Or they could change text-to-speech to "speech synthesis" ...