MusicLM: Generating Music From Text

I'm skeptical. Without telling a computer what jazz is, and programing it ahead of time, there is no way a computer will know what jazz is. In fact, most people who think they play jazz don't know what jazz is.
 
From what I understand. It gets "jazz' (in your example) input from the text description. Then it matches 'audio sets' to the text.

"AudioSet consists of an expanding ontology of 632 audio event classes and a collection of 2,084,320 human-labeled 10-second sound clips drawn from YouTube videos. been pre-described by musicians."




And I've found three music genre's I now want to explore:
berlin 90's house!!! peruvian punk!!! accordion death metal!!!!!!1
 
  • Like
Reactions: Gomez Adams