After using it for a few coding projects recently, I find that ChatGPT is a great way to lower the barriers to building smaller, self-contained projects—things that have been hiding in your to do list that take a bit too much effort to attempt but is still a good idea.
ChatGPT makes getting started easy. With a few prompts, you get semi working code to try out. Small projects don’t need much context, so there is very little context to explain (large amounts of context is a problem for LLMs). Using ChatGPT also offloads things you might not enjoy doing—writing documentation, product copy, or even writing basic tests.
I find pairing with ChatGPT to write code is ideal for internal tools and small projects to improve workflows. In just a few hours I was able to capture output of chatgpt-shell in org-mode and make an org export backend for Notion—both I would have said is nice but not worth the effort.
- Knowledge capture loops make for good systems and ChatGPT speeds up each iteration
- ChatGPT is a great fit for one of one software like customizing your Emacs init file