June 17th, 2024 | RESEARCH
The increasing presence and importance of Artificial Intelligence (AI) in our society has led to calls for its inclusion at all levels of education. However, the field is only beginning to understand what how AI learning experiences may be designed to be effective and developmentally appropriate, especially for young children. One challenge children encounter is in conceptualizing the “intelligence” of AI while they are still developing a metacognitive model of their own human intelligence. To investigate potential ways to address this, we developed a strategy, metacognitive embodiment, through which children are supported to (a) elicit a mental model of their own intelligent performance on a task and (b) compare that elicited model to an AI designed to accomplish the same task. From this study we found evidence suggesting that engaging children in metacognitive tasks in coordination with AI learning experiences (where the AI performs an analogous task) better positioned them for sensemaking about the AI’s intelligence.
Document
Team Members
Eric Greenwald, Author, University of California Berkeley’s Lawrence Hall of ScienceAri Krakowski, Author, University of California Berkeley’s Lawrence Hall of Science
Timothy Hurt, Author, University of California Berkeley’s Lawrence Hall of Science
Kelly Grindstaff, Author, University of California Berkeley’s Lawrence Hall of Science
Ning Wang, Author, University of Southern California, Institute for Creative Technologies
Citation
Publication: IDC '24: Proceedings of the 23rd Annual ACM Interaction Design and Children Conference
Funders
Funding Source: NSF
Funding Program: AISL
Award Number: 2116109
Related URLs
Tags
Audience: Elementary School Children (6-10) | Learning Researchers | Museum | ISE Professionals
Discipline: AI
Resource Type: Conference Proceedings | Research
Environment Type: Museum and Science Center Exhibits