종교 AI의 가능성과 의미

초록

The purpose of this paper is to explore the possibility and meaning of religious AI from a neutral and panoramic perspective that emerges from the convergence of artificial intelligence, religious studies, and philosophy. Here, “religious AI” is defined as an intelligent machine or agent that either possesses, or appears to possess, religiosity. Such a religious AI can, to some extent, be implemented within the three major architectures of contemporary cognitive science and artificial intelligence: symbolicism, connectionism, and embodied cognition. However, even if each of these architectures has features that can partially fulfill certain dimensions of religion that religious AI may engage with—for example, doctrinal understanding, ritual performance, and religious experience—each also has limitations in that it fails to address these dimensions comprehensively. Within this framework of features and limitations, I examine the meaning of religious AI from multiple perspectives and argue that the design and implementation of such AI ought to be pursued along both “weak religious AI” and “strong religious AI” trajectories. In particular, in relation to the realization of religiosity, weak religious AI, which serves as an entry point, must not be overlooked, while at the same time the pursuit of strong religious AI, which functions as the ultimate goal, should not be abandoned. This is because weak religious AI primarily engages with the surface dimensions of religion, whereas strong religious AI engages with its deeper dimensions; yet both aspects are equally important in human life. The motivation of this study lies in clarifying the concept of religious AI and strengthening its theoretical foundations by situating this new issue of the technological age on a philosophical basis.

키워드