OtterMind Krystian
Wydro
PL
Painterly AI image: a mother holding her child in the doorway of a stone house
Knowledge base Manifesto - On tools that rewrite the rules

A Donkey, Monet
& Midjourney

In 1910, a painting made by a donkey fooled art critics. Over a century later, history repeats itself - and once again, everyone draws the wrong conclusions.

- AI Creative Director, OtterMind  ◆   ◆  7 min read
01 - The Donkey of Montmartre

A Brush Tied to a Tail

Paris, 1910. At the foot of Montmartre, outside the Lapin Agile cabaret, stands a donkey named Lolo. Journalist Roland Dorgelès ties a brush to its tail, slides a canvas underneath, and waits. The animal swishes. Paint hits canvas. A painting is born.

Dorgelès titles it - "And the Sun Fell Asleep Over the Adriatic" - and submits it to the Salon des Indépendants under the fictitious name Joachim-Raphaël Boronali. The painting is accepted. Exhibited. Praised by critics for its bold palette and expressive energy.

When the truth comes out, Dorgelès claims victory. Here is proof, he says, that modern art is a sham. If a donkey can do what the avant-garde does, then the avant-garde is worthless.

Except Dorgelès didn't prove what he thought he proved. He didn't expose art as fake. He exposed criticism that can't tell the difference between intention and accident.
02 - The Pattern of Panic

Every new tool triggers the same panic

Thirty-six years before Lolo, the Impressionists heard the same accusation: this isn't art. Monet "doesn't finish his paintings." Renoir "paints carelessly." The very word "Impressionists" was coined as an insult - invented by critic Leroy to ridicule a group of painters who dared to leave the studio and work in the open air.

What did they actually do? They understood light more deeply than anyone before them. Their "sloppiness" was a precise decision - a conscious rejection of academic finish in favor of perceptual truth. It wasn't taking the easy way out. It was going deeper, with tools the establishment refused to accept.

A few decades later, Duchamp went further - he took a urinal, signed it "R. Mutt," and submitted it to the Society of Independent Artists exhibition in New York. The organisers rejected it. The Fountain of 1917 required zero craft. It required one thing: a decision. And that decision redefined art for the next hundred years.

And this pattern repeats with surgical regularity:

Repeatable pattern
~1870
Impressionism
"That's not painting. It's scribbles."
~1900
Photography
"That's not art. The machine does it for you."
~1917
Ready-made
"That's not art. You didn't make anything - you just picked it."
~1990
Photoshop
"That's cheating. The computer retouches it."
~2023
Generative AI
"That's the easy way out. AI does it for you."
03 - Shift, Not Disappearance

Difficulty never disappears

Every time, the accusation sounds the same: "it's too easy." And every time, it's wrong - because it rests on the assumption that creative value is measured by manual effort. That the more sweat, the more art.

But the truth is different. Difficulty doesn't disappear - it relocates. From the hand to the eye. From the brush to the decision. From execution to thinking.

A photographer doesn't need to mix paint, but must understand light, composition, and timing. A designer in Photoshop doesn't draw from scratch, but must know what to change and why. A person with AI doesn't paint - but without understanding creative patterns, they'll generate noise.

This is the key that critics keep missing: the tool lowers the barrier to entry, but it doesn't lower the ceiling. Anyone can type a prompt. Not everyone knows why one prompt produces a masterpiece and another produces generic slop.

04 - The Manifesto

Dorgelès Was a Prompter

Let's go back to the donkey. Because there's something in this story that Dorgelès never admitted.

The donkey didn't create the painting. The donkey was a tool. Dorgelès chose the paints. Dorgelès positioned the canvas. Dorgelès gave it a title, framed it, invented a fictional artist, and constructed a gallery context. Without that - it's a stain on fabric. With it - a work that entered art history.

Dorgelès was the prompter.
The donkey was the model.
The value came from decisions, not brushstrokes.

And that's exactly what happens today. AI is the donkey - it generates motion, produces output. But it's the human who decides on direction, context, meaning. It's the human who turns noise into signal.

The difference between Dorgelès and a real artist? Dorgelès wanted to prove that randomness is enough. An artist knows that randomness is only the beginning.

05 - Creative Patterns

Creativity is repeatable

This is the question I've been asking for years: is creativity repeatable? The answer is yes. But not the way most people think.

Creativity isn't about waiting for inspiration. It's about recognizing patterns - in composition, narrative, emotion, contrast - and applying them consciously. The Impressionists didn't paint randomly. They had a precise system for observing light. Their genius was building a repeatable process where others saw only talent.

AI doesn't eliminate that process. AI reveals it. When you have to describe in words what you want to see - you first have to understand what you're looking for. A prompt isn't an instruction for a machine. A prompt is proof that you understand the rules of the game.

Give someone Midjourney and ask them for a campaign-grade Key Visual. Watch how fast "the easy way out" falls apart.

Building that repeatable process with a team is what my AI Training for Companies is about.

06 - Ethics & Responsibility

The machine watched other people's work

But there's something we can't skip. And here, the critics are right.

The camera was empty. Photoshop was empty. AI is not. AI has "seen" billions of images, texts, works. It absorbed other people's output without asking permission. And the real revolt of artists isn't about fear of a new tool. It's about the fact that nobody asked them if they wanted to be fuel for an algorithm.

That's a fair accusation. And it needs to be acknowledged.

But the market is maturing. Ethical models are gaining ground - Adobe Firefly, Shutterstock AI, Moonvalley, even Poland's own Bielik - trained on public domain and licensed assets. Compensation systems are emerging. Artists are finally starting to be paid for the work that builds the technology. It's not perfect. But it's a direction.

And ethics isn't just about models. It's our daily choice. This is what I teach in workshops: never put an artist's name in a prompt. Yes - "style" isn't legally protected. But copying someone's name is intellectual laziness that borders on plagiarism.

Do what directors have done for decades. When cinema draws from painting - it doesn't copy a work 1:1. It breaks it down to first principles: light, palette, texture, composition, mood. When you can translate inspiration into your own precise words - you stop being an imitator. You become a creator.

Because behind every model, every algorithm, every generated image - there's a human. A human who built it. A human who fed it data. And a human who decides what to do with the output. AI doesn't replace the creator. AI reveals who the creator really is - and who was just holding the brush.