The Beatles have announced a 'final' record, utilising artificial intelligence to extract John Lennon's vocals from a decades-old demo. The track, 'Now and Then', was begun by Lennon in 1978 and completed last year, with AI technology separating his voice from a low-quality cassette recording.
Paul McCartney explained in a Radio 4 interview that the AI was trained to isolate Lennon's voice and piano, removing other instruments. 'We were able to take John's voice and get it pure through this AI,' he said. The technology, also used in Peter Jackson's 2021 documentary series 'Get Back', marks a significant leap from the 1995 Beatles reunion 'Free as a Bird'.
The development raises ethical questions about posthumous releases and 'deepfake' vocal imitations. While fans often welcome new material from deceased artists, critics argue that such projects can feel exploitative. The music industry has seen a surge in AI-generated tracks mimicking stars like David Bowie and Freddie Mercury, though many lack artistic depth.
Despite the controversies, some artists like Grimes and Holly Herndon have embraced AI voice technology. However, legal frameworks around AI and intellectual property remain unclear globally. The Beatles' final record exemplifies both the creative potential and moral dilemmas of resurrecting voices from the past.
