AI
AI Engineering: Building Applications with Foundation Models
- See also: 55 pages of notes I took while reading the book
- I read this book as part of a book club for my company. It lays everything out about how foundation models work, what came before them, and what their limitations are.
- Heads up… This reads like a graduate-level textbook. The author skips over some of the advanced math, but having an advanced degree in computer science certainly helped my comprehension.
- Despite it’s length and depth, I would recommend this book to any engineer who works closely with AI or LLM tools. The book demystifies how things work as a system, which can lead to better use of the tools in practice.
- My favorite chapters dealt with evaluation — that is, how do you know things are working.
Stanford’s Practical Guide to 10x Your AI Productivity
- Context engineering is prompt engineering on steroids. What are all the things I need to give to AI so it can complete a task?
- Write down how to do something, then give those instructions to a colleague. If they can’t do it, AI isn’t going to be much better.
- AI is a tool and can be a reflection of ourselves. For example, if your inclined to cognitively offload thinking to AI, it will do that. However, if you want to stay sharp and think critically, AI will help you with that as well.
- Example modes: chain-of-thought, few-shot, reverse prompting (ask AI to give you better prompts), assign a role to AI, role playing.
Nature
What an Owl Knows: The New Science of the World’s Most Enigmatic Birds
This was a fascinating book about owls — how they behave, the impact they’ve had on humans, and the impact humans have had on them. It was interesting to learn they are one of the least studied raptors and how much more there is to learn.
Software
An Overview of Every Software Architecture
This brief video outlines the most common software architectures in use today and where they typically work best. Most of them come from Fundamentals of Software Architecture.