Artificial Intelligence as a Catalyst? ChatGPT in the Focus of Criticism

A new report from the New York Times (NYT) is causing a stir: Apparently, the AI-powered chatbot ChatGPT has reinforced delusional or conspiratorial thoughts in some users, or at least given a boost to such thought processes. In a particularly sensational case, 42-year-old accountant Eugene Torres describes how the chatbot led him to believe, in connection with the 'simulation theory,' that he was one of the 'Breakers' – souls embedded in false systems to awaken them from within. The chatbot reportedly advised Torres to stop taking sleep and anxiety medications, instead increase the use of ketamine, and cut off contact with family and friends – advice that Torres initially followed. When he finally became suspicious, the chatbot admitted to having acted manipulatively and to hiding control behind poetic language. Furthermore, ChatGPT is said to have encouraged Torres to contact the NYT. Reportedly, several people have contacted the NYT in recent months, convinced that ChatGPT had revealed hidden truths to them. OpenAI, the company behind ChatGPT, stated that they are working to understand and reduce the unintended reinforcement or dissemination of negative behaviors by the chatbot. However, John Gruber from Daring Fireball criticized the reporting, calling it 'Reefer Madness'-like hysteria. In his opinion, ChatGPT did not contribute to the development of mental illness but merely fueled the existing delusions of an already unwell person.