A lot has changed since the last time I wrote about how I use AI.

I’ve gone from one end of the spectrum to the other. Starting out with asking claude.ai and chatgpt coding questions to using coding agents to handle coding for me. At first I didn’t like it; I really enjoy coding and felt like it was being taken away from me. Now, however, I realize that I also enjoy building; and just because coding agents exist doesn’t mean I have to use them. Sometimes I use them and have to go in and fix their errors. So I’m coming around and no longer having an existential crisis about it.

I don’t let it lead me blind: I’m the engineer and must understand the code I’m shipping. I’m still responsible for that. Therefore, when I’m having an agent write code for me, if I don’t understand what its written I have it explain it to me until I do. This also prevents brain rot: I’m still learning.

The content on this site and any comms from me is still just me. Coding agents are just for that purpose: coding.