iscover
Share this article on:
Powered by
MOMENTUM
MEDIA
Despite its widespread controversy, Spotify has ruled out the idea of banning AI-generated content from its service completely.
Speaking with BBC News, Spotify chief executive Daniel Ek said there are a number of valid uses for artificial intelligence (AI) in the creation of music. But he said the use of AI within music falls into three categories: acceptable, not acceptable and a grey area.
Ek said the use of autotune or other tools to improve a track are an area where AI can be used positively and should be allowed.
On the other hand, AI tools that directly mimic artists without consent should be banned, a view that Spotify had previously upheld when it removed a track from its platform back in April that featured AI-cloned voices of two major artists, Drake and The Weeknd.
Neither artist was aware that their voices were cloned on the track, Heart on My Sleeve, whose creator then tried and failed to have it nominated for a Grammy.
“You can imagine someone uploading a song, claiming to be Madonna, even if they’re not,” said Ek.
“We’ve seen pretty much everything in the history of Spotify at this point with people trying to game our system.
“We have a very large team that is working on exactly these types of issues.”
However, the line between acceptable and not acceptable gets blurred when it comes to AI-created tracks that are obviously influenced by a famous artist without directly mimicking them.
Artists from around the world have spoken out against the use of AI in the music industry, arguing that the technology is a threat to culture and creativity.
“It’s not just a threat to existing works but to future generations of artists and to culture itself,” said famous songwriter and record producer David Porter to Congress in June.
“If all we have is machine-made music copied from existing works, there will be less and less creativity, artistry, and soul to go around. What a penalty to put on future generations.”
However, Daniel Tashian, award-winning songwriter, producer and musician, said that the use of AI can be constructive and beneficial to musicians, citing the use of an algorithm used when writing Believe by Cher in 1998 to determine the ideal pitch for the track.
Tashian also said AI has yet to become a threat to real human musicians.
“I’ve yet to sort of experience an artwork created by a computer that gives me goosebumps or gives me chills because it’s so beautiful,” he said.
“So until that happens, I’m just going to keep my head down and keep doing my best organic music and using computers when they can help.”
Despite this, a number of AI tracks replicating real musicians have emerged already, and while not on mainstream music platforms like Spotify, they can still be found on YouTube in some cases.
An album replicating Oasis, cleverly titled “AISIS – The Lost Tapes/ Vol.1”, has appeared online, effectively mimicking the band’s early sound.
Comments on the video are divided, with one commenter saying he was “mind blown by how good these tunes are”.
Others were more hesitant to applaud the creation, with one saying: “This is horrible. Don’t waste your time. And please, don’t mistake AI for art. Art is human creation. AI is based on existing data. One don’t belong in the other. It’s impossible.”
Be the first to hear the latest developments in the cyber industry.
