AI News

Musicians Must Specialize as AI Transcends Generalist Tasks and Enters Commercial Sphere

Credit: Elevenlabs

Key Points

  • The rise of commercially licensed AI music moves the technology from a legal gray area toward a potential professional tool.
  • Devin Arne, a composer and Assistant Professor of Studio Composition at West Chester University, argues that to stay competitive, musicians should specialize and hone their unique voice, as AI will be a generalist.
  • He distinguishes between AI for "utility" music and human-made art, which fosters deep emotional connections that AI struggles to replicate.
  • While a moral victory, licensed AI models are unlikely to be a financial windfall for artists, comparing the potential earnings to less than Spotify royalties.
Devin Arne - Assistant Professor of Studio Composition | West Chester University
When I was coming up, it was good to be a jack-of-all-trades. AI tools are the generalists now. You need to find your own lane as a musician and go all-in. Being your own unique self is the best path forward.Devin Arne - Assistant Professor of Studio Composition | West Chester University

Until now, AI music generators' legally murky "train first, fight later" model has prevented the use of AI music in mainstream commercial use. But the launch of a new AI music generator from ElevenLabs, built on licensed data partnerships, is changing that. The move toward legitimacy and commercial viability stands in stark contrast to other music makers who allegedly train their models on copyrighted music without permission, and for professional composers, it marks a turning point. A new wave of licensed AI affirms the technology as a professional's tool and opens doors to opportunities for AI-generated tracks in high-stakes commercial projects in film, television, or advertising.

To understand this new reality, we spoke with Devin Arne, an Assistant Professor of Studio Composition at West Chester University of Pennsylvania. As an academic and a working musician, producer, and composer who's been tracking the technology's evolution for years, Arne brings a unique perspective to this conversation—and his music.

  • Uniqueness as a shield: As AI expands into licensed music, there's an opportunity for musicians to make their marks as specialists. "When I was coming up, it was good to be a jack-of-all-trades," Arne explained. "AI tools are the generalists now. You need to find your own lane as a musician and go all-in. Being your own unique self is the best path forward."

  • A moral victory, not a financial one: While ElevenLabs' opt-in, licensed model is a significant ethical step forward for artist consent, Arne cautioned that it's unlikely to become a financial game-changer for most musicians. "This likely still won't become a huge income source for folks," he said. "Similar to Spotify, royalties will likely be in the hundreds of dollars range, not the thousands of dollars. It ultimately won't change an artist's financial situation, but it will bring in a little extra cash."

    Devin Arne - Assistant Professor of Studio Composition | West Chester University
    If the main corpus of the song is AI-generated, I think that's where you should have to label it. And there are algorithms from folks at IRCAM in Paris that are extremely accurate in recognizing AI-generated music.Devin Arne - Assistant Professor of Studio Composition | West Chester University

    For the listener, the rise of commercially safe AI music presents a new dynamic. Arne drew a clear divide between music as utility and music as art. For passive consumption like a study playlist or background music for a YouTube video, AI-generated music is perfectly fine. But for the deeper, more meaningful engagement that defines fandom, it misses the mark by failing to deliver the authentic human experience at the heart of music.

  • Utility vs. connection: "Everyone has songs from ten or fifteen years ago that still spark a strong, intense nostalgia," Arne said. "AI music will probably never prompt that kind of connection." Human experience is something AI can only guess at. It's still a human-to-human relatability that allows music to foster that resonance with listeners. The songs that last tap into that connection to become classics.

  • Transparent consumption: As AI-generated tracks begin to populate streaming platforms alongside human-made music, the question of transparency becomes critical. Arne advocated for clear labeling, but with a nuanced distinction that separates AI as an author from AI as a simple tool in an artist's process. His proposed standard is not just a philosophical one, but a technically enforceable one. "If the main corpus of the song is AI-generated, I think that's where you should have to label it," he proposed. "And there are algorithms from folks at IRCAM in Paris that are extremely accurate in recognizing AI-generated music."

Despite the challenges, Arne expressed optimism about AI's potential to serve as a powerful creative partner. He described its greatest value not in replacing artists, but in augmenting their process. Helping them break creative blocks, explore new directions, and ultimately, finish more music. He pointed to specific applications, like AI-powered stem separation, that his students and colleagues are already using to push their creative boundaries.

"Stem splitting is such a cool aspect of AI in music. It's really changing remixing. It's exciting to see all of the creative ways that my students are applying it," Arne said. "There are really cool things for composers across the spectrum, from your bedroom producer to your classical composer."