I think we’re just adapting. LLM felt kind of magical at first, and now we’re all experts in detecting AI slope.