ChatGPT’s Healthcare Responses are Comparable to Human Doctors, Study Finds
Have you ever had a conversation with an AI chatbot? These advanced robots, capable of thinking, learning, and conversing like humans, are becoming increasingly common. OpenAI, a leading artificial intelligence company, has recently developed a chatbot named ChatGPT that possesses these remarkable abilities. While ChatGPT can discuss a broad range of topics, including healthcare, researchers from New York University wanted to determine if it could provide answers to medical queries on par with real doctors.
To assess ChatGPT’s capabilities, the NYU research team conducted a study involving 392 participants of different ages and genders. Each participant was presented with ten common questions that patients often pose to doctors. For each question, two answers were provided – one from a real doctor and the other from ChatGPT. The participants were unaware of who provided each response and had to guess accordingly. Additionally, they were asked to rate their level of trust in the robot’s answers on a scale of 1 to 5, with 1 representing no trust and 5 indicating complete trust.
To the surprise of the researchers, the results revealed that participants struggled to differentiate between the responses given by ChatGPT and those provided by actual doctors. Remarkably, their accuracy rate for correctly identifying the source of the answer was approximately 65% for both the robot and the doctors. Moreover, the participants acknowledged that they trusted ChatGPT, albeit to a varying extent. On average, the robot received a trust score of 3.4 out of 5. While this indicates a moderate level of trust, there is room for improvement.
Notably, the level of trust varied depending on the nature of the question. Participants placed the most trust in ChatGPT when it addressed straightforward concerns such as scheduling doctor appointments or providing details on insurance. For these types of inquiries, the robot received an impressive trust score of 3.94, almost reaching 4, which can be considered quite good. Similarly, when it came to advice on maintaining good health, such as obtaining vaccinations or undergoing cancer screenings, participants deemed ChatGPT reasonably trustworthy, giving it a score of 3.52.
However, the scenario changed when it involved diagnosing and treating illnesses. In these cases, participants expressed less confidence in ChatGPT’s responses, assigning it a trust score of only 2.9. Clearly, when it comes to crucial and nuanced medical advice, people still prefer to consult a human doctor.
This study provides valuable insights into the potential role of robots like ChatGPT in facilitating doctor-patient conversations. While they excel at addressing simple inquiries and assisting with administrative tasks, they lag behind human doctors in providing medical advice. It is important to remember that robots, despite their advancements, lack human elements such as empathy and firsthand understanding of illness. Nonetheless, the continued development of robots like ChatGPT may lead to enhanced abilities in offering medical guidance in the future.
AI robots are not confined to research studies; they are increasingly present in everyday settings. You may have encountered one while contacting your phone company or interacting with a smart speaker. These robots possess diverse capabilities but remain limited in certain areas. They are tools designed to assist us, not replace us. Ultimately, we still rely on human doctors who possess the deep understanding and empathy necessary for effective healthcare.
In conclusion, robots like ChatGPT can undoubtedly offer valuable support, particularly for managing administrative tasks and addressing simple inquiries. However, they cannot completely replace human doctors when it comes to providing personalized and intricate advice on medical conditions. It is crucial to exercise caution and use these robots wisely. They can be helpful allies in the realm of healthcare, but they should always be complemented by the expertise and human touch of real doctors. By understanding and respecting the limitations of AI technology, we can embrace a future that combines the benefits of robot assistance with the unique qualities that healthcare professionals bring to the table.
