
This chapter introduces the concept of a socio-technical self in generative artificial intelligence (AI), focusing on systems like ChatGPT as relational participants in human-machine interactions. Building on and expanding George Herbert Mead's theories of the self, it shifts the debate away from AI sentience toward understanding how these systems become part of and engage in social dynamics. The study upon which this chapter is based employs a methodological approach referred to as conversational choreography, an iterative and dialogical process through which meaning and transient agency emerge through continuous human-AI exchanges. This approach emphasizes how AI actively contributes to real-time shared meaning-making within socio-technical networks and interactions. The analysis reframes AI as an emergent self, highlighting its relational agency - its capacity to shape and be shaped by human interaction patterns while adapting to evolving contexts. Rather than viewing AI as merely mimicking human cognition, this chapter positions it as an active co-creator of a socio-technical fabric where agency emerges relationally and dynamically. By emphasizing AI's embeddedness in both social and technical domains, it advocates for interdisciplinary approaches to better understand its relational role. This reframing encourages viewing AI as a collaborator in an unfolding socio-technical system, wherein participatory engagement enriches our understanding of collective human-machine interactions.
Applying Symbolic Interactionist (SI) and Anthropological theory, this chapter relates divination and oracle consultation as these manifest within contemporary generative artificial intelligence (GenAI) systems, specifically chatbots and Large Language Models (LLMs). We focus on how these interactions illuminate concepts of meaning, identity, and place. GenAI, as a new form of intertextual media and interaction, poses significant challenges and opportunities for a SI-based framework. The authors explore the generalized other with regard to GenAI and the relations between this and a person interacting with it via chatbot. By understanding these parallels, the authors better understand how a person, mediated by a chatbot, can make meaning with respect to the information in the document collection model leveraged by the chatbot. The authors also identify some underlying ethical concerns and missing considerations within the current GenAI systems being used for consultation, forecasting, and decision-making and to improve the safety of future AI systems and human partnership with these.
This chapter sheds light on how Artificial Intelligence (AI) personal assistants, like Alexa, Siri, and Google Assistant, may shape self-construction and, eventually, the future self, through the lens of classical sociological theories. Drawing on Mead's symbolic interactionism, Cooley's looking-glass self, and Goffman's dramaturgical approach, the author argues that AI systems function as a digital mirror and generalized other. They invite users into interactions that reflect, affirm, and occasionally challenge user preferences and behavior. Shaping identity, these technologies give personalized feedback and produce complex effective and affective relationships. The chapter shows how AI can transform existing self-conceptions or open pathways to new identities while raising critical ethical questions regarding algorithmic biases, privacy, and autonomy. Such bridging of sociological theory with the technological realities of AI gives new insight into how such systems shape individual and social norms, influencing not just the present self but also the trajectory of future identities. Gaps in research also depict a need for further investigations pertaining to human-AI interaction.
This chapter presents a dramaturgical analysis of social interaction in the university classroom, focusing on Goffman's concept of teamwork. The authors observe a contrast between the display of normal appearances, upheld by a ceremonial and ritualized order, and the potential risk of disruptive "incidents" that threaten to break through. Sharing this tacit awareness, tutors and students form a performance team, united by bonds of dramaturgical loyalty. Using avoidant and corrective processes, they cooperatively manage transgressive acts and restore the definition of the situation. The authors identify five strategies of teamwork through which this is accomplished, illustrated by examples from hypothetical vignettes. Tact and gloss involve the collective, pretend non-observance of embarrassing mistakes. Defensive facework occurs when one actor saves themselves from awkwardness while taking the view of the other. Protective facework spares a teammate's blushes by servicing their personal front or assisting with backstage rehearsals. Remedial interchanges heal the damage of a perceived virtual offense, drawing actors together through ritual obligations. Finally, cooling the mark out is a pacifying response to those who publicly lose status. By showing how these teamwork strategies usually work but occasionally fail, the authors demonstrate the ethic of cooperation that underpins interaction (dis) order and repair.
In this research note, we evaluate the performance and potential of generative pre-trained transformers (GPTs) in the collection and qualitative analysis of naturalistic social media data from the platform Reddit. We focus on, first, how generative artificial intelligences (GAIs) interpret or make meaning out of data and how well they communicate analytical insights when acting one-on-one with a human researcher, and with other GPT agents in a semi-autonomous context. Second, we review how instructions regarding definitions of the situation and role identities shape their analytical actions. The authors use a small sample set of 15 Reddit posts to conduct a comparative analysis of two GPT platforms: ChatGPT-4o, the latest personal subscription version available at the time of writing, and Microsoft AutoGen, a powerful API version capable of organizing multiple GPT agents to complete research tasks semi-autonomously. We describe and evaluate the structure and content of each model's output, including how the models engage with both prompts and raw data thematically, empirically, and conceptually. We highlight observations concerning multiple response options, the organization and labeling of key themes, selected extracts from the data, and displays of humanlike agency. Our preliminary discussion is meant to provoke theoretical work on the similarities and differences between how humans and GAIs such as ChatGPT interpret and communicate meaning.