WORKSHOP //
Generation of AI Images
Boris Eldagsen
The intensive workshop focused on Promptology—the discipline that studies the optimization of prompts to make interaction with generative models effective, safe, and strategically useful—was led by German multimedia artist and educator Boris Eldagsen. A pioneer of Promptography, Eldagsen is internationally renowned for turning down the Sony World Photography Award after winning it with an AI-generated work, sparking a global debate on the boundary between photography and generative art.
The course was conducted in English through an intensive three-hour session, specifically designed to explore advanced visual generation through an interactive methodology. The program was tailored to the specific and pre-identified needs of the participants, a group composed of creatives from diverse backgrounds, including photographers, artists, designers, photo editors, and art directors, with less than fifty hours of previous experience in image generation. The main focus of the event was on image generation from visual inputs (Image-to-Image) using Midjourney, and included dynamic "prompt battles" on the Zoom platform, a practical and competitive exercise aimed at consolidating the skills acquired in the field.
During the hours of activity, the most advanced digital visual panorama workflows were analyzed and applied. Experimentation with Midjourney provided insight into the evolution of the Blend command and the Image-Prompt, Describe, and Remix techniques. Post-processing management was addressed through Inpainting and Outpainting processes, including tools such as Generative Fill, Expand, Vary Region, and Zoom, along with an introduction to Faceswap and the combined use of key parameters such as Personalization, Character Reference, and Style Reference.
The software integration included the use of Magnific AI for Style Transfer and Relight, the transfer of lighting conditions from one image to another, alongside Photoshop and Adobe Firefly for managing structures and styles. Participants also saw a comparative study and benchmarking of the leading upscalers on the market, including Gigapixel, Magnific, Upscayl, and Krea. Finally, a critical demonstration led by Eldagsen highlighted the potential and technical and expressive advantages that Stable Diffusion offers compared to Midjourney.
This initiative explored the convergence of photography, art, and contemporary languages, concretely demonstrating how AI can be used by visual professionals not as a substitute for creativity, but as an extension of intellect and visual design.