Is coding still worth learning for kids?

Yes — but not the way it was taught five years ago, and anyone telling you otherwise is selling something.

This is a question we get from parents constantly, and it usually arrives with an edge of anxiety: if AI writes the code, is my child learning a dead skill? Was the coding camp a waste of money?

What actually changed

Writing syntax used to be the barrier to entry. Before you could make anything, you had to memorise how to structure a loop, where the semicolons go, what a variable declaration looks like. Months of it. Most kids quit somewhere in that stretch, and they quit for a sensible reason: they were paying a large upfront cost with nothing to show for it.

AI removed that toll. A child can now describe what they want and get working code back. The gap between having an idea and seeing it run has collapsed from months to minutes.

That's the whole change. It's significant. It is not the same as saying coding is over.

What didn't change

Everything that was actually hard is still hard:

•     Knowing what to build. AI will happily build a bad idea, perfectly.

•     Breaking a big goal into steps small enough to actually finish.

•     Reading what came back and spotting that it's wrong.

•     Understanding why it's wrong well enough to describe the fix.

•     Deciding what to cut when the thing is too ambitious.

These were always the skills that separated people who ship things from people who don't. The syntax was never the point — it was the toll booth in front of the point. Now the toll is gone and what's left is the part that mattered.

So should they learn syntax at all?

Some, eventually, and not first.

A child who has built three working projects has a reason to care how the code works. A child who has built nothing does not. Motivation runs in that direction, not the other way, and the old curriculum had it backwards — two years of theory in the hope that a project might follow.

What we see in practice is that kids start reading the code the AI writes because they want to change one small thing and it's faster than asking. That's the right on-ramp. It's curiosity-led, and it sticks.

The uncomfortable version

Here's the part that gets left out of most articles on this topic: nobody knows what these tools look like in ten years, including us. Any confident claim about which specific skills will pay off in 2036 should be treated with suspicion.

What we'd bet on instead is more general. A child who can take a vague idea, break it into steps, direct a powerful tool toward it, evaluate the result honestly, and finish the thing — that child is in good shape regardless of what the tools become. That capability isn't about code. Code is just a convenient place to practise it, because the feedback is immediate and the result is real.

That's what we teach. The Roblox game is the excuse.