There are plenty of GenAI-powered music editing and creation tools out there, but Adobe wants to put its own spin on the concept.
Today at the Hot Pod Summit in Brooklyn, Adobe unveiled Project Music GenAI Control, a platform that can generate audio from text descriptions (e.g. “happy dance,” “sad jazz”) or a reference melody and let users customize the results within the same workflow.
Using Project Music GenAI Control, users can adjust things like tempo, intensity, repeating patterns and structure. Or they can take a track and extend it to an arbitrary length, remixing music or creating an endless loop.
Developed in collaboration with researchers at the University of California and Carnegie Mellon, Project Music GenAI Control may be made publicly available at some future date. But for now, it’s firmly in the research stage, said Adobe’s head of audio and video AI research Gautham Mysore during a panel at Hot Pod — the platform doesn’t even have a UI yet.
“So this really gets the idea of, AI is generating music with you in the director’s seat and there’s a bunch of things you can do with it,” Mysore added. “[The tool is] generating music, but it’s [also] giving you these various forms of control so you can try things out. You don’t have to be a composer, but you can get your musical ideas out there.”
GenAI music tools — and GenAI tools broadly — are raising ethical and legal concerns as AI-created music, artwork and text proliferates.
Homemade tracks that use GenAI to conjure familiar sounds, lyrics and vocals that can be passed off as authentic, or at least close enough, have been going viral. Music labels have been quick to issue takedowns, citing copyright. But there’s still a lack of clarity on whether “deepfake” music violates the IP of artists, labels and other rights holders — particularly in the case of GenAI music tools trained on copyrighted content.
A federal judge ruled in August that AI-generated art can’t be copyrighted. However, the U.S. Copyright Office hasn’t taken an exceptionally firm stance yet, only recently beginning to seek public input on copyright issues as they relate to AI. Also unclear is whether users could find themselves on the hook for violating copyright law if they try to commercialize music generated in the style of another artist.
Mysore said that Adobe, as a general rule, develops its GenAI tools against data that’s under license or in the public domain to avoid potentially running afoul of IP issues. (Mum’s the word on whether that’ll be the case with Project Music GenAI Control, though.) He added that Adobe’s working on watermarking technology to help identify audio produced by Project Music GenAI Control, but admitted that it’s a work in progress.
“Adobe takes a particularly responsible approach to [these things,]” Mysore added. “There’s a lot of really great musicians making this content … I think [they and tools like Project Music GenAI Control] are going to coexist. There’s going to be new musical ideas that come out.”
