What Kids Should Learn While AI Rewrites the World of Work — and Why the Basics Still Matter
The question now comes up at every parents' evening: are times tables still worth it when every device in the house can do the maths? Is the slog of learning to read still worth it when AI reads aloud and summarises? And does our child need to learn to code now, just to have a job in 2040?
The short answer to the first two: yes — more urgently than before. The answer to the third: not necessarily. Because the goal cannot be to predict which jobs will exist in fifteen years. Nobody can do that. The goal is to raise a child who can learn new things, think independently, work with technology, and adapt when the world turns again.
Don't optimise your child's education for a job title in 2040. Optimise it for capability, curiosity and adaptability.
🧭 Six pillars that hold
We would build a child's school years around six abilities. None of them is tied to a particular technology, and all of them can be developed across childhood and adolescence.
- 1. Reading, writing, communicating. Read books, understand complex texts, write clearly, explain ideas, ask good questions. That includes telling facts from opinions — and talking to people, not only to machines.
- 2. Mathematics and logical thinking. A solid foundation in arithmetic, fractions, percentages, algebra, geometry, probability and statistics. Maths teaches children to reason, estimate, spot errors and read data — all of which stay valuable even when the machine does the calculating.
- 3. Technology, AI and computational thinking. Understand at a basic level how computers, software, AI and robots work. Code, automate, handle data, use AI tools responsibly. The aim is to direct and create technology — not merely consume it.
- 4. Problem-solving and creativity. Real problems with no obvious answer. Design, experiment, build, fail, improve, invent. AI produces plenty of answers; recognising which problem is the right one, and what is worth building at all, stays with the human.
- 5. Human and social skills. Empathy, listening, teamwork, negotiation, leadership, patience and handling disagreement. Working with people from different backgrounds and taking responsibility for a shared outcome.
- 6. Adaptability, health and independence. Learn independently, manage time, absorb setbacks, look after body and mind, understand money, and keep picking up new skills. Our children will probably change career several times.
That list lines up broadly with what employers report in the World Economic Forum's Future of Jobs Report 2025: analytical thinking first, then resilience and flexibility, creative thinking, technological literacy, AI and data skills, and lifelong learning. The survey runs to 2030 — a useful signal, then, but not a forecast for the 2040s.
⚖ Why the basics are becoming more important, not less
The common reflex goes: if the machine reads and calculates, the child can skip that and learn AI instead. That is the most expensive trade available.
Reading and writing
AI can summarise a contract, write a report and explain a book. But a child who cannot read deeply cannot judge whether that summary is accurate, misleading, or missing the one thing that mattered. And that judgement is exactly the new core task: not producing the text, but checking it. A child who can't do that is entirely dependent on the machine being right — and it sounds just as convincing when it isn't.
Priority: fluent reading, comprehension, vocabulary, writing and critical evaluation.
Mathematics
AI calculates faster than any human. Even so, somebody has to understand the order of magnitude, the risk behind it, what something costs — and whether a result is plausible at all. A person who understands maths can check a machine-generated calculation and make better decisions. A person who doesn't will never notice the error.
Priority: number sense, mental estimation, fractions, algebra, statistics and mathematical reasoning.
Why automaticity is still required
There is a sober cognitive reason for this. Working memory is narrow. A child decoding every word one at a time has nothing left at the end of the sentence for what the sentence meant — that has been well described since the 1970s. A child who has to stop and work out 7 × 8 loses the thread of the word problem. Only once the basic craft runs automatically does the head come free for what you actually wanted: understanding, checking, arguing. Automaticity isn't the opposite of thinking. It is its precondition.
And one important distinction
"Basics" does not mean "stop at the basics". A child should build a solid foundation and then keep going, as far as interest and ability carry them. Anyone heading for engineering, medicine, research or robotics will need substantially more than confident mental arithmetic. The World Economic Forum lists reading, writing and maths among the comparatively stable skills — which is not a reason to neglect them, but a reason to combine them with the new ones.
🤖 What children actually need to learn about AI
Not every child needs to become an AI engineer. Every child should become AI-literate. Four things matter:
- Understand what AI can and cannot do. Patterns, training data, made-up answers, bias, privacy — and why a system can sound confident and still be wrong.
- Use AI as a tutor, not as a substitute for thinking. Have it explain a hard topic, ask it for practice questions, have it challenge your answer, get feedback. But attempt it yourself first, then ask.
- Learn to build things. Start with visual programming, later perhaps Python, electronics, sensors, simple robots, small automations. The point is the feel for how instructions turn into actions.
- Understand the real-world consequences. Privacy, misinformation, copyright, fairness, safety — and the question of who is actually responsible when an automated system gets it wrong.
The OECD and the European Commission have published a joint framework for exactly this, aimed at primary and secondary education. It sorts AI literacy into four stages: engage with AI, create with AI, manage AI responsibly, and help shape where AI goes. The order is worth noticing — operating the tool does not come first.
📅 A learning path by age
Flexible guard rails, not a curriculum. Alongside all of this, children need time to play, to have friends, to read for pleasure and to follow their own interests.
5 to 8 — build the foundation
- Read aloud together and encourage reading alone.
- Practise numbers in daily life: games, cooking, shopping, puzzles.
- Build with Lego, cardboard, magnets, simple mechanical toys.
- Get outside, ask questions, take curiosity seriously.
- Use technology creatively and in moderation.
9 to 12 — learn how things work
- Grow reading comprehension and mathematical reasoning.
- Scratch, first programming, robotics kits, science experiments.
- Search for information and verify the source.
- Use AI occasionally for explanations — with an adult alongside.
- Finish projects: a sensor, a game, a plant experiment, a small website.
13 to 16 — depth and independence
- Carry on with algebra, geometry, statistics and science.
- Programming, spreadsheets, data analysis, AI concepts.
- Build larger projects — and document them.
- Speak in public, write, work in a team, present.
- Look into fields: engineering, healthcare, design, business, trades, arts, environment.
16 to 20 — meet real work
- Choose further education or vocational training by interest and strength.
- Internships, apprenticeships, work experience, serious projects.
- Learn the professional AI tools of the chosen field.
- Build a portfolio of things actually built, researched or achieved.
- Personal finance, professional communication, responsibility at work.
🏠 What parents can do at home
The most effective preparation is less about buying technology than about establishing a few habits:
- Protect reading time. Let them choose — stories, science, history, biographies. Then talk about it.
- Bring maths into daily life. Work out costs, compare prices, measure ingredients, estimate travel time, read charts.
- Give real responsibility. Cooking, repairing, organising, running a budget, caring for animals or plants, solving household problems.
- Projects instead of consumption. A small robot, a garden experiment, a story, a model, a game — worth more than endless videos about technology.
- Allow productive struggle. Don't hand over the answer immediately — neither you nor the AI. Learning to persist through difficulty is an enormous advantage.
- Find out what they enjoy. No child has to be pushed towards coding. The working world will still need healthcare, trades, teaching, design, science, business and the arts.
The other factors that shape how school goes — sleep, food, routines — we collected in what actually helps kids succeed at school.
✅ The homework rule for the AI age
One simple family rule carries surprisingly far:
First think. Then ask. Finally verify.
- First: try the problem yourself, read the text yourself, write your own answer.
- Then: have the AI explain, give hints, answer questions, suggest another approach.
- Finally: check the result, explain it in your own words, and solve a similar problem without help.
In maths and reading especially, that order decides everything. AI should increase understanding, not manufacture the illusion of it — and the difference between the two only shows up in the exam.
🎮 Where ABC Smash and Math Fighter fit
The awkward thing about foundational skills is that they need volume. A word doesn't become automatic through explanation but through many successful encounters spread over weeks; the same goes for 7 × 8. Distributed practice is one of the best-established learning principles there is — and simultaneously the thing that least often happens at home.
A game can take over that part. Short daily rounds, instant feedback, a repetition schedule that brings material back on its own: in a randomised trial with maths apps, children doing short daily sessions learned measurably more than the comparison group. If screen time is happening anyway, this is one of the few ways to point it at the foundation.
📖 ABC Smash — reading until it sticks
Letters, syllables, everyday words, the 250 most frequent words, adventure lessons and "read and answer" — in short daily rounds, in eight languages, offline in the car and on the plane. For children from roughly 5 to 12.
⚔ Math Fighter — mental arithmetic that sticks
Addition, subtraction, multiplication and division in short fights, with a belt system from white belt to black. Exactly the number sense that later tells you a result cannot be right. For children from about 6.
And the honest framing we attach everywhere: these games take over the repetition — letters, words, mental arithmetic. They do not take over the understanding, the conversation about a book, or the question of whether an answer is plausible. How much screen time makes sense overall is covered in gaming and screen time.
💬 Our recommendation in one sentence
Raise children who can read deeply, reason mathematically, understand technology, build useful things, work well with others — and keep learning for the rest of their lives.
We don't know which job titles will exist in 2040. That these six things will still count is about the safest bet available on the future.
📚 Sources
- Cepeda, N. J., Pashler, H., Vul, E., Wixted, J. T., & Rohrer, D. (2006). Distributed practice in verbal recall tasks: A review and quantitative synthesis. Psychological Bulletin, 132(3), 354–380. doi:10.1037/0033-2909.132.3.354
- LaBerge, D., & Samuels, S. J. (1974). Toward a theory of automatic information processing in reading. Cognitive Psychology, 6(2), 293–323. doi:10.1016/0010-0285(74)90015-2
- OECD & European Commission. AI Literacy Framework for Primary and Secondary Education (AILit). ailiteracyframework.org
- Outhwaite, L. A., Faulder, M., Gulliford, A., & Pitchford, N. J. (2019). Raising early achievement in math with interactive apps: A randomized control trial. Journal of Educational Psychology, 111(2), 284–298. doi:10.1037/edu0000286
- Sweller, J., van Merriënboer, J. J. G., & Paas, F. G. W. C. (1998). Cognitive architecture and instructional design. Educational Psychology Review, 10(3), 251–296. doi:10.1023/A:1022193728205
- Willingham, D. T. (2007). Critical thinking: Why is it so hard to teach? American Educator, 31(2), 8–19. aft.org
- World Economic Forum (2025). The Future of Jobs Report 2025. weforum.org
The Future of Jobs Report is based on an employer survey with a 2030 horizon. It describes expectations, not certainties — and certainly not a forecast for the 2040s.