A belief is an idea with an emotion attached – John Ewbank 2017
Last year, I gave a keynote speech at Art-house-life https://www.art-house-life.com/ewbank-2024 on the Isle of Wight about the nature of AI-generated art. My goal was to explain how AI creates art, rather than explaining the artwork itself. To achieve this, I explored how the various components that went into AI art generators were trained, how they function, and their limitations. One of my primary goals was to convey that “Artificial Intelligence” isn’t magic—it’s mathematics.
The Nature of AI Art Creation
To demonstrate this concept, I wanted to illustrate that the process of creating images with AI is deterministic. Two people using the same parameters on different computers would generate the same image. The art generator takes in instructions and outputs a result. If you do exactly the same thing twice, you get exactly the same result.
While AI systems are deterministic—the same inputs will produce the same outputs—the complexity of these systems means that subtle changes in initial conditions will cause dramatic changes in the output. This leads to effectively infinite possible image variations. Layers and layers of complexity can be added to the process, creating endless opportunities for creative expression.
The Spectrum of AI Art Creation
There’s a crucial distinction in how people approach AI art creation. At one end, you have simple text-to-image generation—typing prompts into ChatGPT or Midjourney and accepting whatever comes out. This is akin to asking someone else to paint something for you based on a brief description.
At the other end, you have tools like ComfyUI, where artists are building their own creative workshops. Here, the artist isn’t just describing what they want—they’re selecting from an almost limitless number of parameters, choosing AI models, setting up complex chains of image processing, and combining multiple techniques. It’s more like an artist selecting their brushes, preparing their canvas, and mixing their colours.
The creativity emerges from how these tools are combined and deployed. Each decision adds another layer of creative possibility. The artist might start with one image, feed it through various models, adjust it, combine it with others, and continue refining until they achieve their vision.
The Question of Intelligence and Creativity
When we talk about AI art, discussions inevitably turn to questions of intelligence and creativity. These debates often become circular, tangled in semantic arguments. When we ask “What is intelligence and creativity?” the answers dissolve into ambiguity—not just because we’re being imprecise, but because our language lacks the specificity to capture these concepts.
Consider how we judge human creativity: a child drawing with crayons, a photographer capturing a moment, and a painter creating a masterpiece are all engaging in creative acts—but in fundamentally different ways. Which represents “true” creativity? The question itself becomes meaningless.
The same applies to AI art. When an artist uses ComfyUI to build intricate chains of image processing, is this fundamentally different from a traditional artist selecting brushes and mixing colours? Both are using tools to express their vision. The tools differ, but the creative intent remains.
This brings us to the central question: Is AI art really art? But to answer this, we must first understand what gives any art—human or machine-made—its value. And this leads us into more contentious territory, because art’s value often has little to do with the artwork itself.
Is AI Art Really Art?
The Nature of Artistic Value
Art is ego. Its worth derives not from anything intrinsic to the piece itself—it is entirely extrinsic. Who made it defines the value. Art has notional worth that is fulfilled by the wants of others. What the artwork actually is provides no relevance to its value.
Da Vinci
Consider Salvator Mundi by Leonardo da Vinci, which sold for £350 million in 2017. It is objectively a poor artwork in its current state. The depth and detail that Leonardo painted have long since been extinguished by time, degraded to a visage of its former glory. Moreover, the technical skills of contemporary painters have far surpassed da Vinci’s abilities. Yet its value exists because of human ego and desire for rarity—it’s a statement that says “I have the money to buy this.”
The worth of art parallels that of family heirlooms: it stems not from the object itself but from the irrationality of humans attaching emotion to an object. Expensive artworks derive their worth from sentimental value at scale. It is purely societal conditioning. A story underpins the artwork, the story becomes well known, and people buy into that story.

Rothko
Rothko is another example of an artist where the values of his works are grossly disproportionate to their quality. He is known for paintings of geometric shapes. He was very good at painting them. But that isn’t to say they couldn’t be recreated with ease.
You might ask why would an indistinguishable copy of a Rothko be worth less than the original? If they look the same, what makes the original different? If the value of art was rational, then they would be worth the same. Therefore, it isn’t the art that defines its worth. It is human irrationality that defines the value of the artwork.

Biographer James Breslin writes that “Rothko wanted the ‘ego’ in his paintings to remain difficult to grasp.”
An artwork possesses an arrogance, a sort of self-importance that the creator attaches to the work. It is all a game: if the artist pretended what they created had no worth, then it wouldn’t and there wouldn’t be a market or interest. Unless the artist was famous, then it would probably have the opposite effect.
Banksy
Like Rothko, Banksy is an artist where you could question whether his art is really art. The messages he conveys are zeitgeist. They are popular because they directly reference popular culture and current events. Mostly with a political slant which appeals to those in the art world. This leads to the virtuous circle which hypes his work.
Many of his works contain a supercilious theme which portrays culture or people as being less intellectual or enlightened than he is – or for that matter those who appreciate his work. The power of virtuousness is obvious for everyone to see, it sells art, ideas, and politics.


Value or Worth
That isn’t to say that his ideas are worthless or his art is bad. It is more of an observation that if someone off the street put some traffic cones and a shopping trolley in a Monet, then it wouldn’t attract much attention. Let alone sell for £7.5 million. The value comes from the story, the context, the reputation – not the work itself.
This is the heart of art’s value paradox – it’s a collective fiction we all agree to believe in. If you took a priceless masterpiece and showed it to a nomad in the desert, it would have little value to them. When lost Constable paintings are discovered in attics and sold for millions, their value only materialises when an ‘expert’ recognises who made it. The painting didn’t change – it sat forgotten in that attic for years. Our perception of it changed.
“But you didn’t make it” – so it isn’t Art
This brings us to a common criticism of both AI art and traditional art – the question of authenticity and creation. Most artists don’t really make their artwork. They don’t make the canvas, they don’t make the frame, they don’t make the paints or pencils. They don’t make the tools. They mostly don’t even make the message that they are trying to convey, instead they borrow the messages from other places in society and history and create a work based on that message.
Either that, or as Rothko did with his squares, they don’t tell you the meaning and let you derive your own meaning so essentially they borrow your meaning – this approach is smart. It is smart because it enables the observer to create their own attachment to it. Their ego creates what it represents, when oneself creates the meaning that meaning is ever more powerful.
The art world’s obsession with provenance – who made something – rather than what was actually made, reveals the truth about artistic value. When Constable paintings are found in attics, they’re worthless until someone recognises the signature. It can’t be that the painting itself is extraordinary – after all, it sat ignored for years. The value comes from our collective agreement about who Constable was, what his work means, what owning it says about the owner.
In this context, the criticism that AI artists “didn’t make” their work becomes particularly hollow. No artist makes everything they use. The question isn’t about making – it’s about meaning, context, and the stories we choose to value.
AI Art Isn’t Art. But Nor Is Normal Art.
In reality, the argument against AI art is that ‘it isn’t very good’. This is the case with most art produced, it is objectively awful. This is true for AI art as much as normal art. The only difference is, it is much faster to make AI art, so a lot more bad art can be made more quickly.
This speed and accessibility threatens the art world’s carefully maintained system of artificial scarcity. When anyone can create technically proficient art, the emperor’s new clothes start to look rather threadbare. This might explain why the art world is so defensive – AI art exposes the arbitrary nature of how we assign artistic value.
The Evolution of Artistic Tools
Every major artistic tool has faced the same criticism. When cameras appeared, they were “cheating.” When digital art emerged, it wasn’t “real art.” Yet each new tool ultimately expanded artistic possibilities rather than diminishing them. AI is just the latest in this evolution, but it’s particularly threatening because it forces us to confront how we value art.
As AI capabilities improve, determining whether something was AI-assisted will become impossible. This might finally force us to stop obsessing over how art was made and instead focus on its impact. After all, most human artists use references, combine existing ideas, and build on others’ work. AI does this more explicitly, but the process isn’t fundamentally different.
Art in Itself Is Meaningless
I created the image below while exploring different ideas. The initial concept I had didn’t match up with what I ended up creating. However, there are many meanings that can be taken from the work. But those meanings are derived from your own knowledge, and what you choose to infer about the artwork.

For me, the creation of the image, and the ambiguity of how it was created and whether it is real is part of the challenge and fun. As you zoom more details emerge. The oil paints look real. But are they? Did I really paint this? Does it matter?




The Future of Artistic Value
If AI can replicate any style, what becomes of artistic originality? Perhaps the future isn’t about unique styles but unique visions and ideas. The artist becomes more curator than executor, more director than technician. This isn’t new – it’s just more obvious now.
We stand at the forefront of AI-assisted art, exploring the convergence of human imagination and mathematical precision. My work isn’t about proving AI can make art – it’s about revealing how arbitrary our definitions of art have always been.
AI art is the democratisation of creativity. Just as the camera freed artists from the need for perfect reproduction, AI tools free us from technical constraints. But this democratisation threatens the very system that gives art its monetary value. The art world relies on exclusivity and artificial scarcity – AI makes this harder to maintain.
What matters isn’t whether AI can create art. What matters is that we’re finally forced to confront the fiction of artistic value we’ve maintained for centuries. When we say “this isn’t art” or “that’s not real art,” we’re really just expressing our biases about what stories we choose to value.
Perhaps the real question isn’t whether AI can make art, but whether this moment will make us reconsider what we mean by art in the first place. After all, if a degraded da Vinci can be worth hundreds of millions while identical Rothko copies are worthless, maybe it’s time to admit that art was never about the art at all. It was always about the stories we tell ourselves, and who gets to tell them.

Entropy Drift
printed on Giclée Hahnemühle Photo Rag Inspired by the summer landslides in Ventnor on the Isle of Wight, I sought to create an image that showed the shifting impermanence of our world.