The Grimes AI project has released in beta, enabling users to make vocals sound as though they’ve been recorded by the Canadian artist. Photo Credit: Jordan Uhl
As a growing number of artificial intelligence tracks – including more than a few unauthorized works – continue to release, Grimes is officially pushing forward with an AI project through which others can create music with her voice and pocket a piece of the resulting royalties.
Vancouver-born Grimes (real name Claire Boucher) just recently rolled out a beta for her artificial intelligence concept, which she first outlined on social media last month. Of course, it’s hardly a secret that an avalanche of AI music, from a multitude of extremely similar “songs” to an array of works made to sound like Drake tracks, has debuted as of late.
While absent from Spotify and other leading streaming services, the Drake-soundalike AI creations – “Winter’s Cold” (1.1 million YouTube views on just one video), “Talk To Me Nice” (129,000 YouTube views since being uploaded two days ago), and the much-publicized “Heart on My Sleeve” among them – have racked up millions of views on YouTube and social media. Additionally, some music companies have already confirmed that they’re taking steps to license AI songs.
It’s against this backdrop that Grimes has directed fans and creators to a web platform called Elf.tech, with which interested individuals can use artificial intelligence to make vocals sound as though they’ve been crafted by the “Genesis” artist herself.
35-year-old Grimes in a series of tweets described the basics of the offering and the royalty-distribution plans for the songs at hand, relaying in part: “we ask for 50% splits on master recording royalties in exchange for a grimes feat and distribution. There’s a *small* chance we can organize getting you publishing $ as well but we can’t guarantee this yet.”
Meanwhile, Elf.tech itself requires little more than a click to transform vocals (pre-recorded or recorded directly in the browser) “into a GrimesAI-1 voiceprint,” the platform shows. “Users will receive the transformed Grimes AI – 1 voiceprint back in a WAV file,” said platform’s text explains. “They can incorporate these GrimesAI-1 vocals in the creation of new sound recordings.”
Elf.tech can likewise handle distribution (for $9.99 yearly, per the website), and “GrimesAI-1 does not claim any ownership,” according to the fine print. However, those who utilize the AI tool “agree to share an equal % of the master recording royalties distributed on streaming services, onchain marketplaces (primary and secondary sales m) and other DSPs with GrimesAI-1.”
Bearing in mind this split, label-signed acts and/or persons who wish to self-distribute must receive permission beforehand, Grimes emphasized. The Miss Anthropocene creator encouraged users to forward related requests to an email address tied to CreateSafe, a self-described “web3 music studio developing the sovereign smart contract protocol for music IP metadata.”
Users can upload completed tracks via a seemingly straightforward “submit” section, identifying the song title, artist name, producer(s), writer(s), and other pertinent details in the same form. Grimes has also made nearly 20 gigabytes’ worth of stems available through Elf.tech and touched upon plans to drop music of her own in the near future.
Needless to say, it’ll be interesting to monitor the Grimes AI platform moving forward – if only because of the massive amount of music that it seems poised to produce. But as mentioned, ample takedown notices and firmly worded public statements have thus far failed to stem the tide of artificial intelligence releases in any event.