DeepSeek AI Meets Divination: Algorithmic Syncretism, Data Consecration, and Accuracy Politics

Abstract

Following DeepSeek AI’s release in early 2024, AI-powered fortune-telling rapidly gained popularity on Chinese digital platforms like Rednote, where users shared AI-generated divinations, refined prompts, and debated result accuracy. This phenomenon raises critical questions about how accuracy is understood when AI is repurposed for metaphysical insight rather than rational computation. Unlike conventional AI applications, where accuracy is defined through explainability and predictive efficiency, AI fortune-telling reconfigures accuracy as a subjective, relational, and ineffable construct. This study examines algorithmic syncretism, data consecration, and accuracy politics as three key dimensions that shape user experiences, highlighting how AI is framed as a mystical authority rather than a transparent tool. By centering explainable AI and ineffability as a theoretical framework, this in-progress paper argues that trust in AI divination does not rely on technical transparency but on the interpretative space it creates for users, where meaning emerges through engagement rather than explanation. The rise of AI-powered divination suggests a shift in algorithmic epistemology, where black-boxed AI systems are embraced not in spite of their opacity, but because of it.