I was talking with a friend recently, someone in the know, an old hand in media and editorial. They saw the future of music services in the future like this: “You’ll just describe a mood or a vibe and the robot will just serve you up music exactly for that situation,” they suggested. It wouldn’t be an artist you knew, but it would be perfect for the moment. Designed to go along with what you are feeling at the time.
I thought: “On what fucking planet.”
And then the news about the Velvet Sundown hit.
In case you’ve been hiding under a rock, let me catch you up. There’s been a spate of reports about a few AI acts becoming viral hits on music listening platforms in the past couple of weeks. Perhaps the biggest is the Velvet Sundown, a band with now over a million monthly listeners on Spotify (it was around 350k before the news broke). The band’s tunes had all the signs that it was created on one of the AI music apps like Suno and Unio. Middle of the road, faintly Laurel Canyon ‘70s mellow country-tinged rock. Indistinguishable artist pics. And a lot of cuts with dust in the titles, for some strange reason.
The reaction to this band and others was unsurprising. Many were pointing out the mechanics of how the act got its listeners, by including the band’s entire catalog on classic rock playlists—a classic scammer method. Many wondered if it was fair for a fake band to take away market share and therefore money from actual real live bands. Several wondered if we were going to get flooded with an overwhelming amount of AI content from others, making these services a cesspool of fake content, somewhat how Pinterest has become nearly unusable by the amount of AI slop making its way on the platform.
Some thought that maybe this is the tipping point that my friend suggested, where soon, we won’t care a whit about what’s playing, just as long as it fits the mood. But I find it really hard to believe that music fans are going to warm up to AI acts anytime soon.
Context Matters
Of course even hardcore fans of artists will end up using AI music, because it will be everywhere. Throughout the day, we are surrounded by music, from the elevator, to car commercials, to the gym, to buying a ticket and going to a show. I can see some AI music seeping into some of these specific contexts, if not completely dominating some of those products.
There are certain circumstances where AI may have a role. Like background music for studying or plowing through work, or to wind down and fall asleep. Generally music that goes with other things. Instead of licensing music or hiring a professional creator to generate music, I can see producers turning to AI instead. Of course it will have an impact on those artists. In the past the car commercial producer would hire a composer who could create a sound that is so close to John Williams or Bob Dylan or Bon Iver, without having to license those songs. For better or for worse, those composers are now going to be competing with (if not totally replaced by) AI programs like Suno.
Of course there is a massive number of people who could care less about musicians and the act of music making and they will be quite happy with the output of AI music, thank you very much.
This is not the case with music fans.
But let’s not confuse low-priority music needs that AI will fulfill with what music fans want. Sure the music is a huge part of what drives people’s love, but just as importantly, it’s who made it and how that person connects with their audience. Fans tell us that they want to know more about their favorite musicians. What inspires them and why they make music. On top of that is the way fans connect with artists live, which to point out the obvious, isn’t something AI music can deliver.
That’s not changing. Not today. Not tomorrow. Not ever.
Will other AI acts be able to replicate the gimmicks of Velvet Sundown? Without a doubt. But to me this speaks more about the mechanics of going viral on platforms these days as much as it does about AI taking over. Over the past couple years, a few early-stage indie country acts saw their Spotify stream numbers explode for a month or two. After investigating, it became clear that a viral video on TikTok used the band’s material to accompany a video, leading to a spike in numbers and nice payouts for a couple months. But the discovery was more about the video than it was the artist; and generally the band’s numbers came back down to where they were previously. It’s almost like music served a specific purpose for the video producer. Is that the way to really get new fans?
Platforms are having and will have a much more difficult issue to deal with from AI, just like Pinterest is facing. By its nature, AI tends to flood the zone with content. There may be so much new and copycat AI content on a platform like Spotify that it makes the service unusable. Emily White has been writing that the streaming era is ending and we might be on the verge of something new. You could see a flood of AI content start to make people flee from streaming services.
Out of the Dust
But flee to what? There has been a whisper of a trend recently of fans quitting streaming because they don’t think the big streamers are fair to artists. And of course there are a bevy of startups offering more equitable payouts to artists from streams, however none have seen much traction.
Any new platform is going to need to focus on the needs of music fans first and foremost. As I’ve written about before, today’s platforms take real music fans for granted. Most of them are designed for mass consumption and not deeper experiences. If you’re interested in discussing a framework for building the next generation of music services, let’s talk.
Great summary and insight Jon. No Bot Radio 🤖