I'm a full-stack developer feeling really conflicted about using AI in my work. On one hand, I find it incredibly convenient for getting tasks done, but it's impacting my confidence in the code quality and taking away the joy of coding. It used to feel great when AI was just a fancy autocomplete tool, but now I'm using the agent mode in my coding software and I feel miserable about it. I'm looking for advice on how to strike a balance; how do you all integrate AI into your workflow while still enjoying the coding process?
11 Answers
Honestly, I just use it as a more advanced autocomplete. It streamlines some tasks, but I still prefer the satisfaction of writing my code from scratch.
I use it for quick research or reference. For coding, I rely on it for simple tasks but do the heavy lifting myself. It’s better for my learning. Also, those coding exercises? I steer clear of using AI for those as they don’t help in the long run.
I’m a full-stack engineer and I don’t use AI much because it slows me down. And let’s not forget the ethical issues surrounding it.
I typically use AI for the tasks I find boring or tedious, like writing unit tests. For bigger refactoring projects, I’ll set up a pattern and then ask it to replicate that in other areas. Sometimes when I hit a wall, I just ask AI for help. Just don't let it take full control; it's all about finding that balance!
But seriously, don’t use AI for writing tests! AI can’t know your real intentions; better to write the tests and then have AI assist with code.
I suggest trying to find a middle ground. Remember that it's a tool to help you, not take over. Automate the repetitive tasks so you can focus on the fun parts. Try to keep it enjoyable!
If you can, try to avoid using AI entirely. I find that doing things manually helps me grow as a developer, and AI just complicates it for me.
I mainly use AI for boilerplate code I already understand. The rest, I prefer to code by hand—it feels more rewarding that way.
I like using agent mode too. It really cuts down on context shifts, but I still do about half my coding by hand. I’ll write my own stuff first then get AI to review it. It’s good at finding optimization issues I might miss.
I treat AI like an intern. It’s great for menial tasks or brainstorming, but I always make the final call on my projects. Don't let it diminish your skills!
If your code quality is dropping, it might be because you're not explicit in your requests or fully reviewing what AI generates. Use it for guidance but stay in control of your projects.
My job pushes me to use AI as much as possible, but I’m thinking of leaving because it’s hurting my skills. I just let AI do the heavy lifting while I try not to care as much about it.
So 100% of the time?