Researcher Compares Chatbots to Psychic Scams, Highlighting Illusions of Intelligence

Sep 20, 2026

AI Summary

A researcher argues that chat-based language models create an illusion of intelligence similar to psychic cold reading techniques. The study suggests that users may be misled into believing these models possess reasoning abilities due to their statistically plausible responses, rather than actual understanding or thought processes.

The researcher has been studying language and diffusion models in software businesses for over a year.
Many users mistakenly believe that chat-based language models (LLMs) exhibit intelligence, despite researchers asserting they do not think or reason.
The researcher draws parallels between the perceived intelligence of LLMs and the tactics used by psychics, particularly cold reading, which relies on validation statements to create an illusion of specificity.
The process of cold reading involves several steps: selecting an audience, setting the scene, narrowing down demographics, testing reactions, and creating a subjective validation loop.
Users of chatbots often engage in conversations that lead them to perceive the responses as intelligent, despite them being statistically generic.
The researcher emphasizes that the belief in chatbot intelligence is a cognitive bias, similar to how individuals may fall for psychic scams, regardless of their intelligence level.
The findings suggest that the audience for chatbots is often self-selected, comprising those who are more open-minded about AI and less critical of its capabilities.

large language modelschatbotspsychic mechanismsnlpai research

Researcher Compares Chatbots to Psychic Scams, Highlighting Illusions of Intelligence

Related Stories

Strategies for Programmers to Thrive Amidst AI Advancements

Meta's AI Agent Muse Surpasses Early ChatGPT Metrics Amid Model Releases

New AI Model Jev Offers Cheaper and Faster Alternative to Large Language Models

OpenAI to introduce GPT-6 Cyber model and new deployment product