Orkhan Mammadov

Orkhan Mammadov was born in 1990 in Ganja, Azerbaijan, and works as a new media artist and creative technologist. He taught himself graphic and web design by the age of ten and was competing internationally as a teenager. At eighteen he started working as a web and graphic designer, moved through design studios into art and creative director roles, and at twenty three put up his first public art installation. His education runs across computer science, visual communication design, and fine arts, and he moved from audiovisual and interactive work into machine learning once he saw what the two could do together. In 2019 he represented Azerbaijan at the Venice Biennale with Circular Repetition, his first AI driven installation, and the same research has since travelled to the Moscow International Biennale, Art Dubai, and Art Basel Miami Beach. His material is cultural heritage as data. He has processed well over a hundred thousand archival images of carpets, rugs, kilims, miniature paintings, and ornament from museums and libraries, and trained models on them to make new work that is recognisably of that tradition without copying any single object. He has had more than a dozen solo shows, and OpenAI licensed a work from his Visions series.

He builds datasets out of Islamic and Azerbaijani visual heritage, trains GAN and diffusion models on them, then pushes the output through procedural algorithms and particle simulation in Houdini to make moving carpets, miniatures, and ornament. The work lands as NFTs, as large immersive installations, and as physical or phygital data sculptures. Named projects include Muraqqa with its Data Miniatures, Neural Impressions, and The Process collections, plus Singularity in Heritage, Revival of Aesthetics, Carpetdiem, Visions, and Echoes of Alhambra.

• Representing Azerbaijan at the Venice Biennale in 2019 with Circular Repetition
• Singularity in Heritage with GAZELL.iO
• The Muraqqa project and its Data Miniatures and Neural Impressions collections
• Training on more than 150,000 archival images of carpets, kilims, and miniatures
• The Visions series, one work of which was licensed by OpenAI
• Carpetdiem Experiments

On the show