Refik Anadol is a Turkish-born, Los Angeles-based media artist who has become one of the most visible figures in AI art, known for turning enormous datasets into immersive, moving installations he calls “data paintings” and “data sculptures.” He traces the approach to a 2016 residency in Google’s Artists and Machine Intelligence program, where, in his own words, he “coined the terms ‘AI Data Painting’ and ‘AI Data Sculpture’” out of the question “if a machine can learn, can it also dream? Can it hallucinate?”
His best-known work, “Unsupervised,” ran at the Museum of Modern Art in New York from November 19, 2022, to October 29, 2023. For it, his studio processed 138,151 images from MoMA’s digitized collection through NVIDIA’s StyleGAN2 ADA model, creating embeddings in 1,024 dimensions, and asked, in Anadol’s framing, “What would a machine mind dream of after ‘seeing’ the vast collection of The Museum of Modern Art?” Left running unsupervised, the model continuously generates new colors, forms, and motion derived from the archive, projected at architectural scale. The piece sits within his larger multi-year “Machine Hallucinations” project.
Anadol’s work has also drawn criticism over how much is genuine novelty versus spectacle, making him a useful focal point for debates about what AI art is.
Why business readers should care: Anadol turned generative models into blue-chip museum and commercial commissions, demonstrating an enterprise-scale market for AI-driven visual experiences well beyond the single-image generators most people first encountered.