So I was thinking about this again, and the default AI in SF is a being of numbers and logic and precision that has difficulty with feelings (ie autism coded), while a LLM is all about (imitating) vibes and bad with facts and figures. I think that may be part of why some people react so differently.