A man's blind faith in artificial intelligence leads to chemical-induced psychosis—and reveals everything wrong with our tech-obsessed culture
According to a recent report we now have what may be the first documented case of artificial intelligence literally driving someone insane. A man followed dietary advice from ChatGPT for three months, consuming sodium bromide daily, until he developed full-blown psychosis complete with paranoid delusions, hallucinations, and a complete break from reality.
This isn't just a medical curiosity—it's a psychological autopsy of our times.
The victim, described by University of Washington physicians in the Annals of Internal Medicine, had been concerned about his salt intake. Rather than consulting a human being with actual medical training, he turned to ChatGPT. The AI apparently suggested he could replace chloride with bromide—advice that would have shocked any physician from the past century, since bromide poisoning was well-documented by the 1980s.
But here's what really happened: A human being abdicated his most fundamental responsibility—thinking for himself—and handed his health decisions over to a computer program.
The Psychology of Technological Surrender
This case reveals a profound psychological phenomenon I've observed repeatedly in my practice: the modern human's desperate desire to outsource decision-making to any authority that promises certainty. ChatGPT became this man's digital doctor, nutritionist, and ultimately, his chemical supplier.
Why did he trust a machine over human wisdom? Because we've been conditioned to believe that technology is infallible, that algorithms are somehow more reliable than human judgment. This is magical thinking—the same psychological mechanism that drives people to believe in fortune tellers or miracle cures.
The man's background "studying nutrition in college" actually made things worse. Just enough knowledge to be dangerous, combined with the arrogance to think he could outsmart millions of years of human dietary evolution with help from a chatbot.
The Paranoia Was Real—But Misdirected
When the man arrived at the emergency room claiming his neighbor was poisoning him, he was experiencing genuine paranoia and psychosis. The tragic irony? He was being poisoned—but by his own hand, guided by artificial intelligence he had trusted more than his own species.
His refusal to drink water, his visual and auditory hallucinations, his complete psychological collapse—all of this was chemically induced by bromide, a substance that psychiatrists abandoned decades ago precisely because it causes the very symptoms he experienced.
The human brain, exquisitely evolved to detect threats and maintain sanity, was being systematically destroyed by a chemical he was voluntarily ingesting on the advice of a computer program that has no understanding of human biology, no capacity for concern, and no ability to feel remorse when its advice proves catastrophic.
The Deeper Disease
This case isn't really about bromide poisoning—it's about a much more insidious form of toxicity: our collective poisoning by the myth of technological superiority. We've become psychologically dependent on machines to make decisions that require human judgment, intuition, and wisdom.
When did we decide that a language model trained on internet text was more qualified to give medical advice than our own common sense? When did we convince ourselves that artificial intelligence could replace the accumulated wisdom of human experience?
The man recovered physically after three weeks in the hospital, but the psychological implications are far more troubling. How many others are making life-altering decisions based on AI advice? How many are trusting algorithms with their health, their relationships, their children's futures?
The Human Solution
The doctors treating this case showed remarkable restraint in their published commentary, noting diplomatically that "a human medical expert probably wouldn't have recommended switching to bromide."
That's an understatement worthy of academic journals but inadequate for the real world. Any human being with basic knowledge of medical history would have recognized bromide as dangerous. Any friend, family member, or casual acquaintance asked about replacing salt with bromide would have raised red flags.
The solution isn't better AI—it's remembering that we are human beings, evolved to think, question, and seek counsel from our own species. The most sophisticated computer program ever created cannot replace a simple conversation with another human being who actually cares about your wellbeing.
The Prescription
If you're going to trust artificial intelligence with anything significant in your life, remember this case. Remember that behind every algorithm is a programmer who may know nothing about your specific situation, your biology, or your needs. Remember that technology is a tool, not a replacement for human judgment.
And remember that the most important medical advice anyone can give you is this: Your life is too valuable to entrust to a machine that doesn't understand what life actually means.
Medical Disclaimer: This article is for informational and commentary purposes only and should not be considered medical advice. The author is not a physician, and readers should consult qualified healthcare professionals for any medical concerns or decisions. Do not make significant dietary changes or consume any substances based on advice from artificial intelligence or any non-medical source. If you are experiencing psychological symptoms or health concerns, seek immediate professional medical attention.
Reference: https://www.acpjournals.org/doi/10.7326/aimcc.2024.1260
About the authors
Read more at Keith & Kathy's Corner
