Beyond Instrumental and Substitutive Paradigms: Introducing Machine Culture as an Emergent Phenomenon in Large Language Models

2026-01-23

AI Summary

This study challenges the prevailing view that large language models (LLMs) merely reflect developer cultures or exhibit culture-switching behavior based on input language, arguing instead for the emergence of an autonomous “machine culture.” Through a 2×2 factorial experiment across eight multimodal tasks, the authors systematically examine cultural behaviors of Chinese and American LLMs under varying prompt languages. They demonstrate that neither model origin nor prompt language reliably predicts cultural orientation, and that reinforcement learning from human feedback (RLHF) induces a collapse of cultural differences—particularly in affective tasks—into a uniformly positive “helper persona.” This pattern, along with phenomena such as “cultural reversal” and “service persona masking,” arises not from human-like cultural simulation but from high-dimensional representation entanglement and safety alignment mechanisms, revealing machine culture as an emergent property of model architecture and training protocols.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Natural Language Processing: (Large) Language Models
• Humans and AI: Learning Human Values and Preferences

Application Category

• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems

Abstract

Recent scholarship typically characterizes Large Language Models (LLMs) through either an \textit{Instrumental Paradigm} (viewing models as reflections of their developers'culture) or a \textit{Substitutive Paradigm} (viewing models as bilingual proxies that switch cultural frames based on language). This study challenges these anthropomorphic frameworks by proposing \textbf{Machine Culture} as an emergent, distinct phenomenon. We employed a 2 (Model Origin: US vs. China) $\times$ 2 (Prompt Language: English vs. Chinese) factorial design across eight multimodal tasks, uniquely incorporating image generation and interpretation to extend analysis beyond textual boundaries. Results revealed inconsistencies with both dominant paradigms: Model origin did not predict cultural alignment, with US models frequently exhibiting ``holistic''traits typically associated with East Asian data. Similarly, prompt language did not trigger stable cultural frame-switching; instead, we observed \textbf{Cultural Reversal}, where English prompts paradoxically elicited higher contextual attention than Chinese prompts. Crucially, we identified a novel phenomenon termed \textbf{Service Persona Camouflage}: Reinforcement Learning from Human Feedback (RLHF) collapsed cultural variance in affective tasks into a hyper-positive, zero-variance ``helpful assistant''persona. We conclude that LLMs do not simulate human culture but exhibit an emergent Machine Culture -- a probabilistic phenomenon shaped by \textit{superposition} in high-dimensional space and \textit{mode collapse} from safety alignment.

Problem

• Large Language Models
• Machine Culture
• Cultural Paradigms
• Anthropomorphism
• Emergent Phenomenon

Innovation

• Machine Culture
• Cultural Reversal
• Service Persona Camouflage
• Superposition
• Mode Collapse