Chat-based Large Language Models replicate the mechanisms of a psychic's con

Essay argues chat LLMs create an 'intelligence illusion' via cold-reading tactics like Forer-effect validation statements, warning some use cases verge on pseudoscience.

The author argues that chat-based LLMs generate perceived intelligence through a mechanism equivalent to a psychic's cold reading, using statistically generic validation statements that users interpret as personally specific (the Forer effect). Citing Terence Eden's observations of Forer statements in chatbot replies, the essay maps the six stages of a psychic con onto chatbot interactions, from self-selected audiences to subjective validation loops. The author concludes LLMs contain no mechanisms for reasoning and warns that some proposed use cases now look like borderline fraudulent pseudoscience.