The Fugatto AI model was fed millions of audio samples by Nvidia employees in order to be trained to create a range of sounds that had never been heard by human ears ...