Complacent Innovation

Last month, the Beatles won a Grammy award for best rock performance for their song, “Now and Then.” They had also been nominated for record of the year for the same song.

The Beatles last performed together on August 29, 1966, and last recorded together on August 20, 1969. This “new” song is built on a demo recorded by John Lennon in 1977, with some overdubs made by George Harrison in 1995, new material added by Paul McCartney and Ringo Starr in 2022, and a few snippets from earlier Beatles songs recorded in 1966 and 1969. This is the third (and likely, the last) posthumous Beatles song to be released, after Free as a Bird and Real Love made their appearances in the mid-90s as part of the Anthology series.

Image generated by Dall E.

The song’s production relies heavily on technology that wasn’t available even a few years ago. The original demo had audio problems that prevented its use in 1995, but Peter Jackson’s magic AI software was able to overcome those shortcomings. While nothing on the track is AI generated or enhanced, the technology removed extraneous noise from the tape machine and the room in which the demo was recorded. It separated Lennon’s vocal from the piano he was playing so they could adjusted individually. Similarly, the recordings from 1995 were isolated and mixed in, along with the new stuff. It’s literally a record that was 35 years in the making.

But is it any good? Sure. It’s fine. It’s very Beatle-y. Like the two other new songs (that are now 30 years old), they sound like Beatles records. John has a wispy, piney voice consistent with much of his other work. Paul is distinctly Paul, laying down a melodic bass and harmonizing with John in a warm, melancholic nostalgia befitting the circumstances surrounding the song’s development. The instruments and arrangement are very typical for a Beatle song. Is it the best Beatles song? Definitely not. Is one of the best songs of 2024? Probably not. But it’s a NEW BEATLES SONG! Just listen to it. It’s just like they’re back from the dead!

And that’s the problem.

See, the Beatles weren’t Beatle-y. When a new song came out, it didn’t sound anything like the old stuff. Beatles for Sale was nothing like A Hard Day’s Night. It’s hard to believe that Revolver and Sgt. Pepper are the same people, released less than a year apart. She Loves You, Penny Lane, Back in the USSR, and Here Comes the Sun are all the same band. The new stuff never matched the established pattern. They were always breaking new ground. They were always pushing the boundaries. They never colored inside the lines.

So maybe the thing that rings hollow with Now and Then is that it fits in so well with the things that came before. What are the characteristics that define a Beatles song? The rhythm is familiar. The bass style is something we’re used to seeing. The string arrangement reminds us of Eleanor Rigby (and so many others). And we have one person singing the verses, another on the bridge, and a duet on the chorus. I’ve got a feeling. A feeling deep inside.

This isn’t an AI-generated song. But it very well could be. Technology is really good at echoing us back to us. It can take the things we give it, and give back other things that are like it. I’ve said before that AI can write a Taylor Swift song, but it can’t invent Taylor Swift. But it goes beyond that. If we ask the algorithm to review resumes for a job opening we have, it’s going to pick people who fit in with the staff we have now because we’ve trained it on our previously successful candidates. It won’t recommend people who will challenge our existing thinking or push us to grow in a new direction. If I ask AI for help on selecting the best dog food for our new puppy, it’s going to recommend the brands that lots of people are buying and talking about, without necessarily weighing the nutritional characteristics that are much more important. When I go to YouTube and it recommends videos for me to watch, it’s going to give me things that are very similar to the ones I’ve already watched. The YouTube algorithm is designed to get me to watch more videos, not better videos.

Sometimes, that’s okay. There’s comfort in familiarity. Maybe I do want to just watch Youtube videos that are like the ones I’ve already seen. But there’s a danger in the stagnation. And every time we click, we’re further cementing the idea that we just want to stay where we are. That’s not how we got here, and it’s not how we’re going to get out of here.