AI Is ''Fun and Useful but Emotionally Still Distant''
Children Attracted to Practicality While Simultaneously Feeling Emotional Distance and Vague Fear

Paper: "Analysis of the Relationship Between Elementary Lower-Grade Students'' AI Concept Formation Level and Perceptions" (Seo A-ri, Han Seon-gwan, 2025). Research problem: AI education research predominantly focuses on program development and application (practical perspectives) while neglecting how young children who encounter AI devices daily actually conceptually understand what AI is. Key finding: children perceive AI as very useful (responding instantly to instructions, helping daily life), but understanding of AI''s core — "autonomy" and "self-learning function" — is relatively weak. Research results: average AI concept score 14.23 (71% correct rate); concept formation level significantly predicts interest (R² = 26.6%, β = .516) and usefulness perception (R² = 35.2%, β = .593); BUT concept formation had only LIMITED effect on emotional intimacy (R² = only 7%, β = .265). The cognitive-emotional asymmetry: knowing AI conceptually (head) vs. feeling emotionally close to AI (heart) are completely separate dimensions — simple concept education alone doesn''t naturally lead to emotional intimacy with AI. This means children can understand AI accurately while still feeling uncomfortable or distant with it. Policy implication: AI education for young children requires both cognitive curriculum (how AI works, what it can and cannot do) AND emotional-relational experiences (positive, safe interactions with AI systems that build familiarity). The "vague fear" finding deserves attention: children''s fears often center on loss of control or replacement — addressing these anxieties explicitly in age-appropriate terms may be as important as teaching AI concepts for creating healthy human-AI relationships in future generations.