Man is poisoned after following dietary advice from ChatGPT
While ChatGPT may have some uses, it should not be used to seek dietary advice.
Admittedly, ChatGPT is a cheaper alternative to seeking diet plans from trained nutritionists, but it comes with a sizeable risk attached, as one 60-year-old man has discovered. After allowing ChatGPT to decide what he eats, the man in question poisoned himself and ended up in the hospital.
The ChatGPT user wanted to eliminate salt from his diet and used AI to decipher how to do so, which turned out to be a catastrophic decision. He swapped out sodium chloride, which is everyday table salt, for sodium bromide and wound up in a hospital bed after three months on his newfound diet.
The story came to light in the latest journal by Annals of Internal Medicine, which noted that ChatGPT likely thought the substitute wasn’t for dietary purposes but for tasks such as cleaning.
In the newly published case study, researchers said, “This case also highlights how the use of artificial intelligence (AI) can potentially contribute to the development of preventable adverse health outcomes.”
They explained further: “Unfortunately, we do not have access to his ChatGPT conversation log and we will never be able to know with certainty what exactly the output he received was, since individual responses are unique and build from previous inputs.”
The researchers also confirmed that it’s “highly unlikely” that medical professionals would ever recommend a patient takes sodium bromide as a substitute for sodium chloride.
Researchers concluded: “It is important to consider that ChatGPT and other AI systems can generate scientific inaccuracies, lack the ability to critically discuss results and ultimately fuel the spread of misinformation.”
Another symptom of the dietary swap was paranoia, and the man believed that his neighbours had poisoned him. Other symptoms included acne and insomnia, which were all brought on due to the bad advice.
While it shouldn’t need to be said, do go to a trained professional rather than AI for medical advice, unless you want to become a paranoid insomniac in a hospital bed. Otherwise, your most embarrassing moment could wind up being used as a case study in a research journal.