Artificial Balad comes from an early 2020 StyleGAN project exploring Jeddah’s historic Al-Balad through generative image synthesis. Built from a dataset of architectural images, textures, facades, windows, doors, and urban details, the work uses machine learning to form an unstable visual memory of the old city.

The resulting images do not reconstruct Al-Balad literally. Instead, they produce partial architectures: buildings that feel familiar but cannot be located, surfaces that carry traces of coral stone, wood, shadow, and ornament, yet remain structurally uncertain. The work sits between archive and hallucination, where the machine preserves atmosphere and pattern while losing geography, scale, and lived context.

The project combines generated stills, latent-space walk videos, training-progress grids, and process videos. Together, these forms reveal how the model absorbs and reassembles the visual language of Al-Balad through computational systems. In this sense, Artificial Balad treats early AI image synthesis as a form of latent-space archaeology, where the city returns as memory, distortion, and architectural residue. The latent-space video was later shown at Misk, where the work entered an early public moment of engagement with AI-generated imagery in the Saudi art scene.

Nasser Alshemimry is a Saudi audio-visual artist whose alias, DesertF!sh, reflects his Najdi roots and life on the Red Sea coast. Based in Jeddah, his practice moves between sound, generative visuals, interactive systems, and computational image-making. Through his studio “Desert.Dream Audio Visual” he develops immersive works and provides technical direction for artists and cultural institutions, bridging the languages of composition, code, installation, and spatial experience. After initially serving the contemporary art scene as a consultant for technologically complex artworks, Nasser began producing standalone audiovisual installations that examine perception, memory, and the human encounter with machines. Since his debut at the Saudi Art Council’s 21,39 exhibition in 2019, his work has been presented by Misk Art Institute, Ithra, Noor Riyadh, the Royal Commission for AlUla, the Saudi Pavilion at Expo 2025 Osaka, and Diriyah Art Futures’ inaugural exhibition Art Must Be Artificial. Works such as Transcendence, Sympathetic Strings, Inner Light, Digital Anemone, and his early StyleGAN projects explore responsive environments, biomimicry, machine vision, and the shifting image of Saudi identity through computational systems. His current research revisits early AI image synthesis as a form of cultural mis-recognition and latent-space archaeology. Through images of Al-Balad, AlUla, and Saudi material culture, he treats generative systems not as neutral tools, but as unstable archives: machines that remember, distort, and reassemble cultural forms. His work seeks a precise balance between technical experimentation and poetic residue, where sound, gesture, image, and heritage become signals within a larger field of transformation.