Rico Carranza, Eduardo; Huang, Sheng-Yang; Besems, Julian; Gao, Wanqi; (2023) (In)visible Cities: What Generative Algorithms Tell Us About Our Collective Memory Schema. In: Koh, Immanuel and Reinhardt, Dagmar and Makki, Mohammed and Khakhar, Mona and Bao, Nic, (eds.) HUMAN-CENTRIC - Proceedings of the 28th CAADRIA Conference. (pp. pp. 463-472). CAADRIA: Ahmedabad, India.

Preview

Text 264_.pdf - Accepted Version Download (961kB) | Preview

Abstract

The last decade has witnessed a turn in AI technologies working with differentiable neural network architectures learning the embedded functions between data points and performing generative operations synthesising unseen data. The move to a continuous and generative AI paradigm aligns with ideas in the field of cognition and psychology, where a growing body of authors are beginning to conceptualise memory and our representation of the past as a dynamic, malleable and ultimately generative field. So, how effective are generative algorithms in supporting and enabling this creative process of remembrance? To answer this research question, we propose an experiment on how the spatial movement and exploration of maps of real and imagined images can help our brain reconstruct its memories in a dynamic yet accurate manner. We develop an application allowing visitors to dynamically explore real and AI-generated images of a given site clustered by similarity in a virtual 3D space. Analysing visitor paths and observed images helps us understand visitors’ perspectives on real and AI-generated data such as an increased preference for synthetic images by visitors familiarised with the site. We conclude with recommendations on how to approach visitor experience in generative AI-powered applications for engagement with historical and archival data.

Download activity - last month

Export as JSONCSVXML

Download activity - last 12 months

Export as CSVXMLJSON

Downloads by country - last 12 months

Export as JSONCSVXML

Archive Staff Only

View Item