AI Music: What’s coming & Why it isn’t happening yet [Bill Werde]

Music is on the verge of an AI-driven revolution, writes Bill Werde. But concerns about creative control, copyrights, and much more are holding it back… for now.

By Bill Werde

A version of this essay first appeared in his free, weekly Full Rate No Cap email. Werde is a former Billboard Editorial Director and Director of The Bandier Program for Music and the Entertainment Industries of the Newhouse School of Public Communications at Syracuse University. Taylor Swift played the Billboard cafeteria for Werde and a small handful of other editors when she was 16 years old and just starting to break.

AI Music: What’s Coming and Why It Isn’t Happening Yet  

AI will lower the barrier of entry for music creation, both in terms of who can create music, and how easily established musicians can create. But music offers specific tech and licensing challenges that other mediums do not. 

In the context of “we read everything so you don’t have to,” I thought these LinkedIn posts from two different leading thinkers in the AI music space combined to make one great “state of things” in the sector.

The first is from Mansoor Rahimat Khan, CEO and co-founder of Beatoven.Ai. Beatoven is a promising entrant in the field of generative music; I created an account and generated a new, passably-good production music-sounding track based on my direction (tempo, instruments, style, etc) in the space of roughly three minutes.

And the second is from Cherie Hu, whose Water & Music collective of music technologists and industry types is literally redefining the notion and value of an industry trade, while simultaneously leading creation in the very tech its writing about.

Khan breaks down the impact he expects from AI in the music space, while Hu explains why none of it is happening quite as quickly as we might like.

Here’s a left-field idea:

The major labels could expedite solutions for some of the challenges that Hu identifies (meaningful parts of her 1, 2, and probably 3) by deciding to cooperatively license/own the AI Music space. VEVO for AI anyone? Or at least a one-stop B2B shop to license the necessary data for the purposes of AI development, in exchange for equity?

Back in the early days of streaming, the majors were great at suing perceived threats, but very slow to offer viable, legal alternatives. Hopefully, they move more quickly with the current wave of AI/web3 tech possibilities. Certainly, the evidence is that the majors consider anyone training AI with major label music is a no-no for now.

I keep thinking about the interview outgoing Warner CEO Steve Cooper gave at a Goldman Sachs conference in September, where he actually seemed to brag that “what we’ve done over the last number of years is reduce our [financial] dependency on superstars.”

I have to imagine that major labels and their key shareholders would love to edge towards a world with less dependence on those pesky, you know, ARTISTS, and total control/ownership of music.

Also

The courts keep saying AI can’t generate copyright. This will lead to some fascinating grey areas, and yet-to-be-answered ethical questions. Somewhere in between the dudes who used AI to generate 68 billion melodies, and artists who will use AI-generated stems and sounds to then create human-edited/composed works is a line of authorship and ownership that needs to be drawn. Would love to publish some thoughts from readers about how and where to qualify these choices. 

MORE: How to design free cover, poster art using AI [Artificial Intelligence]