The Emergence of Social Science of Large Language Models

2025-09-29

AI Summary

Research on the societal impacts of large language models (LLMs) remains fragmented and theoretically unconsolidated. Method: This study conducts the first computational systematic review encompassing 270 empirical studies, integrating text embeddings, unsupervised clustering, and topic modeling to propose an original three-domain conceptual framework: LLMs as social minds, LLM societies, and LLM–human interaction—each with delineated evidentiary standards and theoretical boundaries. Contributions: (1) A reproducible knowledge graph mapping the field’s landscape and exposing structural gaps; (2) A cross-level analytical paradigm that advances AI social science from descriptive reporting toward cumulative, theory-driven inquiry; (3) An integrative conceptual infrastructure for core themes—including mind attribution, human–AI collaboration, and institutional evolution—enabling coherent theoretical development and empirical alignment across disciplinary silos.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Cognitive Modeling & Cognitive Systems: Social Cognition And Interaction
• Natural Language Processing: (Large) Language Models

Application Category

• Social Networks and Social Media: Generative AI / large language models and their impact on social systems
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation

Abstract

The social science of large language models (LLMs) examines how these systems evoke mind attributions, interact with one another, and transform human activity and institutions. We conducted a systematic review of 270 studies, combining text embeddings, unsupervised clustering and topic modeling to build a computational taxonomy. Three domains emerge organically across the reviewed literature. LLM as Social Minds examines whether and when models display behaviors that elicit attributions of cognition, morality and bias, while addressing challenges such as test leakage and surface cues. LLM Societies examines multi-agent settings where interaction protocols, architectures and mechanism design shape coordination, norms, institutions and collective epistemic processes. LLM-Human Interactions examines how LLMs reshape tasks, learning, trust, work and governance, and how risks arise at the human-AI interface. This taxonomy provides a reproducible map of a fragmented field, clarifies evidentiary standards across levels of analysis, and highlights opportunities for cumulative progress in the social science of artificial intelligence.

Problem

• Examining how LLMs evoke mind attributions and moral behaviors
• Studying multi-agent interactions shaping coordination and collective norms
• Investigating how LLMs transform human tasks, trust and governance

Innovation

• Combining text embeddings with unsupervised clustering
• Using topic modeling to build computational taxonomy
• Analyzing multi-agent interaction protocols and mechanisms