Large language models may be standardizing human expression
I think it is important to distinguish "human expression" from copying a response from an LLM. Someone who outsources their thinking to an LLM is only offering an AI's expression. It's not human expression.