Lin and Huang explore how to capture AI models in a modern bestiary, in which they approach the models as unknown organisms. This will result in an interactive ‘field guide’ to be published in March 2024.
Artificial intelligence models are now present in many applications. The most familiar are ChatGPT, DALL-E and Midjourney. The working of AI models is often a mystery to users, and how we relate to AI is rapidly evolving. Lin’s and Huang’s project is in line with those current events.
‘As AI grow more complicated, new properties and characteristics are emerging to the extent that it is no longer apparent at the level of algorithms’, according to Lin and Huang. ‘Their influences permeate from the virtual world into the physical world and become peculiarly sophisticated so that they are at the borderline of being sentient. As humans, we are again faced with unknown semi-living beings so we need to develop new ways to understand them.’
To better understand AI models, Lin and Huang approach them as unknown organisms. They develop methods to categorise AI models and depict them as a contemporary bestiary. In doing so, they use methods originating from artistic and scientific research.
About Pei-Ying Lin and Yi-Ching Huang
Pei-Ying Lin (born in 1986, Taiwan) holds an MA in Design Interactions from the Royal College of Art, UK and a BSc of Life Science from National Tsing Hua University, Taiwan. She has exhibited her projects worldwide. The current exhibitions including Gene Cultures at MIT Museum in Boston and My Body, A Coral Reef? at the Rudolf Scharpf Gallery of the Wilhelm-Hack-Museum, Ludwigshafen am Rhein, Germany. Both exhibitions are devoted to the Virophilia project, which explores the possibilities of using viruses for culinary purposes. Pei-Ying has been working and living in Eindhoven, the Netherlands, since 2016.
Yi-Ching Huang (born in 1983, Taiwan) has been an Assistant Professor in the department of Industrial Design at TU/e in the Netherlands since 2020. She was awarded her PhD at National Taiwan University in 2018. Her current research is exploring human-AI co-learning for facilitating new ways of creativity and productivity. Specifically, she strives to design and develop toolkits for supporting designers to familiarise themselves with AI as design materials, and to play with data, computation and AI to solve diverse problems in the various contexts.
About the Mingler Scholarship
The Mingler Scholarship is intended for artists and scientists who want to launch a research project together. The scholarship is awarded by the Society of Arts and The Young Academy. The sum of 10,000 euros is made available by the Niemeijer Fund Foundation.
Image: Image generated by Midjourney AI. The Midjourney AI was asked to generate an inkblot image. Then Midjourney was asked to 'describe' the image. This description was used to generate another image: a bird. The bird is to some extent the imagery representation of what Midjourney has 'seen' from the inkblot image it generated.