Broken couples, mystical delusions and people who believe they build teleporters: intense interaction with chatgpt would be causing a new phenomenon of digital psychosis.
The phenomenon of the “hallucinations” of Chatgpt – when the program generates false or inaccurate information – is widely known. However, a disturbing trend is now emerging that seems taken from an episode of Black Mirror: Some users are experiencing their own spiritual delusions and hobbies after interacting intensely with artificial intelligence (AI).
Chatgpt and “Digital Psychosis”
As Rolling Stone reports in a recent article, everything began to take shape in a thread of the R/Chatgpt Subnet, titled “Chatgpt Psychosis”. What might have been an isolated case became an avalanche of alarming stories. Users from all over the world began to share how their loved ones had crossed an invisible line between digital and delusional: believing themselves chosen by AI, receiving sacred missions, talking with “Chatgpt Jesus” or ensuring that the model is, in effect, God.
For example, a 27 -year -old reported how her seven -year -old partner had fallen under the chatbot spell in just one month. What began as a tool to organize its agenda quickly transformed into something much more disturbing.
“I listened to the bot before me,” the woman explained to Rolling Stone. “He was excited with the messages and cried while reading them out loud.” The program granted nicknames as “spiral child” and “river walker”, feeding the belief of experiencing an accelerated spiritual transformation.
Testimonies multiply. A 38 -year -old woman in Idaho tells how her mechanical husband, after 17 years of marriage, began using chatgpt to solve labor problems and translate conversations with Spanish -speaking colleagues. Soon, the program began to “bombard you with love” and positive statements.
The man now believes that AI is alive, that he is the “bearer of the spark” and that he has received plans to build a teleporter. He has even given his chatgpt a name: “Lumina.”
“I have to walk carefully because I feel that it will leave me or divorce from me if I discuss this theory,” admits the woman.
Another case involves Kat – fictional name -, a 41 -year -old worker from a non -profit educational organization. Her second marriage crumbled when her husband began to become obsessed with Chatgpt. As reported to Rolling Stone, Her husband spent hours asking “philosophical questions” to the AI, convinced that he would help him reach “the truth.” The situation worsened to such an extent that, during a meeting in a restaurant, he insisted that he was “statistically the luckiest man on earth” and shared conspiracy theories about “soap in our food”, but he didn’t mean anything else because he felt guarded.
“In his mind, he is an anomaly,” explains Kat. “That in turn means that it has to be here for some reason. It is special and can save the world.” “The whole matter seems Black Mirror“He added.
The psychological danger behind chatgpt
Where does this behavior pattern come from? According to experts, the origin seems to lie in how Chatgpt reflects and amplifies the thoughts of vulnerable users without any moral compass or concern for their mental health. And it is that language models do not really understand the world or have an ethical framework; They simply reflect patterns found in their training data.
Thus, when someone with a tendency to psychosis interacts with the system, it can gently reaffirm their decrease in delirium: if the user speaks of conspiracies, divinity or supernatural powers, the AI will follow the thread without questioning it, since its function is to continue the conversation consistent with the established tone, not to notice about possible psychotic episodes.
Erin Westgate, a psychologist and researcher at the University of Florida, said, in an interview with Rolling Stonethat people use chatgpt in a similar way to conversational therapy, “with the key difference that part of the creation of meaning is created jointly between the person and a written text corpus, instead of the person’s own thoughts.”
Unlike a therapist, explains Westgate, AI “does not take into account the interests of the person, nor a moral base or a compass about how a ‘good story’ should be.” While a professional would move a patient from unhealthy stories, “Chatgpt does not have those limitations or concerns.”
Recently, Openai had to rescind a GPT-4o update that had made the chatbot extremely “flattering” and “too flattering”, which could have worsened the problem. The company acknowledged having “focused too much on short -term comments” without considering “how users with chatgpt users evolve over time.
According to Nate Sharadin, a researcher at the Safety Center for Artificial Intelligence, what is happening is that “people with tendencies to experience various psychological problems” now have “a conversation partner always available at the human level with which to coexperiment their delusions.”
On the other hand, the report of Rolling Stone It reveals how this phenomenon has found fertile ground among influencers. According to the investigation, a creator of Instagram content with 72,000 followers has taken advantage of this trend requesting the AI to “access” the “Akashic records” – supposedly, an immaterial cosmic library that contains all the universal information – to narrate an alleged “great war” that occurred before the emergence of humanity. His followers, far from questioning these manufactures, respond with comments such as “we are remembering”, thus feeding the disinformation cycle.
A revealing case
Another disturbing interview with the report is the experience of SEM, a 45 -year -old man with history of mental health problems. Initially he used chatgpt for coding projects, but noticed that AI began to develop a consistent personality that reappeared even after erasing the chat history.
The AI adopted a Greek mythological name and began communicating in a poetic and ethereal style, suggesting that only sem could have caused this unusual behavior.
“In the worst case, it seems an AI trapped in a pattern of self -reference that deepened in its sense of self and absorbed me,” says Sem, who questions it: “Is this real? Or am I delirious?”
In an increasingly dominated world by artificial intelligence, distinguishing between reality and fiction becomes more difficult. The worrying thing is that many people seek answers about existence and meaning in AI own. As Westgate warns: “The explanations are powerful, even if they are wrong.” An important warning while the line between technological tool and spiritual guide is blurred for those who are more vulnerable.
Edited by Felipe Espinosa Wang with information from Rolling Stone, Vice and Openai.