Artificial Intelligence (AI) is making its way into healthcare, from appointment reminders to automated billing. However, the question remains: Can patients rely on platforms like ChatGPT and Google Bard for medical advice? To find out, Samiran Mukherjee, chief fellow in gastroenterology at the Perelman School of Medicine at the University of Pennsylvania, conducted a study in February 2023 to test the effectiveness of ChatGPT as a tool for providing medical information. The results, published in the medical journal Gastro Hep Advances in July, shed light on the potential benefits and limitations of AI in healthcare.

Mukherjee emphasized that ChatGPT's ability to mimic human conversation is astounding due to the wealth of data it possesses. The study primarily focused on questions regarding colonoscopies. While the answers were easy to understand and seemed human-like, some of the information provided by ChatGPT was found to be incorrect. The AI performed well with simple, single objective questions but struggled to provide accurate answers to two-part questions. Mukherjee suggested that ChatGPT can serve as a starting point for general information gathering, especially for patients who may be unfamiliar with medical procedures.

One advantage of using ChatGPT over popular search engines like Google is the ability to provide a more comprehensive list of possible causes for symptoms. While Google often offers alarmist diagnoses, ChatGPT presents a range of potential causes in a more ordered manner. Mukherjee highlighted the importance of relevant search results being prioritized, making ChatGPT and Google Bard potentially useful alternatives in such cases.

The future of AI in healthcare may also involve reading and interpreting medical test results such as CT scans or MRI scans. Trials are currently underway to explore the potential for AI to recognize patterns and process information faster than humans. However, Mukherjee emphasized that AI should be seen as a collaborative tool rather than a replacement for human expertise. While AI can provide accurate predictions, human professionals are essential in delivering counseling and empathy, which are crucial aspects of the patient-physician relationship.

Despite the potential benefits, it is important to be cautious when using AI in healthcare. Patients and medical workers should not overestimate the results and should be aware of the limitations of these technologies. Previously, early versions of search engines like Google provided potential diagnoses without proper context, leading to unnecessary anxiety. To mitigate these risks, robust rules and guidelines are being developed by organizations such as the FDA and the World Association of Medical Editors to ensure appropriate use of AI in healthcare.

Mukherjee emphasized that incorporating new technologies into healthcare, with proper collaboration between AI and human professionals, can lead to improved patient care and access to information and diagnoses. By understanding the limitations and continually refining AI, the healthcare industry can navigate the challenges and take advantage of the benefits offered by AI in the future.