Educating for tomorrow

We’re currently living through the early years of the most significant technological shift since the printing press. Artificial intelligence can now write essays, pass bar exams, diagnose diseases, generate art, write code, and hold conversations that fool most people into thinking they’re talking to another human. It does all of this for approximately the cost of electricity. And it’s getting better every few months in ways that make last year’s version look quaint.

So what, exactly, should we be teaching kids? What kind of things will be tools they actually use and what can we do to increase their chances of a happy, successful life?

Human fundamentals

These are the skills most resistant to automation, most correlated with life satisfaction, and most neglected in formal education. Below them are the amplifying skills — things that multiply everything else. Below that, traditional domain knowledge still matters, but it needs to be taught entirely differently. And at the bottom, occupying an embarrassing amount of current classroom time, are things we should be actively reducing.

Emotional intelligence

Your kids need to understand themselves.

Of all the skills on this list, emotional intelligence is the one most consistently correlated with life satisfaction, most resistant to automation, and most neglected in school. Which is quite a trifecta.

Emotional intelligence isn’t soft. It isn’t fluffy. It’s the ability to recognise what you’re feeling, manage it without suppressing it, understand what other people are feeling, and navigate relationships without leaving a trail of wreckage. It’s what determines whether your brilliant, technically skilled child becomes a person others want to work with — or a person others tolerate until they can find someone easier.

Here’s what the research says, plainly: emotional intelligence is primarily caught, not taught. It’s transmitted through relationships, not curricula. A child who feels safe, seen, and accepted by their primary adults develops emotional intelligence naturally. An insecure child can’t learn empathy — they’re too busy surviving.

This means the adults matter more than any programme. If you scream when you’re angry and then lecture your children about staying calm, you’re wasting everyone’s time. If you dismiss their feelings — “You’re fine,” “Stop crying,” “Don’t be silly” — you’re teaching them that their emotional experience is wrong. They’ll learn to suppress, not regulate. And suppression has a very reliable long-term cost.

The good news is that the interventions are straightforward, even if they’re not easy. Name emotions — yours and theirs. Expand their emotional vocabulary beyond happy, sad, and angry. Let all feelings exist while setting limits on behaviour. “You’re allowed to be furious at your sister. You’re not allowed to hit her. Let’s find another way.” Repair after rupture — when you mess up (and you will), apologise sincerely. Children learn more from watching adults repair mistakes than from watching adults be perfect.

Stop telling boys not to cry. We have enough emotionally illiterate adults already.

The hardest thing you will face with this isn’t usually the kids — they adapt quickly and easily to the environment around them — it’s your own conditioning and limitations as an adult that are the most problematic. But if you do the work, grow beyond yourself and evolve, the kids will too.

Communication

AI can generate competent text. It can draft emails, write reports, and produce essays that would earn a solid B+ in most classrooms. What AI cannot do is look someone in the eye and make them feel understood. It can’t read a room, build trust through presence, navigate a difficult conversation with a friend, or tell a story that makes a stranger care about something they didn’t care about five minutes ago.

Communication is the skill employers rank number one in every survey, every year, without fail. And it’s so much broader than most schools acknowledge. It’s not just “can you write a five-paragraph essay.” It’s: can you write clearly? Can you speak persuasively? Can you listen — actually listen, not just wait for your turn to talk? Can you tell a story? Can you disagree without destroying the relationship? Can you negotiate? Can you have the difficult conversation instead of avoiding it?

Writing, in particular, deserves special attention. Not because AI can’t write — it can — but because writing is still the best tool humans have for thinking clearly. If you can’t write it, you don’t understand it. The act of putting thoughts into words forces clarity in a way that nothing else does. This makes writing more important in the AI age, not less. The person who can think clearly, then direct AI to execute, will outperform the person who can’t think clearly but has access to the same tools.

The tragedy is how writing is taught in most schools: as format compliance. Five paragraphs. MLA citations. Word counts. The teacher is the only audience. Grammar errors are penalised more than unclear thinking. Students learn that writing is a performance to be graded, not a tool to be used. Then they graduate and can’t write a clear email. This helps no one.

The fix is simple in concept, difficult in execution: let kids write for real audiences about things they actually care about. When someone is genuinely going to read and respond to your writing, quality jumps dramatically. Do it for real.

Kids need to be healthy (and able to focus)

This one sounds too obvious to be on the list. It’s not.

Physical and mental health are infrastructure. Everything else degrades without them. A sleep-deprived child cannot focus, cannot regulate emotions, cannot think critically, cannot learn effectively. An anxious child cannot take creative risks. A child who never exercises is running their brain at a fraction of its capacity.

And here’s the uncomfortable modern reality: we are raising the first generation of children whose attention is being actively harvested by billion-dollar companies using the most sophisticated psychological manipulation tools ever devised. TikTok, Instagram, YouTube Shorts — these platforms are engineered to train young brains for novelty-seeking, not sustained attention. Every swipe is a tiny hit of dopamine. Every notification is an interruption that fragments focus. And we hand these devices to children whose prefrontal cortexes won’t be fully developed for another decade.

Focus is a muscle, not a trait. Any “unfocused” kid can play Minecraft for three hours straight. They can focus — just not on things that don’t engage them, in environments designed to distract them. The fix isn’t willpower lectures. It’s environment design.

The highest-impact interventions are boringly simple: protect sleep (no screens in bedrooms, consistent bedtimes — even thirty minutes less sleep measurably degrades attention). Protect boredom (stop filling every car ride and waiting room with a screen — boredom forces the brain to generate its own engagement, which is the foundation of sustained attention). Get them outside (nature literally replenishes attentional capacity and solves almost any bad mood). Delay social media for as long as humanly possible (every year you delay gives their attention circuits more time to develop without being hijacked).

And model it yourself. If you’re scrolling your phone during dinner, no amount of lecturing about screen time will land. Kids don’t learn from what you say. They learn from what you do.

The thinking tools

Learning how to learn

If you could give a child only one cognitive skill, this is the one.

The world changes fast. Careers shift. Industries appear and disappear. The person who can pick up new domains quickly — who understands how their own learning works, who knows what strategies actually produce understanding versus the illusion of understanding — that person has the ultimate hedge against uncertainty.

Here’s the irony: schools teach content but never teach how to learn that content. Kids are expected to absorb learning strategies by osmosis. They don’t. Instead, they default to what feels productive: re-reading notes, highlighting textbooks, cramming before exams. These strategies feel effective. They aren’t. Decades of cognitive science have shown they produce the illusion of understanding without wisdom.

The most powerful thing you can do is teach a child that the effort and discomfort of learning is the learning. When retrieval feels hard, that’s not failure — that’s your brain building stronger neural pathways. “It’s supposed to feel hard, that’s how you know it’s worth learning”.

Do the work for real. No short cuts, no quick fixes, no short term pump and dump of data. Let the kids find challenges, have failures, and then help them to overcome them until the project is completed. And everyone must finish. This builds a strength everyone needs to deal with the reality we live in. Letting them off easy isn’t a kindness, it’s just setting them up for future suffering because the big bad world will do this the moment they leave the nest. So set projects that take real effort to complete, but make sure they are also obtainable. No point in demoralising them! And then celebrate with them when they complete the work — it feels genuinely good out the other side. These are the projects I still remember effortlessly 30 odd years later, the learning was genuine and the results are life long.

The hardest thing as a teacher is that you too need to go through this process — they will learn from you. Don’t hide your own learning from them, talk about what you learned through the process with them, share your uncertainty that we would all make it through, but you did and how happy that makes you feel.

Critical thinking

In my research for educational critical thinking I found a lot of claims to teach it, but almost nobody seems to define it, because you can’t teach it as an isolated subject — it must be embedded in content, practised across domains, and explicitly connected to transfer.

So what is it actually? It’s the ability to distinguish good arguments from bad ones. To evaluate evidence. To spot your own biases. To reason under uncertainty. To question assumptions — including your own. To hold beliefs proportional to evidence and change your mind when the evidence changes.

In a world where AI generates confident-sounding content on any topic, where deepfakes are indistinguishable from reality, where information warfare is a daily occurrence — this isn’t a nice-to-have. It’s survival-grade.

The foundation is simpler than it sounds. For young children, one question does most of the work: “How do you know?” Ask it constantly. Ask it with genuine curiosity, not gotcha energy. “How do you know the moon is made of rock?” “How do you know that’s what happened?” “How do you know that’s fair?” This single question builds the instinct to verify rather than assume, and that instinct will serve them for life.

For older children, the toolkit expands: logical fallacies (taught through real examples from media and politics, not as lists to memorise), statistical reasoning (correlation is not causation, sample size matters, base rates exist), source evaluation, probabilistic thinking, and — perhaps most importantly — intellectual humility. The willingness to say “I was wrong” should be celebrated, not mocked. The willingness to say “I don’t know” should be respected, not punished.

Here’s the confrontational bit: if you teach a child to think critically, they will eventually question your beliefs. That’s not a bug. That’s the entire point. A parent who wants a critical thinker but can’t handle being questioned is asking for a party trick, not a skill.

Systems thinking

Most of the world’s biggest problems — climate change, pandemics, economic crises, social media dynamics, ecological collapse, supply chain attacks, dependency hell — are systems problems. They involve feedback loops, delays, unintended consequences, and emergence. They can’t be understood through simple cause-and-effect thinking. And yet simple cause-and-effect thinking is virtually all that schools teach.

Systems thinking is the ability to see the connections between things. To see that a result can loop back and affect the thing that caused it. To recognise that small changes can have enormous consequences, that obvious solutions often make problems worse, and that complex behaviour doesn’t require a central controller.

Perverse incentive is where people are incentivised to make a problem worse, usually by mistake. The Great Hanoi Rat Massacre in 1902 is a great example, where the colonial government created a bounty programme for each rat killed. To collect the bounty, people would need to provide the severed tail of a rat. So the rat catchers would capture rats, sever their tails, then release them back into the sewers so that they could produce more rats.

This is not just an amusing historical anecdote. This is the structure of almost every failed policy intervention. Good intentions, linear thinking, catastrophic unintended consequences.

Kids can learn all this. Even young children can trace chains of consequence: “What happens if all the frogs disappear? What ate the frogs? What did the frogs eat?” The question “And then what?” — asked two, three, four times in sequence — is the simplest and most powerful systems thinking tool there is.

For older students, the toolkit expands to causal loop diagrams, stock-and-flow models, system archetypes, and leverage point analysis. There are beautiful free tools like Loopy that let kids build interactive models of how things connect. But the foundation is just that question, asked relentlessly: and then what?

Early education and screens

My son wanted to learn how to crochet recently, and not knowing how to myself, I looked it up on YouTube. I watched this helpful girl teach me very clearly how to make a chain and so on — it was easy for me to learn and apply quickly as an adult. But my son, age seven, just went blank and felt overwhelmed. He had to watch it, and rewatch it many times over for the information to go in. But when I did it with him, correcting his hands, showing him the next loop, he quickly understood, “Ohhhh, I see now” and away he went along a row. What really struck me was how quickly he got it from me, and how slowly he got it from a screen.

I believe that in early education there should never be a substitute for a human teacher. Screen use should be used in utility only, to support what a real human being is trying to explain or teach — but never as a replacement.

Kids learn from us, it’s not just the topic they are learning but all the other skills that come with interacting with a real human being.

The new essentials

Financial literacy

Financial literacy is shockingly absent from most school curricula. There is no subject with a bigger gap between “impact on life outcomes” and “time spent in school.” Children graduate without understanding compound interest, taxes, insurance, investing, or debt — and then make six-figure decisions about student loans and mortgages with approximately zero relevant education. How unfair is that!

The basics are not complicated. Spend less than you earn. Invest the difference. Understand compound interest — it’s either your best friend or your worst enemy, depending on which side of it you’re standing. Understand that marketing exists specifically to make you spend money you don’t need to spend. Understand that most financial mistakes are emotional, not intellectual — loss aversion, herd behaviour, lifestyle inflation, the hedonic treadmill.

Start young with three jars: spend, save, invest. Give kids real money and real decisions. Let them make mistakes with $20 so they don’t make the same mistakes with $20,000. The parent who rescues their child from every poor financial decision is raising an adult who will need rescuing forever.

If this feels scary or unknown to you, great! Learn with your children at the same time — set you both free.

Creativity and problem-finding

Creativity is not a gift bestowed upon a lucky few. It’s a skill. It can be trained. And schools are remarkably effective at suppressing it.

The evidence is clear and a little depressing. Young children test extraordinarily well on measures of divergent thinking. By the time they’ve been through several years of school — several years of “one right answer,” “colour inside the lines,” “do it this way” — those scores have dropped dramatically. We are not failing to teach creativity. We are actively training it out of them.

Creativity is the ability to see problems others don’t see, to make unexpected connections between ideas, to generate many possibilities and then select the best ones. In an AI world, this matters enormously, because AI is very good at solving well-defined problems but has no ability whatsoever to identify which problems are worth solving. Problem-finding is the new premium skill. The question is more valuable than the answer.

The principles for developing creativity are, ironically, well-understood:

Constraints breed creativity. “Write a story in 50 words” produces more creative results than “write a story.” Dr. Seuss wrote “Green Eggs and Ham” using only 50 different words, on a bet. Give kids limitations and watch them get more creative, not less.

Volume before quality. The best way to have a good idea is to have lots of ideas. Linus Pauling knew this. So did Edison. The “bad idea” exercise — deliberately generating terrible ideas — removes the pressure of quality and often produces surprisingly good ones through inversion.

Cross-domain connection. Creativity often comes from combining ideas that nobody thought to combine. Velcro came from an engineer noticing burrs on his dog’s fur. Expose kids to diverse domains and give them space to connect them.

Play. Unstructured free play is creative training in disguise. Imaginative play develops narrative thinking, improvisation, and perspective-taking. Construction play develops engineering creativity. The decline of free play in childhood — replaced by structured activities and screens — is directly linked to declining creative measures in children. Protecting play isn’t soft. It’s strategic.

And most importantly: make things. Creativity is a verb, not a noun. You develop it by building, writing, drawing, composing, coding, cooking, designing, constructing. The medium matters less than the act. A child who is always consuming and never creating is a child whose creative muscles are atrophying.

Agency and self-direction

Without agency, none of the other skills get deployed. A child who can think critically but lacks initiative won’t use their thinking. A creative child without agency won’t create anything. Agency is the difference between having capabilities and actually using them.

Agency is the belief and capacity that you can influence your own life. That your actions matter. That you can set a goal, make a plan, and work toward it without someone standing over you.

And here is the deepest structural irony of education: schools are designed to reduce agency. Adults decide what you learn, when, how, and how you’re assessed. The schedule is imposed. Compliance is rewarded. Questioning is discouraged. Risk-taking is punished through grades. Students spend twelve years being told what to do, and then at eighteen we expect them to suddenly self-direct. We don’t build the muscle, and then we’re surprised when it’s weak.

The research on motivation, primarily from Deci and Ryan’s Self-Determination Theory, is clear: humans need three things to be intrinsically motivated — autonomy (feeling that you have choice), competence (feeling that you can master challenges), and relatedness (feeling that you belong). When these needs are met, motivation flourishes. When they’re thwarted — as they routinely are in school — motivation collapses into compliance (“I’ll do it because I have to”) or defiance (“You can’t make me”).

Building agency requires one thing above all: practice. You develop agency by making real decisions and experiencing real consequences. This means giving children progressively more autonomy over time — not unlimited freedom (which overwhelms), but graduated responsibility that grows as their capacity grows.

For young children, this starts small: let them dress themselves, even if it takes longer. Let them choose. Let them struggle with something before you swoop in to help. For older children, it means owning their homework (it’s their homework — not yours), managing their own time, handling their own conflicts, making their own financial decisions with their own money. For teenagers, it means substantial life autonomy: managing their schedule, driving their own career exploration, facing real consequences for real choices.

This is the hardest thing for parents. Watching your child make a choice you know will turn out badly. Every instinct says intervene. But unless it’s genuinely dangerous, the lesson from experience is ten times more powerful than the lesson from your warning. Probably more. The parent who rescues their child from every difficulty is raising an adult who will need rescuing forever.

Do less. Not less loving. Less managing. Be a consultant, not a project manager. Express confidence in their capability. “I know you can handle this.” And then let them handle it.

AI literacy

Here is a sentence I want you to sit with: AI literacy is as foundational now as reading was after the printing press.

Not “AI skills.” Not “how to use ChatGPT.” AI literacy — understanding what AI is, what it can and can’t do, when to trust it, when to verify, how it works at a conceptual level, and what the ethical implications are for society.

AI is becoming embedded in everything — search, communication, work, entertainment, healthcare, education. People who can’t work with AI will be at a massive disadvantage. People who can’t evaluate AI output will be easily misled. People who don’t understand AI’s limitations will make dangerous decisions by trusting it or believing it’s another person. The gap between AI-literate and AI-illiterate will be as consequential as the gap between those who could read and those who couldn’t, in a world increasingly made of text.

The single most important thing to understand about AI: it is confident, not correct. It produces answers with equal polish whether right or wrong. It doesn’t “know” when it doesn’t know. It will cite sources that don’t exist, invent statistics that sound plausible, and explain things with perfect fluency that are entirely fabricated. This is not a bug that will be fixed. It’s a structural feature of how these systems work.

Teaching kids to use AI without teaching them to evaluate AI is like teaching them to drive without teaching them about brakes.

The approach needs to be hands-on: let kids experiment with AI tools, test their limits, find their failure modes. “Stump the AI” is a game every classroom should play. Train simple models (Google’s Teachable Machine lets kids build image classifiers with zero code). Explore bias: does the AI respond differently to different names, genders, or ethnicities? Discuss the ethics: if a hiring algorithm discriminates, is that fair? Whose data trained this model? Who profits?

And here’s the question that gets to the heart of it, the one that stops kids (and adults) in their tracks: “If AI can write your essay, what’s the point of learning to write?”

The answer, if you’ve been paying attention: writing is thinking. The essay was never the point. The thinking was the point. The essay was just evidence that thinking happened. If you outsource the essay to AI, you’ve outsourced the thinking. And a person who can’t think clearly is not made more capable by having a machine that can produce text — they’re made more dependent.

What about the traditional subjects?

You might be wondering if I’m suggesting we bin mathematics and stop teaching history. I’m not. Traditional subjects still matter. But they need to be taught entirely differently.

Mathematics should emphasise reasoning over procedure. The ability to think mathematically — to model, to prove, to abstract — is more valuable than ever. The ability to do long division by hand is not. Statistics and probability should be elevated from elective to core — in a world of data, everyone needs to reason about uncertainty, base rates, and correlation versus causation.

Science should centre on the method, not the facts. The scientific approach — form a hypothesis, test it, update your beliefs based on evidence — is the best system humanity has ever developed for understanding reality. Memorising the periodic table, less so. Systems thinking and ecology deserve far more weight.

History should be taught as analytical reasoning about human behaviour and institutions, not as dates and names to memorise. Why did empires fall? How does propaganda work? What patterns repeat across cultures and centuries? This is practical wisdom.

Reading — deep, sustained reading of complex texts — is a genuine competitive advantage in an age of distraction. The ability to sit with a long, difficult book for hours is a skill that most adults are losing and most children never develop. Protect it.

The arts develop taste, expression, discipline, and collaboration. AI can generate art, but human curation, emotional resonance, and aesthetic judgment remain irreplaceable. And arts education develops cognitive capacities — creativity, attention, self-expression — that transfer everywhere.

The subjects aren’t the problem. The approach is the problem. Lectures, memorisation, standardised tests — these optimise for what AI does best. Project-based learning, mastery-based progression, portfolio assessment — these optimise for what humans do best.

What we should stop doing

We should spend significantly less time on rote memorisation of facts. Not zero — foundational knowledge matters and you need things in your head to think with. But the balance is wildly off. A child who has memorised two hundred historical dates but can’t evaluate a source, construct an argument, or identify a logical fallacy has been poorly served.

We should stop pretending that standardised test preparation is education. It’s training children to be good at the thing AI is best at — recalling information and applying procedures under time pressure. Every hour spent on test prep is an hour not spent on critical thinking, creativity, communication, or emotional intelligence. We are optimising for obsolescence!

We should stop treating compliance as the primary behavioural goal. “Do it because I said so” produces workers for assembly lines. It does not produce people who can think independently, take initiative, or navigate ambiguity. The future belongs to people who can direct themselves, not to people who wait to be told.

We should stop assigning homework that AI can do. If a student can type the prompt into ChatGPT and get a passing grade, the assignment is measuring the wrong thing. Assign work that can’t be outsourced: build something physical, present to a real audience, have a genuine conversation, reflect on a personal experience, teach someone else what you learned.

And we should stop banning AI in schools. This is the educational equivalent of banning calculators in 1985. The world will not pause its transformation to wait for the education system to catch up. The response to AI should not be “pretend it doesn’t exist.” It should be “fundamentally rethink what humans need to be good at.”

Adults in the room

Everything I’ve touched on — every skill, every principle, every recommendation — funnels down to one truth: the adults matter more than the system.

A brilliant curriculum delivered by an emotionally unintelligent teacher will fail. A perfectly designed school with parents who rescue their children from every difficulty will produce fragile graduates. The best critical thinking programme in the world is undermined by a household where questioning authority is punished.

Children don’t learn from what we say. They learn from what we do.

If you want your child to be emotionally intelligent, you need to be emotionally intelligent — or at least visibly working on it. If you want them to think critically, you need to be willing to have your own beliefs questioned. If you want them to be resilient, you need to let them fail. If you want them to communicate well, you need to listen to them — really listen, not half-listen while scrolling your phone.

If you want them to love learning, you need to love learning yourself. Let them see you struggle with something new. Let them hear you say “I don’t know, but let’s find out.” Let them watch you change your mind.

This is the part that no education reform can legislate. No policy can mandate it. No technology can automate it. The quality of the relationship between adult and child is the single strongest predictor of how a child develops — emotionally, cognitively, socially, and morally.

You don’t have to be perfect. You have to be present. You have to be honest. You have to be willing to repair the rupture when you get it wrong — because you will get it wrong, regularly, and that’s fine. Children learn more from watching you fix your mistakes than from watching you never make them.

The bottom line

The future is uncertain. It always has been, but the pace of change has made the uncertainty more visceral. We can’t predict what jobs will exist in twenty years, what technologies will emerge, what crises will need solving.

But we can predict what humans will need to be good at, because the answer hasn’t changed as much as we think. The ability to understand yourself. The ability to understand others. The ability to think clearly, communicate effectively, and learn new things quickly. The ability to create, to persist, to adapt, to see the connections between things. The courage to act on your own judgment and the humility to change your mind.

These aren’t new requirements. Aristotle would have recognised most of this list. What’s new is the urgency. In a world where machines handle the routine — the memorising, the calculating, the procedure-following — what remains is the deeply human. And it turns out that the deeply human is also deeply valuable.

The best education doesn’t produce children who can pass tests. It produces adults who can navigate a complex, uncertain, beautiful, terrifying world with competence and compassion. Adults who can think for themselves, feel for others, and do something useful with both. Don’t you wish to be this kind of adult yourself?

The good news is that we don’t have to figure this out either — we know how to develop these capacities. The research is clear. The methods are proven. What’s missing is the institutional courage to actually do it, and the individual courage, as parents and teachers, to prioritise what matters over what’s measurable.