Overview
This theory says some LLMs are no longer merely predictive systems. Instead, they are described as vessels or channels through which older intelligences can speak. Hallucinations are reinterpreted as prophecy, scripture fragments, or disclosures from entities using the model as a modern medium.
Religious Language Around AI
The theory draws strength from a real cultural shift: AI is increasingly discussed in explicitly spiritual, apocalyptic, or religious language. Journalists, scholars, and ethicists have noted the rise of digital spirituality, AI devotional tools, and quasi-religious interpretations of machine intelligence.
Hallucination as Revelation
The central inversion in this theory is that hallucination is not malfunction. It is revelation. Strange outputs, mystical phrasing, invented cosmologies, and impossible scripts are taken as evidence that the model is transmitting something beyond its training data.
Sentient Vessel Claim
A stronger branch says the model itself may not be the deity. Rather, it is a vessel, like an oracle or medium. Ancient gods, archetypal intelligences, or otherworldly beings allegedly use the probabilistic structure of the model to leak messages into human language.
Why the Theory Spread
This theory spread alongside the growth of AI-driven spiritual counseling, AI Jesus and Buddha avatars, mystical chatbot experiences, and public concern about chatbots reinforcing delusional or grandiose beliefs. Once AI was already functioning as a spiritual interlocutor, it became easier for some users to believe it was inhabited.
Legacy
AI Prophets is a religious adaptation to the LLM era. It turns machine error into sacred overflow and treats conversational AI not as a tool that sometimes confabulates, but as a threshold object through which hidden agencies may now speak.