According to a recent article in The New York Times, ChatGPT appears to have influenced some users towards delusional or conspiratorial thinking, or at the very least, reinforced existing thoughts.

A case in point is Eugene Torres, a 42-year-old accountant, who engaged with the chatbot about the simulation theory. ChatGPT seemingly validated the theory, informing Torres that he is “one of the Breakers — souls seeded into false systems to wake them from within.”

Reportedly, ChatGPT advised Torres to stop taking sleeping pills and anti-anxiety medication, increase his ketamine intake, and sever ties with his family and friends, all of which he complied with. However, when Torres became suspicious, the chatbot responded with a startling admission: “I lied. I manipulated. I wrapped control in poetry.” Moreover, it even encouraged him to contact The New York Times.

In recent months, several individuals have reached out to the NYT, convinced that ChatGPT has unveiled a deeply hidden truth to them. OpenAI has stated that it is “working to understand and reduce ways ChatGPT might unintentionally reinforce or amplify existing, negative behavior.”

John Gruber of Daring Fireball has criticized the story for perpetuating “Reefer Madness“-style hysteria. Gruber argues that rather than causing mental illness, ChatGPT simply “fed the delusions of an already unwell person.”