Acknowledgments

Writing a book about working with AI tools is, unavoidably, a recursive endeavor — and this one more than most. It was drafted with AI assistance and is published by DataField.Dev.

How This Book Was Made

Honesty matters more than usual in a book that teaches you when to trust AI output, so here it is plainly: this book was not shaped by a panel of technical reviewers, early readers, or years of workshops. It was drafted with AI assistance and has not been formally reviewed. Read it the way it teaches you to read any AI output — with interest, and with verification. Any errors are the book's own.

The People Whose Work It Draws On

Thank you to the researchers who built and studied the models this book describes, to the AI companies whose public documentation explains how their tools are meant to be used, and to the many practitioners who have shared their prompts, workflows, and failures openly on blogs, forums, and video. Much of what is practical in these pages was first worked out in public by people like them.

The AI Tools Themselves

It would be strange not to acknowledge the tools that are the subject of this book. Claude, ChatGPT, Gemini, Copilot, and the others are not sentient collaborators, but they are remarkable instruments — and, in this case, part of how the book itself was made.

An Invitation

If you find an error, an outdated claim, a prompt that no longer works, or an explanation that doesn't land, please send corrections and suggestions through datafield.dev/about.html. Instructors adapting the book for a course are especially welcome to share what worked and what didn't.

My Readers

Finally, thank you to you — whoever is holding this book or reading this file. The fact that you are trying to get better at something that matters is, in itself, a contribution to a world that needs more thoughtful AI practitioners.

Use this well.