AI for Documentation and Code Comments

When I first started Boomering I used a third party partner to write the original code. They are a good team and I can recommend them for MVP building. However, one of the issues I had was they didn’t comment their code. As I have taken a more direct role in the technical side of boomering, it is a pain to parse the code and figure out what is going on. Enter Chat-GPT.

I created a personalized GPT called Flutter Kick that is prompted to provide feedback on an app with boomering’s technical stack. Then, I’ve gone through pieces of the code and shared them with Flutter Kick. It spits out recommendations, documentation, and even new code. It has been a superb help in untangling a lot of the spaghetti code that I inherited. It’s even found some silly security flaws that I didn’t know existed.

I am not a computer scientist, nor a programmer, but at this rate I will be a junior programmer by the end of the year. Chat-GPT has absolutely democratized coding, and if this existed when we started boomering we would have likely expended fewer resources on the initial build. I still think it’s necessary to have programmers to write and maintain the initial code, to plan the architecture, and to provide a devops service. But now, there is no reason founders and product people shouldn’t understand what is going on behind the scenes.

I suppose the interesting question here is whether or not we even need code documentation anymore? If AI doesn’t require it and only spits it out for human use, once AI starts writing all the code, why would we need documentation? Maybe that’s what my original third party partner had in mind when they first wrote their code. Can’t be sure, but probably not.