All you need is an AI use case.

The Beatles’ use of technology on their new song carries some useful lessons for the rest of us.

Peter A. McKay
3 min readNov 5, 2023
The Beatles landing at New York’s JFK International Airport in February 1964. Photo by United Press International via Wikimedia Commons

Gen Xers like me remember cassette tapes as a rather crappy tool for listening to music, let alone making it. Cassettes would inevitably either wear out or get “eaten” — which is to say hopelessly mangled — in the guts of your Walkman.

It makes The Beatles’ new song, released Thursday, all the more miraculous to learn it started life as a humble cassette recorded by the late John Lennon as a demo in the 1970s. However, as an excellent making-of video on the band’s official YouTube channel explains, the song was ultimately finished with a big assist from a much newer technology: artificial intelligence.

Thus the new song, titled Now and Then, is an important contribution to the broader conversation we are all having right now about potential impact and use cases for AI. In particular, I think two comparisons are important to discuss regarding the Beatles’ use of AI.

The first is to compare Now and Then to two songs the Beatles released in the late 1990s using technology of that era to flesh out Lennon’s parts. Those songs — Free as a Bird and Real Love — started on the same demo cassette as Now and Then.

Paul McCartney, George Harrison, and Ringo Starr tried to complete all three songs in the studio in the ’90s. But they ultimately decided not to release Now and Then with the other two because they couldn’t get Now and Then to sound “right” at the time. In particular, McCartney says in the making-of video that Lennon’s voice and his piano playing were too muddled together on the demo version of the song.

More than two decades later, AI helped the band solve this problem, isolating and sharpening the quality of Lennon’s vocal, without the piano. The film director Peter Jackson suggested this solution to the Beatles because he had used AI to improve images and sound on Get Back, his sprawling 2021 documentary about the band for Disney+. (That work is a technical marvel in its own right, by the way. It’s based on archival footage from 1969 but looks and sounds like it was shot much more recently.)

The result of the Beatles using Jackson’s AI to clean up Lennon’s vocal on Now and Then is stark. Frankly, the song sounds way better — and much less like a demo — than the two ’90s singles. The comparison isn’t even close, which underscores how much of a leap forward AI is compared to the audio tech available in the ‘90s.

Listen to all three songs for yourself here, here, and here. Or, if you prefer, just skip near the seven-minute mark of the making-of video for Now and Then, where they play back Lennon’s AI-isolated vocal track, clear as a bell.

As Ringo puts it: “Since Peter took John off and gave him his own track, it’s like John’s there. It’s far out.”

The second big comparison worth making is how Now and Then differs from other AI-influenced songs that have been in the news lately. In its coverage of the Beatles’ release, Decrypt observes:

Up until recently, generative AI audio models were making headlines primarily due to unofficial “collaborations” and remixes, including the infamous ‘Heart on My Sleeve’ AI-generated song emulating artists Drake and The Weeknd, and a remix of The Weeknd’s ‘Starboy’ featuring an AI-generated Selena Gomez.

Now and Then isn’t like those tracks at all because it’s rooted in the Beatles’ actual, flesh-and-blood performances. AI wasn’t used to replace any of the Beatles. It just helped them do specific tasks to make them more effective at their jobs in the studio.

This is probably where AI is headed in a lot of other fields as well, I think. The sensational narrative of people being replaced outright gets a lot of press attention. But the real sweet spot is probably to use AI to enhance the human touch in the work we do, to wed the two together, sometimes beautifully so.

This post was adapted from my free email newsletter #Web3Weekly. To get the full version in your inbox every Sunday, including additional headlines about decentralization from around the internet, please subscribe here.



Peter A. McKay

Storyteller, thought leader, and marketer focused on blockchain/web3. I publish #w3w, a newsletter about decentralization. Ex-reporter for the Wall St Journal.