This is my current job exactly. Now I have to show piles of receipts(Generated by a LLM too) to say that I used a ton of AI in my work or I will get a bad performance review. The reviews are AI judged too.
I protested against it, but I can't anymore if I want to feed my family.
I get why programmers are reaching to LLMs in corporate settings, especially given how many companies/higher ups have shifted the goalposts to either "10-100x productivity since we have LLMs now!" or "your sole performance evaluation is how many LLM tokens you use" and firing/laying off people who don't meet these absurd new requirements (which, side note, don't even usually align with the goals of the business lol), but I really don't like it. I also keep seeing people who should be (or, even just a few years ago, were) extremely good programmers outsourcing all their thinking to LLMs and who can't even explain their own code now. I'm also seeing a major rise in security issues and unstable software in places that have been using LLMs heavily.
I've got no clue how to actually solve any of these problems on a larger scale (since the people who pay the paychecks have decided that LLMs must be both good and the future, apparently), but I do generally find myself agreeing with the takes you keep making around LLMs and similar technology.