About Generative AI in Music
Generative AI in Music than just a cool new toy. It’s an incredibly powerful tool that could completely change how songs are written and produced.
Fully generative apps can “scrape” huge databases of existing music to learn musical structures, harmonies, melodies, rhythms, dynamics and timbres – then use those patterns to create entirely new content. The result is often utterly surprising and strikingly beautiful. The winning entry in this year’s third international AI Song Contest by Thai musician and engineer Yaboi Hanoi, for example, used sounds from native Thai instruments to recreate the sound of a traditional woodwind instrument called the pi nai, which has a non-Western system of intonation.
Shaping the Future with Pixels: A Deep Dive into the History of Generative AI
More interestingly, generative AI can also reproduce the distinctive voice of an artist, right down to the vocal tics and styles they’re known for. This technology has become a hot topic of interest among major music companies, with many, including Spotify, BandLab’s SongStarter and Authentic Artists, investing in and developing generative AI music creation tools for artists.
But these systems are generating some serious controversy over copyright issues. When a generative AI system authored a song named “Heart on My Sleeve” that used voices similar to those of Drake and The Weeknd, the track raked in millions of views across multiple streaming platforms before it was pulled from circulation. The music industry’s reaction to these systems, and the legal questions they raise, shows how much of a game changer this type of technology is.