At this point I don't think there's any point arguing with this belief. If you haven't found a way to make the models useful you will have a lot of difficulty staying relevant

I wouldn't hire anyone who doesn't use LLMs and I specifically screen for people who are good at it