Embodied Voice and AI: a Techno-Social System in Miniature

One of the original motivations for the practice-based research presented in this paper is rooted in the premise that embodied knowledge is relevant to the field of AI, not only in terms of application to design but also in terms of critical practice. A key proponent for integrating embodied knowledge into Human-Computer Interaction, Schiphorst (2009: 225) reckons “embodiment in the context of designing for technology” matters, and “concepts such as embodied computing and embodied interaction” require “design strategies that take advantage of our senses, accessing a richer and more fully articulated form of human being.” In terms of critical practice, feminist technoscience scholars such as Leurs (2017: 137) regard the researcher as an embodied subject, and stress the importance of materiality, by emphasizing the relational entanglement of bodies and infrastructures in the creation of knowledge. In her seminal book, *Artificial Knowing: Gender and the Thinking Machine*, Adam (2006) argues that “the body plays a crucial role in the making of knowledge.” The aforementioned scholars draw on feminist research to counter the dominant data science discourses that still equate intelligence with rationalistic problem solving, in which claims for objectivity and universal truths propagate what feminists call a “view from nowhere.” Anthropologist Suchman (2007), in her critique of the “disembodied intelligence in AI,” also points out that “feminist theorists have extensively documented the subordination, if not erasure, of the body within the Western philosophical canon,” an erasure Suchman claims extends to the field of AI and Robotics.

While addressing the wide literature and theory in both application design and feminist critical practices lies outside the scope of this paper, these are the kinds of perspectives that provided the original motivations for embarking on our research project. Continued research will further explore the implications of these theories for this practice-based setting. What we can do here is to acknowledge, as Suchman (2007) posits, following Haraway (1997: 11), that “technologies [...] are forms of materialised figuration.” They are brought into existence by assemblages of humans and machines, bringing meaning and values, including dominant assumptions and biases that subordinate, into the process of new knowledge creation. We believe this calls for more critical introspection of one’s own practice following Leurs (2017: 145), who states, “it is through being critically self-reflexive about one’s positionality, locating research practices within wider power relations and structures, that we can begin to destabilise the normalised politics of knowledge production.” Therefore, in presenting the framework below, we aim to also reflect on the positionality and the ethics of our research, in part through revealing the individuals and the corresponding assemblages in the development of tools and processes.

• Organisation