Health Man 60 poisoned himself after taking medical advice from ChatGpt

[CharlieH]

 

A 60-year-old American man faced severe health issues after swapping table salt for sodium bromide, a chemical used in pool cleaning, on the advice of ChatGPT. This decision resulted in hallucinations, paranoia, and severe anxiety, leading to a three-week hospital stay.

 

Sodium bromide, once used in sedatives, is now rare due to its toxic effects, which include bromism. This condition causes psychosis, delusions, and nausea, highlighting the risks associated with misusing such compounds. The man reported his neighbour was poisoning him, a claim dismissed by medical staff.

 

Doctors, curious about the AI's advice, found that ChatGPT still recommended sodium bromide as a salt substitute without mentioning potential health risks. Published in the Annals of Internal Medicine, this incident emphasizes the risks of AI-generated misinformation, particularly regarding health.

 

This isn't the first AI blunder. A Google bot previously suggested consuming rocks for health—incorrect advice potentially sourced from satirical content. OpenAI, the company behind ChatGPT, advises against using its tools as the sole source of factual information, encouraging users to seek professional guidance.

 

Clinical psychologist Paul Losoff warns of the growing dependency on AI, stressing that it can exacerbate mental health issues. Particularly for those with conditions like anxiety or depression, reliance on AI might worsen symptoms such as distorted thinking or pessimism. Losoff highlights the potential for AI interactions to be misinterpreted, especially in individuals with thought disorders like schizophrenia.

 

He cautions that AI's potential for error and harmful advice underscores the necessity of human oversight and professional consultation. As AI technologies continue to evolve, balancing their use with expert guidance becomes increasingly crucial to prevent adverse health outcomes. The case serves as a reminder of the importance of caution and critical thinking when integrating AI into daily life.

 

 

  Adapted by ASEAN Now from The Daily Mail 2025-08-13

 

 

[James105]

AI is not a cure for stupidity and until such time a brain implant is invented that fuses the human brain with (genuine) AI then stupid people will continue to do stupid things.   I'm reminded of similar kinds of stories when GPS was first implemented of people driving into rivers etc as they relied too much on the computer maps instead of simply using their eyes and coming to the conclusion that the map might not be 100% accurate.