Stability AI releases a sound generator

Called Stable Audio Open, the generative model takes a text description and outputs a recording up to 47 seconds in length.