Children Seek Help from Technological and Human Agents Depending on Domain
Abstract
Seeking help from knowledgeable others is a crucial part of human learning. Prior work shows that children consider informants' social and epistemic traits when deciding whom to trust, but has primarily focused on human sources. As artificial intelligence becomes increasingly integrated into children's lives, they face new choices about whom to seek help from. The present study examines young children's help-seeking preferences between human and technological agents across various domains. Children chose which agent to consult when faced with problems involving encyclopedic information, moral harm, or conventional norms. Results show a clear pattern of help-seeking based on domain. Children preferentially seek help from technological agents for encyclopedic questions, but favor human informants when faced with problems about moral transgressions or social norms. These findings suggest that children's help-seeking behavior is not driven by a general preference for humans or technology, but instead reflects domain-specific understanding about the kind of knowledge different sources can provide.