KAIST Brain Scientist: The One Skill Kids Must Have to Thrive in the AI Era

Here's a question that's been keeping me up at night. If a machine can now write essays, solve equations, and code better than most adults, what exactly are we preparing our kids for? The honest answer is that nobody really knows. But a neuroscientist at KAIST — one of the top science universities in the world — says he does.

His name is Professor Jae-Seung Jeong, and he studies how the human brain actually learns. He's issued a pretty stark warning to parents: there is one specific skill that matters just as much as reading, writing, and maths. If your child doesn't develop it, they will hit a wall later. Not maybe. Will.

It isn't coding. It isn't a second language. It isn't even creativity, exactly. It's something quieter than that, something most of us were never taught ourselves. And once you understand what it is, you start noticing where it's missing — in your kids, in your co-workers, honestly, in yourself.

The skill that outlasts every subject

Before we get into the how, Professor Jeong wants parents to reset something more basic. Most of us assume the goal of raising a bright child is getting them into a good university. He pushes back hard on that. The actual purpose of education, he argues, was never the acceptance letter. It was raising someone who can contribute to society and stand on their own two feet.

That distinction matters urgently right now. The old model of schooling was simple: memorise the knowledge, reproduce it fastest, get the grade. That's the game we built for decades. But AI now does exactly that, better and cheaper than any student could. So what does society actually need from young people? Professor Jeong's answer is: something AI cannot supply. Their own perspective. Real, lived experience. And human skills like cooperation, working through conflict, reading a room.

This is where metacognition comes in — not as the finish line, but as the tool that gets a child there. It literally means thinking about your own thinking. In Professor Jeong's words, it's the ability to observe yourself from the outside, almost like an alien studying a human. It sounds abstract, but it's incredibly practical. A child with metacognition can say, "I thought I knew this chapter, but when I tried to explain it out loud, I couldn't." A child without it just keeps highlighting the textbook and feels vaguely confused about why the test went badly.

In the old world, you could power through school on sheer memorisation. AI has now flattened that entire skill. What a machine can memorise, your child no longer needs to. What AI cannot do — yet — is know what it doesn't know. That gap is where humans still win, and Professor Jeong is saying this is the muscle you have to start building in childhood.

Why even the top students are hitting a wall

Here's the part that made me sit up. Professor Jeong warns that children who never develop this capacity will absolutely hit a limit later. Not because they aren't smart, but because they can't correct course. And his clearest warning isn't aimed at struggling students — it's aimed at the opposite end.

He says the students who get into Korea's most prestigious universities, the ones every parent points to as proof they did it right, are now some of the ones struggling the most. Why? Because they were raised to execute instructions perfectly. Told what to study, when, and how, they did it flawlessly. But nobody ever asked them what they actually wanted. The moment parents and schools stop telling them what to do — at university, in a first job, in real life — they don't know what to do with themselves. They look around at what everyone else is doing and blindly follow.

The pattern often plays out like this. In primary school, everything looks fine. Grades are okay, the child seems bright. Then somewhere around the tween years, the material gets abstract — algebra, essays, science that requires real reasoning — and suddenly the wheels come off. Parents panic. Tutors get hired. But the real issue isn't the maths. The child never learned to monitor their own understanding, or their own wants.

Most of us weren't taught this either. We were taught to get the right answer, not to notice when our own thinking was off, and definitely not to ask what we wanted underneath it all. So if this feels foreign to you as a parent, that's normal. You're not behind — you're just noticing something your own education skipped.

The 'alien perspective' trick

So how do you actually build this in a child? Professor Jeong uses a beautiful image. He talks about looking at yourself the way an alien would look at a human — with curiosity, without judgement, just observing.

That framing matters. If you ask a child, "What are you bad at?", they get defensive. Nobody likes feeling exposed. But if you ask, "If an alien watched you do your homework today, what would they notice?", suddenly it's a game. They can be honest. They might say the alien would notice they got distracted every time their phone buzzed, or that they skipped the hard problems. That's metacognition, live.

In practice, this looks like small, low-stakes conversations after the fact. What was hard? What did you actually understand, versus what did you just kind of nod along to? What would you do differently next time? You're not correcting them. You're teaching them to notice themselves. A child who sees themselves objectively and stays curious about their own mind is, essentially, self-updating software — and in the AI era, that's the whole ballgame.

Curiosity is the fuel, not the finish line

Here's where a lot of parents get it slightly wrong, and I've been guilty of this too. We treat curiosity like a personality trait — some kids have it, some don't. Professor Jeong pushes back on that. He treats curiosity as a trainable habit, tightly linked to how the brain forms new connections.

The practical version for us: protect the questions. When your five-year-old asks why the sky is blue for the fourteenth time that week, the temptation is to say "just because" and keep unloading the dishwasher. But every time a question gets shut down, a tiny bit of that muscle atrophies. You don't have to know the answer. "Huh, I actually don't know. What do you think?" does two things at once — it validates the question and hands the thinking back to them.

Now, here's something that might surprise you, especially if you're anxious about the AI race. Professor Jeong actually recommends avoiding AI tools at young ages. Not embracing them early — avoiding them. If you've been worrying that your child is falling behind because they aren't using ChatGPT for homework yet, let that go. That's not the gap that matters. What young children need is hands-on experience: touching, building, getting things wrong with their own hands. They need to form their own opinions, out loud, before an algorithm hands them a polished one. And your child's worth is not tied to how well they perform in school. Curiosity doesn't grow in a kid who feels like their value is on the line every time they don't know something.

In a world where AI handles the answers, our job as parents shifts. We're not raising kids to know things. We're raising kids who know themselves.

— K-Parenting

What this actually looks like on a Tuesday night

Let's ground this, because "raise a metacognitive child" sounds like a lot when everyone's tired. It's smaller than you think, and it really comes down to three habits woven into ordinary evenings.

The first is narrating your own thinking out loud. When you're deciding something — what to cook, how to handle a work problem — say the messy middle part where your child can hear it. "I'm not sure if this will work, but I'm going to try it because…" You're modelling the internal voice they need to build. The second is asking reflection questions instead of performance ones. Instead of "Did you get a good grade?", try "What surprised you today?" or "What did you get wrong, and what did you learn from it?" Getting things wrong has to be safe in your house, or none of this works.

The third is resisting the urge to rescue. When your child is stuck, wait a beat longer than feels comfortable. Struggle is where metacognition is actually built. If you jump in with the answer every time, they never get to practise noticing their own confusion. Professor Jeong's core message really is this: in a world where AI handles the answers, our job as parents shifts. We're not raising children to know things. We're raising children who know themselves — how they think, when they're stuck, and how to move.

---

Watch, Listen & Subscribe

Watch the full episode on YouTube, or if you'd rather have it playing in the background, listen on Spotify or Apple Podcasts under K-Parenting.

New episodes land every week — subscribe so you don't miss the next one.