Effectiveness of a Home-Based Video-Guided Hip Muscle Exercise Program in Elderly Patients Following Bipolar Hemiarthroplasty for Femoral Neck Fracture: A Randomized Controlled Trial.

Artificial intelligence (AI) refers to intelligence exhibited by machines, distinct from natural human intelligence (). With advances in computing power and data availability, conversational agents have moved beyond laboratory settings and become integrated into smartphones and computers as chatbots and voice assistants, providing on-demand information, scheduling, and entertainment services (; ). These dialogic systems, when designed with greater human-likeness and utility, are more likely to receive favorable user evaluations and promote continued engagement (). Understanding why users like, trust, and increasingly rely on such AI has emerged as a critical focus in human–computer interaction (HCI) research. According to the need to belong theory, humans possess an innate motivation to form and maintain positive and lasting social bonds (). When this need is thwarted, individuals may experience loneliness and turn to “social surrogates,” such as favorite television shows or digital pets, as alternative sources of socioemotional experience (). In the rapidly expanding context of conversational AI, this substitution effect may become more salient: always-available social and companion chatbots can serve as sources of companionship and emotional support (; ). Perceived anthropomorphism plays a pivotal psychological role in this surrogate process. Individuals tend to attribute human-like qualities to non-human agents (). According to the three-factor model proposed by , anthropomorphic tendencies are driven by knowledge of human characteristics, effectance motivation, and sociality motivation. When social motivations are activated due to a lack of connection, users are more likely to perceive computers, robots, or chat agents as “quasi-human companions.” The resulting liking of AI and perceived emotional support may serve as psychological channels through which anthropomorphic perceptions of AI are associated with belongingness-related motivation. Enhancing anthropomorphic cues such as appearance or emotional expression, can significantly boost users’ trust in, empathy toward, and positive attitudes toward AI agents ().