Filmmaker Steven Soderbergh has detailed his use of generative AI in the upcoming documentary John Lennon: The Last Interview. The film, created with support from Meta, pairs AI-generated visuals with Lennon's final interview audio. It is set to premiere at this month's Cannes Film Festival.
Steven Soderbergh defended his decision to incorporate generative artificial intelligence into John Lennon: The Last Interview, a documentary based on the audio from Lennon's interview with Yoko Ono on December 8, 1980, hours before his murder. Announced last year with the backing of Lennon's estate, the film mostly uses archival material but employs AI for about 10 percent of its content, specifically during abstract philosophical discussions. Soderbergh described this as 'thematic surrealism' to create metaphorical imagery that enhances the audio without misleading viewers. He compared the AI to VFX or CGI, insisting it is not intended to 'fool' or 'manipulate' audiences into thinking the images are real. Examples include one-year-old babies in Sixties outfits bawling and cavemen depicting male behavior, scenes he said are impossible or less effective to shoot practically. The AI sequences were developed in partnership with Meta, which provided technology and funding after producer Michael Sugar suggested the collaboration. Meta viewed the project as a way to stress-test their video generative tools, with Soderbergh agreeing to finish the film using them. Sean Ono Lennon supported the approach, telling Soderbergh that his father 'would’ve wanted to engage' with the technology, as the Beatles loved experimenting with new tools. Soderbergh emphasized transparency as key, stating his 'moral obligation' is to deliver the best version of the film honestly. He dismissed misconceptions like recreating Lennon, saying, 'Do I look like somebody that would do that?' John Lennon: The Last Interview is scheduled to premiere at the Cannes Film Festival this month.