I think a lot of the focus on AI output poisoning future models' training has left their effects on our reasoning/ vocubulary neglected, which is why I brought it up.

How we define words, what connotations they evince and the qualities they exude, simultaneously limits and enables thought.