Art, AI and Pollock’s denial of accident (Part II)
In part I (read it here), we established a contrast between Jackson Pollock's denial of accidents in his artistic process and the inherent randomness of AI-generated art. AI models rely on probabilistic decisions and randomness, introducing unpredictable elements to any generated output. Unlike Pollock, an artist working with generative AI cannot fully determine what is accidental and what is intentional, making randomness an intrinsic part of the medium. Inevitably, certain elements fall outside the creator’s control.
In traditional art, the artist retains full control over creation. Every components later kept or edited out were put there by action of the artist. Pollock’s approach again exemplifies this.
This level of control simply doesn’t exist in AI-generated art. The accidents introduced by the AI are not editable, as they cannot be precisely identified or isolated. This fundamental difference means the artist cannot fully shape or refine the outcome as Pollock could.
Obvious. Edmond de Belamy. © Obvious, 2018 - The AI-generated portrait was sold for $432,500 at a Christie's auction in New York in 2018
Yes, we can agree that the artistic piece doesn't appear spontaneously, and it requires the intention and action of a creator. And AI is as much a medium as it is a tool. But the core of the issue is that it is not a passive medium. As the creator sets parameters and guide the direction of the creation, they must also accept that the AI will introduce elements beyond their control. That's an irrefutable fact and an unavoidable effect of the nature of generative AI, at least in its current state.
To make a concrete example, let's take prompting. In prompting, the artist can guide the AI but never fully control the result. Even in editing, randomness remains, as the artist cannot fully identify or control which elements to keep or discard. Control in this case is elusive.
In Pollock's case, he had the ability to modify, refine, and perfect each element of his work. If an accident occurred—a drip that went astray or a splatter that didn’t align with his vision—he could either integrate it into the composition or discard it entirely.
So, an important difference between traditional art and AI-generated art is the level of control the artist can exert over the final outcome.
One might argue that this lack of precise control can be both a limitation and a source of creative freedom, or that it can lead to "unexpected" and "innovative results" that push the "boundaries of what art can be". But to recognize AI-generated work as art would require us to rethink the definition of art itself. Yet redefining 'art' to encompass AI creations is a slippery slope, as it opens the door to classifying anything (even non-art) as art. That of course is a broader discussion, not for this article.
Intentionality at the heart of Art
So this is not really a debate on whether AI generated art is art. Rather, I want to drive home the point that intention is indispensable for art creation. because without intention we have only output. While it might be true that AI opens up new creative possibilities, this leaves open the question of whether those possibilities hold the same artistic value if intentionality is diminished.
As the author points out in this piece on intentionality,
"intentionality is not just a human phenomenon, but a deeply embedded principle in our understanding of meaning and action."
This concept of intentionality being inseparable from subject and object also extends beyond art. He continues:
"Intentionality also establishes an inescapable duality in reality, because all intentions require a subject and an object. What results is not merely a duality but a trinity: The intentional act is a third thing, along with the subject and object."
Note that in AI-generated art, this intentional act is absent. The machine may generate an object, but without an intentional subject guiding each decision, the crucial link that binds creator, creation, and process seems broken. That could be one of the reasons that traditional art and AI art resonate differently with us.
Moby Dick, 1943 by Jackson Pollock
Pollock’s mastery of his medium beautifully illustrates the trinity of subject, object, and intentional act. He is the conscious subject, fully aware and in control of the process, shaping the object (his artwork) through deliberate action. This intentionality is what makes Pollock’s art resonate on a deeper level.
Pollock famously said:
"I want to express my feelings rather than illustrate them."
His work was a conscious expression of emotion and intention, something AI cannot replicate.
In that spirit I would say that a true artist creates, not just generates.
The randomness that Pollock denied is, in AI, a core feature. But is it fair to say that generative AI leaves behind a trail of creations that exist without the intentional guidance that defines true artistic mastery? At what point does the original intention of the creator dissolve in the process? Are AI generated artworks mostly expressions of randomness, and does it make them less meaningful creations?
I think these are questions worth exploring.
References:
jackson-pollock.org, website dedicated to the life and work of the artist.
MindMatters.ai - How Is Intentionality Embedded in the Universe?
