For DNØR creator Doniell Clay, generative AI has changed more than how music can be produced. It has also made it possible to build visual worlds around Deni$e, Noah and Reg without the budget normally associated with professional music video production.
“AI didn't give me the dream. I already had that. It gave me a way to make the dream possible,” Clay says. For him, the technology has lowered the financial barrier to turning songs and characters that previously existed mostly in his imagination into finished visual stories.
His workflow is also more deliberate than simply generating clips and arranging them afterward. Clay begins with the song and the character who should express it, then uses the lyrics and personality to decide the setting, clothing, body language and atmosphere. In Deni$e's case, for example, her emotional identity shapes how relationship-focused songs are presented visually.
Only after that does the generation process begin. Clay describes repeatedly adjusting scenes when the face is right but the movement is wrong, or when a strong camera angle produces an inconsistent character. He then selects the usable clips himself, edits them to the music and checks elements such as pacing and lip-sync.
That process points to a wider change in music economics. YouTube says more than two billion logged-in users watch music videos each month, while the platform paid more than $8 billion to the music industry between July 2024 and June 2025 through advertising and subscriptions. Spotify is also moving further into video, and reported in May 2026 that more than two thirds of Premium subscribers had watched music videos.
AI could dramatically increase the supply feeding that demand. A video that once required performers, locations, cameras, lighting and substantial post-production can increasingly be produced by a much smaller team or even one creator.
For DNØR, that does not mean visual identity becomes disposable. Clay still keeps the original images of Deni$e, Noah and Reg as references, while deliberately allowing their appearances to evolve. Noah has become more mature in appearance, Deni$e has changed hairstyles, and Reg's bandana has become part of his visual identity.
That difference between accidental inconsistency and deliberate evolution could become increasingly important. If generating an attractive performer becomes easy, maintaining a recognizable character across songs, videos and years of releases may become much harder, and therefore more valuable.
The same applies to production quality itself. Clay argues that professional-looking AI video still depends on storytelling, pacing, editing, movement and knowing how to communicate a visual idea to the model. Access to the tool does not guarantee a coherent result.
The consequence may therefore not be that music videos simply lose their value. Instead, the source of that value could change.
If almost every independent release can eventually have cinematic visuals, an expensive-looking video will no longer signal the same level of financial investment or industry backing. Attention, recognizable identity and creative direction may become scarcer than the images themselves.
This continues a question we previously explored in The Story Behind AI Music Artists You Might Have Missed, where DNØR showed how fictional performers can still carry the writing, ideas and perspective of a human creator.
The future is also unlikely to divide neatly between traditional music videos and AI-generated ones. Grimes' recent Accelerando video used AI-generated elements, but Business Insider reported that the production still involved more than 70 people, including performers, choreographers, makeup artists and production staff.
AI was therefore one layer of a much larger production rather than a replacement for it. The emerging model may be a spectrum, ranging from independent creators producing entire videos with generative tools to larger productions using AI selectively alongside conventional crews.
For music videos, the shift may ultimately resemble what is already happening to music itself. AI lowers the cost of producing another option, which means far more creators can participate.
What it does not lower is the difficulty of building a visual identity people recognize and creating something they actually want to watch.
Watch and listen to DNØR
Readers can explore the DNØR project directly through the videos discussed in this article. Deni$e appears in What Happened to Love, while Noah's music video explores a different side of the project's visual storytelling. Reg's video addresses the debate around AI music from the perspective of the character himself.
More of DNØR's music is available through the project's official SoundCloud.
