Until someone fine tunes their LLM to match whatever the trend is and then the contrarians buck that trend for something else. An endless game of cat and mouse.

I personally don’t care if an output is from LLM or human, as long as someone has validated it and determined that it’s correct and useful.

True, but I don't think we have to underestimate the power of defaults. I doubt that defaults will be like that