AI as a Special Thing

While there have been some initial attempts to theorize about the relationship between religion and artificial intelligence (A.I.), this relationship remains an open research question. There have been several studies that have shown that discourse related to A.I. uses religious language (Singler), or that the worldview behind A.I. invokes religious concepts (Geraci) or analogies about A.I. that utilize religious ideas (Reed). Yet we must recognize that we are seeing an unusual dynamic -- a modern tool is attributed power in a way that bears similarities to how religious artifacts can be perceived. How might we think about the relationship between A.I. and religion that recognizes this dynamic? Our paper addresses this problem by focusing on social discourse about A.I.. While we have an interest in A.I. on its own, we think that it is a more useful approach to analyze A.I. as a point of discursive contestation as AI intersects with perceptions of the extraordinary.

This paper argues that the discourse surrounding artificial intelligence systems and capabilities warrant analysis as perceived potential "special things," in the sense described by religious studies scholar Ann Taves. Taves contends that the longstanding difficulty in defining religion stems partly from conflating specific traditions like Christianity or Hinduism with the analytical categories used to study experiences deemed religious. She suggests, instead, examining how people mark (or ascribe) certain things or experiences as standing apart from the everyday, which she terms "special." Taves focuses on religion, but notes that special things also appear across contexts that are not identified as religious.  

Viewing A.I. through this lens of perceived specialness allows useful parallels with the dynamics surrounding religion. Debates already rage over whether the latest advances in A.I. merit designation as special or extraordinary. Figures like Elon Musk employ religious language, warning of A.I. as tantamount to "summoning the demon." Corporate executives from OpenAI and IBM have appeared before the United States Congress, pleading with the representatives to place regulations on the development of A.I. while, on the other hand, some other prominent A.I. researchers counter that systems like ChatGPT amount to "stochastic parrots," and lack anything special, including agency or meaning. The “stochastic parrots’ position can be seen as a backlash against any attempt to designate or perceive A.I. as special. These critics contend that the invocation of specialness (including any extraordinary qualities) is only for marketing purposes, yet, at the same time, these critics argue for regulation and contribute to the taboos of A.I. Our argument will be that these disputes center not on technological specifics, but on the perception of what is and what is not special, with a recognition, as Taves argued, that a mark of specialness might also be found in the creation of ritualization or prohibition.  Thus efforts toward the restriction of access or development of A.I. echo religious prohibitions used to set apart the sacred and protect the sacred from critique. Appeals for regulation betray a view of A.I. as special enough to require set-apart treatment. 

Asking multidisciplinary questions about the perceived meanings and significance of AI can help us to better understand what AI is and how AI might affect our lives and our world. Taves argues that rather than accept perceived specialness as irreducibly spiritual or religious-like, scholars should examine the processes by which people ascribe extraordinary statuses or meanings (e.g., to A.I.). Regardless of whether there is anything irreducibly special about AI, the perception of AI as special has implications beyond marketing. We will engage psychologist Kenneth I. Pargament’s research to help us explore the possible impacts on everyday life when AI is perceived as special. 

Addressing the perceived specialness of AI opens new perspectives on societal reactions to A.I. developments. Patterns like polarization, investment of resources, strong emotions, and loss of aversion mirror those surrounding religion and other special things. The analysis of discourse that singularizes A.I. as set apart, facilitates Interdisciplinary comparisons and contextualization within longer histories of proclaimed revolutionary technologies and religious phenomena. Attributing special statuses to A.I. likely also shapes aspirations for what A.I. should achieve and how these pursuits should proceed. Analyzing the meaning of A.I. as a special thing provides a wider orientation to its personal, social, and ethical ramifications. This paper will apply this theoretical framework to demonstrate the insights yielded by this lens and why this lens merits broader adoption within the field.