I use GPT4 like I would ask a question on StackOverflow. While my prompts are relatively long compared to a web search, I’ve found the results really good. I u

I use GPT4 like I would ask a question on StackOverflow. While my prompts are relatively long compared to a web search, I’ve found the results really good. I usually state the problem, provide code context, motivations, and end with restating what I want as a result,why and a question. Here is short example from my history: —- I have the following docker command to help me backup data from some containers. ```sh (omitted for brevity) ``` I want to improve it by not only copying and zipping all the related files in the container volume, but using pgdump to take a copy of the database before doing the copy and zip. How might I change the above to achieve this? —- This takes a lot longer than starting a web search but the quality of answers is high, and I find faster than wading through maybe semi related content farms,cookie dialog a, prompts for news letters etc. One thing to remember is like StackOverflow answers, the code might not be up to date, or have bugs etc, as I test I feed issues back into it with any relevant context. I’ve started building a Jetbrains IDE plug-in around this workflow for myself with the ability to use self hosted models, improving it as I learn new tricks and find what workflows I prefer.