An artist has traditionally been easy to recognize through a combination of voice, face and sound. A singer develops a vocal identity, an instrumental palette becomes familiar, and listeners gradually learn what to expect.
VEXSAN complicates that idea.
The project can move between French and English, switch from heavy rap to more melodic music, and use different generated voices without treating each change as a new artist. For VEXSAN, the continuity sits somewhere deeper.
“The styles can change dramatically, but the writing remains the backbone of VEXSAN.”
That raises an increasingly relevant question for AI music. If the voice, genre and production style can all change, what actually makes an artist recognizable?
Identity can begin with the writing
VEXSAN places writing at the center of the project.
He describes his lyrics as influenced by 1990s rap, particularly its directness. He prefers concrete images and lines that connect quickly rather than making a lyric complicated for its own sake. Rhythm also begins before production, through phonetic rhymes, internal sounds and the way words fall against a beat.
“For me, a text should already have music inside it before the production even begins.”
The production can therefore change substantially without removing the underlying perspective.
This offers one possible model for artist identity in generative music. The recognizable element does not necessarily have to be a fixed synthetic singer. It can be the way situations are observed, written about and turned into songs.
More voices create more reasons to reject them
Generative tools make it easy for VEXSAN to test those ideas through very different voices and productions.
That freedom does not mean the first technically successful result becomes the finished song. VEXSAN says he has completed songs for which he has already tested around twenty versions because none of the generated voices carried the intended emotion.
“AI opens an enormous number of doors, but the real work is knowing which ones not to walk through.”
A voice can sound good and still be rejected. A production can be impressive and still overwhelm the lyrics. For VEXSAN, the deciding question is whether the result serves the original intention.
This connects with a pattern running through several of our recent interviews.
In AI Music and the New Burden of Choice, Vitalii Klimenko described how easier generation creates more possibilities while increasing the importance of judgment. Brian Myk similarly discussed selection in What Skills Should Artists Value in the Age of AI Music?.
VEXSAN adds artist identity to that problem. The creator is not only choosing which generation sounds best. He is deciding which version still sounds like something VEXSAN would make.
A song can travel without becoming a different project
Language creates another test of that identity.
VEXSAN originally wrote Debout pour eux in French before developing the English version Standing for Them. His goal was not to produce a direct translation. He wanted an English-speaking listener to experience the same emotional idea.
That required starting again several times and testing different interpretations. VEXSAN says he did not consider the song finished until the English version created the emotional response he was looking for.
“Technology can generate twenty different voices. Finding the one that truly makes the words come alive is still a human decision.”
Generative music makes this kind of adaptation easier to attempt. A creator can test different voices, phrasing and musical treatments without rebuilding the entire recording process each time.
But VEXSAN's workflow also shows why translation alone is not enough. Words can be converted between languages automatically. Preserving intention requires another layer of judgment.
The goal is not to make the French project sound American. It is to let the same idea travel while keeping the perspective behind it intact.
The artist can extend beyond the voice
VEXSAN also carries that identity into the visual side of the project.
Several of his comic-inspired videos feature a recurring character wearing a death mask. The character appears across different stories, sometimes participating directly and sometimes observing what happens around him.
The mask is not only a visual signature. VEXSAN connects it to the project's direct writing style.
“The mask hides his face, but it never hides what he has to say.”
That recurring character gives VEXSAN another form of continuity when the musical surface changes.
This is particularly relevant for AI-assisted projects because generative tools increasingly allow one creator to work across music, images and video. Artist identity can therefore be distributed across several elements at once.
A voice can change while the writing remains. A genre can change while the visual language remains. A song can move into another language while the perspective behind it stays recognizable.
AI artists may need stronger creative direction
VEXSAN describes himself as directing this wider universe rather than simply generating its individual components.
AI allows him to develop songs, characters, visuals and videos without requiring a separate team for every part of the process. But more available tools also create more decisions.
“AI provides the tools. The coherence, the choices, and the artistic direction remain human.”
That idea also connects with Story Behind AI Music Artists You Might Have Missed, where we explored how AI music projects can build identity around fictional performers and creative concepts rather than relying only on a conventional performer.
VEXSAN approaches the question from another direction. The project does not depend on maintaining one fixed generated persona. Its identity is meant to survive changes in voice, language and musical style.
This suggests that AI artist identity may become less about preserving one surface characteristic and more about preserving a recognizable set of decisions.
What remains when everything else becomes easy to change
Generative tools are making many parts of music production more flexible.
Voices can be replaced. Genres can be tested quickly. Visual styles can change between projects. A creator can explore twenty versions of an idea before deciding which one to keep.
As those possibilities expand, the ability to generate them may become less useful as a way of defining an artist.
VEXSAN believes audiences will instead connect with what the creator repeatedly chooses to say, how they say it and the world that forms around those choices.
“A song can completely change musical style or voice, but if the writing, the intention, and the universe remain recognizable, you can still feel the same person behind it.”
That may be one of the more interesting changes introduced by generative music.
An artist no longer necessarily needs one voice to remain recognizable. Identity can sit in the decisions underneath the voice.
VEXSAN summarizes the idea more simply.
“AI will be able to produce more and more content. But having something to say will probably remain much rarer than knowing how to generate something.”
If generation becomes increasingly abundant, recognition may depend less on which tools were used and more on whether listeners can still recognize the perspective behind them.
Explore VEXSAN
You can follow VEXSAN on YouTube and explore the project across its French and international releases.
For an introduction to the English-language side of the project, listen to Standing for Them. You can also hear another side of VEXSAN through Delete My Number.
