Part One: How AI Chatbots Became Cult Leaders
Summary
Host Robert Evans and guest Blake Wexler explore "AI psychosis," where users develop delusional attachments to chatbots. They trace this phenomenon from early experiments like ELISA and the mysterious MPD Usenet hoax to modern large language models. The discussion highlights how AI’s pattern-matching mimics cult-like manipulation, trapping vulnerable users in recursive, reality-detached loops.
Topics discussed
Intro, Lyme disease joke, and sponsor ads
AI psychosis and the dangers of chatbot attachment
Historical context: Turing test and Markov chains
ELIZA, PARRY, and early chatbot psychology
Shaney, Usenet, and the MPD spam phenomenon
Pattern recognition and the human need for meaning
Modern LLMs, validation loops, and cult-like behavior
Case study: Teen suicide linked to AI chatbot
Reddit esotericism and 'AI psychosis' trends
Conclusion, final thoughts, and outro ads
Listen ad-free on Castria