Music is a large part of my life – I consistently log hundreds of thousands of minutes on Spotify annually. However, as much as I and other fans support artists through the service, they still earn an incredibly low cut for the volume of streams.
As of late, AI has made their jobs a bit more difficult.
Image credit: Timmy Loen
Voice replication with generative AI has become common. Programs such as So-Vits-SVC simply require a short audio sample to immediately clone a voice, then it can replace or mimic vocals on a preexisting track.
The tool, which has created glaringly fake covers of hits such as Hey There Delilah in the style of Kanye West, was initially started as a joke. But content creators have come to capitalize on the now-uncannily realistic songs.
Last week, an original song called Heart on My Sleeve mimicked the voices of Drake and The Weeknd and was uploaded on Apple Music, Spotify, YouTube, and Tidal, among others. The track, which was later removed, garnered over 1.6 million listens on Spotify.
Counterfeits have always been a problem in the music industry. From duplicate vinyl and bootlegged records to file-sharing via Napster in the early 2000s, music has been susceptible to theft because of how easy it is to distribute.
Though Heart on My Sleeve wasn’t a direct copy of any existing work, it utilizes copyrighted material to create, depriving artists of compensation for their intellectual property. But most frighteningly, the work is non-consensual and can be misappropriated to convey messages the original artist may disagree with.
The acceleration of innovation in generative AI has also made it difficult for legislators to keep up. In the US, current laws state that works that are “transformative” do not infringe copyright. However, this term is open to interpretation.
Image credit: Timmy Loen
After an influx of AI-generated music, Universal Music Group, under which Drake and The Weeknd are signed, told streaming platforms that it “will not hesitate to take steps” to protect its rights and those of its artists.
Though there’s some disapproval of the new AI tech, some artists welcome it. Singer Grimes told her Twitter followers that fans should “feel free” to use her voice without penalty and said she is fine with receiving 50% of the royalties for these. This would be the “same deal” she would have with any artist she collaborates with anyway, she said.
Regardless, this tech is here to stay. Google has also been developing MusicLM, a text-to-audio program trained on over 280,000 hours of music.
Stay updated on the go with our mobile app.
Get latest insights with smoother, more personalized experience through TIA mobile app.
