Petra Jääskeläinen's 80% seminar - More-than-Human Re-Imaginaries As an Approach for Re-Designing More-than-Human Creative AI

Opponent: Laura Forlano.

Title

More-than-Human Re-Imaginaries As an Approach for Re-Designing More-than-Human Creative AI

Abstract

This doctoral dissertation explores the question of how more-than-human (MTH) commitments could be brought into the design of Creative AI technology. The work is theoretically aligned with more-than-human (feminist posthumanities), and frames MTH design as a design paradigm that is emerging as an alternative to human-centered design approaches in HCI. The central contribution lies in the design methodology of more-than-human re-imaginaries, which aims at re-configuring the design practices of Creative AI on an ontological level by examining the imaginaries of Creative AI through their socio-material manifestations in aesthetic design materiality. More concretely, the dissertation introduces three critiques of current design practices in Creative AI (1) rationalist conceptualization of ecological relations in the design of Creative AI, 2) human-centred conceptualization of creativity in the design of Creative AI, and 3) values of self-interest and human-centered design of Creative AI) and describes three design pathways for MTH design of Creative AI (1) MTH onto-epistemology, 2) MTH aesthetics, and MTH data practices). For example, the MTH onto-epistemology argues for the need for finding knowledge representations that align with MTH commitments, such as more-than-human conceptualization of creativity (which is introduced in this dissertation). More-than-human aesthetics encompasses two areas: aesthetics of interaction and materiality, as well as visual aesthetics, which work as material vehicles to convey the onto-epistemologies discussed earlier. Lastly, more-than-human data practices are brought forth as the third pathway and leverage point to MTH of design of Creative AI, encompassing a reflective framework that can help designers to engage with questions related to MTH data practices and ethics, as well as brings forth three concepts that can be utilized in the analysis of MTH data practices in the context of Creative AI: cultural representations’ normativity, constituency of the representation, and situatedness of the representation.