You are correct the weather channel’s voice started out human, but the computer is pulling up sound files. (No humans required after the original recording). I hope “suits” at one of the “C’s” (CC, CBS, Citadel, and Cumulus) does not read this: but I have always wondered why some one has not taken a “voice” and recorded the typical phrases, intros, song sweeps, time and weather conditions and do what the weather channel does with local forecasts. The RDS could trigger the correct sound file for that song. You could have one voice on ALL stations. After the first sets of file are recorded you could even “part time” this voice to just come in and pronounce the new song titles. Just think the executives who say anyone can do a radio show could even do it themselves. No need for expensive talent!!