This research paper provides insights into how children (aged 7-13) value and design AI interfaces, specifically teachable machines. The study involved a co-design session where children and adults participated in AI problem formulation activities to imagine their own teachable machines. The findings reveal that children’s proposed ideas require advanced system intelligence, such as emotion detection and understanding the user’s social relationships. The children also cared about family and expected the machines to understand their social context before making decisions. This study has implications for the design of future technologies, particularly in the field of Child-Computer Interaction.
Publication date: 28 Sep 2023
Project Page: https://doi.org/XXXXXXX.XXXXXXX
Paper: https://arxiv.org/pdf/2309.15839
