[GavinMcG]

I think one factor is the lack of variation. Sure, a copywriter might use those techniques as a hook, but there’s far more content using them paragraph after paragraph after paragraph than I’ve ever seen before.

You might also reframe how you read those comments. Perhaps when people are labeling a piece as “written by AI,” they’re just conveying that they perceive it to use the same “voice” that LLMs use, and judge that voice negatively. Sometimes people say things non-literally and don’t need proof.