The History of Attention:
Messy Milk Maid by Joshua Adam Risner, Oil on linen aluminum panel.
Why AI is the Mirror the Art World Needs
Artificial intelligence has officially entered the high-stakes arena of art attribution, promising to settle age-old debates with cold, hard data. Recently, The Guardian reported that AI analysis cast "91% negative" doubt on a Jan van Eyck masterpiece in Philadelphia. At first glance, this feels like a crisis—a digital assassin killing off a national treasure. But as a painter, my optimism is tempered by a different reality: AI isn’t just a forensic tool; it is a whistleblower. It is a diagnostic mirror that exposes a fundamental truth about our culture: we have become lazy viewers who value the certificate more than the canvas.
The "Author Function": Art’s Invisible OS
To understand why a percentage point can tank the value of a masterpiece, we have to look at what Michel Foucault called the "Author Function." Foucault argued that "The Author" isn’t just a person; it’s a functional tool society uses to categorize and commodify thought.
We don’t hunt for "Van Eyck" because we want to connect with a 15th-century human; we do it because the name acts as a classification system. It is the "Intel Inside" sticker for a canvas. It allows a museum to file a painting under "Masterpiece" rather than "Anonymous Study," instantly moving the decimal point six places to the right. As long as art is an investment, attribution equals money. We have outsourced our aesthetic judgment to a label.
The Expert’s Blind Spot: The Lost Leonardo
The 2021 documentary The Lost Leonardo provides a perfect autopsy of this phenomenon. It follows the Salvator Mundi, a painting purchased for $1,175 that eventually sold for a world-record $450 million once it was branded as a lost Leonardo da Vinci. The film reveals how the "Author Function" overrides physical reality when financial stakes are high.
Despite serious scholarly doubts regarding its stiff composition and inconsistent optics, the market required the "Leonardo" brand to be absolute to justify its price. As a painter, I see the expert’s blind spot here; historians often over-interpret a sitter’s expression as a profound psychological choice, when a practitioner can see where a jawline was simply adjusted to make a composition work. The Lost Leonardo shows us that when art is treated as a financial asset, the "Master" is often just a rationalization we overlay onto the messy, inconsistent reality of a human life.
The Diagnostic Mirror
This is where AI performs its most valuable service. By introducing "data-driven doubt," AI makes the "Author Function" unstable. It does this in the same way it makes art: by identifying and replicating "acceptable patterns." If an algorithm can successfully uncover a pattern that suggests a painting isn’t by a specific master, it doesn't mean the work has lost its inherent greatness; it means our criteria for "greatness" might have become dangerously superficial.
I recently finished an inspired copy of Vermeer’s The Milkmaid. If this painting were hung in a gallery under Vermeer’s name, it would command millions; under mine, it is a "study." But the work itself hasn't changed. The weight of the ceramic jug and the specific glow of the yellow fabric are the results of my own intense observation. If the work is suddenly "lesser" because my name is on the back instead of Johannes’, then we must admit that the art market isn't trading in beauty—it’s trading in provenance. AI is simply pointing out that if "greatness" can be distilled into a statistical average, then the greatness we were celebrating was never about the soul; it was about the algorithm of the market.
Beyond the Pattern: The Labor of Attention
This is why the fear of AI "replacing" artists is a misunderstanding of what art is. We fear the machine only because we have spent too long treating art as a "product"—a finished image to be consumed, valued for its connections and not its testament. If art is just a product, then an algorithm wins. But if art is more than its lineage, market value, or brand, then artists still have a job to do.
If Foucault exposes the flaws in the system, the theologian Rowan Williams reveals what the system has been missing. Williams describes art as a "labor of attention"—an intense, unselfish looking. To paint a subject is to perform a kind of "dispossession of the self"; you look at the world so deeply that your ego disappears.
As a painter, I know this labor is a record of the good days and the bad. It is found in the "bad marks"—the hesitant lines and frantic corrections that happen when the eye and hand fail to sync. Under a rigid AI audit, my own work might even be flagged as a forgery of my past self because an injury changed my mark-making, but the machine’s "correctness" is irrelevant. It cannot replicate the witness. Even an identified "study" can encapsulate this evidence of attention.
Every work of art is a history of attention. It is a record of these battles. A masterpiece isn't "perfect" because of its recognizable patterns; it is powerful because it reveals the artist’s refusal to look away, even when the observation was difficult. This record of human problem-solving—the honest grit of trying to see clearly and dispossess—leaves a physical fingerprint on the work that no algorithm can recreate or generate. In turn, real appreciation occurs when a viewer performs their own labor of attention, recognizing that specific struggle on the canvas and meeting the artist in that space of shared attention.
Conclusion: The Soul in the Grit
We should thank AI for bursting the bubble of attribution. By using data to disrupt the "certainty" of the label, AI is doing the viewer a profound favor. It is stripping away the "Algorithm of Attribution" and forcing us to ask a much harder, more honest question: "Does this work speak to me, regardless of the name on the plaque?"
AI provides a veneer of scientific certainty that may disrupt lucrative markets, but for the human being, it serves as a clarifying lens by which we see ourselves. It points out that we have been looking at art with a "checkbox" mentality and gives us permission to return to the grit. When the "Author Function" finally breaks, we are forced to look at the work itself. We return to the only thing that matters: the record of a human being’s attention—the struggle, the resolution, and the communicative power of the art, soul to soul.