TY - GEN
T1 - Making AI Understandable by Making it Tangible
T2 - 34th Australian Computer-Human Interaction Conference: Connected Creativity, OzCHI 2022
AU - Ghajargar, Maliheh
AU - Bardzell, Jeffrey
N1 - Publisher Copyright:
© 2022 Owner/Author.
PY - 2022/11/29
Y1 - 2022/11/29
N2 - The embodiment of Artificial Intelligence (AI) in everyday use products is raising challenges and opportunities for HCI and design research, such as human understandings of AI's functions and states, passing back and forth of control, ethics, transparency and user experience. To respond to these challenges, HCI and design researchers have been working on research areas such as Explainable AI (XAI); Fairness, Accountability, and Transparency (FAccT); Human-Centered AI, as well as guidelines for Human-AI interaction design. Consequently, the interest in studying interaction modalities and their contributions to understandable and transparent AI has been growing. However, the tangible and embodied modality of interaction and more broadly studies of the forms of such everyday use products are relatively under-explored. This paper builds upon a larger project on designing graspable AI and it introduces a series of concept cards that aim to aid design researchers' creative and critical exploration of tangible and understandable AI space. We conducted a user study in two parts of online sessions and semi-structured interviews and found out that to envision physicality and tangible interaction with AI felt challenging and "too abstract". Even so, the act of creative exploration of that space not only supported the participants to gain new design perspectives, but also supported them to go beyond anthropomorphic forms of AI.
AB - The embodiment of Artificial Intelligence (AI) in everyday use products is raising challenges and opportunities for HCI and design research, such as human understandings of AI's functions and states, passing back and forth of control, ethics, transparency and user experience. To respond to these challenges, HCI and design researchers have been working on research areas such as Explainable AI (XAI); Fairness, Accountability, and Transparency (FAccT); Human-Centered AI, as well as guidelines for Human-AI interaction design. Consequently, the interest in studying interaction modalities and their contributions to understandable and transparent AI has been growing. However, the tangible and embodied modality of interaction and more broadly studies of the forms of such everyday use products are relatively under-explored. This paper builds upon a larger project on designing graspable AI and it introduces a series of concept cards that aim to aid design researchers' creative and critical exploration of tangible and understandable AI space. We conducted a user study in two parts of online sessions and semi-structured interviews and found out that to envision physicality and tangible interaction with AI felt challenging and "too abstract". Even so, the act of creative exploration of that space not only supported the participants to gain new design perspectives, but also supported them to go beyond anthropomorphic forms of AI.
UR - http://www.scopus.com/inward/record.url?scp=85153363252&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85153363252&partnerID=8YFLogxK
U2 - 10.1145/3572921.3572942
DO - 10.1145/3572921.3572942
M3 - Conference contribution
AN - SCOPUS:85153363252
T3 - ACM International Conference Proceeding Series
SP - 74
EP - 80
BT - Proceedings of the 34th Australian Computer-Human Interaction Conference
A2 - Martin, Charles
A2 - Mckay, Dana
A2 - Rogerson, Melissa
A2 - Cumbo, Bronwyn
A2 - Wadley, Greg
A2 - Sweetser, Penny
A2 - Taylor, Jennyfer Lawrence
A2 - Hespanhol, Luke
A2 - Tsimeris, Jess
A2 - Xi, Mingze
A2 - Turner, Jane
A2 - Yoo, Soojeong
A2 - Cooper, Ned
A2 - Rahman, Jessica
A2 - Andres, Josh
A2 - Pillai, Ajit G.
A2 - Kutay, Cat
PB - Association for Computing Machinery
Y2 - 29 November 2022 through 2 December 2022
ER -