People struggle to get useful health advice from chatbots, study finds

Many people are turning to AI-driven chattabs like ChatzPT for self-development, with long wait list and growing health care system. About 1 in American adults already use chatbot for minimum monthly health consultation, According to the recent one surveyThe

However, it may be risky to have excessive confidence in the outputs of chatbots, because people fight to find out what information will give for the best possible health recommendations, According to a recent study led by OxfordThe

The director of the Graduate Studies of the Oxford Internet Institute and co-authors of the research, told Techcunch, “The study of two-way communication has revealed the breakdown.” “Those who are using [chatbots] The online search or their own judgment did not make a better decision than the participants who depend on the traditional system. “

For study, writers appointed about 1,300 in the UK and gave a medical situation written by a group of doctors. Participants were given the responsibility of identifying potential health situations and using chatboats as well as to determine potential activities (such as a doctor to visit or go to the hospital) to use their own methods.

Participants used the default AI Model Powering ChatzPT, GPT -4O, as well as quo’s command R+ and Meta Lama 3, which once included the company’s meta AI assistant. According to authors, chatbots simply do not reduce the chances of identifying any relevant health conditions of the participants, but it makes the chances of reducing the severity of their identified conditions.

Mahdi said that the participants often received the chattobs when asked or answers that were difficult to explain.

“[T]She reacted to receiving [from the chatbots] Often good and weak recommendations are combined, “he added.” For current evaluation methods [chatbots] Do not reflect the complexity of conversation with human users. “

Technology companies find out the growing AI as a way to improve health results. Apple is Report Develop an AI equipment that can provide practices, diets and sleeping advice. Amazon is searching for an AI-based way to analyze medical databases for “social determinants of health”. And Microsoft is helping AI to create tragedies messages for care providers sent from patients.

However as TechCrunch ReportBoth professionals and patients are mixed up about whether both are ready for higher risky health applications. The American Medical Association suggests against the physician use of chatbots such as ChatzPT for clinical conclusions, and the main AI companies, including OpenAI, recommend a warning against the outputs of their chatbots.

“We would suggest relying on trusted sources of information for healthcare decisions,” said Mahdi. “The current evaluation method [chatbots] Do not reflect the complexity of conversation with human users. Like clinical trials for new drugs, [chatbot] The systems should be tested in the real world before deploying. “