prototyping/diffracting AI

Scurto, Caramiaux, Bevilacqua, 2019—2021
research-action-creation method

A diffractive method for machine learning prototyping, to resist normative framings of AI. I propose to craft machine learning, using it as creative material in art and design projects, involving diverse bodies and beings. Reflecting on my experience with this practice in interdisciplinary contexts, I identify five socio-technical conditions that enables to interfere design and engineering in machine learning prototyping: situational whole, small data, shallow model, learnable algorithm, and somaesthetic behaviour. I suggest that these interferences are vital in reshaping practices and discourses in AI.

Year
2019—2021
Credits
The project was developed with Baptiste Caramiaux and Frédéric Bevilacqua, in collaboration with the Reflective Interaction group of EnsadLab, the HCI Sorbonne group of ISIR, and the ISMM group of IRCAM, in the context of a postdoctoral fellowship program at Faculty of Medicine, Sorbonne Université.
Publications
Paper at DIS (2021)

This slideshow requires JavaScript.