A Practical Field Guide to AI Coding

I feel like I only ever see super extreme positions on AI coding online (AI-everything or AI-nothing), so I decided to write up a guide on how I’ve been using AI in my own coding workflow.

What’s been working:
– Green field tasks (e.g. Create a rate limiter class)
– Pattern expansion (e.g. Create another API endpoint based on these 3 examples)
– Documentation (e.g. write docstrings for all the functions in this file)
– Code explanation (e.g. what does this module do)
– Tech spec review (e.g. review my plan for adding session auth to my app)

What’s not been working:
– debugging (why is this useEffect being called infinite times)
– multi-file edits (e.g. add this feature to my frontend and backend)

Would love to learn what’s been working for others…