BORIS ELDAGSEN: "THE PHOTOGRAPHY OF THE FUTURE IS WHAT AI LEAVES BEHIND"

Boris Eldagsen: Good, because it means that I have thought ahead and seized an opportunity to initiate a worldwide discussion that was overdue for photography, art, and society at a time of historical upheaval and technical disruption. In retrospect, this will have greater social significance than painted water lilies.

BE: The water lilies have not led to a global discussion about the future of authentic photos and how we can still assure ourselves of facts in democracies. Any information we see on screens can be fake. Does this mean that we are drowning in “alternative facts” and can no longer even agree on a factual basis that we can discuss democratically?

BE: None of the responses surprised me. Both the AI haters and the AI worshipers were predictable. I see myself as a realist who tries to make the advantages and disadvantages transparent. I love both photography and AI. Both are great tools for my artistic vision.

BE: Related to images and art: Haters think that AI is dumbing down the world; it is evil because it is not only unpredictable but also theft. It doesn’t require any skills, is not creative, and is soulless. Simultaneously, haters fear that AI is going to replace them, making creative jobs and artists superfluous. They overlook that it is a skill to work professionally with AI and set up creative workflows in which the human is the director, using his or her professional knowledge.

AI worshippers think it is the democratization of art. They love that everybody, including AI itself, can now be an artist. However, their definition of art is very basic. Often, it just means "pretty pictures", for example, Kitsch. I am old enough to remember that the camera and computers also promised the democratization of art and that this utopia was also nonsense.

For me, art is an expression of the human condition. To achieve this, you need to delve within yourself, becoming conscious of the world inside you and how it interacts with the world around you. Not everybody is able or willing to undertake this journey. And AI? It desires nothing! It has no intention and no consciousness.

BE: AI is a knowledge booster. I work with my knowledge and experience. The stronger my background in image making is, and the more I know about the technical aspects of AI, the stronger my lead becomes. The collaboration has three steps: prompting, generating, and evaluating. The more informed the prompting and evaluation, the better the outcome.

When collaborating with AI, my role as an artist changes: from being a solo singer to becoming a conductor. A conductor needs to make sense of the gigantic, anonymous choir that the training data represents.

BE: Unfortunately, many jobs will disappear, especially those that are generic. Others will evolve: yesterday's professionals will become tomorrow's quality control. It is important to understand how AI works in order to actively participate in reshaping professions.

BE: My rejection of the prize has indeed sparked the necessary discussion about the nature and future of photography, art, and human creativity, as well as the issue of disinformation. In the first few weeks following the rejection, I was the one still pointing out the new problems to many journalists, but soon new podcasts and journalist teams specializing in AI were emerging everywhere.

Many people have asked me about the meaning and message of THE ELECTRICIAN I have always answered that the question "What is the artist trying to tell us?" is mundane. I do not aim to promote a political message through art, as many other contemporary artists do. I am not interested in addressing a specific political issue but rather in exploring the timeless human condition that underlies these contemporary problems. If I wanted to do that, I would write an editorial for a newspaper or attend a demonstration.

For me, a good work of art serves as a catalyst for an inward journey. The question that viewers of a work of art should ask themselves is: "What thoughts, feelings, or memories does the work evoke in me? Am I drawn to it or repelled by it? Why?" This is why I also reject any interpretation of my work. Providing interpretation would deny the viewer the opportunity to experience the work without bias. Instead, they would only seek visual confirmation of my words. Visual art should never be directly translatable into words; otherwise, its essence would be lost, and one could simply hang a text on the wall.

BE: I first encountered the term AI in the old sci-fi series I watched as a child, such as Star Trek and Battlestar Galactica. Science fiction was always filled with it. Sometimes AI was portrayed as a tool, and other times as an evil entity like HAL 9000 in Kubrick’s 2001: A SPACE ODYSSEY.

In 2014/15, I became aware that an AI called Deep Mind could generate images using AI. The AI was trained with GAN (Generative Adversarial Network), which consists of two components that improve each other. For example, one component acts as the producer, generating human faces, while the other serves as quality control, filtering out faces with too many eyes, ears, or noses. If one component becomes too dominant, the entire model collapses.

For nearly a decade, I have been keeping an eye on the GAN scene, but I never worked with it myself. I wasn't a nerd who coded; I was a creative individual who captured peculiar images at night.

BE: As a child, I felt awe and anticipation that I would experience all of this. In 2001, I was 31 years old. I wasn't in a spaceship, but in a new economy startup, watching the news of September 11.

BE: With the shift from GAN to the diffusion model in 2021, I foresaw a significant change on the horizon. Suddenly, one no longer needed to be a nerd to work with AI. Through mainstream platforms like DALL-E or Midjourney, one could achieve results that were becoming more and more photographic. My "wow" moment came in July 2022, and I was the first in Germany to deliver lectures on this topic (Photopia Messe Hamburg / September 2022) and offer workshops (starting in December 2022 with DGPh).

Following my presentation at Photopia, the audience, consisting of professional photographers and stock photo providers, was shocked by the potential impact that AI would have on their business.

BE: I was amazed that with text-prompting, I could work purely from my imagination. As a photographer who was never interested in a documentary approach, I had to transform what was in front of my lens into a timeless symbolic image. Now, I could do this without compromises. I could work like a painter and create images that looked photographic. These were images that I could have never captured through photography alone, without a movie set and film team.

BE: I have never been afraid of AI. I have always described myself as an artist who has worked with photography, video, and installation. Now, I work with AI. I embrace AI because I see it as the tool I have always been waiting for without knowing.

As an artist, AI is a liberation from material constraints. I can create images purely from my imagination. I utilize the knowledge and experience I have gained over 30 years in prompting and evaluating. It's cool and exciting. The last time I felt this way was at the beginning of my studies.

However, as a citizen of a democratic society, I see the gigantic potential of AI in spreading disinformation. That concerns me. For the past 1.5 years, I have been a member of the AI working group of the German Photo Council (the umbrella organization of over 30 professional photographer associations). We are in direct contact with Germany's picture editors to discuss what we can all do to ensure more transparency.

BE: It is not the photo artists who reject AI. I have had many curious colleagues in my workshops who hold professorships. It is the commercial photographers who rightly fear that they will become redundant. This is where the real issue lies. Any form of photography that is not documentary will be replaced by AI.

BE: The future of photography will be defined by AI. What will remain of photography is documentary photography. Only when authenticity is needed will we photograph. Everything else will be generated.

BE: In 2022, many Instagram users used the hashtag #AIphotography. I think it's sloppy thinking to see AI as the logical development of a medium. Photography is done with light and requires going out into the world and usually interacting with what's in front of the camera. You experience ‘a’ world. The journey in promptography goes inward. I can generate AI images in a dark basement; all I need is power, WiFi, a laptop, or a smartphone.

I need to interact with myself. To distinguish between photographs and AI-generated images, it is helpful to have a new term. I prefer "promptography" because it goes beyond the photo bubble.

"Promptography" can look like photography but also like drawing, painting, and 3D graphics. Promptography can be text, moving images, or sound. The creative workflow always starts with a prompt.

BE: If you refer to Fontcuberta’s THE MANIFESTO-DECALOGUE OF POST-PHOTOGRAPHY, this is more of a media-theoretical description of the changing uses and tasks of photography. He is not talking about photography as art.

There have been so many post-whatever terms in my lifetime (e.g., post-modernism, post-gender, post-race, post-feminist, post-colonial), that it is important to have a look at what 'postness' actually means. I support Nishant Shah’s definition that questions "have either been resolved or have been arbitrated enough for us to move on." But is that really the case with photography? And what kind of photography are we talking about? The possible applications are as varied as the text as a medium.

I don't really care about the label post-photography. But what I can say about the future of photography is that it will not be determined by the medium itself, but by AI. The photography of the future is what AI leaves behind: documentaries. I see an irony in history here. When photography took jobs away from painting, painting was able to develop away from depiction and towards 20th-century painting. Now AI is taking jobs away from photography, and what's left is depiction.

BE: I am happy to speak about the structural workflow. In the past, I mostly started with text prompts, then I used the resulting images as material for Midjourney’s blend, which fuses 2 to 5 images to a new one. The blend results became material for image prompts, which combine a reference image with a text prompt.

When I saw a promising image, I continued with inpainting (erasing parts of the image and using a new text prompt to generate this part again) or outpainting (expanding the canvas, adding pixels, and telling the AI with text prompts what to fill in the new space).

Today I often use my older AI images to start with. I use them to transfer the style; I also change images while upscaling (enlarging the pixel size). Upscaling has become a creative discipline of its own, offering additional text prompts and a hallucination scale. I also use 'nudify' apps (originally made to undress figures) for inpainting/outpainting, as they are less censored.

Photoshop Beta recently added an inpainting/outpainting option that uses image references and is very helpful in the process. But each work has a different workflow: Generally speaking, I am always experimenting with new approaches and tools.

BE: You need to have a professional background in any image-making. It will help you to create informed text prompts, as the majority of possible text prompt elements refer to your knowledge of a certain medium, its history and technology, your knowledge of composition, the use of colors and light, and the use of visual references and genres.

A professional background will also help you in evaluating the generated images and refining the prompts. I have staged photos and videos my whole life, thus I know how to compose an image from scratch. It also helps that I have always collaborated with other artists. You need to combine the strengths of the team members; otherwise, you will fail. Collaborating with AI is similar. You need to have a sound knowledge of the setup and the features of the available technology.

Only when you know the strengths and weaknesses of the current AI platforms will you be able to combine skills in a meaningful way and be the leader of this collaboration between man and machine.

BE: I see many comparisons in the creative workflow to photography, drawing, and painting. The main difference is my role: As an artist working with AI, I stop being the solo singer and become the conductor of an anonymous choir: The training data of the AI.

BE: PSEUDOMNESIA III was a creative experiment in the summer of 2023. I was invited to show at Marion Gallery Panama, one of the leading commercial galleries of abstract art in Latin America. I wanted to find out if I could create a fusion of the histories of abstract art and photography. I succeeded. The series has become a classic AI work and is shown in museums and festivals around the world.

HUNGER was the first work I created with Stable Diffusion, shortly after its official release in the summer of 2022. I wanted to see how far I could push the AI. To produce crazy body mutations, I used text prompts like "a worm is giving birth to a wild boar is giving birth to a woman is giving birth to a man who is vomiting a bird who is eating the worm". My social media followers hated it; they found it revolting. I said, "Stop being babies, I am only posting harmless images on meta; otherwise, I will be banned immediately".

BE: You feel it when an image is finished. I learned how to choose in photography.

BE: I have a psychological approach. I am interested in the workings of the human mind. My message is a very old one: travel inside and become more conscious of who you are. The ancient Greeks called it Γνῶθι σαυτόν ('know thyself') and attributed it to the oracle in Delphi.

So next time, when you are visiting an exhibition, don’t ask yourself, 'What does the artist want to tell me?' Ask yourself what the artwork triggers in you: What emotions, thoughts, and feelings?

BE: There are colleagues whose work I love, such as AI.S.A.M., Ben Millar Cole, LeMoon, Forbidden Toys or Placenta Shake. But they don’t inspire me. My inspiration is the human mind, as seen in psychology and philosophy.

BE: No particular artists influenced my AI work. I am also not working with style references or reference images of other artists. But I have my favorites in the history of art, like Rembrandt, Hieronymus Bosch, Rene Magritte, Roger Ballen. Or filmmakers such as Peter Greenaway, Alejandro Jodorowsky, Quentin Dupieux, or Bertrand Mandico.

BE: I don’t think that I will create too much work to digest. Since I am critical in my selection and take my time with the official release of work, nobody will be overdosing on it. However, you can expect 3-4 bodies of work per year.

BE: Soon, it will be normal to use AI as an artistic tool. My advice is to identify your unique skills and knowledge and use them for prompting. Remember that AI is a knowledge booster. Experiment and create your individual workflow. This is an exciting time for creatives. I truly believe that ‘Augmented Creativity’ is possible. By identifying the unique strengths of AI creativity, human creativity will become more focused on reaching a new level.