Perspectives on How Sociology Can Advance Theorizing About Human-Chatbot Interaction and Developing Chatbots for...

Area of Science:

• Sociology
• Human-Computer Interaction
• Artificial Intelligence Ethics

Background:

• Research on chatbots (conversational agents) is rapidly expanding across disciplines.
• Sociology has not kept pace with other fields in publishing research on human-chatbot interaction.
• Existing models of chatbot use often overlook sociological factors like social structure and systemic discrimination.

Purpose of the Study:

• To demonstrate how sociological theories can advance the study of human-chatbot interaction.
• To offer four specific sociological theories to enrich current research on chatbots.
• To guide the development of chatbots for social good, ensuring safety, equity, and human agency.

Main Methods:

• Application of resource substitution theory and power-dependence theory to understand drivers of chatbot use.
• Integration of affect control theory and fundamental cause of disease theory for chatbot intervention design.
• Theoretical analysis of how sociological frameworks can improve chatbot alignment and social impact.

Main Results:

• Sociological theories provide novel insights into why individuals use chatbots, including potential for problematic emotional dependency.
• These theories highlight the influence of social structures, such as discrimination and resource distribution, on chatbot adoption.
• Theories can inform the creation of chatbot interventions that minimize risks and promote equitable access to resources.

Conclusions:

• Sociology offers valuable perspectives to deepen the understanding of human-chatbot interactions.
• Sociological theories can guide the development of chatbots that address societal challenges while upholding human agency.
• Integrating sociological insights is crucial for creating responsible and beneficial AI applications.