On 13th June, Paul McCartney brought some good news: The Beatles are going to release a new song. How is this possible when the group’s emblematic member, John Lennon, died in 1980?
Paul McCartney revealed during the announcement that it was thanks to artificial intelligence that the group would be able to release new material. To achieve this, John Lennon’s voice was reconstructed from an audio cassette where the singer had recorded his voice for a song before his death. So, more than 50 years after the legendary band split up, fans will be treated to a new song.
“We managed to take John’s voice and purify it using AI to mix the recording,” he said. “We came to make what will be the last Beatles recording, it was a demo of John’s that we worked on,” explained Paul McCartney. “We’ve just finished it, and the song will be released this year,” he added.
The song wasn’t generated by artificial intelligence, but AI did enable the band to recreate the sound of its former member, John Lennon. This recreation technology enabled the Beatles to create an original track, incorporating the voice of a late member.
Paul McCartney was quick to point out that he found the technology “cool but scary”. And it’s not the first time artificial intelligence was used to create a Beatles song. In 2016, artificial intelligence Flow Machine, created by Sony’s Paris lab under the direction of François Pachet, caused a stir with the release of “Daddy’s car”.
Daddy’s car was entirely created by artificial intelligence, which was fed with Beatles music data. The result: music based on the band’s repertoire that’s a perfect illusion and seems to have won over some people, with the track recording over 2.9 million views on YouTube.
It wouldn’t come as a surprise if this type of projects multiplies in the future, with the expansion of artificial intelligence.