The Child Is the Curriculum

I have two boys. They are ten, both in Grade 4, and I teach them at home, every day, using AI to build and personalise every lesson.

That last part matters for this piece. I am not writing about what AI might mean for education in theory. I have spent a full term watching it happen, in real time, on two specific children whose results I track weekly. So when I talk about what still holds its value, I am not guessing. I am reporting.

For a long time I answered the "what should I teach my child" question the way most parents do. Get the basics right. Push maths. Steer them toward something technical, because that is where the work is going.

I have had to revise that. This is what changed my mind — and what a term of actually testing the revised version on my own sons confirmed.

The advice I was giving was already out of date

In July 2026, the mathematician Terence Tao described an experiment in which four AI systems were given ten genuinely new research-level mathematics problems. Seven of the ten were solved to publication standard by at least one of them, checked by expert referees.

Not exam questions. Not problems with answers in the back of a book. New ones.

Around the same time, Jensen Huang — chief executive of Nvidia, the company whose chips most of this technology runs on — was asked in a podcast interview who the smartest person he had ever met was. He said he could not answer, and explained why. What most people call intelligence — technical problem-solving — is becoming a commodity, and it is the part AI handles most easily. Software engineering, he pointed out, was supposed to be the ultimate proof of a sharp mind. It was among the first things to fall.

"Learn to code" was good advice for about fifteen years. It is now advice to compete directly with the thing that does it faster, cheaper, and without getting tired.

That does not mean do not teach it. It means stop treating it as the destination.

What Huang says holds its value instead

His own definition of smart is someone technically capable who also carries real empathy, and who can read what has not been said out loud. People who are able to see around corners — someone who senses a problem before it arrives, because they can feel the vibe in a room.

That instinct, he said, comes from a combination of analysis, first principles, life experience, wisdom, and reading other people.

And then the line that stopped me: a person like that might score badly on a standardised test.

I have now watched this play out directly in my own homeschool. My twin sons are on the same curriculum, in the same house, taught by the same person, and they are almost opposites in exactly the way Huang was describing.

One of them — I have come to call him The Storyteller in our lesson notes, because that is simply what he is — needs time before a topic lands, and when it does, it lands completely. Without being asked to, he wove the emotional register of a novel we had been reading for weeks into an unrelated writing assignment. He consistently finds the human meaning inside a fact before anyone points it out to him. None of that shows up as a number on a quiz.

The other — The Inventor — is fast and systems-minded. He invented his own shortcut for the eight-times table before I taught him one, and then independently extended the same logic to the sevens and nines. That kind of instinct does not show up cleanly on a standardised test either, but it is exactly the pattern-recognition intelligence that will matter for the rest of his life.

Both would score well on a normal test. But the test would tell you almost nothing about which of my sons is which, or what either of them is actually good at. I only know because I watch closely enough to see it.

So why am I still teaching maths?

This is the question I get asked most, and I want to answer it properly, because the obvious conclusion is wrong.

If AI can solve research-level mathematics, why should a ten-year-old grind through long division?

Because we never taught maths so that children could compute. Calculators removed that reason forty years ago and nobody stopped teaching it. We teach it because of what it builds underneath: the ability to hold a structure in your head, to follow a chain of reasoning without losing the thread, to work in steps, and — this is the part that matters most now — to know when an answer smells wrong.

That last one used to be a nice extra. It has become essential.

I saw exactly why this term. My boys' quiz scores — the recognition-based kind, pick-the-right-answer — were consistently strong. Their fully written answers, the ones requiring them to construct an explanation from nothing, were noticeably weaker. Not because they did not know the material. Recognising a correct answer and producing one from a blank page are different skills, and AI-generated practice is extremely good at testing the first one while quietly letting the second one go soft.

These tools also produce wrong answers with total confidence. They do not hesitate, do not hedge, and do not tell you when they are guessing. A child with no feel for whether an answer is plausible will simply accept whatever appears on the screen. A child who has built that feel — through exactly the kind of grinding, step-by-step maths I am describing — will stop and say: that cannot be right.

So we still do maths. But I have changed what I am listening for. I am less interested in whether the final answer is correct and more interested in whether my sons can tell me why they believe it, whether they would notice if it were wrong — and now, practically, whether they can say the full explanation out loud before they are allowed to write it down. That last habit came directly out of watching the gap between their quiz scores and their written work. It has done more for the quality of what they produce than any amount of extra content.

The same principle applies to reading. Not "can you read this," but "does this make sense, and what is this person actually trying to do to you?"

I have started observing my two the way a scientist studies a market

Here is where being an entrepreneur genuinely changes how I do this.

When you build a business, you do not decide in advance what the market wants and then insist on it. You watch. You test. You pay attention to what people actually do rather than what you assumed they would do, and you revise.

Most of us do the opposite with our children. We decide early what they should become, then spend twelve years pushing them toward it and calling the resistance a discipline problem.

So I observe my two the way a scientist would study a market - deliberately, in writing, every week. Four questions I keep returning to.

What does this child actually love? Not what he says when asked. What he returns to when nobody is watching and there is no reward attached.

What is he genuinely good at? Not what I wish he were good at. This one requires more honesty than it sounds like it does.

What will the world need — that AI cannot do? This is the question that has changed, and it is the reason I am writing any of this. The old version was "what jobs will exist." That is no longer answerable with confidence. But "what will still require a human" is answerable, and the test I use is simple: work where a machine can check its own answer is work that gets absorbed. Work that requires judgement about people, trust, and what matters does not.

And how does that become a living? This is the entrepreneur's question, and I do not apologise for asking it. Loving something and being good at it is not enough. There has to be a path where somebody will pay for it.

There is a warning worth stating plainly, because this term taught it to me directly. A child is not a startup. A business begins with an intention of what it will become, and this approach goes badly wrong the moment you impose that intention on a child - forcing them into your own idea of what success looks like. I keep a discipline for exactly this reason: before I record anything from a lesson, I ask myself what happened today that I would not have predicted. That is what keeps the observation scientific rather than confirmatory: the experiment has no predetermined outcome - I am finding out who they are, not proving who they should be. If the honest answer is "nothing," that itself is a signal — a run of days with no real surprises usually means I have stopped genuinely looking and started confirming a story I have already decided is true about my son. I revisit my read of each boy monthly, because a child who is developing can quietly outgrow a description that was accurate six weeks ago. The point of close observation is the opposite of a fixed plan. It is noticing who is actually in front of you today.

What we actually do differently now

Some of this is small and unglamorous, and all of it came directly out of watching what actually worked and what did not.

They use the tools, and they argue with them. Not banned, not treated as cheating. But nothing gets accepted because a machine said it. The habit I want is checking, not asking.

Everything gets tested more than once, in more than one way. A fact learned once and tested once in one format is a fact that will be gone by exam time. We test the same core knowledge across a quiz, a full written case study, and an oral explanation said aloud before writing — because each format catches a different kind of gap, and the gap between quiz confidence and written production is exactly where real understanding either holds up or falls apart.

One subject, properly, for the whole day. I used to split each school day across two subjects, the way most schools do. I changed it after noticing both boys disengaged every time we switched. One of our subjects — values and character — covers several topics: a physical activity component, a creative expression component, and others. Taught across a full day instead of split into separate lessons, my sons made a connection neither had made before: that courage looks the same whether you're playing fair in a game you could win by cheating, or being honest in front of an audience when perfect would be easier. That only surfaced because the day stayed inside one subject long enough. I'd assumed variety would help concentration. For my two, depth beat variety every time.

Real work with real consequences. Case studies with an actual character in an actual situation who has to make a real decision. Content without a story attached simply does not stick — not for either of my sons, and I suspect not for most children.

People, deliberately. Time with grandparents, with neighbours, with people who are not like us and who require some effort to understand. Reading a person is a skill, it is learned by practice, and it is close to the top of the list of things that will still matter in twenty years.

And I have stopped pretending I know what they will do for a living. Nobody does. What I can do is send them into it as people who notice things, who can tell when something is off, who can be trusted, and who are not frightened of the tools.

Why I think personalised learning is about to become the norm, not the exception

This is a prediction, and I am making it as someone who has now run the experiment on two children for a full term.

I think we are about to see a flood of small, personalised learning setups — micro schools, structured homeschool pods, families combining resources around exactly the kind of close, individualised attention I am describing.

The reason is not that mass schooling was badly designed. It is that it was well designed for a different economy. Standardised schooling was built to produce people who could follow instructions accurately and reliably at scale — because that is what factories, and later offices, needed. It did that job properly for a century.

That economy is the one being absorbed fastest. Following instructions accurately is precisely the checkable work machines now do without tiring. A system optimised to produce reliable instruction-followers is optimising for the wrong thing, at exactly the wrong moment.

What replaces it, I think, looks less like a smaller classroom and more like an inversion: the child as the curriculum, rather than the child fitted to one. Instead of pulling a child into a fixed box and measuring how well they fit, you start with who they actually are — the way I had to start with the real, specific differences between my two sons — and build outward.

I want to say this part plainly, because it is easy to assume otherwise: this does not have to be an expensive, elite thing. What it requires is a room, a child, an adult who pays close attention, and access to tools that generate the content, track the gaps, and catch the curriculum before an exam does. Those tools are now genuinely inexpensive. The barrier to personalised education used to be the cost of one adult per handful of children preparing everything by hand — the lesson plans, the differentiation, the revision materials, the tracking. AI does not remove that adult. It removes a great deal of what used to make that adult's job impossible to sustain at small scale.

What I actually tell them

Not much, honestly. Ten-year-olds do not need a lecture about labour markets.

But when it comes up, it is roughly this: the machine is very good at the parts that can be marked right or wrong. It is not good at people, and it does not know what matters. Get good at those, stay curious, learn to use the tools without being impressed by them.

And be someone others can rely on. That has never once gone out of date, and after a full term of watching what actually holds up in two real ten-year-olds, I do not think it is about to start.