and sorting (once degrees became employment requirements).

Each shift made sense in its moment. But the cumulative effect was that nobody could articulate what school was for anymore. Citizenship training? Vocational prep? Intellectual development? Childcare? The mission became incoherent. The structure stayed the same.

Now, the system serves too many functions to reform. Parents need childcare. Employers need credentials. Teachers need stable jobs. Politicians need something to campaign on. Real estate agents need good school districts to sell houses. Even people who recognize the system is broken have no obvious alternative that serves all these constituencies simultaneously.

Meanwhile, inside the classroom, the constraint is brutal. One teacher. Forty students. One pace.

The struggling student doesn't need the same material delivered more slowly, they need a completely different entry point into the material. The advanced student doesn't need to wait, they need new threads to explore. Both are failed by the same system for the same structural reason: the assumption that there's one path through any subject and the only variable is how fast you walk it.

Teachers aren't the problem. Teachers are victims of this system. The job has been reduced to information delivery and crowd management—the two least human parts of education.

There's simply no time for individualized attention when you're managing forty kids through a standardized curriculum on a standardized timeline.

Bloom's Ghost

In 1984, the educational psychologist Benjamin Bloom published a paper called "The 2 Sigma Problem." His finding was staggering: students who received one-on-one tutoring performed two standard deviations better than students in conventional classrooms. The average tutored student outperformed 98% of the classroom group.

We've known this for over forty years. The education establishment's response has been to search for group instruction methods that approach tutoring's effectiveness. They haven't found them.

I know why, because I've been on the other side of it.

When I was a post-grad in Edinburgh, I tutored five students aged thirteen to seventeen through their GCSEs and A-Levels (the UK's standard curriculum).

One of them, a brilliant kid with a deep interest in the humanities, was falling behind in Biology. To him, the material was a disconnected pile of facts: molecular structures, cell processes, biodiversity categories. None of it cohered.

The standard fix would have been more practice papers. More revision. More repetition of the same content at the same angle. Instead, I took him to the university and we spent a day in the sciences library going deep on evolutionary theory, which, absurdly, occupied only a tiny fraction of his actual syllabus.

He was fascinated. And once he saw evolution as the organizing principle (the trunk from which molecular biology, cell structure, and biodiversity all branch) the disconnected facts became a coherent system. I didn't teach him Biology. I taught him how Biology thinks. I could only do that because I had one student, I knew his interests, and I had the freedom to throw out the syllabus for a day and find the door that opened for him specifically.

No classroom teacher with forty kids can do that. That's the supply problem Bloom identified.

But what Bloom didn't measure is that the supply problem creates its own demand problem. When students have no agency over what they learn, no ability to follow their curiosity, they stop wanting to learn at all. School becomes something that happens to them. A chore imposed by adults, not a pursuit they'd choose.

My student proved this in reverse. Once the material became his, once he saw the organizing principle and felt the fascination, he started demanding more. He asked questions the syllabus didn't cover. The curiosity was always there. It had just been suffocated by a system that never gave it a foothold.

Think about your own education. The subjects where you had a teacher who made the material come alive, who connected it to something you already cared about. You looked forward to those classes. Homework wasn't a burden. But other subjects? They were pure drudgery. Something you struggled through out of a sense of duty, or in many cases, a fear of disciplinary action.

A personalized system doesn't just deliver better learning outcomes. It creates students who demand higher standards through their own inquiry. Solve the supply problem and the demand problem begins to resolve itself.

The Door, Not the Room

We've known the solution to the education problem for forty years. Bloom proved it in 1984: students who received one-on-one tutoring performed vastly better than students in conventional classrooms. The constraint isn't knowledge. It's supply. We can't afford enough human tutors for every student.

AI changes this equation, but only if we're precise about what we're asking it to do.

When I tutored that struggling student, I didn't just explain the material differently. I diagnosed where his understanding was breaking down and found a different entry point. He needed the organizing principle before the details made sense. Another student might need the opposite: concrete examples first, theory later. A third might need to dissect something with their hands before any of it clicks. Different doors into the same room.

A classroom teacher with forty students cannot do this diagnostic work for each mind simultaneously. An AI system can. It can figure out that this student learns through narrative, that one through visual patterns, another through building physical models. It can map existing interests (this kid loves history, that one loves video games, another loves cooking) and route new material through those doorways. The personalization isn't about making things easier. It's about making them accessible to each specific mind.

To that end, AI can be a wonderful syllabus architect.

Right now a teacher plans one lesson for forty students. With AI as the planning layer, that teacher could have forty customized curricula but the actual teaching still happens human-to-human.

The AI handles what computers are good at: pattern matching, optimization, infinite patience for iteration. It generates the reading list, the problem sets, the sequence of topics tailored to each student's starting point and learning style. Then it assigns the actual work: read this book, conduct this experiment, have this conversation with your teacher, build this thing with your hands.

Much of the learning itself could happen off-screen. The AI just makes it possible for one human teacher to mentor multiple students deeply instead of managing a crowd superficially.

The teacher uses the AI the way a carpenter uses power tools—they make the work faster and more precise, but they don't replace the craftsman's judgment. The Socratic questioning, the character formation, the mentoring; that still requires a human.

The fear that AI will make education more alienating is justified if we're talking about parking kids in front of YouTube videos or chatbot tutors. Instead, technology could be the mechanism that makes the Greek model economically viable for the first time since antiquity.

This points to another crisis hiding inside the first: we've built a system that repels great teachers. Smart, high-character people largely don't go into teaching today. The pay is poor, the autonomy is nonexistent, and the job has been hollowed out into information delivery and discipline.

If technology reclaims those functions, teaching becomes what it was supposed to be: mentorship, dialogue, formation. That's a vocation worth choosing. That's a job that might attract the kind of people who currently go into tech or consulting or finance because those fields offer the intellectual challenge, autonomy and financial incentives that teaching no longer does.

The Question You're Not Asking

Everything I've described so far is structural. Policy-level. Easy to nod along with. But structure becomes personal the moment you're choosing a school for your own kid.

You've probably given some thought to schools. Maybe you've looked at test scores, browsed neighborhood rankings, asked friends where they're sending their kids. Most of us stop there.

But try a different question: how much of what you learned in school do you actually use? Not the credential; the knowledge. Think about the skills that matter most in your daily life. The ability to reason through an unfamiliar problem, communicate clearly, manage risk, stay focused when things get hard. Where did you pick those up? For most of us, the honest answer is not in a classroom. We learned them later, on the job, through relationships, through failure. School gave us the diploma. Life gave us the education.

Which raises another concerning point: the system that produced you is worse now than when you went through it. Classroom sizes have grown. Chromebooks have replaced conversation. Teachers have less autonomy. This is what you're about to hand your children to.

There's a reason homeschooling is the fastest-growing segment of American education right now. The families pulling their kids out aren't fringe. They're educated, affluent, and deliberate. They've looked at the system closely and decided to exit. You don't have to reach the same conclusion they did. But the trend is diagnostic. When the people with the most optionality are choosing to leave, it tells you something about what they're leaving.

Paideia

The ancient Greeks had a word for what we've lost: paideia. Not job training. Not test preparation. The deliberate formation of a person's character, physical capability and capacity for flourishing. It encompassed everything: the intellectual and the physical, the individual and the civic. It was what education meant before we narrowed it down to credentials and childcare.

We have, for the first time, the technological capacity to return to that model. Not to replicate the ancient academies; they were small, elite, and exclusionary. But to take what made them work and make it economically viable at scale. The method exists. The technology exists. What's missing is the institutional will to dismantle a system that too many people depend on for too many reasons.

My wife and I want kids, and I'm looking at the same options you are: public school, private school, homeschool, some hybrid we haven't imagined yet. Each comes with tradeoffs I can't fully evaluate because we're all making these decisions inside a system that hasn't been redesigned since the Industrial Revolution.

But I know this: the question matters. What you're handing your children over to, what those 16,000 hours are actually for, whether the system produces flourishing humans or just sorts them into categories. These are decisions we'll all make in the next few years that will shape our children's mind, character, and capacity for a life well-lived.

The current model isn't the answer.