Codex for the Rest of Us AI agents explained for normal people
Getting real work done with AI agents can feel like magic. Once you see what they can do, your work life changes forever. For a long time, this territory seemed reserved for other people: the ones who write code and speak the language of technology. This book starts from a different belief: Codex can be for the rest of us too. Not because AI is the next big thing, but because more people deserve the space to think, create, and take part in the world that is coming.
Written by Alonso Astroza Tagle.
Licensed under a Creative Commons Attribution-NonCommercial 4.0 International license.
Contents
- Introduction
(Part 1: Getting Oriented)
- What an AI Agent Is
- Files, Folders, and Where Work Lives
- Giving Context Without Overexplaining
- A First Useful Delegation
- Checking the Agent's Work
(Part 2: Everyday Work)
- When the Tool Breaks
- Turning a Messy Document into a Useful Summary
- Cleaning a Spreadsheet
- Drafting Email Without Losing Your Voice
(Part 3: Workflows)
- Building a Small Internal Tool
- Automating Repeated Administrative Work
About Me
I have worked on artificial intelligence projects for more than 15 years. In recent years, I have taught more than 10 courses on generative AI to hundreds of students and professionals.
Today my work centers on helping people learn Codex. Some of them are engineers and technical teams. Many others come from law, medicine, nursing, kinesiology, general studies, psychology, dentistry, journalism, education, design, film, architecture, and business. That range has forced me to find a simple, practical way to explain a tool that is often presented as if it were only for programmers.
This book comes from that experience. It is my attempt to write down the method I use to teach the Codex app to intelligent, capable people who do not necessarily have a technical background, but who want to use these tools with judgment and confidence.
Get Chapter Updates
This site is a first public shape for the book. The manuscript will grow chapter by chapter, with small exercises, visible outputs, check steps, and recovery patterns for when tools fail.
Leave your email to hear when new chapters are published.