Developer Educational Content Probably Needs to Change

Developer educational content used to have a fairly obvious audience: developers.

Documentation, tutorials, code examples, API references. Different formats, but broadly created for people who understood the fundamentals of software development and wanted to learn how to build with a particular platform.

AI is changing that.

Because developer education increasingly isn't just about teaching people how to build.

It's also about teaching AI how to build, and teaching a new generation of builders what's possible and how to translate their ideas into something AI can build.

I think that means developer educational content now has three quite different audiences.

1. Developers

Developers still need much of what they've always needed: accurate documentation, examples, APIs, architecture, best practices and explanations of how things work.

But now there are two new audiences.

2. AI agents

Increasingly, the thing consuming your developer resources isn't a developer.

It's Claude, ChatGPT, Cursor or another coding agent trying to build something.

That creates a very different set of requirements.

AI agents need information that is structured, predictable and comprehensive. Clear schemas, consistent terminology, good examples and explicit relationships between things become incredibly important.

In that world, developer resources aren't just educational content.

They're also infrastructure for AI.

3. Vibe coders

Then there's a rapidly growing audience that needs something quite different again.

Vibe coders don't necessarily need to learn how to write the code themselves. But they do need enough technical understanding to know what's possible and what to ask AI to build.

Knowing that an API exists isn't enough.

They need to see what can be done with it. They need ideas, examples, patterns and enough foundational knowledge to understand which pieces of technology might solve which problems.

Because there's an important translation happening:

Problem → Idea → Technical approach → AI

“I want customers to get a message when this happens.”

“This API gives us the data, this hook can detect the event, and this service can send the message.”

Now AI can take it from there.

So perhaps developer education for vibe coders should focus less on step-by-step implementation and more on building a mental map of what's possible.

What can this API do? What problems could it solve? What can it connect to? What are some interesting things people could build with it? What are its limitations?

The goal isn't necessarily to turn vibe coders into developers.

It's to give them enough technical understanding to translate a problem or idea into the right technical approach, then communicate that to AI.

For years, developer education was largely about helping developers learn how to build.

Now it also needs to help AI understand how to build, while helping a much wider group of people understand what can be built and how to ask for it.

One platform.

Three audiences.

And probably three quite different approaches to developer education.