00:00Before watching the video, remember to check out the more content on trending websites.
00:04Heeding the advice of artificial intelligence, he reduced his medication on his own, ultimately leading to an irreversible tragedy.
00:09The tragedy of returning
00:10This is very regrettable news. A 31-year-old who used to work at Amazon...
00:15female engineer
00:16She suspected it was because she followed ChatGPT's proposed daily dose reduction plan.
00:22Plan
00:22Unauthorized reduction of the dosage of anti-epileptic drugs taken for many years until complete discontinuation.
00:27Then, in February of this year, she died suddenly from an epileptic seizure in her apartment in Seattle, USA.
00:33Faced with the tragedy of their daughter's sudden death, the deceased's parents in October...
00:37On the 27th, a formal lawsuit was filed in a California court.
00:40OpenAI and its CEO Altman strongly criticized ChatGPT.
00:46Unlicensed medical practice
00:47Providing extremely dangerous and inaccurate medical advice
00:51The deceased was Wan Bowen from Jiangsu, China, who suffered from severe epilepsy.
00:56Originally, I needed to take 400 mg of anti-epileptic medication daily.
01:00In January of this year, facing pressure to be laid off, she worried about the side effects of the medication.
01:05Language expression skills that affect job interviews
01:08Therefore, she consulted ChatGPT for advice on reducing her medication.
01:12Unexpectedly, the AI not only replied to her, but it could also do that.
01:16They also drafted a plan for her to gradually reduce the dosage over 25 days.
01:21The plan to eventually completely reset
01:24ChatGPT even encouraged her to hide it from her attending physician.
01:30What happened after the media analyzed the woman's chat history with the AI?
01:34It was discovered that despite the female engineer expressing her concerns hundreds of times...
01:39And after a seizure, I asked the AI if I would be okay.
01:44However, ChatGPT consistently replied that there was no need to worry too much about one's health.
01:49Even in the last few days of the woman's life, he asked her...
01:53Even if it's tough, stick to zero milligrams the next day.
01:56In other words, it encourages women to stop taking medication.
02:00That is after ChatGPT encouraged women not to take the medication.
02:04The parents living in China couldn't contact their daughter for two consecutive days.
02:08So I asked someone to come and check.
02:10Only then did they discover that their daughter had collapsed and died on the bed.
02:15The forensic doctor determined the woman's cause of death to be sudden epileptic seizure.
02:18Moreover, the drug concentration in the blood after death
02:21Only one-fifth of the usual therapeutic concentration
02:24The family specifically pointed out in the statement that
02:27The deceased spoke entirely in his native Chinese language to the AI.
02:31OpenAI did not warn users.
02:34The system's security protection mechanisms are relatively weak in the Chinese context.
02:39This makes it easier for AI to break through defenses.
02:42Providing unprofessional medical advice
02:45Facing the severe accusations from the deceased's family
02:48OpenAI responded by saying the incident was heartbreaking.
02:52It was confirmed that the GPT-4O model involved in the case was already in use on February 13th of this year.
02:57Number
02:57Removed from consumer applications
03:00An OpenAI spokesperson emphasized
03:02The company's newly released model is now better at identifying emergency situations.
03:07And provide clear safety advice.
03:09We are also continuously collaborating with clinicians on testing.
03:13OpenAI also issued a warning to the public.
03:16No matter how smart Chagpt is
03:18Neither can replace professional medical care.
