Take a source WAV, make a MP3 from it. Invert the mp3. Subtract the mp3 from the WAV. The resulting “compression artifact” will seriously shock you by the remarkable amount of ‘content’ removed; not only will the song be entirely wholly instantly recognizable, it will in fact not sound very much worse than the mp3. mp3 has been a [devastatingly destructive] hollowing-out of sound quality akin to how transistorized audio destroyed most of recordings made during the early-transistor era.
When MIDI was iniitially hyped similarly to how AI is being initiailly hyped, everybody screamed about the imminent demise of the drummer. LOL. Well, we all see how that went…didn’t.
“AI is a magic trick. You say oh neat and then you never care about it again.” is a perfect explanation.
My favorite quote about ai music, “why should anybody listen to music you couldn’t be bothered to write?”
