Image via the algorithm
As president Harley Geffner is back like he never left.
AI is getting alarmingly good. Would you ever guess this was AI-generated? pic.twitter.com/M2Fiw1qv7A
— Pigeons & Planes (@PigsAndPlans) March 20, 2023
Can you tell this isn’t Kanye? I can. It’s an eerie approximation of his Sunday Service music with a little bit of the slurry autotune of 808s. The voice is almost there, and I’m sure it will get even closer with each new iteration of an AI tool replicating itself (and sometimes even hiring humans) to become more powerful. But it still feels like it’s lacking soul, which is ironic since it’s approximating Kanye approximating gospel.
Not to be all doomsday, but there are some scary and easy conclusions to draw from these advances. First, the homogenization of popular music will continue to spiral to a place that’s already more out of control than it is with the advent of algorithmic playlisting. Second – the job of the producer will largely go by the wayside, and possibly the singer to follow. With enough data inputs from labels and other companies continuing to buy up catalogs, they will be able to rip an even larger slice of the profit pie from real artists with their in-house beat-making algorithm churning out whatever the most sonically pleasing 3-chord melodies are. The beats are much easier than the full songs, so the producer will go first. But eventually, who needs the artist when you can create fandoms around digital avatars? I discussed this very topic in a Rap -Up almost 3 years ago.
There will spawn a feedback loop between multiple algorithms – labels will tune the song-making A.I. algorithms to meet the preferences of the Spotify playlisting algorithms, then those playlists will lean further and further into those sounds. The consumer, us, will suffer the consequences as popular music becomes more and more like every American city’s main street, giving the appearance of character where there is no actual individuality. You’ll find a Blue Bottle Coffee on every other corner, which is cute and minimalistic without giving the appearance of being too corporate like seeing Starbucks everywhere. There is a Capital Bank Cafe (yes this is real and there is one in the Topanga mall that’s always packed). There is a fast casual [insert your favorite cuisine] that feels like a more expensive Chipotle, and ooh look a cute little boutique shop. But wait, it’s overpriced, and it feels the same as every other little Main Street boutique shop across the country when you walk in.
This A.I.-controlled future feels bleak when you look at it this way, mostly because A.I. has no taste! It figures out what is popular and replicates it with excruciating accuracy. When asked to rank the 10 best rap albums of the 2010s, ChatGPT spit out this list. No character. When I asked it to rank the most influential rap albums of the 2010s, thinking maybe it would pull something with more verve, it gave me an almost identical list, but subbed out Chance the Rapper for J. Cole.
There are still reasons to remain hopeful amidst this depressing morass of algorithmic control. Though A.I. is really good at knowing what we like, and feeding us more of it, it still hasn’t yet been able to predict new trends and innovations. A Rap Caviar sponsored A.I. beat-maker fed a bunch of Metro Boomin, Zaytoven, and Pop Smoke type beats would not be able to predict the success of a Milwaukee song with an elementary beat. It can see a Milwaukee artist break through and absorb that style into its amorphous creation brain, but in that sense, it’s always going to be playing catch up.
Rap music was innovated by working class and poor people without access to funding. Most every subsequent movement in rap, every push forward, has come from lower income artists with limited means to industry infrastructure. Even as cities and popular music and playlists coalesce around specific markers, there will always be an underground. Maybe they won’t be able to flourish as easily via communal gathering spaces, but they will exist and there will still be forward-thinking people breaking new ground, even if the experiences around it become more insular.
There will be people who aren’t plugged into the algorithm because paying 10 dollars a month for Spotify is less important than the meal that money might provide for their kids. These are the people through whom innovation will continue. People insulated in their regional bubbles, messing around on pro tools to try to come up with something that makes them laugh, something different that subverts expectations. They may not always get the credit as the algorithms become faster and faster at absorbing them into the mainstream (see the Kelces talking about “the Lil Uzi Dance” that really should be credited to the Philly Goats) but there still will be forward pushes. The boring popular music of today was once the interesting underground music of yesterday.
There’s insane music being made all over the globe from West Bengali bass competitions (h/t Tayler) to meditative ASMR leaning into alt-metal, and my favorite, the weird and chaotic Milwaukee rap scene referenced earlier. Though it’s depressing to look at the charts and popular playlists, there ARE people with an appetite for something different. The fact that this site still exists is proof of concept. It’s on us, the type of people to write for or read POW, to spread the gospel. We are the likely 1% of people who are actively searching for new and interesting sounds, whereas (no shade) most of the general public doesn’t have time and acquires new music through it assimilating into their brain from the radio, popular playlists, or other means of public consumption.
We can be that other means of consumption – but it means we need to keep sharing cool music with anyone who will listen. It doesn’t need to be in a soapboxy way like I do it, but it really is on us to push back against the sameification of popular music and taste. We are the antidote to this, so instead of getting depressed about it, go find a new rabbithole to burrow into.
Emptying the Chamber