My 9 year old Atlas has been practicing his multiplication tables. We have found that when he’s not actively working on it, for instance, during spring break, his skills can get a bit stale. We have flash cards and workbooks to help keep the rust off, but like most kids his age, he typically loses interest after a few minutes.
Meanwhile, he has shown a genuine interest in coding. What started with Scratch a few years ago has since grown into building small Python projects using ChatGPT, VSCode or Replit.
As a strong believer in the Montessori philosophy of “follow the child”, I had an idea.
I asked him to “vibe code” himself a math quiz application and handed him a piece of paper with the following requirements.
1) It should ask the user for the highest multiplier.
2) It should ask the user for a number of questions for the quiz.
3) It should ask the user for their name.
4) It should randomly quiz the user on a set of multiplication problems.
5) It should keep track of a user’s answers and calculate their grade.
I thought he’d be busy for at least an hour, so I was caught off guard when he came bursting into my office 15 minutes later with excitement. He wasn’t just proud to show me what he built; he was even more thrilled to show off how he had scored an A+ on multiple runs of the game.
For the next hour or so, it had his full attention as he live-demoed it for anyone and everyone nearby.
Will it hold his interest? Hard to say. Probably not much longer than his version of rock‑paper‑scissors or the question bot he built to demo for his class.
But that’s OK.
While the initial goal may have been to reinforce some math skills during a week otherwise occupied by far too much YouTube and encore presentations of the Minecraft Movie, I realized that there was a much more important lesson in all of this – we all now have the ability to build software that solves our own, unique problems.
Atlas, along with his brother and sister, is growing up in a world where building a software tool to solve a problem will be just as natural as using Google Docs or sending an email.
Is this the same as “real coding”?
Probably not. But honestly – who cares?
It’s holding his interest, teaching him self‑reliance and “real world” problem solving, and most importantly, fueling his curiosity.
I’ve always considered code a tool of the software developer to get an idea from their head into the real world. Like a painter’s brush or a musician’s guitar, it’s just a means to create.
So what’s happening now?
The tools will continue to become more accessible, less intimidating, and open the door for more people to use them.
Today, Atlas may be painting by numbers – but he’s learning the strokes, getting a feel for the medium, and refining his taste. As his skills continue to develop, he’ll naturally reach for more advanced tools that help bring what’s in his head to life with more precision and depth.
So what if he doesn’t?
Honestly, I still don’t care. Vibe coding is in its infancy, and the tools are only getting better from here. The ability to quickly build disposable, personalized apps is already becoming a superpower for anyone who leans into it. So even if my 9 year old never takes it further than this, I’m good with that.
I’d rather have my kids building skills for the “real world” than obsessing over what counts as “real code.”
Later that afternoon, I checked in with Atlas to see how things were going and if he’d made any updates to his game. He’d moved on from multiplication, but had used the same framework to build a new quiz—this one on Minecraft trivia for his brother and friends. Then he asked how he could put it online so others could play.
With a little help from me (and ChatGPT), we converted the project to HTML and JavaScript, added some basic Tailwind styling, and deployed it to GitHub.
Does he fully understand what any of that means yet?
Probably not.
But still, my 9 year old just pushed his first code to GitHub.
Pretty cool – check it out.