A Seventh Grader Lost His Job to AI This Week

Okay, not really.

But something did happen that sparked a lot of reflection. At my son’s school, there’s a seventh grader who’s earned a bit of fame for building the best Scratch game anyone’s seen. He’s respected. Known. Maybe even feared (in the way only middle school coding legends can be).

Last weekend, my 11-year-old son used the AI platform Lovable to build a game. In just a few hours, and with a little help from me, he created something that possibly blows the reigning champ’s game out of the water.

And he didn’t write a single line of code.

Lovable is one of a growing number of AI tools that lets people design full games just by describing what they want in plain English. My son had been using Scratch for years, and he’s learned a lot about coding through trial, error, and the occasional meltdown when a sprite wouldn’t behave. We’ve even used AI to help him debug and improve his Scratch projects, with decent success. But when Lovable offered a free weekend, I figured we’d see what it could do.

What we found was something very different: a tool that bypassed the entire learning curve and let him build something he was genuinely proud of, without any of the frustration.

Every image was generated with ChatGPT. Every idea he had became part of the game almost instantly. And he loved it.

But he also learned very little about how it was actually built.

So here’s the question I’ve been wrestling with. When it comes to kids and AI, what matters more: learning how to build something from scratch, or learning how to bring your ideas to life?

We’ve always put a premium on learning to code because coding has been the gateway to building anything digital. But what if that gateway is about to change? Given how quickly AI is advancing, is it really so far-fetched to think that in a few years, 95% of people creating digital products won’t be coding in Python or JavaScript, but simply describing what they want in natural language and letting AI do the rest?

If that’s where we’re headed, and I think it is, then maybe what’s most important right now is helping our kids learn to work with AI as a creative partner. To imagine boldly. To tinker. To revise. To build. Not necessarily to code.

Still, there’s something undeniably valuable about struggling through code. It builds frustration tolerance. It teaches problem solving. It forces you to sit with a problem and wrestle with it. That skill will never go out of style, even if the tools change.

So maybe the real challenge is balance. Give them access to powerful tools, but don’t abandon the long road. Let them build quickly, but also make sure they know what it means to build slowly. Help them create, and help them care about how it all works.

Here’s the game he made. Try it out and let me know what you think.