Cheeky title aside, I'm going to contribute another take to the AI discourse. Useful or not, I can share my reflection that's been running through my head lately.
I am noticing myself becoming more impatient and learning fewer programming-related-things due to my usage of LLMs.
There's really not much more to say than that. What might have taken me more time and mental work in learning how to do something previously now simply doesn't take any work at all. My brain is not learning new things. Not only that, when I try to stop and ask myself, "Hey, would you like to try to learn this and actually read the documentation?" I am too impatient to do so.
This is fine and good if your entire goal is to get that result. But as everyone and their cross-stitched pillow keeps telling me, it’s the journey that is more important than the destination. But I’m starting to not have a journey at all. I’m too impatient to even leave the house.
I sort of realized this as I took on setting up a home server, my first computer-y hobby project I’ve done in my free time in ages. You should see the long Claude conversations I had in order to orchestrate Docker, set up backup scripts, and learn enough about Arch Linux to be dangerous. But if you ask me how the nuanced parts work, I'd come up short. What's more, nearing the end of theproject, I was growing so impatient to have my slick home server setup that I was barely reviewing the python code that Claude was spitting out. It may be a me problem but it's still a problem.
Still, it was pretty amazing for me to be able to set up a home server with such little effort. I see now that with each project like this that I take on, there is an opportunity to assess: is this one of those moments where I want to learn everything, you know, the old way? or, do I want to get results faster and risk getting impatienter and dumberer?.