Robin Mudge – These Are Not Snapshots. They Are Conversations Between Image and AI

Review by Gerhard Clausing •

Robin Mudge offers a great definition on his website: “a photobook is an exhibition in your hands.” In this photobook he has compiled an exhibition of his images that reflect everyday observations, and he has paired those photographs with ChatGPT descriptions. The ‘machine’ system can thus be considered a collaborator for the human creator of the images; a new type of comprehensive dialogue is created that redefines what an image may mean. We have well-observed authentic pictures produced by a photographer, subjected to curatorial commentary by AI-supported software. This could be considered a bridge between human understanding and algorithmic logic.

It is clear that ChatGPT recognizes and mentions more things than the average human can observe at one time. Not only is it capable of gathering data from more sources than any person can access instantly, but it is also capable of phrasing the resulting conclusions in a much more elegant and smooth-sounding manner. It seems that ChatGPT more rapidly combines and integrates several stages that curators would have to contemplate individually and in a more time-intensive manner.

Nevertheless we have to remember that the machine learning on which the image viewing commentary written by AI is based contains an enormous amount of human input and human observations. It is an abstract of all that is available, a concisely phrased summary and concentrated review of human existence and impressions. Human experience and human perception are the background on which the machine response and phrasing are based. Our human observations might be more partial, or, when spontaneously given, they might be less concise. Our human perceptions might also be less perfectly phrased, not immediately ready for printing or for display on a museum wall; the reactions of humans are nevertheless the basis of ChatGPT’s comments. Any innovations at this time would seem to occur in the combinations and the conclusions reached.

Some additional points are worth mentioning. Unlike humans, who are trained to be precise whenever possible, the machine has no qualms about dwelling in the arena of creative ambivalence, which is so central for art interpretation. We need to learn to listen to AI’s ability to extract, compound, and extend our visual imagination, even when the statements are sometimes over the top. The result becomes a richly meditative space, one in which clarity can give way to interpretive imagination to become part of an expanded discourse. While the process also implies a kind of destabilization of authorship, it can also be seen as an expansion of horizons and as a synthesizing force that can widen our comprehension through its greater amount of data and objectivity.

So let us look at some examples. I decided to subject the prose curation of the pages reproduced below to the latest version of ChatGPT, which is GPT-5. The results were surprising, to say the least. In general, the response was that the ChatGPT commentary in Mudge’s book, while most interesting and properly layered most of the time, did not take into account the photographic realm sufficiently and went over the top with some of the relatively modest subjects, such as the images of rubber ducks: “The invocation of ‘adult performativity’ and commodification, while accurate, risks sounding overly serious for such a cheeky subject. The commentary doesn’t engage the photographic aspect of the work as strongly as the sculptural/arranged-object aspect.”

Without going into too many details, I will present just one alternative curation commentary by GPT-5, in this case for the interesting double page labeled “Urban Palimpsest #17” (= something superimposed on something else):

Urban Palimpsest #17 (Revision)

This photograph captures not just a wall but a surface in flux — a fragment of the city that records its own history. Layered graffiti, torn posters, and fading paint combine into a dense visual sediment, but the image resists becoming merely an inventory of marks. The camera’s framing emphasizes the junctions where one texture collides with another: the hard edge of spray paint against peeling plaster, or the soft paper fibers curling away from brick.

Light plays a crucial role here, grazing across the wall at an angle that heightens relief, making abrasions and cracks read almost as engravings. Shadows carve depth into the surface, and the shallow depth of field isolates this patch of urban skin, momentarily lifting it out of its broader street context.

Rather than documenting urban decay, the photograph operates as a study of attention — how the lens can transform what might otherwise be dismissed as visual noise into an intentional composition. The work aligns with traditions of the vernacular and the readymade, yet its force comes less from theory than from perception: the ability to make viewers pause before a wall they would likely pass without notice.

If the city writes over itself endlessly, this image is a pause in that writing — a momentary arrest of an ongoing script. The palimpsest is not only historical but photographic: each exposure fixes a layer of time, allowing the eye to dwell in what is usually fleeting.

From this revised assessment we can see that the dialogue between human-created images and AI commentary is a work in progress, and that the results will depend on instructions given and refinements evolving over time.

While your opinion in all these matters may differ, it is nevertheless the case that artificial intelligence can serve as an expansion of our perception of images, whether they are traditionally photographic, as in the case of this photobook, or partly or wholly created with the help of AI. This photobook makes an important contribution by showing in great detail the comprehensive enhancement that AI image description and curation can add to our understanding. An expanded discourse on visual material with both human and machine perception can become part of a more reflective inquiry in the arts.

____________

Gerhard Clausing, PBJ Editorial Consultant and Editor Emeritus, is an author and artist from Southern California and Franconia.

____________

Robin Mudge – These Are Not Snapshots. They Are Conversations Between Image and AI

Photographer: Robin Mudge (born in the UK; lives in Plymouth, UK)

Text: ChatGPT 4.0 and Robin Mudge

Language: English

Paperback, spiral-bound; 66 pages; 10 x 7 inches (25.5 x 17.7 cm)

____________

Articles and photographs published in the PhotoBook Journal may not be reproduced without the permission of the PhotoBook Journal staff and the photographer(s). All images, texts, and designs are under copyright by the authors and publishers.