AI set to upend melody making

Machine learning tools are already quietly transforming the music industry

The use of AI will undoubtedly lead to the revolution of music composing, but robots will not replace humans in the art of making melodies. Or at least that is the opinion of Francois Pachet, a scientist, composer and the director of the Spotify Creator Technology Research Lab.

“AI will not replace good artists and composers. AI will change the way people make art, but it won’t replace them,” Pachet told participants of the TechnoArt 2019 conference in Tel Aviv.

Pachet is considered a pioneer of computer music, and specifically its interaction with AI. At Spotify he currently leads development of AI-based tools for musicians. Before that he was director of SONY Computer Science Laboratory in Paris, where he and his team created artificial intelligence software - called Flow Machines - to make new kinds of music. As a result of the experiments, which involved the feeding of the software with 45 Beatles songs, Flow Machines managed to produce a brand new song, “Daddy’s Car,” based on the Fab Four’s style, by studying the notes and patterns of the Beatles songs and “then using them in a new context to create something new.” Flow Machines was also able to create an album of 15 songs called “Hello World” - the first music album ever composed with AI - together with French songwriter and producer Benoit Carre last year.

Outside of composition, AI tools are used to gain insights into how consumers listen to music, as well. In a study published in March, Pachet and two other researchers talk about the phenomenon of “music skipping.”

As we all know, online streaming services are a preferred way to consume music. But because there are so many songs available via these services, consumers tend to “skip” quickly from one song to another. This has allowed the researchers to create a “skip profile” - which indicates how long people actually listen to music before skipping to the next song. Research data presented proves that a quarter of all streamed songs are skipped within the first 5 seconds, 29% in the first 10 seconds, and 35% in the first 30 seconds, and only some 48% of all songs listened to in their entirety, Pachet said.

“Skipping is a crucial feature in understanding modern listening behaviours,” the researchers say in the paper. “For the first time in the history of musicology, researchers can systematically collect and analyse massive amounts of data about music listening behaviour.”

Most interestingly, the skip profile for each song was proven to be the same every day, every week and every month for everyone included in the study.

“This means that skipping is not people centred, but it is because of the song, it is the signature of the song,” Pachet said.

The research also showed a correlation between skipping and the structure of the music piece. With this information, he said, musicians will be able study how people react to their music and will be able to develop compositions that are better-received by the audience.

In addition, Pachet said, with the advancement of the use of AI tools in music, regulators may need to redefine the concept of what is considered “original” content, and perhaps create new copyright laws and a new royalties system for music that is reinvented by AI, but based on the remixing and reworking of originals.

Similar articles