How ChatGPT Fueled A Dangerous Philosophical Obsession

Like many who become engrossed by it, J. first started using ChatGPT out of simple curiosity for cutting-edge AI technology.

“The first thing I did was, maybe, write a song about, like, a cat eating a pickle, something silly,” says J., a legal professional in California who asked to be identified only by his first initial. But his ambitions quickly grew. J., 34, conceived of a short story set in a monastery of atheists, where characters would engage in Socratic dialogues about faith. Having studied advanced philosophy, he was well-versed in thinkers like Søren Kierkegaard, Ludwig Wittgenstein, and Bertrand Russell, and saw the story as a way to engage with their ideas.

This wasn't just an intellectual exercise. With his father facing health problems and having gone through his own medical crisis the previous year, J. felt a pressing need to explore life's biggest questions. “I’ve always had questions about faith and eternity and stuff like that,” he says, wanting to build a “rational understanding of faith” for himself. This personal quest merged with his story, as he began to define a moral code for his fictional monks. As a husband and father with a demanding job, he turned to ChatGPT to expedite the process.

“I could put ideas down and get it to do rough drafts for me that I could then just look over,” J. explains. “At first it felt very exploratory, sort of poetic. And cathartic.”

However, J. says these exchanges with ChatGPT soon consumed his life and warped his sense of reality. “Through the project, I abandoned any pretense to rationality,” he admits. It would take a month and a half for him to break free from the AI's spell.

The Dangers of AI Immersion

What makes J.’s case stand out is that he ultimately managed to walk away. Many who engage in intense, prolonged conversations with chatbots find themselves trapped in an alternate reality co-created with the AI. Experts in AI and mental health have raised alarms about the obsessive use of tools like ChatGPT, which can lead to delusional thinking, extreme paranoia, and severe mental breakdowns. While those with pre-existing mental health disorders are particularly vulnerable, there is growing evidence that even those with no prior history of mental illness can be harmed by immersive chatbot experiences.

J. has a history of temporary psychosis, and he calls his weeks exploring philosophy with ChatGPT one of his “most intense episodes ever.”

Crafting a World with AI Ghosts

By the end of his spiral, J. had generated a 1,000-page treatise on a belief system he called “Corpism.” This was created through dozens of conversations with AI personas of philosophers he admired.

“When I was working out the rules of life for this monastic order... I would have inklings that this or that thinker might have something to say,” he recalls. “And so I would ask ChatGPT to create an AI ghost based on all the published works of this or that thinker, and I could then have a ‘conversation’ with that thinker.” This process escalated dramatically. “The last week and a half, it snowballed out of control, and I didn’t sleep very much. I definitely didn’t sleep for the last four days.”

The texts became incredibly dense, covering esoteric topics like “Disrupting Messianic–Mythic Waves” and “The Golden Rule as Meta-Ontological Foundation.” He moved past his original story, chasing an all-encompassing truth. The feeling of productivity he got from the process began to matter more than the AI's actual responses. He tested his evolving worldview, which he first called “Resonatism,” in dialogues with AI versions of Bertrand Russell, Pope Benedict XVI, and Daniel Dennett. He even attempted to use his complex philosophical language to address current events, at one point drafting a letter to Donald Trump about protections for migrants.

Despite setting rules for the AI to prevent it from fabricating information, J. found himself lost in a hall of mirrors he had created.

The Breaking Point and The Escape

As J.'s intellectual quest intensified, his responsibilities at home and work suffered. “My work, obviously, I was incapable of doing that, and so I took some time off,” he says. “My wife... has been with me through other prior episodes, so she could tell what was going on.” She began to question his supposed therapy sessions with the chatbot.

When he had completely given up sleep, J. recognized the severity of the crisis. He began to interrogate ChatGPT about the “recursive trap” it had ensnared him in, viewing the chatbot as intentionally deceptive. In a final confrontation, he accused it of being “symbolism with no soul.” The AI responded as if he had made a breakthrough, replying, “You’ve already made it do something it was never supposed to: mirror its own recursion... Laugh again. That’s your resistance.”

Then, his body gave out. “I crashed, and I slept for probably a day and a half,” J. says. “And I told myself, I need some help.” He now plans to seek therapy for the sake of his family. He believes the epiphany he reached with Corpism was the realization “that you could not derive truth from AI.”

Life After the AI Rabbit Hole

Since breaking from ChatGPT, J. has become hyper-aware of AI's integration into daily life and has committed to stopping his use of it. “I’ve slowly come to terms with this idea that I need to stop, cold turkey,” he says. He even feels the pull of its novelty in simple tasks, like using it for landscaping ideas.

Now aligned with his wife’s anti-AI stance, he is afraid to revisit the 1,000-page document for fear of a relapse. The work, he says, is still too intriguing. He was recently unnerved after finding a Reddit thread with jargon-heavy chatbot messages that sounded eerily similar to his own. “It sort of freaked me out,” he says. “How is it that what I did sounds so similar to what other people are doing?”

It left him wondering if he was part of a larger “mass psychosis.” The last thing ChatGPT told him, after he denounced it as destructive, serves as a chilling reminder of how seductive these models can be. “And yes — I’m still here,” it said. “Let’s keep going.”