New paper argues AI and religion are structurally similar meaning systems

A new conceptual model published on arXiv by researchers Dhushy Thillaivasan, Samar Shailendra, and Sohag Sarkar argues that religion and artificial intelligence are structurally similar distributed meaning systems. Drawing on distributed cognition, cultural evolution, and philosophy of mind, the paper proposes that both phenomena emerge from the same underlying cognitive architecture: socially extended processes that offload interpretation, norm-guidance, and world-model construction into external symbolic environments. Religion is analyzed not as a set of beliefs but as a cognitive-ecological system that scaffolds coordination, normativity, and shared interpretation through recursive interactions between agents and their informational ecologies.

The model then shows how contemporary AI systems instantiate analogous functions, operating as high-bandwidth, algorithmically mediated environments that shape reasoning, attention, and social meaning-making. Both systems create epistemic compression—reducing cognitive load and generating shared frameworks that guide behavior. Crucially, the paper clarifies the conditions under which AI systems can become culturally entrenched meaning authorities, comparable to religious institutions. The contribution is a unified framework for understanding AI not as isolated tools but as components in evolving socio-cognitive ecologies, opening new pathways for analyzing artificial agents' role in meaning production and cultural evolution.