The Beatles’ “Now and Then” Was Created Using Artificial Intelligence (and That’s Okay)
I attended the Grammy Awards last night, and while pop, rap, and country definitely took center stage, it was the Best Rock Performance category that caught my attention the most. The award went to The Beatles for their song “Now and Then “, which is a confusing sentence in 2025. You may know that The Beatles were active mostly in the 1960s, not the 2020s, and considering only half the group is still alive, winning a Grammy for a new song might seem a little strange.
Still, the song is new —at least as new as a Beatles song can be. The track was released at the end of 2023 and is based on a demo recorded by John Lennon shortly before his death. In the 90s, the three living members of the Beatles (Paul McCartney, Ringo Starr and George Harrison) tried to finish the song but never completed it because the original sound quality was too poor. However, modern technology has made it possible to create the product you hear today.
I love The Beatles, but the song itself wasn’t exactly what caught my attention. (I didn’t think much about it until Sunday.) Instead, the announcer made it clear that the song was written using artificial intelligence, a statement that in 2025 would cause everything from an eye roll to a heavy sigh from this tech editor. Music generated by artificial intelligence is very real today. While these AI-generated tools provide some compelling results, there are plenty of clues to pay attention to . The idea of The Beatles releasing a song with some level of artificial intelligence didn’t appeal to me, and I’m willing to bet a lot of others didn’t either.
However, it is important to note the difference between created and created AI. They sound the same, but they’re not. I want to be clear: “Now and then” is not a song created by artificial intelligence: no one is using artificial intelligence to recreate John Lennon’s voice and create another award-winning song.
Wrong AI
AI is not just artificially created content that we are all accustomed to. While it seems like AI has taken over our lives with the advent of ChatGPT at the end of 2022 , companies have been incorporating this technology into our products and services for much longer.
AI is perhaps a bit of a misnomer. In this context, machine learning is a little more precise. Simply put, machine learning is when a program is able to adapt and grow based on the data it receives, much like how our minds work. You feed the program training data, and it adjusts its assumptions and results accordingly. Although the actual process is much more complex, machine learning allows programs to do remarkable things.
One of those things is the separation of audio tracks: one of the reasons the Now and Then project was shelved was because they couldn’t mix the song properly because Lennon’s original recording was so rough. But with the help of an audio editing tool based on machine learning , producers were able to separate Lennon’s vocals from the piano. Neither the piano nor the vocals were created by artificial intelligence—rather, the tool could break down these tracks so that producers, along with the two living Beatles, could use them to record, mix, and release the finished song. McCartney posted this on X the summer before the song’s release:
This tweet is currently unavailable. It may be downloading or has been deleted.
For many reasons, I’m glad this “AI Beatles song” didn’t artificially bring back John Lennon. It’s truly gratifying to know that a song Lennon started decades ago was duly completed by his former bandmates, with Lennon’s son accepting the award on behalf of the band.
The music video, on the other hand, definitely takes things a little further, juxtaposing archival footage of deceased Beatles John Lennon and George Harrison with current footage of McCartney and Starr. (It was produced by Peter Jackson, who produced Now and Then as well as the 2021 Beatles documentary Get Back .) Still, it’s weirder than anything else, and certainly not a product that claims to represent reality – unlike most AI-generated content you encounter in the wild.