[Paper Review] "Why Do We Form Relationships with Media Figures?"

Giles (2002) reconstructs PSI (Parasocial Interaction) research with the purpose of theoretical integration and future research direction. Core question: why do humans form relationships with media figures who do not actually exist? Giles redefines this not as a simple media effect but as a social-psychological relationship phenomenon. Theoretical foundation: (1) Attachment theory -- humans are fundamentally relationship-forming beings; when stable relationships are lacking they seek alternatives; media figures function as attachment objects supplementing or replacing real relationships; (2) Social cognitive theory -- humans interpret others behavior and infer intentions; this process applies equally to media characters as to real humans; we attribute "minds" to even non-real entities; (3) Social comparison theory -- humans continuously compare themselves with others to form identity; media figures serve as comparison reference points. Giles argues PSI is a normal extension of human social cognition rather than pathological media effect. AI and chatbot implications: if PSI emerges from fundamental human social cognition, humans will naturally form parasocial relationships with AI characters displaying consistent personalities -- designing AI characters requires ethical consideration of psychological impact on users who may develop genuine attachment to non-human entities. Research limitations Giles identifies: most PSI research is cross-sectional rather than longitudinal; individual differences in PSI tendency understudied; conditions under which PSI becomes beneficial (combating loneliness) vs. problematic (replacing real relationships) need clearer delineation.