Aesthetics and DALL-E 1-3: Bibliometric Analysis of Papers Published in 2021–2024

The aesthetic aspect of the outputs of image-generative AI models can be considered elementary arbitrary criteria when evaluating output quality and model effectiveness. We processed a bibliometric analysis of a total of 163 papers that were published between 2021 and April 2024 on the topic of aesthetics and the DALL-E 1-3 text-to-image models. The structure of publishers, journals, proceedings, and archives is described using Google Scholar and the Web of Science data. Bibliometric indexes regarding the representation of the publishing countries and the number of authorships, including the first authorships, are presented. Statistics on the number of pages and citations, including the citation timeline, are also involved. The thematic focus of studies shows the concentration of words such as “art” and “design.” Our study is most likely the first bibliometric analysis on the given topic and can serve as a reference for bibliometric material on the given topic.