Ryan Murdock, working under the name Adverb, is an AI artist and machine learning researcher whose early text-to-image experiments helped shape the aesthetics of AI-generated art. In 2021, alongside Katherine Crowson, he built the CLIP-guided image-generation notebooks BigSleep and VQGAN+CLIP, tools credited with pushing the aesthetics of CLIP-based art forward. He creates and presents this work under the name Latent Visions.
Their CLIP notebooks produced some of the first free and open text-to-image models, at a time before such tools became dominated by corporate platforms like DALL-E 2 and ChatGPT. Murdock has worked as a machine learning researcher at organizations including Adobe, Canva and Leonardo.AI, and held a residency at Stochastic Labs. In a 2021 interview for the podcast Interdependence, he discussed Latent Visions together with an idea he calls Promptism and his outlook on the future of AI art.
