Blurred lines of A.I. require musical clarity (1/3)
For new musical artists, breaking into the streaming business is becoming more difficult every day.
The problem is, quite simply, volume. As of last count, there were an estimated 120,000 new tracks being uploaded to major streaming services every day. With that much audio content out there, bringing attention to a single new song can just be about impossible for an independent artist.
The problem has become so acute now that, according to data released in July 2023 by Luminate (luminatedata.com), about 42% of all tracks uploaded to streaming services (about 67 million songs) have had fewer than ten plays, and nearly a quarter – 24%, or about 38 million tracks – have had no plays at all.
And with the arrival of mass-market artificial intelligence (“AI” or “A.I.”) music tools that boast of their ability to create millions of songs, the problem could get much worse in the coming months, let alone years.
Artificial intelligence tools that generate text, images, and music are moving creativity into new territory, raising sensitive questions in law.
AI has become a hot topic in the music industry in recent months, with new examples each week of astonishing AI-generated music, and concerns voiced about the “widespread and lasting harm” of such AI tools to music creators, stake holders, and rights holders.
For the moment, let us consider merely three examples in the music industry regarding artificial intelligence. One is that of Giles Martin, son of Sir George Martin, producer for the Beatles. Last year, in order to remix the Fab Four’s 1966 album, ‘Revolver’ (which, for the trivia-minded, featured one Anil Bhagwat playing tabla on George Harrison’s “Love You To”), Giles utilized AI to learn the sound of each band member’s instruments (e.g., John Lennon’s guitar) from a mono master tape so that he could separate the instrumentation, and reverse engineer them into stereo. On October 28, 2022, a remixed and expanded edition of ‘Revolver’ was released. It included a new stereo remix of the album by Giles Martin, with the help of de-mixing technology developed by movie director Peter Jackson’s WingNut Films.
Second, in April 2023, a new song purportedly by Drake and The Weeknd, “Heart On My Sleeve”, appeared on TikTok and Spotify and quickly spread like wildfire across the internet. The song was apparently met with rave reviews and a high degree of excitement among hip hop fans, not only because of the track’s infectious melody, but because of the featured artists when, in reality, they had not created the song at all!
Instead, the tune had been created using artificial intelligence by a TikTok user who went under the guise of Ghostwriter977, having sufficiently trained AI on Drake and The Weeknd’s works, and generated this “new” song, which impeccably mimicked the artists’ voices, lyrics, and musical styles. Within days, however, the creator’s video, which had gained over nine million views, was removed from TikTok, Spotify, and other digital platforms in response to claims by the artists’ record label, Universal Music Group, home for both artists, who successfully petitioned to have the song removed from streaming services. Expectedly, the track sent shock waves surrounding intellectual property throughout the global media and entertainment industry.
As a third example, in June 2023, Beatles founder Sir Paul McCartney announced that he had employed AI technology on an unreleased Beatles demo from the ‘70s, informing BBC Radio 4’s ‘Today’ programme that AI had been used to “extricate” John Lennon’s voice from a cassette recording containing a demo. “We were able to take John’s voice and get it pure through this AI,” he said. “Then we can mix the record, as you would normally do. It gives you some sort of leeway.”
McCartney has since expounded on the process in a social media post, following widespread coverage amid concerns about how AI will affect the livelihood of artists in future. “We’ve seen some confusion and speculation about it,” the musician subsequently clarified. “Seems to be a lot of guess work out there.”
McCartney further added: “Can’t say too much at this stage but, to be clear, nothing has been artificially or synthetically created. It’s all real and we all play on it. We cleaned up some existing recordings – a process which has gone on for years. We hope you love it as much as we do.”
However, McCartney has not revealed the title or any lyrics from the song, which is due to be released later this year, although it is widely believed to be a 1978 Lennon composition titled “Now And Then”, which was included on a cassette labelled “For Paul” that Lennon had recorded shortly before his death in 1980.
Lennon’s widow, Yoko Ono, apparently provided the cassette to the three surviving Beatles in the ‘90s when they were working on their ‘Anthology’ project – a retrospective of the Beatles career that included three albums, a documentary, and a book, wherein two songs from that tape, “Free As A Bird” and “Real Love”, were officially released as part of ‘Anthology’, recorded by the surviving Beatles and utilizing Lennon’s original voice recording. But “Now And Then” was considered unsuitable for release at the time, with any recording attempts quickly abandoned by the band.
“It didn’t have a very good title, (and) it needed a bit of reworking, but it had a beautiful verse and it had John singing it,” McCartney has said. “[But] George didn’t like it. The Beatles, being a democracy, (so) we didn’t do it.” In a 1997 interview with ‘Q Magazine’, McCartney revealed that the late George Harrison had called the song “rubbish”.
Meanwhile, McCartney has referred to AI as both “scary” and “exciting” adding, with wisdom, that “it’s something we’re all sort of tackling at the moment and trying to deal with…it’s the future. We’ll just have to see where that leads.”
The idea to use AI to reconstruct Lennon’s initial demo came from the filming process of ‘Get Back’, Peter Jackson’s eight-hour Beatles docuseries which used similar AI technology to clean up the audio from archival Beatles footage by separating voices from background noise.
…to be continued