Google presented MusicLMan innovative tool that leverages artificial intelligence (AI) to generate original music and songs starting with a short text description. The user just needs to write some keywords, such as the musical genre or the instruments to use, and MusicLM will create a personalized song in seconds.
MusicLM is available online, on Android and iOS, but only to a limited number of users who have signed up for a waitlist (you can join the waitlist here).
Music created by artificial intelligence: the copyright problem
The project was anticipated by a scientific paper published a few months ago, in which Google researchers showed that MusicLM surpassed previous music generation systems “both in sound quality and in fidelity to textual description”.
Also try Amazon Music with 30 days of free use
However, the same article also highlighted some ethical problems associated with this type of technologysuch as the ability to create songs that mimic other artists’ voice or style without their consent.
“We are aware of the risk of potential creative rights infringement associated with this scenario,” the researchers wrote, adding that “we strongly emphasize the need for further studies in the future to mitigate these risks related to music generation” and that “we do not intend to release these models to the public at the moment”.
Three months later, however, MusicLM has become accessible, albeit to a lucky few. The official announcement does not mention how Google intends to address ethical issues surrounding content creation without acknowledging the work of the people or groups from which the model learned to make music.
Just last week, under pressure from Universal Music Group (UMG), Spotify deleted thousands of songs generated with Boomy, another service that allows you to generate music using artificial intelligence.