Code Health Guardian: The Old-New Role of a Human Programmer in the AI Era

by Artie Shevchenko, Lana Todorovic (editor), Titus Winters (foreword)

Millions of us will soon need to compete with AI. As it reshapes the tech industry, how will you ensure your skills remain indispensable?

It’s unlikely that attempting to pivot and rapidly become AI experts, as some suggest, is the best strategy. After all, AI development and integration appear to be the tasks perfectly suited for AI itself.

Artie Shevchenko, a former Google software engineer and lecturer at ITMO University, believes a far better approach may be to grow faster and deepen your expertise in managing code complexity. That’s because complex reasoning is hard for AI, and in programming, the most intellectually challenging problem is keeping codebases reasonably simple.

Although the skill of keeping code healthy has been a hallmark of seniority for decades, it’s going to become even more crucial in the upcoming AI era, as an average programmer, instead of coding, will soon be mostly working on code health and the overall architecture. This book will prepare you for that future. It will accelerate your growth and equip you with just enough theory to understand the relative value of all sorts of techniques, principles, and best practices so that you can navigate and prioritize them well.

Master managing code complexity today and position yourself as the best candidate for the old–new role of human software engineers in the upcoming AI era—the role of a Code Health Guardian.

Genres: Technical, Nonfiction, Technology

213 pages, Paperback

Published October 27, 2024

Reviews

[Sebastian Gebski · Rating 3 out of 5]

I struggled to finish this one. It's not that it's very bad or something - the issue here is that it barely tells the reader anything new/truly revealing. There are some very good points (especially on the true nature of complexity, importance of hiding the implementation details behind well thought-trough contracts, etc.), but they should be very familiar to you if you're familiar with the basics of scaling complexity and designing non-trivial architectures.

2.3-2.4 stars, rounded up to 3 as there's nothing outrageously bad here.

P.S. The suffix "in the AI Era" is definitely a sales stunt - there's very little on how LLMs/Gen AI change anything related to complexity/lowering the cognitive load thresholds. A promise not fulfilled.

[Caroline · Rating 5 out of 5]

Great reference and very approachable, particularly enjoyed reading about the author's real-world experiences. If you have some years of experience at a company with a great software engineering culture, you likely already know most of the ideas in the book, but I believe you will still gain something out of it. If you don't have this experience then the book is very valuable to learn about processes to ensure healthy and quality code. I keep it on my desk to quickly reference some ideas or concepts and the author includes many references for additional learning. Learning and following the lessons in the book is even more important than ever to guard against AI slop.

[Ernestas Poskus · Rating 2 out of 5]

Nothing particularly novel or new if you have read Unlce Bob or Fowler work.

[Frederic Desgreniers · Rating 5 out of 5]

Some of the book is a summary of other books and concepts, but I came across things in it that made it worth the read

One of the more pragmatic books on the subject.

[Oscar Fuentes · Rating 4 out of 5]

It is truth that the ideas of this books are not new but the way that they are presented, at least for my way of thinking, are so much better than others on the same topics. Really loved the book.