Ask Me Anything: Exploring children's attitudes toward an age-tailored AI-powered chatbot
Project Overview
The document examines the use of generative AI in education through the development of a child-friendly AI-powered chatbot called 'Ask Me Anything' (AMA), aimed at enhancing children's engagement with various educational subjects like astronomy, sneakers, and dinosaurs. The study investigates children's perceptions of the chatbot, revealing that they found AMA to be knowledgeable and trustworthy, which fostered a deep level of engagement and curiosity. Children actively interacted with the chatbot, exploring its features and assessing the credibility of its information. The findings underscore the significance of designing age-appropriate AI tools and incorporating AI literacy into educational curricula to foster positive attitudes towards technology. Overall, the research highlights the potential of generative AI to enrich educational experiences for children by promoting trust and curiosity in learning.
Key Applications
Ask Me Anything (AMA) chatbot
Context: K–8 educational setting for children aged 7–14
Implementation: The chatbot was tested in a public school, allowing small groups of students to interact with it for 3-10 minutes, followed by a post-survey.
Outcomes: Students expressed wonder, built trust in the chatbot, and engaged in testing its responses to evaluate reliability, indicating a positive attitude towards AI in education.
Challenges: The chatbot occasionally responded off-topic despite attempts to maintain relevance, leading to potential trust issues.
Implementation Barriers
Technical limitation
The AI model occasionally produced off-topic responses despite being instructed to stay on topic.
Proposed Solutions: Implementing stricter context management and using alternative models that perform better in staying on-topic.
Cognitive challenge
Younger children struggled with typing their questions, affecting interaction quality.
Proposed Solutions: Simplifying the interface and providing assistance for younger students during interactions.
Project Team
Saniya Vahedian Movahed
Researcher
Fred Martin
Researcher
Contact Information
For information about the paper, please contact the authors.
Authors: Saniya Vahedian Movahed, Fred Martin
Source Publication: View Original PaperLink opens in a new window
Project Contact: Dr. Jianhua Yang
LLM Model Version: gpt-4o-mini-2024-07-18
Analysis Provider: Openai