Google researchers have developed an application that uses a new artificial intelligence system to create music from written commands.
Google has developed an artificial intelligence technology called MusicLM that turns text into music. MusicLM can create minute-by-minute music tracks from text prompts, transforming a whistled or humming melody into other instruments, similar to how DALL-E creates images through texts. You can find examples of the application here.
THERE ARE FIVE MINUTE PIECES
Although there have been similar studies in the past, it has been reported that MusicLM outperforms previous systems in terms of both sound quality and adherence to text description. There are 30-second tracks that look like real songs, built from paragraph-length descriptions that predict atmosphere and even specific instruments, as well as five-minute tracks made from a word or two like “melodic techno.”
MusicLM can even simulate human vocals, but the sounds still remain static compared to the human voice. The application has not yet been made available to people. The company did not comment on when it will open.