Skip to main content
Raising Children for the AI Era: 7 Skills That Matter More Than Memorising Answers
Home / Blog / Human Intelligence & AI, Learning Science & AI Tutors
Human Intelligence & AI, Learning Science & AI Tutors

Raising Children for the AI Era: 7 Skills That Matter More Than Memorising Answers

The scariest thing about the AI era is not a bad report card. It is raising a child with the playbook that worked twenty years ago, in a world where answers now cost almost nothing.

That idea opens a short talk by a Taipei-based educator known as Teacher Lin (林老師), recorded after a two-day youth AI bootcamp in Taipei. By his own account he has worked with young people for about 25 years. After the camp, parents kept asking him the same question: with AI arriving, what should my child actually learn so that he or she is not replaced?

His answer is a list of seven abilities. It is a good list, and it deserves more than a translation. So this article keeps his seven-skill framework, adds what the research says about each skill, flags the places where the evidence calls for a footnote, and turns every skill into something a parent can try this week. Singapore readers also get one surprise from the OECD data, in the section on creativity.

Credit where it is due: the framework is Teacher Lin’s. The research, the caveats, the exercises, and the Singapore and workplace angles are ours.

What Skills Do Children Need in the AI Era?

Seven: asking good questions, filtering information, creating original work, developing taste, collaborating, assembling resources around a goal, and sustaining interest through setbacks. None of them is a school subject, and all of them can be practised at home in about ten minutes a day.

The short version, if you only have a minute:

  1. Ask better questions than the answers you are given.
  2. Filter what you read before you believe it or share it.
  3. Create things that have no single right answer.
  4. Develop taste, the ability to say what is good and why.
  5. Collaborate, because trust is an asset that compounds.
  6. Integrate resources by starting from the goal, not from what you happen to have.
  7. Sustain interest long enough to get good at something.

Four Shifts Behind the Seven Skills

Teacher Lin opens with four statements. Each is useful, and each needs a small correction to be accurate. Here they are, restated with the evidence.

1. Answers Are Getting Cheaper, So Questions Are Worth More

The talk says knowledge is losing value quickly, and illustrates it with the idea that what once took sixteen years to learn can be organised by AI in sixteen seconds. Treat that number as rhetoric rather than a statistic. The direction, however, is right. Producing a fluent answer to a well-formed question now costs almost nothing. What remains scarce is knowing which question is worth asking, and that is where the first skill comes in.

2. Information Is Abundant, So Judgement Is Scarce, and Judgement Needs Knowledge

This is the statement that needs the biggest footnote. If knowledge is cheap, it is tempting to conclude that children should stop learning facts. The cognitive science points the other way. Daniel Willingham, a cognitive scientist who has reviewed the research on critical thinking for educators, argues that critical thinking is not a free-floating skill. It depends on knowing the subject at hand, and readers who lack background knowledge struggle to understand a claim, let alone judge it.

So the sensible reading is this: knowledge is no longer the finish line, it is the raw material for judgement. A child who knows nothing about how interest rates work cannot tell whether an AI summary of a loan offer is sensible. Teach the facts, then teach what to do with them.

3. Routine Work Is Getting Cheaper, So Human Skills Rise

The World Economic Forum’s Future of Jobs Report 2025, based on a survey of more than 1,000 employers across 55 economies, gives a useful picture of what employers say they need. Analytical thinking is a core skill for about 69 percent of employers, resilience, flexibility and agility for about 67 percent, creative thinking for about 57 percent, motivation and self-awareness for about 52 percent, and curiosity and lifelong learning for about 50 percent. The report also expects creative thinking, resilience, and curiosity to grow in importance.

Look at that list again. Almost all of it is Teacher Lin’s seven skills in different words. Technological literacy is also on the list, so learning to use AI is part of the picture, but only part.

4. “People Who Use AI Will Replace People Who Do Not” Is a Slogan; How You Use It Is the Finding

The talk’s third statement is a popular slogan. It is plausible, but it is not a measured result. What has been measured is more interesting. In a field experiment published in PNAS in 2025, Hamsa Bastani, Osbert Bastani and colleagues gave nearly 1,000 high school maths students in Turkey access to a GPT-4 tutor during practice. Students with a standard ChatGPT-style assistant scored 48 percent higher on practice problems. Students with a tutor designed to guard learning scored 127 percent higher. But when access was taken away, the standard-assistant group scored 17 percent worse than students who never had AI at all, while the guarded tutor largely removed that harm. The authors describe students using the standard tool as a crutch.

The gap between the two groups came from how the tool was designed and used, not from the model. One honest limit: this was one school, one subject, and teenagers, so treat it as strong evidence of a mechanism, not a universal law. The lesson for parents is practical. Using AI is not the skill. Using it so that you still learn is.

Skill 1: Ask Better Questions

Many parents quietly dread the child who keeps asking why. Teacher Lin’s advice is to welcome it. The most competitive children in the future will not be the ones who answer best, because AI can answer almost anyone’s question. They will be the ones who ask the question that nobody else thought to ask. AI cannot decide which question matters to your child.

What a Good Question Looks Like

A good question narrows ignorance. It reveals what the asker knows, what they do not, and what would change their mind. A useful way to coach this is a four-rung ladder:

  1. What is it? A fact question.
  2. Why does it happen? A cause question.
  3. What if it were different? A hypothesis question.
  4. How would we know? An evidence question.

Most children get stuck on rungs one and two, because school rewards answers to those. The skill grows when someone helps them climb.

What Parents Say Matters

Not knowing the answer is fine. Teacher Lin’s advice is to ask the question together, including asking AI together. The phrase that works is “good question, how could we find out?” The phrase that kills it is “stop asking so many questions.”

Try This at Home

  • Question of the day. At dinner, everyone brings one question they cannot answer. Pick one and spend ten minutes investigating it with a book, a search, or an AI tool.
  • Climb the ladder. Take any question your child asks and help them upgrade it by one rung.
  • Question notebook. Write questions down and mark the ones still unanswered after a month. Those are usually the interesting ones.
  • Predict first, then ask AI. Have the child guess the answer, then ask the AI, then ask: what is the best question this answer did not address?

Warning sign: a child who says they do not know what to ask, or who only asks whether something will be on the exam.

Skill 2: Filter Information Before Believing or Sharing It

Teacher Lin says the biggest future competition is not intelligence but judgement, and that more information also means more rubbish. His habit for children is to hold back before believing or forwarding anything, and to ask three questions: what is the source, where is the evidence, and is there a different point of view?

Add One Technique: Read Laterally

Those three questions are a strong start. Research gives a practical way to apply them. In a 2017 Stanford study, Sam Wineburg and Sarah McGrew compared 10 historians, 10 professional fact checkers, and 25 Stanford undergraduates as they judged live websites. The historians and students tended to stay on the page, and were swayed by official-looking logos and domain names. The fact checkers did something different: they left the page quickly and opened new tabs to find out what others said about the source. They reached more reliable conclusions in a fraction of the time. The technique is called lateral reading, and it is easy to teach a child: before trusting a page, find out who is behind it by asking somewhere else.

The AI Twist

AI tools can state wrong facts, and invent references, in a confident and fluent tone. That makes verification a core skill, not an optional one. It also connects back to the footnote above: a child can only filter well in areas where they know enough to notice something is off. Knowledge and filtering grow together.

Try This at Home

  • The three-question habit. Before sharing anything, the child answers: source, evidence, other views.
  • The two-tab rule. No claim is accepted until the child has opened a second, independent source about it.
  • Family group chat detective. Every family has forwarded messages in a WhatsApp group. Pick one a week and investigate it together. It is real, low-stakes practice.
  • Check the AI. Ask an AI tool for sources on a topic, then verify that each source exists and says what the AI claimed.
  • Steelman. Ask the child to state the best version of a view they disagree with, in a way its supporters would accept.

Warning sign: a child who shares first and checks later, or who trusts whoever sounds most confident.

Skill 3: Create Things With No Single Right Answer

Teacher Lin’s argument is that AI is excellent at reproducing patterns, and the most precious human contribution is creating something new. Yet many children spend every day completing tasks set by others: tuition, tests, homework, with no time left to daydream, think, or explore. His point is that creativity does not grow in a classroom of worksheets. It grows through trying, making mistakes, and exploring. Children need permission to do things with no standard answer, and to finish an imperfect piece of work, because finished matters more than perfect.

The Singapore Surprise

The OECD gave the first-ever international test of creative thinking in its PISA 2022 round, covering 64 countries and economies. Singapore came first. Singapore students averaged 41 out of 60 points against an OECD average of 33, and 58 percent reached the top two proficiency levels, against 27 percent across the OECD.

Now the surprise. Singapore’s Ministry of Education reported that the same students were less likely than their OECD peers to believe in their own creativity. Fewer than half thought they could produce good stories (47 percent against 61 percent across the OECD), good drawings (42 percent against 55 percent), or invent new things (43 percent against 57 percent). In other words, Singapore children are good at creative tasks and are less convinced that they are creative people.

Belief is something families can influence through what they notice and praise. The OECD also found a small positive link between scoring well and holding a growth mindset about creativity, meaning the view that creativity can be developed. Creativity is also not magic: like critical thinking, it draws on knowledge, because new combinations need something to combine.

Try This at Home

  • The unscheduled hour. One hour a week with nothing planned and no screen. Boredom is the point.
  • Make before you search. For any creative task, the child produces a first attempt before asking AI or the internet.
  • Ship small. Each week the child finishes one small thing, a comic, a recipe, a gadget, a song, and shows it to the family. Praise the finishing and the process, not the talent.
  • Constraint games. Build the tallest tower from ten household items, or write a story in exactly fifty words.
  • AI as critic, not author. Ask the AI to critique the child’s draft and suggest three questions, not to write it.

Warning sign: every hour of the week is scheduled, the child dislikes tasks with no right answer, or says “I am not creative.”

Skill 4: Develop Taste

Many people treat art and music as hobbies. Teacher Lin disagrees: he argues that aesthetic sense will become a competitive strength. His advice is to take children to read a good book, see an exhibition, and listen to good music, and above all to ask what they like and why. In his view, taste is the start of value judgement, and a child who can tell what is truly good is harder to lead by the nose with rubbish information, low-quality entertainment, or poor values.

Why It Matters More Now

When AI can generate fifty versions of a logo, a story opening, or a slide in seconds, the scarce act is no longer producing options. It is choosing among them and improving the one you pick. That is taste at work. Taste is built by two things: exposure to good work, and the habit of explaining a preference in words.

Try This at Home

  • Like, because. Make this the family sentence stem: “I like this because…” Accept any honest reason, then ask for a second one.
  • Pick three, defend one. Ask an AI tool for ten options for something small, such as a name for the family pet or an opening line for a story. The child picks the best three, then defends one.
  • Library and museum visits with one question. At each, ask: which one would you take home, and why?
  • Make it better. Take something mediocre and improve one element at a time, saying what changed and why.

Warning sign: everything is “okay,” and the child cannot say why they prefer one thing over another.

Skill 5: Collaborate and Build Trust

Teacher Lin’s fifth point is that the future is less about going it alone and more about groups creating value together. The people least likely to be replaced are not necessarily the strongest individuals, but the ones who collaborate best. His suggestion is to change the question parents ask. Instead of “what did you score today?”, ask: who did you help today, did you understand your teammate, and did you explain your idea clearly? His reasoning is that the biggest future asset is not only ability but trust.

Why AI Raises the Stakes

AI makes solo work faster, which makes the human parts of teamwork more distinctive: listening, explaining, disagreeing without damage, and keeping promises. Employers notice. In the WEF survey, at least half of employers listed empathy and active listening as a core skill.

Try This at Home

  • The three dinner questions. Who did you help today? Did you understand your teammate? Did you explain your idea clearly?
  • Family project with roles. Plan a trip or a meal where each person has a defined role and a deliverable.
  • Teach-back. The child explains what they learned to a sibling in two minutes. If the sibling cannot follow, the explanation needs work.
  • Disagree well. Argue one side of a family decision, then switch sides.
  • Group project reflection. After any group assignment: what did I contribute, and what did a teammate contribute?

Warning sign: group work is seen as extra effort, one child always takes over, or one always lets others do the work.

Skill 6: Start From the Goal, Then Assemble the Resources

The most valuable person in the future, says Teacher Lin, is not the one who can do everything. It is the one who knows who to ask, which tool fits, and how to combine them. He teaches children one habit: do not start by asking what you have, start by asking what you want to accomplish. The goal decides the method, the method decides the resources, and the resources create the result. His summary is goal first, resources second, and do not wait for resources before you set a goal.

Why It Matters in the AI Era

Tools are now abundant and cheap. The scarce capability is choosing the right ones and wiring them together around a real problem. It is the same lesson we draw for businesses in our piece on the AI implementation gap: tools alone do not solve problems, a well-designed workflow does.

Try This at Home

  • The goal sentence. “I want to ___ by ___ so that ___.” Only then list what is needed, who can help, which tools to use, and the first small step.
  • A real mini-project. Plan a family meal on a fixed budget, or run a small charity sale. The child sets the goal, then finds the resources.
  • Ask well. Have the child write a short message asking an adult for help, with context and one specific request.
  • Tool menu. List the people, books, tools, and skills the child can already reach. Most children underestimate this list.

Warning sign: the child starts with “I do not have X, so I cannot.”

Skill 7: Sustain Interest Through Setbacks

Teacher Lin calls this the most important skill. AI can replace knowledge and many skills, but it cannot replace a person’s real interest. In his view, great achievements come not from being clever but from being willing to invest over a long time. Parents often say “do not learn that, it is useless,” yet the most valuable abilities of the future may look useless today. His chain is: interest produces focus, focus builds ability, and ability eventually creates value.

The Footnote: Do Not Oversell Grit

This is the skill where research asks for the most care. A 2017 meta-analysis by Marcus Credé, Michael Tynan and Peter Harms pooled 88 independent samples covering about 66,800 people. It found that grit is only moderately related to performance, and that it overlaps very strongly with conscientiousness, an older and well-studied trait. It also found that the perseverance-of-effort part of grit was more useful than the consistency-of-interest part. The lead author concluded that other factors, such as study habits, adjustment, and low test anxiety, were more strongly related to academic performance.

The practical reading is balanced. Perseverance is valuable, and so is real interest. But “just be gritty” is not a plan. Build the habits that make persistence possible, protect the things the child loves, and avoid turning passion into pressure. A forced passion is not a passion.

Try This at Home

  • Interest log. Keep a simple list of what the child chooses to do when nobody is asking. Look for patterns over months, not days.
  • The season rule. A family rule that a child finishes the term or season of an activity before deciding to quit, followed by an honest conversation about what they learned. This is a suggestion, not a research finding, so adapt it to your child.
  • Failure review. After a setback, ask three things: what happened, what did you learn, and what will you try next?
  • Habits over speeches. Short, regular practice sessions with a clear plan beat motivational talks.
  • Model it. Let the child see you work at something difficult over a long time, including the dull parts.

Warning sign: constant switching with no depth, or doing everything only for parental approval. If a child shows persistent distress, speak to the school counsellor or a doctor. Resilience training is not a substitute for support.

The Seven Skills on One Page

Print this and put it on the fridge.

Skill Daily question Weekly practice Warning sign
1. Ask What did you wonder about today? Question of the day, climb one rung Only asks if it is on the exam
2. Filter Source, evidence, other views? Investigate one family group chat forward Shares first, checks later
3. Create What did you make today? Finish and show one small project Every hour is scheduled
4. Taste What did you like, and why? Pick three, defend one Everything is “okay”
5. Collaborate Who did you help today? Teach-back, group reflection One child always takes over
6. Integrate What do you want to get done? A goal sentence for a real mini-project Starts with what is missing
7. Sustain What was hard, and what next? Failure review, interest log No depth, or only for approval

How to Use AI With Your Child Without Weakening These Skills

The Bastani study is the reason this section exists. The same tool can build skills or hollow them out, depending on how it is used. Six rules follow from it.

  1. Attempt first, then ask. The child tries the problem before opening the AI. Struggling for a few minutes is where much of the learning happens.
  2. Ask the AI to coach, not to answer. A prompt such as “Do not give me the answer, ask me questions that help me work it out” mirrors how the guarded tutor in the study was designed.
  3. Verify every fact. Names, numbers, quotations, and sources all need a second check.
  4. Keep personal data out. Do not enter a child’s full name, school, photos, or other personal details into tools. Check each tool’s age rules and prefer school-approved tools. The data-handling discipline we describe for businesses in our Shadow AI piece applies at home as well.
  5. Make it yourself first. For creative work, the first draft belongs to the child.
  6. Talk about it. Show your own use of AI, including its mistakes, and agree on family rules together. Review them every few weeks as the tools change.

These Are Workplace Skills Too

If you run a small business, notice how closely this list matches what you want from a hire. The WEF figures above are about employers, not children. The seven skills are the ones that decide who can work well with AI: who asks the right question of a tool, who checks the output, who can choose between ten options, who can work with others, and who can keep going when a project hits trouble.

There is also a training lesson. When AI absorbs the routine entry-level tasks, the practice that used to build judgement can disappear. We wrote about this risk in Will Your Staff Forget How to Do Their Jobs?, and the same logic applies to children: if the practice disappears, the skill has to be built on purpose. Judgement is also what makes a good checkpoint work, as we explain in our guide to human-in-the-loop design, because a human approver who cannot judge is only a rubber stamp.

Interview questions worth borrowing:

  • Tell me about a question you asked that changed a project.
  • Describe a time you found out that something you trusted was wrong. How did you find out?
  • Show me something you finished that nobody asked you to make.
  • What did you like about a piece of work you saw recently, and why?
  • Who did you help last month, and who helped you?
  • Describe a goal you reached with resources you did not start with.
  • What did you keep working on after it got hard?

Five Myths Worth Dropping

Myth 1: Keep AI away from children until they are older

Age limits and supervision matter, and some tools are not meant for young children. But a blanket ban gives you no chance to teach judgement, and children will meet these tools anyway. Supervised use with clear rules teaches more.

Myth 2: Knowledge no longer matters

It matters as the raw material for judgement, as Willingham’s review explains. What has changed is the goal: understanding and use, not storage.

Myth 3: Learning to code is the only future-proof skill

Technological literacy appears on the WEF list, but alongside analytical thinking, creativity, resilience, and curiosity. It is one skill among several.

Myth 4: More tuition equals more safety

Tutoring can help close a gap in a specific subject. The problem arises when the schedule leaves no room for the unstructured time in which creativity, taste, and interest grow.

Myth 5: Passion plus grit guarantees success

The grit research suggests the popular version of this story is oversold. Interest, habits, support, and opportunity all matter, and no single trait is a guarantee.

A 30-Day Starter Plan

Ten minutes a day is enough to begin. The plan below puts two skills in each week and ends with a review of your family’s AI rules.

  • Week 1: Ask and filter. Run a question of the day at dinner. Investigate one forwarded message together.
  • Week 2: Create and taste. Schedule one unscheduled hour. Have the child finish one small project, then play pick three, defend one.
  • Week 3: Collaborate and integrate. Ask the three dinner questions every night. Run a family mini-project from a goal sentence.
  • Week 4: Sustain and review. Do a failure review after a setback. Start an interest log. Then sit down together and rewrite your family AI rules in light of what you noticed.

At the end, ask each other which skill grew most and which was hardest. The answer tells you where to focus next.

Frequently Asked Questions

What skills should children learn to prepare for the AI era? Seven stand out: asking good questions, filtering information, creating original work, developing taste, collaborating, assembling resources around a goal, and sustaining interest through setbacks. Employer surveys such as the WEF Future of Jobs Report 2025 point to the same cluster of abilities.

Will AI replace children’s jobs in the future? Nobody knows precisely. What the evidence supports is that routine, predictable tasks are the easiest to automate, while judgement, creativity, and collaboration are what employers say they need more of. The safest preparation is to build those skills and to learn to use AI well.

Should children use ChatGPT for homework? It depends on how. A 2025 PNAS field experiment found that students using a standard chatbot scored higher during practice but worse on a later test without it, while a tutor designed to guard learning largely avoided that harm. A good rule is to attempt first, ask the AI to coach rather than answer, and verify what it says.

Is knowledge still important if AI can answer anything? Yes. Cognitive science suggests that good judgement depends on knowing the subject. Without background knowledge it is hard to understand a claim, let alone check an AI answer.

How can parents teach critical thinking at home? Use the three questions of source, evidence, and other views, and practise lateral reading by checking a claim in a second independent source. A forwarded message in a family group chat is a good place to start.

How early can children start building these skills? From the age at which they ask why. The exercises scale: a young child can climb the question ladder and play like, because, and a teenager can run a project from a goal sentence and a budget.

Sources and Further Reading


Willie is the Managing Director of Mxgsoft Pte Ltd, a Singapore-based digital transformation company specialising in ERP implementations, workflow automation, and AI-powered business solutions.

Next step: Pick one skill from the one-page table, ask the daily question tonight, and notice what your child says. If you found this useful, share it with another parent or with the person who hires your next team member.