Empowering end-users in co-designing AI: an AI literacy card-based toolkit for non-technical audiences
File(s)013_Smith_BCSHCI23.pdf (543.24 KB)
Published version
Author(s)
Smith, freya
Sadek, malak
Mougenot, Celine
Type
Conference Paper
Abstract
While there are growing calls to involve end-users in designing AI, the limited understanding of AI’s
capabilities among the general public hinders their effective participation in AI co-design. This study
aims to facilitate the participation of non-technical audiences in AI co-design through the
development of a card-based AI literacy toolkit. The toolkit was designed (1) to inform users about
AI by presenting 16 competencies extracted from the AI Literacy conceptual framework by Long and
Magerko (2020), and (2) to support idea generation and questioning about AI through "What if?"
prompts, as done by designers. The effects of the informative and generative toolkit were assessed
using a mixed methods approach. First, its impact on AI literacy was evaluated in a design task
where non-technical participants (N=50) generated questions about future AI systems. Results
showed a statistically significant increase in critical feedback, breadth, and frequency of AI-related
keywords in participants' responses after using the toolkit, indicating improved AI literacy. Second,
the effects of the toolkit were examined in a co-design workshop involving a group of six
participants, half of whom without an AI background. Post-workshop interviews revealed positive
effects on collaboration, including the development of a shared vision and common ground within
the group. Both studies highlight the potential of AI literacy in enhancing the participation of non-
technical audiences in co-designing AI with practitioners.
capabilities among the general public hinders their effective participation in AI co-design. This study
aims to facilitate the participation of non-technical audiences in AI co-design through the
development of a card-based AI literacy toolkit. The toolkit was designed (1) to inform users about
AI by presenting 16 competencies extracted from the AI Literacy conceptual framework by Long and
Magerko (2020), and (2) to support idea generation and questioning about AI through "What if?"
prompts, as done by designers. The effects of the informative and generative toolkit were assessed
using a mixed methods approach. First, its impact on AI literacy was evaluated in a design task
where non-technical participants (N=50) generated questions about future AI systems. Results
showed a statistically significant increase in critical feedback, breadth, and frequency of AI-related
keywords in participants' responses after using the toolkit, indicating improved AI literacy. Second,
the effects of the toolkit were examined in a co-design workshop involving a group of six
participants, half of whom without an AI background. Post-workshop interviews revealed positive
effects on collaboration, including the development of a shared vision and common ground within
the group. Both studies highlight the potential of AI literacy in enhancing the participation of non-
technical audiences in co-designing AI with practitioners.
Date Issued
2023-08
Date Acceptance
2023-07-16
Citation
36th International BCS Human-Computer Interaction Conference (BCS HCI 23), 2023, pp.13-22
ISSN
1477-9358
Start Page
13
End Page
22
Journal / Book Title
36th International BCS Human-Computer Interaction Conference (BCS HCI 23)
Copyright Statement
© Smith et al. Published by BCS Learning and Development Ltd. Proceedings of BCS HCI 2023, UK
License:
This work is licensed under a Creative Commons Attribution 4.0 Unported License. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/
License:
This work is licensed under a Creative Commons Attribution 4.0 Unported License. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/
License URL
Identifier
https://www.scienceopen.com/hosted-document?doi=10.14236/ewic/BCSHCI2023.2
Source
36th International BCS Human-Computer Interaction Conference
Publication Status
Published
Start Date
2023-08-28
Finish Date
2023-08-29
Coverage Spatial
York, UK
Date Publish Online
2023-08