Daddy, can we make a video game?

“Daddy, can we make a video game?” The question dropped after lunch, just like that. My son is seven. He just finished first grade. And there was one word in what he said that changes everything.

He said make

He didn't say code. He said make. And too many of us have stopped hearing the difference.

Everything is in that verb. Coding is a matter of technique, of syntax, of years. Making is a matter of creation, and a seven-year-old knows exactly what that means. The reflex, the one we all share, was to say no. Not now, it's too complicated, you have to learn to program first. I didn't say no. I asked: what kind of game? The question sounds like nothing, but it does everything. By asking which game, I had already granted that it was possible. I validated his idea before he had even finished forming it.

He didn't hesitate: Astro Bot. Sony's little robot, the platformer crowned Game of the Year 2024. I explained it to him, honestly, without killing the urge: around sixty people worked on that game for almost three years, a whole studio, to carry it to the top.^[1] To let him feel the size of the mountain without talking him out of the climb. And then I opened another door. What if, instead of remaking Astro Bot, we built something simpler, but something that was ours?

Hands on, not syntax

We would have made him learn syntax for years. He took the wheel in one afternoon.

I didn't teach him to code. I taught him to ask. My job stopped almost there. I set the ground: a few instructions to frame the conversation, support for his ideas, help to unblock whatever stalled. I showed him how to talk to Claude. And since he has only been writing for a few months, he didn't type: he spoke. Voice transcription turned his voice into text, the machine answered, he kept going. Then I handed him the controls.

And he drove. He wanted a scooter, not a bike. Jumps. Tricks. Obstacles, music, a score that climbs. He dictated out loud, the machine built, he watched the result appear on screen. When something broke, when a jump landed wrong, he didn't turn to me to fix it. He told the machine, and he tried again. I won't walk through every step. What matters is the motion: an idea, an attempt, a result, a correction. Again. And again.

He couldn't write the code. He could say what he wanted, and recognize what was wrong. That was all he needed.

By the end, the game ran. A little character on a scooter, jumping, spinning in the air, grinding on rails, against a sunset, with his music. He played it. And he said two things I won't forget. The first: it was the best video game he had ever played. Not because it beat Astro Bot, of course. Because it was his. Because he had made it. With his father, on Father's Day. The second, right after: “Daddy, can we give it to my friends so they can play?” He didn't only want to create. He already wanted to share.

Driving the machine that makes

Here is the new border. No longer those who know and those who don't. Those who dare to speak to the machine, and those who wait for it to decide for them.

I keep thinking about what happened that afternoon. It isn't that my son is a prodigy, or that he “learned to code.” He retained nothing of the syntax, and he didn't need to. What he drew on was something else: form an intention, test it against a result, adjust, start over. That skill has no name in our school curricula. Yet it is the one that now decides what you can produce with these tools. No longer knowing how to do, but knowing how to drive the machine that does. It is the skill I place at the very top of the adoption ladder, the one that separates those who endure these tools from those who command them.

It is the same dynamic I stage in my trainings and my talks: dare the impossible. With one difference, and it is a large one. For a child, it is simpler. He doesn't brake against reality, he doesn't yet know what is supposed to be out of reach, and when the people around him push him to try, nothing holds him back. The adult has absorbed, over a lifetime, a sense of limits, what Bourdieu called habitus, that small voice whispering “this isn't for you.”^[2] It is exactly that voice I watch fall silent, in training, the minute someone dares. It is the individual revolution I describe in The revolution that never happened.

Born curious, staying curious

Every child is born curious. That isn't the question. The question is who will be allowed to stay that way.

If I press the point, it's because one detail haunts me. My son succeeded because no one switched off his question. Curiosity is not his alone. Disability aside, it is a universal disposition: at six months, children already explore, reach out, try to understand. Susan Engel, who spent her career observing it, puts it plainly: there is almost nothing to do to make a toddler curious.^[3] What changes everything, afterward, is not the curiosity at the start. It is what adults do with it. Answer a child who asks why with an explanation, or better, with a “let's go find out together,” and the curiosity grows. Answer “I don't know, stop bothering me with your questions,” and it dies. The research on children's questions says exactly this, and adds something harder: those who would need support the most are often the ones discouraged the most.^[4]

And this is where the next divide opens, for me. Curiosity is the same everywhere at the start, whatever the background. What is passed on unequally is the permission to use it: explore, get it wrong, talk to the machine, start again. If the skill that will matter tomorrow is that one, then failing to prepare our children for it is not a neutral act. It lets a fault line widen, from childhood on, between those who will dare to speak with these machines and those who will endure them. I'll devote a whole Audace to it soon, generative AI and our children, and what we risk by doing nothing. So you don't miss it, you can sign up for The bold newsletter.