Talk: AI and Images in Art

Geoff Davis introduces a discussion of AI image making by distinguishing generative algorithms from systems intended to mimic human responses. He considers accessible tools, creative enjoyment, image overproduction and the uncertain relationship between popular platforms and art institutions. Luba Elliott surveys developments from DeepDream and style transfer through generative adversarial networks to text driven image generation. Her examples emphasise conceptual choices, distinctive aesthetics, training data and uses of AI beyond image production. Anna Ridler explains how photographing and labelling ten thousand tulips makes human labour and judgement visible within machine learning. Myriad presents the dataset itself, while Mosaic Virus links generated flowers to Bitcoin prices, connecting speculative value with historical tulip mania. She compares her planning and relinquishing of control with environmental art. Mark Webster describes Hypertype, which transforms sentiment analysis data into typographic compositions rather than treating automated emotional classifications as reliable descriptions of human experience. Patrick Lichty presents an imagined exhibition catalogue of absent, synthetic artists, using generated studios and statements to investigate institutional conventions and social anxieties. The concluding discussion asks how audiences develop visual literacy, how curators communicate artistic processes, and how changing technologies affect exhibition formats, historical interpretation and collaboration between computational tools and artists.

AI Generated Summary