🔵 Correctly predicting Spotify's AI move

(Not that I can prove it now as Music Business Worldwide beat me to it 😆)

This morning I woke with a clear sense of what I was going to write about in Network Notes. Much has been made of this deal between both major and indie labels and Spotify to develop AI experiences with artist’s music. I’ve seen a few people speculating, including some great (and blunt) thoughts from Midia’s Mark Mulligan, over on LinkedIn.

My plan for today’s piece was to write my own prediction of what would come - specifically, a system that would allow users to mash up one track with another to create what used to be called ‘bootleg’ mixes back in the day (which, I might add, I also contributed to with my own effort back in 2004!).

I was feeling rather pleased about this all… until I checked the news and saw that those clever people over at Music Business Worldwide had gone digging on the US patent office website and, well, proved me 100% correct. Obviously, my claim to have predicted this now rings entirely hollow, so I’ll forgive anyone raising an eyebrow and saying “yeahhhh, sure you did” because I have zero means to prove otherwise, LOL.

Either way, I do feel this will be the tip of the spear of this much-hyped AI deal between the majors and Spotify. When you think about it, it makes a huge amount of sense: mixing two tracks together has been a staple of music culture since the 70s, and has yielded many a wonderful outcome.

Equally, this method allows Spotify to track usage and ensure that all artists and rightholders involved are paid accordingly. Mashing up Eminem and that Thomas The Tank Engine theme tune (er, again)? No problem: everyone involved gets their share of the money, something I’m not even sure happens on YouTube at present.

Quite how that money is figured out is the part I was less clear on. If I recreated my own bootleg mix, for example, can I ‘release’ it - which I would take to really mean, ‘share as some kind of link back to Spotify’? If so, does this have the potential of eventually seeing a proper release, like that Kylie/New Order mashup, or the Sugababes/Gary Numan one? Would I see any credit as the person who originated it if it was officially released? (I’ll put £10 on the table now and confidently state that I wouldn’t.)

When you think about it objectively, it looks like a pretty smart play by all involved. This isn’t AI in the Suno sense, and everyone gets paid. Spotify and the majors get to talk up how they love artists and support driving revenue for them, and the whole AI hype moves (perhaps) into a more sensible realm, where it isn’t attempting to replace all known music, but instead just provides another way to mess with what’s already there.

In principle then, most would call this a positive step. Personally, I think it might be a novelty more than anything; ultimately I still feel far more music fans would prefer to just listen to whats there, rather than mess with it all on some manner - though I’ve no doubt a legion of speeder-uppers and slower-downers over on TikTok might beg to differ.

Ultimately, the whole thing elicits something of a bored shrug from me. It all plays to this continued notion that rather than evolve things further, we should just work to juice more out of what is there. Calling this a new development is a little ridiculous; it’s a very old one just getting made easier thanks to AI. Furthermore thanks to that reduced barrier to entry, this stands every chance of resembling Sora’s insane AI slop frenzy of a feed, spewing out crap at a rate nobody can ever handle, losing the gems amid the deluge.

Crucially though, it is also failing to expand the revenue pie. After all, monies paid out for listens on this just eat into what was already being paid out to Taylor Swift artists. There’s no growth, no expansion: just more ways to split the pie up.

I’ve no doubt share prices for Spotify, Universal and more might jump at this terrific development. That’s how easy Wall St is to please, it seems; just look at how they collectively foamed at the mouth when Spotify signed Joe Rogan for more on that front.

Narratively then, this is a great victory for all involved, and senior people will be doing laps of honour talking up how they made this happen.

In reality though, it just conforms to the ongoing stagnation situation we find ourselves in. Music is not evolving. The experience around music is not evolving. Furthermore, in the face of AI, music is growing exponentially in presence, further tipping the signal/noise ratio massively towards noise, which drowns out the undoubtedly great music that is getting released. One might even argue that putting even more focus on playing with what’s there only reduces the chances of new talent truly breaking through. Again: this simply fragments the pie further, it certainly won’t be growing it. Quite the opposite: it stands every chance of further contributing to Spotify just becoming the equivalent of a skip full of refuse, driving consumers to places that put more emphasis on curated selection of a limited set of tracks, rather than everything anyone has chosen to push out there.

So yes, we can hand it to all involved: they found a way to use AI to do something that might actually create further revenue for some artists. However it simply takes from other artists, redistributing income a little, but not drastically changing anything, whilst further flooding the plain with (for the most part) super low value works.

A victory for rights holders then, maybe, but not a victory for much else. In chasing the revenue, we are arguably further devaluing the very product at the core here, and that cannot be a good thing.

Have a great day,

D.

🎶 Listening to “Discern/Define” by the Poets of Rhythm. I found this in a charity shop for £1, and grabbed it instantly, remembering as I do the incredible track “Guiding Resolution” from this LP. The whole album is a treat; something fans of the Daptone label would instantly take to. Special mention must go to the drummer, whose name I sadly do not know, but who throws down some of the hardest funk breaks I’ve heard in a long while. Stellar stuff.

📖 Reading “All Gates Open: The Story of Can” by Rob Young and Irmin Schmidt. Somehow I was unaware of this book - which is actually two books, one each by the aforementioned authors. The first part/book is a biography of the mighty German band, whilst the second part/book chats with various artists about the band and its influence, among other things. Can’s story is a fascinating one for the modern age: a group expressly focused on improvisation and drawn-out jams that they would edit together, not unlike Teo Macero with Miles Davis. But in an age of immediacy and short-term attention spans, this is precisely one kind of antidote. The book itself is full of wondrous insight; I shall be genuinely sad when I reach the end.

📺 Watching “Why I’m not making any new music” by Stephen McLeod on YouTube. I think anyone making music has hit that rut where they’re not feeling creative, and they then beat themselves up about that. This video should be bookmarked for any such occasion, articulating perfectly why it’s totally OK to be like that. McLeod’s reference to “not always being on transmit, but sometimes enjoy just being in receive mode” really summed it up for me. Wonderful.

---