An A.I.-generated song featuring a “collaboration” between Drake and The Weeknd turned up on streaming services and TikTok last month, quickly racking up millions of listens. In response, Universal Music Group, which represents the musicians, issued an agitated statement, and then, as the song turned up on other platforms, embarked on a legal whack-a-mole campaign in a bid to put the A.I. cat back in a bag.
All of this came even as the legal status of the song remains unclear. Some music business veterans argue the song was functionally no different than a remix, while others decried it as a dangerous new paradigm. Rarely does a video made by a random handle create such uproar.
Then there’s Grimes, who may not be as famous as fellow Canadian Drake, but is a critically acclaimed musician in her own right. Grimes, who is not affiliated with any record label, announced she will use that relative freedom to perform a radical experiment called “Elf.Tech.” Using A.I., Elf.Tech allows users to generate a Grimes-style work with their own inputs. In return, all Grimes asks is for tasteful use of their now-open-source talent and a 50% cut on any master recording royalties.
Because the cost of producing high-quality music has dropped precipitously with advancements in A.I., the race to design a new distribution model is afoot. Currently, platforms must obtain permissions from both publishers and labels to play a song—in most instances, the label owns the sound recording while the publisher controls the underlying music and lyrics. In Grimes’s case, she could be the performer and also own the sound recording.
Tellingly, the first prompt on Elf.Tech asks users to connect to an email account or a cryptocurrency wallet. Presumably, the platform anticipates using blockchain and crypto to help enforce the suggested royalty split.
Unfortunately, as veterans of the NFT space know, while smart contracts have the potential to create a more efficient royalties system, the reality is that the process remains messy since NFT marketplaces circumvent the rules. Nonetheless, many maintain hope a cultural or technical innovation—likely tied to crypto—will emerge to help artists enjoy a fairer compensation system.
Though technology doesn’t exist to enforce it yet, my long-term bet is that Grimes’s approach will prevail over Universal’s. The reason is that the challenges of generative A.I. have striking parallels to those of another disruptive technology: open-source software.
All software began as “open source.” Software vendors would provide their clients with their source code, the human-readable version of the software that’s then transformed into machine code before it’s executed. Tinkerers and computer scientists would modify software, learn from it, and share ideas. As the field commercialized, source code became closely protected, like a trade secret. While companies made handsome profits exerting copyright over their code, allowing for additional investments, the downside was a less cooperative environment.
The open-source movement, a reaction to ever-tighter controls over software, promoted the release of source code under permissive licenses to foster the tinkering, sharing of ideas, and collaboration seen in the field’s early days. Today, open-source projects range from small hobbyist efforts to large-scale projects such as the Linux kernel, which is deployed in billions of devices around the world.
In similar fashion, traditional music publishers and labels thrived in a world where counterparties and IP were tightly controlled. Their moat was legal agreements that could hamstring pirates and, in some unfortunate cases, the artists themselves. As experimentation from pseudonymous creators—including those who deploy A.I.—shifts the balance of power from this model, artists will have to find new ways to engage with the increasing commoditization of their work.
Bitcoin, the first successful cryptocurrency, succeeded in large part because it was open-sourced. Alternatives to centralized institutions that issue credit require a radically transparent, queryable code base to stand apart. Early enthusiasts trusted the outcome of the Bitcoin ledger because they trusted the process, and because they could check it themselves. Today’s music pioneers would be smart to take note.
The next few years will see more explicit separation of designing versus controlling music production. To that end, Grimes’s experiment is a worthy exercise pushing the limits of a multibillion-dollar business and our burgeoning world of open-sourced talent.
Kathleen Breitman is a cofounder of Tezos. The opinions expressed in Fortune.com commentary pieces are solely the views of their authors and do not necessarily reflect the opinions and beliefs of Fortune.
