This article is part of "Minds in the Machine Age" — a companion series to Between Brain & Binary.
← Back to Series Overview | ← Previous: When AI Becomes a Therapist: Boundaries We Need to Draw | → Next: The Habit Machine: Using AI to Build Better Routines
Two mechanical hands shape wet clay on a potter's wheel, precise and untiring, the wheel spinning beneath them. In September 2022, Jason Allen entered an AI-generated image into the Colorado State Fair's fine arts competition - a lush, operatic scene titled Théâtre D'opéra Spatial, produced using the Midjourney platform - and won first place in the digital arts category. The backlash was immediate and visceral. Artists responded with fury. Critics wrote essays. The story was covered internationally, not because an image had won a state fair competition, but because the event had cracked open something that most people had not quite articulated yet: a deep, barely examined assumption that art is something humans do, and that whatever AI produces instead is categorically different, even if you cannot immediately say why.
The conversation that followed was often confused, sometimes dismissive, and occasionally genuinely illuminating. The most interesting thing it revealed was not about artificial intelligence at all.
What We Mean When We Say Creative
Creativity is a word we use freely and define poorly. In her landmark work The Creative Mind, philosopher and AI researcher Margaret Boden proposed a taxonomy that is still among the most useful: creativity, she argued, operates in three modes. Combinational creativity produces novel ideas by combining existing concepts in unexpected ways. Exploratory creativity generates new possibilities by systematically exploring the structural rules of a conceptual space. Transformational creativity changes those rules themselves - it operates outside the existing space and produces something genuinely new.
Most human creativity, most of the time, is combinational. We take what we have encountered - images, stories, sounds, emotions, ideas - and recombine them in ways shaped by taste, skill, intention, and experience. The artist is not creating from nothing. They are translating something - an internal state, a perception, a question without an answer - through the filter of a chosen form.
Mihaly Csikszentmihalyi, whose research on flow mapped the psychological conditions in which creative work is most fully realised, emphasised that creativity is not an event. It is a process embedded in a field - a dialogue between the individual and a cultural domain, mediated by expertise, sustained attention, and intrinsic motivation.
Teresa Amabile's decades of research on creativity and motivation added another layer: the creative process is profoundly sensitive to intrinsic investment. People do their most genuinely creative work when they are driven by interest, curiosity, and the satisfaction of the work itself - not external reward or pressure.
What AI Art Actually Is
AI image generation models like Midjourney, DALL-E, and Stable Diffusion produce outputs through a process that bears only a structural resemblance to human creativity. These systems are trained on vast datasets of images paired with descriptive text, and they learn statistical relationships between words and visual features. Given a prompt, they generate an image by iteratively refining noise toward a configuration that fits the statistical patterns of the training data.
This is extraordinary as a technical achievement. It is also, in a precise sense, combinational - and only combinational. The system has no internal state to translate. No experience that the image is in service of. No curiosity that drove the choice of subject matter. No relationship to the finished work that extends beyond the generation event. It does not feel the satisfaction of having made something. It does not carry an intention that the image either fulfils or fails to fulfil.
What gives an AI-generated image its uncanny quality - its dreamlike coherence, its aesthetic fluency - is exactly what it does not have: the trace of a particular person's encounter with the world. Great art, even when technically modest, is evidence of a mind. It carries the specific weight of someone's specific experience. It was made by someone for whom making it cost something.
What It Reveals About Us
The intensity of the response to AI-generated art is itself data about what we value in creativity. When people say AI art is "not real art," they are not usually making a technical claim about the production process. They are making a claim about meaning. Art matters, in the way that it matters, because it is the trace of a person. Because it is addressed to us from somewhere.
This is why provenance changes everything. A painting by a machine that looks identical to a painting by a human does not produce the same experience - because the experience of art is not just retinal. It is relational. We bring to the object our knowledge of how it came to exist, and that knowledge shapes what we see.
This is not just intuition. Psychologists George Newman and Paul Bloom, across a series of experiments on how people value original artworks versus duplicates, found that value judgements rest on two distinct mechanisms: a belief that the object represents a unique creative performance, and a kind of imagined contagion - the sense that the object carries some trace of physical or intentional contact with its maker. Neither mechanism survives the transfer to a machine-generated image. There was no performance in the sense that matters, because nothing was risked or resolved. And there is no contact to imagine, because there was no one there to have made contact.
What AI art has done, usefully and uncomfortably, is force us to articulate something we had taken for granted: that creativity is not the production of novel objects. It is the expression of a particular kind of subjectivity through a chosen form. The value of the output is inseparable from the humanity of its origin.
This does not mean AI art has no value. It has considerable value as a tool, as a collaborator, as a way of externalising ideas quickly enough to evaluate them. But it is a different kind of value from what it superficially resembles. Confusing them is not an error in aesthetic judgement. It is a category mistake about what art is for.
Copyright law has, so far, arrived at roughly the same conclusion by an entirely different route. In 2023, the US Copyright Office limited the copyright protection available to the graphic novel Zarya of the Dawn on the grounds that its Midjourney-generated images lacked sufficient human authorship, and a federal court reached the same conclusion later that year in a separate case, Thaler v. Perlmutter, ruling that copyright law "has never stretched so far... as to protect works generated by new forms of technology operating absent any guiding human hand." Two entirely different systems - aesthetic judgement and copyright doctrine - converged on the same requirement: for something to count as authored, a human hand has to have been guiding it.
The line usually attributed to Paul Valéry - that a poem is never finished, only abandoned - is technically W. H. Auden's later paraphrase of a longer observation Valéry made in 1933, but the compressed version has outlived the original for good reason. It captures something essential: that creative work is not primarily about the object produced, but about the maker's ongoing relationship with it. What AI produces is never abandoned. It was never begun in that sense at all.
The Question Worth Asking
The advent of AI art does not threaten human creativity. It clarifies it. It strips away the notion that creativity is about producing aesthetically compelling outputs, and returns us to the more interesting question: what is it that we are actually doing when we make something?
The answer has something to do with meaning, with intention, and with the peculiar human need to leave a mark that says - distinctly, irreducibly - I was here.
References (APA style)
Amabile, T. M. (1983). The social psychology of creativity: A componential conceptualization. Journal of Personality and Social Psychology, 45(2), 357–376.
Boden, M. A. (2004). The creative mind: Myths and mechanisms (2nd ed.). Routledge.
Csikszentmihalyi, M. (1990). Flow: The psychology of optimal experience. Harper & Row.
Newman, G. E., & Bloom, P. (2012). Art and authenticity: The importance of originals in judgments of value. Journal of Experimental Psychology: General, 141(3), 558–569.
Roose, K. (2022, September 2). An A.I.-generated picture won an art prize. Artists aren't happy. The New York Times.
U.S. Copyright Office. (2023, February 21). Re: Zarya of the Dawn (Registration # VAu001480196). Library of Congress.
AI Disclosure: Research and organization for this article were assisted by AI tools; all factual claims and citations were independently verified against primary academic sources, and the analysis and conclusions are the author's own. The featured image was generated using Canva AI.
← Previous: When AI Becomes a Therapist: Boundaries We Need to Draw | → Next: The Habit Machine: Using AI to Build Better Routines coming soon, next week
Related reading from Between Brain & Binary: Birth of AI and Early Dialogues (1950s–1970s)
Comments
Post a Comment