Coding with AI

From autocomplete to agents, and how to stay the engineer.

What it is

AI coding tools range from inline suggestions to agents that plan and build whole features. Used well, they speed up the boring parts; used badly, they produce code nobody understands.

Why it matters

Almost every technical field now expects fluency with AI coding tools, and the skill that matters is review.

Key ideas

• Read every change before accepting it.
• Tests are how you check an AI's work.
• Give the agent context: conventions, constraints, examples.

Try it: Build, then review

1. Ask an AI tool to build a small script that cleans a CSV.
2. Read the code and add a comment on every line you don't understand.
3. Ask it to write tests, then break the input on purpose.