Visions of the Impossible
- Alessandro Scali
- May 26
- 5 min read
Kafka, AI, and the Photograph That Never Was
If photography is, by nature, a trace of reality, what happens when it depicts something that never existed?
Kafkaesque Visions was born from this very paradox: where literary language—dense, metaphorical, absurd—encounters a machine capable of generating images, but incapable of dreaming.
I asked myself: is it possible to see Odradek? To photograph Bucephalus ascending the courthouse stairs? To freeze, in a 1920s-style shot, the bridge groaning above the abyss?The answer—partial, imperfect, uneasy—can be found in these nine photographs that look like they’re from a century ago, yet are very much of our present.
They were generated by an artificial intelligence. But imagined first by a human: Kafka, and then myself. And between us, the language—acting as a bridge, a key, a provocation.
Language as a Generative Device
Franz Kafka never owned a camera, as far as we know. But he possessed something more powerful: the ability to evoke the impossible using only words.
Each of his stories is a distorting lens pointed at the world. Kafka doesn’t describe reality—he unravels it, rebuilds it, views it from impossible angles. A horse practicing law. A thread spool endowed with will. A man who is a bridge.Kafka doesn’t tell us what happens. He shows us what might happen if the world surrendered to paradox.
That’s the spirit behind Kafkaesque Visions. Kafka’s descriptions became prompts—textual commands, detailed and complex—addressed to an AI system tasked with generating images that look like they were taken a hundred years ago.Not illustrations. Not comic strips. Vintage photographs of the unreal.
The Photography of the Unreal
Traditionally, photography is considered a document—a reflection of reality, a testimony. But what happens when the photographed subject doesn’t exist? When it cannot exist?
Photography ceases to be evidence and becomes credible invention. A visual artifact that doesn’t portray what was, but what imagination dared to propose. Thus is born a new aesthetic of the impossible—where AI, instead of imitating reality, betrays it to construct an alternate one: more ambiguous, fragile, and poetic.
Selected Visions
Odradek
“It looks like a spool of thread... but it seems to have legs...”
Perhaps the most unclassifiable creature in Kafka’s entire literature. Faced with this semantic paradox, AI must choose. And in doing so, it creates something unprecedented: a shape that is unclear, borderline, simultaneously believable and wrong.
This is the most unsettling image in the series—and also the most philosophical.
The Bridge
“I was stiff and cold, I was a bridge, stretched over an abyss.”
In this photo, a human body is simultaneously being and structure. It does not act—it holds. The AI renders a figure twisted between matter and metaphor. The man becomes landscape. And the result is the most tragic vision in the project.
A report to an Academy
“I, a free ape, have subjected myself to this yoke…”
Here, mimicry is reversed: it’s not the machine that imitates the human, but the animal that becomes human to survive.
The subject—an elegant ape seated in a rocking chair with a bottle on the table—appears normal. But this very normality unsettles: the absurd has donned the mask of civility.
A crossbreed
“I have a strange animal, half kitten, half lamb…”
The most tender and sinister of all images. A creature that seems cuddly but hides something disturbing in its gaze or paws. The AI succeeds here by not choosing entirely—the creature remains ambiguous. And that’s what makes it poetic.
The new lawyer
“We have a new lawyer, Doctor Bucephalus…”
Instead of focusing on the widely known Metamorphosis, I chose this lesser-known, yet equally emblematic tale.
A horse climbing courthouse stairs defies logic. And yet the AI renders it convincingly. The shot looks real. The clerk observing him is believable. The architecture matches. But it’s the subject that creates dissonance: a document of something that never happened.
And then, they move
Many of the Kafkaesque Visions have become brief five-second videos. Micro-sequences where time steps into the frame.
The result is, in a word, disorienting. As long as the image is still, we can hold onto the illusion—it feels like an old photograph, a forgotten relic. But once it moves—even slightly, even a breath—something fractures.
These fragments are no longer mere representations. They’re apparitions. Glimpses of a world that never was, yet now demands to be seen.
In The New Lawyer, for instance, Bucephalus doesn’t just appear—he becomes undeniable. He ascends the marble steps slowly, confidently, as the court usher watches without surprise.
In those five seconds, Kafka stops being an author. He becomes a director. And AI isn’t a machine. It’s a medium.
In The Bridge, there is no action—just a tremble.
A human body lies stretched between two rocks, suspended in space. It doesn’t walk, doesn’t break, doesn’t fall. But it breathes.
In those five seconds, stillness becomes tension. The body is no longer a character—it’s a living structure, a tragic symbol, an architecture of the absurd.
No literary metaphor could render it so visible—or so haunting.
And then there’s the ape. It doesn’t laugh. It doesn’t dance. It doesn’t gesticulate. It sits, composed, like an old-world gentleman. But everything about it is learned.The hands are too stiff, the gaze too rehearsed, the posture too refined.And in those five seconds—where nothing happens—everything becomes clear: there is nothing more Kafkaesque than civilization as effort.
This video doesn’t show animality. It shows the cost of humanity.
AI Is Creative, but Not Imaginative
I often say this: AI is creative, but not imaginative.
It generates new images, but it doesn’t imagine. It responds, but doesn’t propose. It executes, but doesn’t dream.All creativity lies upstream—in the choice of words, the selection of stories, the prompt design, the intuition behind the impossible-possible image.In this sense, AI is a mirror: it reflects our vision of the world, but with an unsettling clarity.
Why Give a Face to the Invisible?
Why photograph what never existed? Perhaps to restore power to language. Perhaps to dissolve the boundary between literature and image. Or perhaps to remind us that even the impossible, if well expressed, can be seen.
In an era where everything has already been shown, we need images that do not document—but disturb, suggest, reveal.
Kafka left us visions. I merely tried to look at them.
Comments