Nearly everyone I suppose. There are always outliers I suppose. I’m going for 90%+ using llm in some capacity even if just a single ChatGPT prompt.