UpdatedSat, October 31, 2026 at 3:37 PM UTC
Artificial intelligence can now create stunning artwork in seconds, producing images that once required years of training and practice.
But something important may be lost when the creative process becomes almost effortless.
A new paper argues that the greatest value of making art has never been the finished product alone. It is how the creative process changes the person behind it.
As AI increasingly separates artistic achievement from human experience, one researcher is raising questions about what creativity should mean in a world where machines can produce remarkable results.
Amit Raphael Zoran is a computer scientist at the Hebrew University of Jerusalem, where he builds craft and fabrication tools in its School of Computer Science and Engineering.
His paper presents a philosophical argument rather than experimental findings, so its conclusions remain open to debate.
“For centuries, creating art meant going through a process of learning, experimenting, and changing as a person,” Zoran said.
How art changes us
Zoran’s answer to what human creativity is for starts by splitting what art does into two jobs.
One is the artwork itself: the painting, the song, the <a href="https://myappsplus.com/how-building-a-cheap-home-server-could-save-you-money-in-the-long-run/” title=”How building a cheap home server could save you money in the long run”>building that carries a meaning from one person to the next.
The other is what he calls artistic becoming, the change that making art, or meeting it, brings about in a person.
In his account, the change happens at three levels. A painter’s eyes and hands are trained over years. A viewer walks out of a gallery seeing differently.
And whole cultures shift: Renaissance perspective reorganized how people saw space, and Cubism unsettled assumptions about objects and viewpoint.
Zoran doesn’t claim the split itself is new. The philosopher John Dewey separated the art product from art as experience in the 1930s.
What Zoran adds is a claim about history: for most of it, the two stayed tied together only because making a sophisticated work required the maker’s years of change.
AI changes the process
Earlier computer art kept the tie. Harold Cohen’s drawing program AARON and the interactive installations that followed it supplied precision and repetition, while people supplied interpretation and meaning.
Generative models work on a different material. They are trained on the recorded traces of human culture, what Zoran calls an unevenly weighted and culturally skewed slice of the human record.
From that material, they produce images and music with qualities that once took years of training. The years are no longer required.
Zoran argues that AI has separated the quality of a creative work from the years of human experience traditionally required to produce it.
Nor does the maker’s change disappear, in his telling; it moves. Prompting, curating, and iterating with a model are real skills.
But they work on symbols instead of clay or wood, and many of them can be picked up without a long apprenticeship.
“Today, AI allows us to produce sophisticated works without necessarily undergoing that journey. But this doesn’t mean human creativity has lost its value. It means we need to rethink where that value lies,” Zoran said.
Judging the tools differently
Image and music generators are usually judged on output: realism, coherence, fluency, how closely a song fits its genre.
Zoran wants a second set of questions. Does a tool widen what a person explores or narrow it? Does it push users toward familiar forms or expose them to unfamiliar ones?
Zoran acknowledges that measuring how creativity changes a person is difficult. A new way of seeing leaves no trace in a model’s outputs, and asking people is a weak instrument.
So he proposes measuring conditions instead. An engineer can count how broad a range a system leads users to explore, and how often they explore instead of merely retrieving.
A tool’s defaults can be checked for how far they drift toward the statistically familiar. Recommendation feeds have already shown how easily software narrows what people encounter.
Experiencing art differently
Zoran describes three projects from his own creative work to illustrate how technology might expand human experience.
However, the projects do not establish whether these approaches actually produce lasting changes.
In the first, a live lecture on perception turns out to be partly prerecorded.
Several versions of the speaker start talking to one another, so the audience feels how easily confidence in a digital presence can be manipulated.
For the second, participants taste an essence reconstructed from an old oak. It was put together from connections among historical alchemy, botany, and chemistry that would have been hard to assemble without computers.
The third puts a tree at the center of a multisensory environment. Visitors smell and taste an essence from the tree and listen to sounds recorded from it.
Generative AI turns those sounds into music that a human musician performs acoustically, while projections show structures a person can’t normally see.
No one measured what any of this did to the people in the room, and Zoran wrote as much.
Zoran names a limit of his own. His examples and the thinkers he leans on are all Western.
East Asian calligraphy has long valued the brush stroke as a record of the writer’s training, and Japanese craft treats repair and imperfection as a continuation of an object’s life.
The tie between a work and its maker may run tighter there, and the thesis has not been tested against those traditions.
Creativity beyond the finished product
He closes with commitments. Artists should treat the effort a model saves as a budget and spend it on experience the model can’t supply.
Schools, museums, and funders should value process and documented engagement alongside finished pieces, since a finished piece no longer certifies them.
Tool builders should design against convergence and make provenance legible, because these systems increasingly shape what people notice and prefer.
Zoran argues that the future of creativity should not be judged solely by what artificial intelligence can produce. The more important question is what people gain, or lose, when they use it.
If technology removes much of the effort involved in making art, that freedom could open new opportunities for exploration, experimentation, and personal growth.
But whether AI ultimately expands human creativity or narrows it may depend on how people choose to use these increasingly powerful tools.
The full study was published in the journal AI & Society.
Like what you read? Subscribe to our newsletter for engaging articles, exclusive content, and the latest updates.
Check us out on EarthSnap, a free app brought to you by Eric Ralls and Earth.com.
