- Artificial intelligence is breathing new life into the Beatles’ music after almost 50 years since the band’s breakup.
- The technology is being used to reunite the Fab Four by combining their solo career songs and even restoring Paul McCartney’s voice from his later works.
- Even though fans are emotionally moved by these creations, these songs raise significant ethical and legal questions
When the Beatles broke up over 50 years ago, their devastated fans were craving for more. But now AI technology has been used to reunite the Fab Four by combining their solo career songs and to restore Paul McCartney’s voice to its youthful peak in his later works. These AI-generated creations showcase the significant progress made in this technology, but they also raise ethical and legal concerns.
However, these AI-generated versions have received quite emotional reactions from listeners. “I’m crying! It’s so beautiful!!!” wrote one listener in a typical YouTube comment about the soaring AI of McCartney’s 2013 single “New,” in which his genius songwriter and friend “sings” the vintage chorus and bridge. Another impressive AI creation is a revamped version of “Grow Old With Me,” one of the last songs written by Lennon that was released posthumously after his murder in 1980. Recently, its new AI-created version has been called “Dae Lims”.
With improved sound quality, an orchestral arrangement and harmonized backing vocals reminiscent of the Beatles’ prime era. These AI-generated tracks have struck an unexpected emotional chord with fans longing for a reunion of the iconic band. “When I hear that, I lose it. I’m going to cry,” said Steve Onotera, also called “SamuraiGuitarist”, a popular musician with 1 million followers on Youtube. After a violent breakup, the most influential band in history left. Fans were left without a definite “happy ending,” he said. “So when we have this reunion artificially but convincingly created by AI, it’s surprisingly emotional”.
Earlier, “Heart on a Sleeve”, the AI-generated song by Drake and The Weeknd, garnered millions of hits on TikTok and other platforms. These covers use scraping technology that analyzes and captures the nuances of a given sound. The creators would probably have sung the parts themselves and then used a cloned voice, like applying a filter to a photo.
While the results can be stunning, they are not easy to achieve and require skilled human operators who combine new AI tools with extensive knowledge of traditional music processing software, Zohaib Ahmed, CEO of Toronto-based audio cloning company Resemble AI, told AFP.
“I think we’re still seeing a very small percentage of the population that can even use these tools,” he said. They have to “jump through hoops, read documentation, have the right computer, and then put it all together.”
Ahmed’s company is one of many offering a platform to make the technology more accessible to entertainment customers — citing that Resemble AI’s voice cloning platform has already been used successfully in a Netflix documentary series, “narrated” made by the late art icon Andy Warhol using the AI-generated voice.
According to Patricia Alessandrini, a composer and associate professor at Stanford’s Center for Computational Research in Music and Acoustics, the recent proliferation of AI tracks represents a significant milestone for a technology that has been rapidly advancing but has remained largely hidden from the public eye.
“This is a great example of what artificial intelligence does very well, which is something like: Train it for something that already exists,” he told AFP. But he added that when it comes to creating new ideas, “There’s really no expectation that it’s going to replace the rich history of humans originating art and culture.”
The impact on the music industry is huge. As AI technology progresses, it could become easier for individuals to transform their own vocals into those of their favourite singers. This raises questions about the rights and compensation of deceased artists whose voices are used in AI-generated music.
The issue of copyright is also complex, with potential claims from songwriters and holders of master recordings. AI creators might argue for “fair use,” citing precedents such as Google’s permission to archive books.
Last month, however, the US Supreme Court tipped the scales the other way when it ruled that a Warhol print of the late pop star Prince infringed the copyright of the photographer who took the original photo. Additionally, celebrities can protect their likeness through the “right of publicity,” which was established when Bette Midler sued the Ford Motor Company in the late 1980s for using a singer who sounded like her in an ad.
Ultimately, “I think there could be voluntary industry standards … or it would be done through litigation,” Ostrow said. Rights holders will also need to consider the potential negative public relations implications of suing over fan-created tributes that are not intended for commercial purposes.