Learning GitHub Copilot

Foreword

I’ve spent a lot of time helping developers learn how to use AI tools as part of their daily work—by mentoring people, running training sessions, and writing about them in books—and I’ve relied on those same tools myself while building real software. One really stands out to me: GitHub Copilot.

It’s a game-changer. I use it all the time.

I’ve been coding for decades, and like every developer, sometimes I get stuck. Maybe it’s when I’m using a new technology or API, and I’m dreading going through poorly written documentation. Maybe it’s when I’m refactoring a particularly nasty bit of code that everyone on my team has been avoiding. Or maybe I just don’t feel like writing dull boilerplate code that has to get written but is going to cost me a half hour of my life that I’ll never get back.

That’s where Copilot can really make a difference. It might give me all the code I need, complete and working. Or it might give me just a starting point—something we’ll iterate on together. Either way, it’s valuable to me, and it almost always gives me enough to get things moving again.

But working with Copilot isn’t a one-way street—it comes with responsibility. Copilot can write code that compiles, runs, even passes tests. But if you don’t understand what it gave you, or why it works—or, most importantly, how you got it to produce that result—you’re not really in control of what’s going into your codebase. This book puts you in control.

Good developers need to do more than just learn ...