Photography in the Light of AI-Generated Images

Generative Adversarial Networks are raising various concerns among photographers and cinematographers regarding the value of their own artwork. With the latest versions of products like Midjourney (i.e., v6) or OpenAI’s Sora, generated images are getting closer to what some may call technical perfection. But is perfection to be sought in photography? Walter Benjamin’s essay – The Work of Art in the Age of Mechanical Reproduction (1935) – seems to astutely approach even today’s debate on artificial intelligence versus photography. While Benjamin’s essay focuses more on photography as a reproduction of painting, it still leads us to a question we raise in this article: will the emergence of AI-generated images shine some new light on the document value of photography? If so, what are the stylistic features of photography that may recall the immediacy of a specific moment? Looking back at the work of some famous photographers (e.g. Bresson, Davidson) and comparing it with AI-generated results, we can further understand that the real value of photography relies on its unrepeatable character. This is being reflected through minor imperfections, either technical or compositional. Can this kind of imperfections become part of a movement that reevaluates form in photography?