Coding with AI tools

Stage 1 of 10 · 6 lessons · ~1 week

What AI coding tools are good and bad at

Learn how the models behind these tools work, the difference between completions, chat and agents, and where they help or hurt, so you use them with judgment from day one.

Start: How a language model writes code

Lessons· one lesson for each topic, in this order

  1. 1How a language model writes code
  2. 2Hallucinations and the context window
  3. 3Completions, chat and agents
  4. 4Where AI helps and where it hurts
  5. 5Automation bias and staying in charge
  6. 6What GitHub Copilot is

Do this

  • Explain in two sentences how a language model produces code.
  • Collect three hallucinations you have seen or caused, and how each could have been caught.
  • List tasks you will hand to your tool, and tasks you will design yourself first.
  • Write the checks you will run on every AI change before you trust it.

Start with· follow one of these

Keep handy

Also worth your time

Before moving on: A one-page note in your own words: how your tool works, three tasks you will hand to it, three you will design yourself first, and the checks you will run on anything it writes.

Tauseef's advice
Copilot meets you in the editor, the terminal and on pull requests. Learn one surface properly before the next: inline suggestions and chat first, then agents, then review and the cloud agent once your testing habits are solid.

Mark this stage as not done?