Genuinely want to hear from the AI skeptics - what keeps breaking for you?

really want to understand why a lot of people complain about Claude/Codex not performing as per their expectations. What are you guys building? What’s erroring out, or what’s frustrating you? In my case (being a DevOps guy), I’ve solved things at record speed since getting Claude/Codex access. When I say solved. here are a few examples of what I was able to achieve: Multi-screen webapps: I was afraid of building UI, but now I’m able to build end-to-end solutions used by both dev and biz teams. Not to forget