Available now
The Boring AI Engineer
Why Boring Technology Wins When AI Writes the Code
Noah's dog-groomer app took him a weekend to build. It took one Tuesday to kill.
AI made building software fun again. It did nothing to make keeping it alive any easier — and the gap between a magic demo and a product that survives real users is where most of them quietly die.
This book is the field map of that gap — which technology the models actually know how to maintain, how to make AI agents behave in your codebase, and why the boring, load-bearing choice keeps winning precisely when code gets cheap.
What the book argues
Readable beats clever
AI can generate clever code faster than teams can safely own it. The durable advantage is code that remains boring enough to inspect.
Architecture is the prompt that lasts
The shape of the system determines what agents can do safely. Good boundaries are more valuable than ever.
Tests become operating instructions
Tests are no longer only regression checks. They are executable context for humans and machines working on the same codebase.
The senior engineer becomes an editor
The job shifts from typing every line to selecting, rejecting, simplifying, and owning the final system.
Andrew Kwak
Andrew Kwak builds software systems and writes about the practical engineering choices that make them last. His patented technology was cited as prior art in Apple's own patent filings. The Boring AI Engineer builds on the deliberately boring stack this book describes.