2023-05-25 22:53:58
In recent weeks there have been many discussions about images generated by artificial intelligence (AI). Software like DALL-E 2, Midjourney and Stable Diffusion are at the epicenter of criticism due to the use that is being given to them. Can a prestigious photography award be given to an image taken without a camera? Is it ethical for an NGO to use images generated by this software to denounce police violence? Can the AI train and use copyrighted photographs without permission? These are the questions that professionals put on the table.
Within the framework of PHotoEspaña 2023 (from May 31 to September 3), elDiario.es talks with the editor and teacher Gonzalo Golpe (Madrid, 1975). On June 5, La Fábrica publishing house publishes When the wind, one of the first books in the world in which AI has been used as a tool to generate images. In this pioneering work, the author of the work, Gonzalo Golpe, visually narrates a dystopian fable –created with the assistance of Martín Bollati and DALL-E 2 (AI developed by the company Open AI, developer of ChatGPT and participated by Microsoft )– which invites us to observe the fears of our society and to question how we use technology. “AI is not intelligent in human terms. The AI is not imaginative ”, assures Golpe.
On April 13, for the first time in history, an image created using AI won a renowned photographic award, the 2023 Sony World Photography Awards in the “Creativity” category. However, the German artist who presented it, Boris Eldagsen, forfeited the $5,000 prize because, as he stated in his speech, “AI images and photography should not compete in an award like this. They are different. AI is not photography.” Eldagsen carried out this performative act in order to show that contests of this type are not prepared for the current boom that AI is experiencing. “Not understanding what it means to call these synthetic images ‘photographs’ is dangerous,” says Gonzalo Golpe.
Two weeks later, on May 2, the Norwegian division of Amnesty International published a report on police violence during the 2021 protests in Colombia. Despite the fact that the country’s National Strike was documented thanks to photojournalism, the organization in defense of human rights accompanied this campaign with several images generated by AI.
The organization reported this at the bottom of the images. However, the use of these was highly criticized on social networks, which led Amnesty International to withdraw the synthetic images. According to Amnesty, the decision not to use photographs by photojournalists in the first instance was taken to protect the identity of the people who suffered police brutality.
What sense does it make for AI to reach a place where humans have already arrived? The noema of photography –theorized in Roland Barthes’ Camera Lucida (1980)– is: “This has been”. In the case of these non-photographs, fabricated by AI models that create images from text, the Barthian “this has been” is meaningless. “The images they generate are the product of a training devoid of critical and moral sense. These programs are heirs and representatives of a visual system based on privilege, discrimination, concealment and derision. That is why it takes the human as masculine, young and white”, analyzes Golpe.
The software that generates this type of content has been trained and feeds on photographs taken by human beings that, on many occasions, are copyrighted. Which is why Getty Images, one of the most popular photo agencies, has sued Stability IA, the company that owns the open source generative AI Stable Diffusion, which, according to Getty, is infringing the intellectual property of Millions of copyrighted images. Illegality that photography professionals from all over the world are also suffering individually.
As Newtral uncovered, in the Spanish framework, at least the work of 55 photographers illegally feeds LAION, the dataset that trained Stable Diffusion. “I think it is necessary, urgent and even mandatory to establish legislative controls that make it necessary to define and identify these images as well as control their use, especially in certain areas”, proposes Golpe.
“These models have been educated based on biases, they respond to segregationist interests that seek to isolate differences in order to execute control and repression policies on them. They are instruments. Not only of human progress, but also of degeneration and the loss of humanity. They were already born addicted. Its current development is being supported by a vulnerable workforce that is exploited and is also disconnected from the agenda of objectives managed by the companies that own their patents. It is alarming”, warns Gonzalo Golpe.
Quoting Susan Sontag, “the photograph in a book is obviously the image of an image.” In the case of When the wind (La Fábrica, 2023) –the first book published in Spain by a human being with the assistance of an image-generating AI– we are not dealing with a photobook. This visual narrative, argued through the texts (prompts) that Gonzalo Golpe typed in DALL-E 2, is an author’s work. A post-apocalyptic science fiction fable that contains an environmental message. In When the wind a synesthetic visual discourse is experienced in a context of sadness. An extraordinary semiotic (and sinister) work with a melancholic halo in the fluorine tones that characterize this type of model.
“I use this technology to execute a visual composition […] While I was creating, editing and sequencing the images, what I was doing was projecting my spirit in a kind of visual symphony, as if I were humming the images”, says Golpe.
Gonzalo Golpe is not a photographer. He has spent a lifetime working in the image environment. His gaze and his rhythm can be found among the pages of important photobooks signed by artists such as Cristina de Middel, Joan Fontcuberta, Alberto García-Alix and Alessandra Sanguinetti. “I am an intermediary between the work and its author, between the community and the works and also within the system”, says the editor. Golpe has published dozens of texts, the most recent being the essays Curso Discurso and Atestado (Cabeza de Chorlito, 2020 and 2021, respectively). Along with Marina Meyer and Sara Arroyo, he is a member of the educational platform Ensambles, which is currently holding debate forums on AI, and since 2014 he has been part of the graphic collective La Troupe. When the wind is Golpe’s first author’s work. “I come from the poetic word, from working obsessively between what mediates between verbal and visual language, so this technology has been a kind of epiphany for me. Now I can name what I see inside and manifest it, materialize it visually”, he confesses.
“I have always felt the desire to hybridize my voice, to go beyond the word and make the image something more than an eloquent sound that can take a visual form in the mind. Before, I used my own and archive photographs to expand my writing, but it seemed extremely violent when people appeared, which is basically what interests me. Making fiction with people I don’t know, using their image to tell my stories always seemed to me an act of violence, a usurpation. Now at least I know that those people I recreate never existed and that frees me ”, analyzes Golpe.
For several years, Golpe has been giving shape to an investigation about the origin of language. It is an essay called Verba Volant, a conclusion to his search. “I consider myself a distance manager, both in the work of editor and in that of a teacher or communicator,” he says.
Photography did not replace painting. The cinema did not replace the theater. Video did not replace cinema. Television did not replace radio. Why fear that AI will replace photography?
Etymologically, the word photography comes from the Greek where “photo” means light and “graphy” means writing. If photography is the art of writing with light, generating an image through text that is sent to an AI, what is it? “They are synthetic images generated computationally from an algorithm. Nothing they show is real”, replies Golpe. With the popularization of AI, new terms have been coined, such as syntograhpy (synthography) and promtography, which have not yet caught on in society. Both refer to the synthesis and to the texts (prompts), respectively, that the human being has to enter for the AI to generate images. That is the process: a computational calculation based on a propositional proposal.
“In artistic terms, I see it as one more tool, I don’t think it’s necessary to distinguish them from other visual technologies, but when these images are taken as photographs, that is, as captures of ‘reality’, their use should be audited by specialists and referenced. . In any case, fake news is just one of the dishonest uses that I can think of, there are many others and probably more disturbing and harmful”, Gonzalo Golpe settles.
#photography #imaginative