I watched a fascinating interview with Smashing Pumpkin Billy Corgan on Bill Maher’s Club Random earlier on. Corgan got onto the subject of the likely effects of AI on the music of the near future. Actually, it wasn’t about the likely effects of AI on music, it was about the certain effects of AI on music. He said: “Five years from now the music landscape will be completely devastated by AI on every level.” A quick skip across the internet suggests he could be absolutely bang on.
In my own case, I was drawn to a new band called The Velvet Sundown and in particular their new record Dust In The Wind. The Velvets are a new band out of America comprising of vocalist Gabe Farrow, guitarist Lennie West, bassist-synth artist Milo Rains, and percussionist Orion “Rio” Del Mar. They have so far released two albums this year, Dust and Silence and Floating on Echoes, and have generated over a million listens on music stealing streaming site Spotify. I’d describe their stuff as generic folk/soft rock (the blurb describes them as ‘indie/psych, whatever the fuck that means). Nothing particularly special but listenable, especially if your musical preference is generic folk/soft rock/psych rock. The only thing is that Messrs Farrow, West, Rains and Del Mar don’t exist. They are one of a legion of AI generated bands. Nothing about this AI-generated video is real. Do you see what Billy Corgan is on about yet?
AI is already here in a big way. If you use Facebook, you are feeding the beast of AI. If you have concerns about AI – and heaven knows, you should – then don’t whinge about it because you are part of the problem. The same goes with YouTube, X, TikTok, Instagram and WhatsApp. If you see something on your social media feed and wonder why on Earth that’s on there, the odds are it will be because of AI. I’d like to think that most users of these platforms would have been aware that AI was hard at work and not everything was as it seemed, but those who run them are often ruthless and cynical operators. The influence of AI has been at work in music for some time.
You might expect someone as vacuous and talentless as Kanye West to employ AI in his “music” but so did Paul McCartney when he helped create the “final” Beatles single Now And Then. Macca said that AI had been used to “extricate” Lennon’s voice from a cassette recording: “We were able to take John’s voice and get it pure through this AI“, adding, “To be clear, nothing has been artificially or synthetically created. It’s all real and we all play on it. We cleaned up some existing recordings.” But that’s different, isn’t it?
Well, it’s still AI but, I would argue, to good purpose. The music is real, played by actual musicians. The AI has been used by humans to enhance something, not as music creation. Indeed, isn’t it A Good Thing that we are able to listen to something we otherwise would have been able to listen to as a scratchy cassette tape version? I can only speak for myself and while Now And Then is not exactly Penny Lane, A Day In The Life or Yesterday, just hearing something “new” by the greatest group of all time has improved the quality of my life considerably.
I will go further. If some of my other favourite bands, like, for example, Steely Dan, R.E.M., The Doobie Brothers and The Beach Boys have music that is on old cassette tapes and is good enough to be “cleaned up” for release, frankly I’d be thrilled. But what I don’t want, and I agree with Corgan that we are going to get it, is made-up bands like The Velvet Sundown.
Listening to “their” music on YouTube, I find it thoroughly depressing. As Google’s AI overview puts it: AI creates music “by analyzing vast datasets of existing songs, learning musical patterns, and then using machine learning algorithms, often neural networks, to generate new compositions based on user prompts, desired styles, or specific musical structures. This process allows AI to produce original melodies, harmonies, and rhythms, acting as a tool to assist or fully generate music for various applications like background scores, theme tracks, or complete songs.” AI describing AI. What a world we live in. Frankly, a music world that comprises of music created by computers, with no human input is of no interest to me and threatens the very basis of everything we listen to and indeed see.
AI is part of the future, a very big part of it, and it will be both good and bad. For music, it sounds totally bad but you just know that, sooner or later, AI will start churning out music that we will find ourselves actually liking because in terms of technology we’re still in the dark ages. And who knows, in years to come live concerts will feature avatars and note-perfect AI enhanced or even produced music, just like the big ABBA show in London. (The music at that show is real and the avatars are created by actual people, but I am sure you will get the drift.)
My fear is that as people get used to AI music, it will become the norm and I am not sure how, or even if, we can stop it, or even control it. People are already more than happy to use the likes of Facebook, despite its extreme use of AI, so I suspect we know where this ends. Give me music by humans anyday, thank you, although I am prepared to accept music if AI is used as it was with Now And Then. I wouldn’t listen to The Velvet Sundown even if they were a real band but hearing their made-up dross I find my inner Luddite being revealed.
I always thought that the 1980s was the decade the music died, but if we allow it AI will finish the job. Is that what we really want?
