Imaging After Photography
AI image generation does not change photography as a practice nor medium, nor category. What AI changes about photography is our presumption of what is distinctly differentiating about it, in our expectations of what it can uniquely do. When those expectations change, the decisions that lead to a demand for it or an application of it also change. It’s the decision-makers that change, not photography.
This month in AI, which will be the Dark Ages by Autumn of 2026, still features a general angst about what photorealistic AI-generated images are doing to the (a.) business, (b.) demand for, and (c.) understanding of what we’re used to calling photographs.
We’re so used to calling them photographs that we’ve long since forgotten that they are photographic images, and that what we’re using calling a photograph is actually a photogram. (You can look up the difference yourself, which will make it more memorable.)
We have a notion of what “photographic” is, and it is heavily loaded by default with something between a preference and a bias for visual “realism” — or said differently, re-presentation of natural visual experience.
But “photo-graphic” literally refers to drawing with light. There are a lot of types of images that can be produced by guiding light waves to create a visible pattern in a field of view. The “graphic” aspect speaks to drawing both literally and metaphorically. But our conventional interest in “photography” generally begins with the further expectation that a drawing will present a form as a subject. Only something identifiable can be a subject, whether it is an actual or virtual item, a physical or conceptual construct.
That observation leads to another: all “subjects” are real. Realism actually refers to any subject with a verifiable presence. Necessarily, then, any image that “visualizes” an idea is doing something that a photograph can do. But if it does it differently, there is a possibility that its method is better for some kinds of ideas than for others.
ORIGINS AND PATHWAYS
Fast-forwarding a bit from that thought: images that are actually photo-graphic have a long history that encompasses far more than documentary realism supporting photojournalism and art. I call out those two highly intentional disciplines because in our current imagination those two have the most to competitively lose by the ascendancy of generative AI as an imaging technique.
To press this issue harder: why should GenAI imaging care about “photography” at all in the pursuit of imaging? It is as different from its predecessor as quartz drives in Swatches were to Swiss Movements in Bulovas. As indifferent as the Word Processor was to IBM Selectrics.
Frankly, there are image-makers who will use whatever they need to use; and there are “photographers” who will use what they need to make a photograph. The only serious question is, when does a photograph solve a need that is not solved by some other form of imaging?
The two dominant arenas of defining and servicing need are curating and marketing. Both are built on an editorial function that addresses an explicit value system. Neither one’s value system defines “photography” as a whole.
Intentional imaging has called on photographic image production for well over a hundred years; but the audience for photographic imagery has also been increasingly diversified and increasingly “visually literate” as consumers. The audience’s idea of the practice of photography easily spans its pervasiveness in their environment because the audience relates to most photography as manufactured imaging. The audience’s main concern, and its main presumption, is about why the photography was used.
Below is a chart wrestling many brainstorming notes into a somewhat chronological sequence of changes in our observation and mastery of imaging as recorded and curated. In my view, the most important aspect of its story is in the curation aspect, which always featured human decision-making.
There is a history of curating photography that is fundamentally about editing, something that spans all optical, electrical, and digital photographic imaging. But of course there’s always a reason for curating, and when there’s no market for your reason or your effort then you’ll just have to do it because it makes you happy.
That marketing refers to some criteria of demand, and if you are a curator/editor, you can set the criteria however you like.
IMAGINING IMAGES
This image: Decorator’s color sample? Crime scene evidence? Therapeutic metaphor? Context rules interpretation, and interpretation imposes criteria as “values”…
Farther than editing: there is a school of thought that calls for custodianship of certain values. That role usually includes requiring that in order to grant “value” to the image, it must be from certain means of production.
In this school’s broader formula for “value”, Specific Production Method + Specified Product Qualities = Distinctive Importance. But the custodianship role further acts on the presumption that what it deems important is the most valuable of all values.
There is also a school of thought that appoints certain custodianships as being more important than others. We’re familiar with exactly this kind of thing from our experience with Clergy. Such a custodial role is a direct manifestation of a culture: the responsibilities of custodianship represent value to the Culture.
Meanwhile, in the formula that designated custodians apply as Curators, the Specific Product Qualities point out what makes a photographic image important to the value of the custodianship. Simply put, the dynamic is that, per the criteria of the Curator, the image somehow represents the Culture.
That dynamic holds whether the culture is, for example, science, history, or art.
There is an argument that over time, these cultures “evolve”. In some examples of that evolution, science expands standing knowledge; history refines the acuity and perspective on standing evidence; and art tests innovation in established perception.
But it is quite common that curators both optimize the recognition of prior exemplars and propose how subsequent works meaningfully extend or change that recognition. In that way, a curator can be a “bridge” across the difference.
Meanwhile, changes can redefine the boundaries of how we evaluate a medium. Said differently: something atypical could come to be recognized as a known prior thing.
But by far, in art, such changes over time do not “redefine” a medium.
Just as photography was not a new form of painting, AI image generation is not a new form of photography.
THAT OBSCURE OBJECT OF DESIRE
It is very easy for an AI-generated image (right) to offer the same qualities of, for example, a Richard Estes super-realist painting (left), and for both of them to offer the same qualities of a camera-generated image. We have to remember, however, that those “qualities” are ones that are preselected as being important. What are they? Verifiable facts? Plausible narrative? Sensory impact?
We can see here that it is entirely possible for different types of media to meet the same selection criteria.
But there are three key factors of differentiation among (for example) painting, photography, and AI as image generators.
One is the human labor. We acknowledge execution; but we empathically sense performance. 12 minutes of prompting; 4 hours of shooting; six days of painting?
Another is the rendering instrument. We engage form; but we interpret presence. What created the observed effects: brushes; computer; camera?
And the third is the anticipated context of the image’s presentation. We acknowledge logic; but we desire preference. Documentary? Advertising? Fantasy? In any specific situation that we care about, why would we especially want a photograph? [Spoiler alert: photography’s role in non-fiction recording anchors its preferred role in “realist” depiction across several industries, but that defines a “use case”; it doesn’t define the medium.]
VALUE IN PHOTOGRAPHS
Appreciation is the term for “attributing value” — and with art, most appreciation is psychological.
We engage form; but we interpret presence.
We acknowledge logic; we desire preference.
We admire execution; we empathically sense performance.
In satisfying our selves psychologically and exercising our appreciation, we can grant curatorial and/or custodial roles to ourselves or to others. The choice we make is mainly a matter of how it supports the position we want to hold in our cultural and social relations, but also in our private world of personality and identity.
Some thinkers in the world of art are especially interested in the notion of how a work has “authenticity” based on its adherence to some definition of a medium. But the only consistent standard for authenticity is the degree of self-awareness that an audience (like a visual art viewer) has about its own requirements when the audience promotes its own selection of works.
In the practice of photography, this kind of difference is often evident in how an image maker categorizes their “personal” work as distinguished from their “commercial” or “commissioned” work. The values pursued in personal work will naturally align to the imagist’s interest in how to support their psychology, regardless of what market (if any) may also have demand that matches it.
Photography’s cultural “clergy” is a market-maker serving a full range of roles — priests, disciples, prophets, believers, preachers, and so on. In their domain, it is accepted that there is orthodoxy certifying how an image has value. It takes on a responsibility for admitting or converting aspirants that are either images or viewers. And it applies interpretation to set the boundary of how its qualifications are recognized.
But new experiences of images — not new types of image production per se, are what expands or replaces our existing boundaries of psychological differentiation. Of course it makes sense to attribute a new experience to whatever produced it. But as shown above, experience is strongly affected by interpretation, and interpretation is also strongly affected by predispositions that are independent of production methods and even of the producer.
SEEING THE LIGHT
What actually preceded AI as a digital imagemaker is CGI. But like CGI, AI image generation does not change photography as a practice, nor medium, nor category.
What AI changes about “photography” is our presumption of what is distinctly differentiating in our expectations of what photography can uniquely do. When expectations change, the decisions that lead to a demand for it or an application of it also change.
The use of a camera (or most specifically, an optical lens) to manipulate light does more to differentiate the “core” of photographic practice than any other single thing that immediately comes to mind. It is fair to ask the question, how far can camera-based manipulation of light go to generate a meaningful image? Probably as far as any musical instrument can go to generate a meaningful sonic work.
In photography this immediately embraces the entire spectrum of imagery from visual non-objective abstraction (bearing no representation of anything other than its own characteristics), to visual concreteness depicting something else through figuration. Thought of that way, there is no more of a limitation on what a photograph can look like than there is on any other medium of graphical effects. We can try to identify something other than that pictorial range, that we would argue is somehow a defining characteristic of photographs — but being able to do that would not change the medium; it would just change who cares.
Likewise, AI imaging just means that we’ve now covered chemicals, dyes, electrons, and pixels as of 2025. Beyond that, what is there? Implants for direct instrumental stimulation of brain cells?
As with most other art forms, an image maker’s attention may be focused on exploring how an audience interprets forms when other factors — including the environment, memory, and compositional strategies — are explicitly manipulated and incorporated as part of the presentation of images. Most of the interest in that will ride on how that is done — within the boundary of an image, externally around that boundary, or both. It is a direct interest in psychological differentiation of the experienced production. And this can correspond with expansion or modification of an audience’s appreciation — its terms for attributing value.
— MR
