A continuation of my reflection on underused experiments—this time exploring how to fold them back into habit and make them feel alive again.
How I bounce between conversational problem-solving in ChatGPT and in-editor flow with GitHub Copilot, plus what’s working and what I’d like to improve.
Reflections on experiments that worked technically, but stalled because I didn’t fold them into habit.
Notes on how I get the most from GPT-5 — what works, what to watch out for, and the tangible ways it improves my daily workflow.