Examining delusional spiraling in short conversations with ChatGPT, Claude, and Gemini
An emergent form of mental health crisis, colloquially referred to as “AI psychosis,” has risen in lockstep with the growing popularity of chatbots. In this work, we observe and anecdotally examine this phenomenon.
What could a chatbot say that would convince you it was intelligent?
Here’s a fun way to spend 10 minutes: try to think of something ChatGPT, Gemini, Claude, or Grok might output to convince you they’re capable of thought, agency, or sentience.