AI Engineering

I can't even imagine the amount of work the author put into writing this book. This feels like an almanac of everything you need to know about AI applications today. From the intro, it sounded like I could be the target audience, especially since we read this with my whole team. While a lot of the topics are interesting and informative, most of the times it felt like they were covered at a level that was too superficial to really stand the test of time.

I'm afraid this book will be obsolete in just a few years, and part of the reason why that is true is not necessarily because of the author but because this discipline is still very much immature. You come to realize this especially when you look at how the system design for building AI applications is described and how a lot of the components are probabilistic. This is because part of the components are based on AI and they are agents which makes your entire architecture too non-deterministic. This is one of the big challenges in AI and how to make your system predictable but it doesn't feel like we solved this yet. We are not in a place that is good enough enough without a lot of iteration, fine-tuning, and refining of the architecture.

In the end, I don't think I'll go back to this book. It didn't give me enough new material to justify the purchase. It has a lot of links to resources, and it's a good starting point to pique your interest, but it doesn't give you enough to become the one-stop shop for AI engineering. There's enough material out there that you don't need to spend $50+ on this book.