Digital Commons@RISD Home > Division of Liberal Arts > Contemporary Aesthetics (Journal Archive) > Special Volumes
Abstract
In generative AI, text-to-image prompting has gained popularity, which calls for image theoretical attention. This article evokes critical image theory, developed by Norman Bryson and W. J. Thomas Mitchell in the 1980s and 1990s, which offers important, relevant nuances when analyzing contemporary AI image-tools. By investigating specific cases of text-to-image AI through this theoretical filter, the article demonstrates how certain aesthetic ideologies—especially perceptualism, the idea of natural images, and splitting and hierarchization of text and image—govern generative AI.