200 Years of Photography: What Do We Believe in the Age of Artificial Intelligence?

From early photographic experiments to generative images — how authorship, trust, and our relationship with what we see are changing.

We instinctively trust photographs. A photograph seems to confirm that this person existed, this place could be seen, this event really happened.

Of course, we know about staging, retouching, and digital manipulation. We understand that a photographer chooses the angle, the light, and the moment. Yet we still tend to give a photograph the benefit of the doubt: we assume there was something real behind the image.

Today, that connection is no longer guaranteed. An image can look photographic — retaining depth, light, movement, and the texture of skin — even when nothing ever took place in front of a camera.

What changes at that point? Photography itself, our understanding of it, or our ability to trust our own eyes?

From Light to Data

One of the earliest surviving photographs, Joseph Nicéphore Niépce’s «View from the Window at Le Gras», is usually dated to 1827. Traditionally, the exposure is thought to have lasted around eight hours. Light formed the image on a metal plate coated with a light-sensitive substance.

It was a slow physical process, almost the opposite of the speed at which images are created today. Photography emerged as a trace of an encounter between light, space, and a material surface.

But even this early image was not a neutral copy of the world. It showed an existing view from a window, yet assembled it through light, time, and a chosen point of view. Over the long exposure, the sun moved, the shadows shifted direction, and several moments came together in a single still image.

Almost two hundred years later, a photographic image can be made without this kind of encounter with reality. A generative system creates an image by drawing on vast amounts of visual data: photographs, illustrations, works of art, and cultural ideas of what a body, a landscape, movement, or human emotion should look like.

Alongside the familiar path from light to image, there is now another: from data to visual form.

Was Photography Ever the Absolute Truth?

Artificial intelligence is often described as a technology that has undermined the credibility of photography. Yet photography was never simply a reflection of reality.

There has always been someone between the world and the image, making decisions: where to point the camera, what to leave outside the frame, which moment to preserve, and how to print the result. Light could conceal certain details and emphasise others. Staging, photomontage, and manual retouching all existed long before digital manipulation.

Photography did more than show the world. It helped shape the way we see it.

Still, there is an important distinction between traditional photography and a generative image. Even a staged photograph retained a physical connection to what stood before the lens. Today, an image can look like a photograph even though the event it appears to show never took place.

That is why the question «is this real or not?» is becoming less useful. Increasingly, what matters is understanding where an image comes from: how it was made, who made the decisions, and who is responsible for the result.

Distrust and a Strange Distance

Distrust does not arise only when someone is trying to deceive us. Sometimes it begins simply because we do not understand how an image was made.

A viewer may not realise that a work was created using AI and may respond to it in the same way as to a photograph: reading emotion, imagining a story, and feeling for its subject. But once they learn how it was created, their response often changes. We begin to evaluate not only the image itself, but also the process behind it.

This creates a distance between feeling and knowledge: the image has already affected us, but new information changes the way we see it.

Where Is the Author?

When artificial intelligence is involved in making a work, our conventional understanding of authorship becomes less clear.

Is formulating a prompt enough? What belongs to the artist when some of the visual decisions are proposed by a system? Where do the artist’s choices end and technology’s contribution begin? What happens to authorship when a model has been trained on images made by other people?

There is no single answer within the art world yet.

In 2025, Christie’s became the first major auction house to hold a dedicated sale of art made using AI. The sale itself showed that generative art had already entered the professional art world.

Yet the market remains cautious. In Artsy’s 2026 survey of more than 300 gallery professionals, just 9% of respondents described AI-generated art as a legitimate new medium. A further 28% saw it as an evolving category whose market value remains unclear.

Perhaps the authorship of a generative work should be located not in a single action, but across a sequence of decisions: the concept, the choice of source material, working with the system, selecting and editing the results, printing the image, and placing it within a broader artistic context.

With this comes responsibility on the part of the artist. Viewers need to understand what they are looking at: a photograph, a digital collage, an image based on the artist’s own photograph, or a fully generative composition.

This does not make the work any less mysterious. It simply means we can speak about it more honestly.

Do We Need to Be Able to Tell a Photograph from an AI Image?

The desire to determine whether an image was captured with a camera or generated by an algorithm is understandable. But looking in this way quickly becomes a search for technical errors: incorrect anatomy, impossible shadows, and repeating details.

As the technology develops, these signs gradually disappear.

Perhaps the more important skill today is not identifying an AI-generated image by appearance alone, but being willing to ask about its origin.

Who made the work? What did the process involve? What existed in front of the camera, and what was constructed digitally? Why did the artist choose this particular tool? Does the origin of an image change its artistic power?

How Do We Look at Photography Today?

Photography is not disappearing because of AI. What is changing is our relationship with it.

Photography was once directly tied to what stood before a camera. Today, an image can be made without a camera, using AI. That is why it matters to understand not only how it was made, but also why the artist chose this particular form of dialogue with the viewer.

Perhaps the central question of the future will not be «is this image real?» but rather: do we understand where it came from — and does that change how we feel?

These are the questions at the heart of the main FRAGILE DISTANCE event, taking place on 16 August at 5:00 pm. The event will include a curator’s talk, a discussion of authorship and the perception of contemporary images, a conversation with invited guests, and a special programme.

The exhibition features works by Lola Graham — AXIS and SEVERANCE; Bulat Asyanov — SKY IN THE OCEAN and FIRE DANCE; Viktoria Sitnik — THE DANCE THAT NEVER WAS, BURGUNDY, and VULNERABILITY; Nadezhda Filippova — FAULT LINE I and FAULT LINE II; and Anastasia Milinets — SHINING.

The exhibition is on view at FOR YOUR HOME until 28 August 2026.

112 Mikheil Tsinamdzghvrishvili Street, Tbilisi.
Admission is free. Advance registration is required for the main event.

Sources:

Harry Ransom Center: The Niépce Heliograph
National Science and Media Museum: Joseph Nicéphore Niépce, creator of the first photograph
Artsy: The Artsy AI Survey 2026
Christie’s: Augmented Intelligence