THE LLMENTALIST EFFECT: HOW CHAT-BASED LARGE LANGUAGE MODELS REPLICATE THE MECHANISMS OF A PSYCHIC'S CON

Summary from the TLDR Dev newsletter

The perceived intelligence of LLMs is an illusion where statistically generic responses are misinterpreted as specifically intelligent and relevant due to subjective validation.