The question parents are asking
It is a fair question. AI tools can now write basic code on demand. Some generate entire web pages from a description. If a child can ask an AI to build something and get a working result in seconds, why spend years learning to code manually?
The question gets the direction exactly backwards. AI has not replaced the need to understand software — it has raised the bar for what 'understand' means. Using AI effectively requires knowing enough to evaluate, direct, and correct what it produces. Without that foundation, AI is just a black box that sometimes gives you what you asked for.
What AI cannot replace
AI can generate code. It cannot think about a problem, decompose it into solvable pieces, recognise which approach is appropriate, or catch a subtle logic error in its own output. Those are human skills — and they are exactly what computing education builds.
Computational thinking — breaking a problem into steps, identifying patterns, designing a sequence that achieves a goal — is the core skill. Code is just the notation. A child who understands how to think about systems will use AI as a powerful tool. A child who only knows how to prompt AI will be limited by whatever the AI decides to produce.
This is not speculative. We are already seeing the gap open up. Adults who write code can audit, correct, and extend AI output. Adults who only use AI tools are dependent on them — and when the output is wrong, they cannot tell.
What changes with AI, and what stays the same
What changes: the syntax barrier is lower. A beginner can get a working Python snippet without knowing Python. This is genuinely useful — it lets learners spend more time on logic and less time fighting semicolons.
What stays the same: the ability to reason about systems, the discipline to design before building, the patience to debug, the judgment to know when a solution is complete versus when it merely appears to work. None of these come from AI. All of them come from making things.
For children learning to code, AI is becoming a powerful collaborator — but only for those who already understand what they are trying to build. The child with two years of structured computing experience will use AI to go ten times faster. The child without that experience will use AI to go in circles.
Our view
We have thought carefully about this at Code Camp. Our curriculum is not about memorising syntax. It is about building the mental model: understanding what a computer does when it runs a sequence of instructions, what a variable is at a conceptual level, how conditionals create branching behaviour.
Children with that mental model adapt to new tools — AI-assisted or otherwise — naturally. Children without it are permanently dependent on instructions, whether those instructions come from a tutorial or an AI prompt.
Coding is still worth learning. The reason is stronger now than it was five years ago.