I was being pressured to make my workflow ai centric when I was let go from my last job and I was one of only two people there who refused to put a high level of trust into what the chatbot was spitting out (it was making very clear errors and often was just hallucinating in superficially convincing ways.) I've heard they have gotten good at code generation but this was a different kind of engineering anyway and one where safety was a huge concern. I ended up being laid off probably in part because of that.