At work we've been heavily encouraged to use LLMs for coding, and I have been productive enough to be able to avoid it for the most part. But I decided to, on a personal project, occasionally ask AI for help since it came as part of my JetBrains subscription. Instead of getting the massive productivity boost LLM proponents swear they get from using it, I continuously find myself horrified at the prospect that this unreliable technology is apparently writing the majority of software now. I have not had a single instance where the LLM was able to give a proper suggestion that helped me resolve something I was stuck on. It frequently suggests incorrect solutions, makes bad assumptions, and gets caught in loops. Often it will insist that you are the one in the wrong when you correct it. In work projects where I've had to deal with code entirely generated by LLMs with no oversight I've found myself horrified at how mangled the codebases get. LLMs will happily shovel in multiple subtly wrong implementations of the same thing into your codebase over and over and over again, making it impossible to know what's actually being used. LLMs will write useless tests, they will write overly verbose code, they will change things that didn't need changing. It makes me feel like I'm out of touch with reality somehow,. What is going on? Am I the only one that only gets slop out of the slop producing machine?