This year’s photography exhibit at the San Diego County Fair excluded images that used AI’s text prompt feature. In short, these photographic artists were told, “These are not your pixels.” Some tweaking of images was permitted using AI (AI assisted), but not with text prompts (generative AI).
The fair began to exhibit photography in 1947. Technology has continued to change how the image was captured and how it was processed. These changes have generally been incorporated into the photographs exhibited at the fair.
Now the question is whether AI technology will be incorporated into the fair’s International Exhibition of Photography in ways that do more than tweak or assist the image making process. Will the exhibit include content created by an AI image generator model rather than solely from a camera?
It was nearly twenty years ago when the fair’s photography exhibit allowed digital art to be displayed in a room adjoining the main photographic exhibit. In those early days, pixels were disdained along with the emergent digital camera technology. Over time, the digital revolution succeeded and the fair’s photography exhibit became primarily digital. Yes, there are likely still photographs made with pinhole cameras without lenses, cyanotypes that expose objects to sunlight and onto material prepared with chemicals, and other moments in the history of photography that have leaned into changing technology.
But what about photographic-appearing images that are generated with AI technology? Or images that unite disparate technologies — hybrid images?
The fair’s Photographic Exhibit assembled a panel of artists who juried submissions and staff coordinators to discuss why this new technology would not be allowed within the hallowed halls of its photographic exhibit. At least, not in the coming year. The panel also included an attorney to explain the legal nuances of copyright.
That’s right. You read the previous paragraph correctly. The panel was concerned about moving too quickly with the generative aspect of AI image making. The main issues involved the fear of litigation and a method to assess how the images were created. Image generator platforms are generally trained on the images of other photographers; the AI generated images are not copyrightable since they lack human authorship; and the fair’s jurors would be unable to verify the content of the images submitted — are they the photographers, do they belong to other photographers and are they imagined, as it were, by a machine and not a human.
Consider the following image. It was created in MidJourney with the referenced text prompt. It is not copyrightable. The only human authorship are the words fed into the AI image generator.
One of the panelists, John Watts, articulated the opposition to including images with AI content in the fair’s photography exhibit. (His blogpost, with minor edits, was echoed by most of the panel discussants.)
“AI content creation is not the same as photographic creation. It is part photographic compositing but it is so much more. It is a machine’s interpretation of a scene, which it then further “creates” by adding elements to an image (and those elements more than likely do not even exist in the real world). That’s the complete antithesis of what I believe photography to be all about. In photography, you create your image in the camera – in a precise moment in both time and place. Call it what you will (AI Art?), but AI Content Creation in Photography – no camera or lens is involved!”
An example of the difficulty of translating this general principle (a camera and a lens) into what images are acceptable as prints was contested. Larry Vogel, a member of the audience, asked whether use of the generative fill tool was in fact creating content.
The contest rule states: Photoshop’s Generative Fill/Expand may be used as a minor corrective tool . . . when not significantly altering the primary photographic content.
However, Adobe’s explanation of this tool is important: Generative Fill is a generative AI tool powered by Adobe Firefly that enables you to add and remove content from images nondestructively. . .
What this means is that this Photoshop tool functions entirely using AI technology, adding pixels to the image that were not created with the camera.
When the fair coordinators were asked to clarify the rule, because it uses vague terms like, minor corrective tool and not significantly altering, they were unable to provide any clarity and stated this is a gray area. This lack of clarity will only continue to confuse the photographers who may enter the photo exhibition next year.
One panelist acknowledged that there were areas of ambiguity with this generative AI tool and required more thought about how much content change should be allowed.
The panel discussion shifted from what AI does, the pros and cons of its use, the fear of deepfakes, and the requirements of statewide and local rules about ownership of submitted images to several artist-panelists expressing their fascination with generative AI and how their artwork was energized by using it.
The panel faced a dilemma. Generative AI was a cultural and technological force that could not be avoided in the long run. And yet the fair’s photography coordinators felt the need to resist it. Their perception of unresolved legal problems and of challenges they would face in their current review process proved to be insurmountable for the near future.
I proposed a solution to this dilemma during the Q&A. The room in which the discussion was being held could become part of an educational forum akin to a university without walls. Next year, this room could be used to experiment and explore what generative AI in photography might be and to test how the problems staff perceived might be resolved. This room had been used for digital art for many years with the same sense of experimentation and exploration of what could and could not ultimately fit into the main photography exhibit.
The answer was no.
A solution offered by a coordinator was that photographers seeking to submit images with generative AI should find another venue. Why impose non-photographic issues on the fair’s photographic exhibit?
That plea for maintaining the integrity of photography within the contours of their rules makes sense for the fair. And so the question arises: Why have a panel discussion about what is outside the scope of the exhibit’s rules?
The arena for an energetic discussion about where generative AI images fits may well be outside the fair and other similar arts institutions. However, there are alternative venues that are venturing into this arena. The Expressions Gallery at the Escondido Arts Partnership will be exhibiting and holding a discussion on an exhibit titled, I am AI, Are You?
This essay is the first part of a two-part essay. The first part is how the San Diego County Fair’s International Photographic Exhibition confronted this dynamic new image making technology. The second essay will be about how several artists and photographers exhibit a range of images created in part or in whole with generative AI and to engage with their reflections about their vision and practice.
Joe Nalven is a San Diego-based digital artist. He is the author of Going Digital: The Practice and Vision of Digital Artists (Thompson, 2005). He has also written about the future of AI.
