I grant your broader point, but extrapolating from this marketing copy is not a great example.

The real question is, if you go back to the bot following this conversation and you challenge it, does it generate the more correct answer?