Is Dr ChatGPT the new Dr Google?
Self diagnosis via "googling" is common practice, but are we now trusting AI to tell us what's wrong?
I heard a snippet of an investigative news story on the radio recently where the presenter was talking about people turning to AI such as ChatGPT for mental health support. This got me thinking about health-related questions posed to AI chats, and how often people are conversing with a chatbot to seek answers to health concerns.
Several studies of user interaction and intention suggest a willingness to engage in AI for health advice1. But why are we so willing? How have we become so trusting of AI to provide a health diagnosis?
Artificial empathy
The answer may lie in AI’s ability to communicate with empathy. In researching for this article, I discovered the term “artificial empathy” in which machines are coded to provide empathetic responses. After all, is this not what we are seeking—to be seen, heard and validated? Studies show chatbots respond more empathetically than physicians, whether in person or online2.
This finding leads to the obvious question—can AI, via machine learning and deep learning, improve our healthcare practice and patient care? Apparently yes, and not surprisingly, this is an area currently being investigated3.
It’s a little ironic that we need to teach machines how to be empathetic, and then these same machines are being used to teach us how to do it in practice.
AI diagnostics
Although AI may not (yet) have the skills to diagnose as accurately as its human peers, it does show promise to both improve its diagnostic capability, and teach human health practitioners.
Does this mean that human health practitioners will become redundant? The easy and less scary answer is no. But I don’t think it’s an easy question to answer.
I’ve seen substantial change in the industry I’m about to enter professionally as a nutritionist. I feel these have been led by a proliferation (and acceleration) of robust evidence, and the rapid advance of machine learning and deep learning.
Health topics of interest
As I sifted my way through research on this topic (read ‘death by database search’) I found references to search query categories, and was perhaps not surprised to learn that “diet, fitness and lifestyle planning” seems to be the number one health-related query category in AI chats.4
Not only is it the most popular, but ChatGPT dietary recommendations have been shown to be indistinguishable to—or even outperform—advice from human nutritionists and dieticians5. This definitely piqued my interest as a student nutritionist!
AI nutritional assessment
AI shows promise in improving data collection used in nutritional assessment, such as recording food intake, which is inherently difficult due to poor client recall and inaccurate measurement of food consumed6. AI can assess and record macro- and micro-nutrients, and calories, by image recognition.
Another area of potential is its ability to predict disease risk, thus providing the tools to mitigate or reduce symptoms of future health conditions. This excites me most, since AI could play an integral role in guiding and improving a preventative health model.
Circling back to the original question I posed for this article—is Dr ChatGPT the new Dr Google?—it seems we are heading that way. And its prospects seem a lot more convincing than internet searches. Although there are limitations, and dangers, the advantages of embracing AI in health and nutrition science are too great to dismiss.
Have you ever asked AI about your health? Leave a comment below and tell me your story!
Shahsavar, Y., & Choudhury, A. (2023). The Role of AI Chatbots in Healthcare: A Study on User Intentions to Utilize ChatGPT for Self-Diagnosis (Preprint). JMIR Human Factors, 10. https://doi.org/10.2196/47564
Armbruster, J., Bussmann, F., Rothhaas, C., Titze, N., Grützner, P. A., & Freischmidt, H. (2025). “Doctor ChatGPT, Can You Help Me?” The Patient’s Perspective: Cross-Section...: EBSCOhost. Ebscohost.com. https://web.p.ebscohost.com/ehost/detail/detail?vid=15&sid=d5824ec7-e32d-41b2-bcb5-e7d777973a13%40redis&bdata=JkF1dGhUeXBlPXNoaWImc2l0ZT1laG9zdC1saXZl#AN=182059066&db=iih
Marrow, E., Zidaru, T., Ross, F., Mason, C., Patel, K. D., Ream, M., & Stockley, R. (2025). Artificial intelligence technologies and compassion in healthcare: A system...: EBSCOhost. Ebscohost.com. https://web.p.ebscohost.com/ehost/detail/detail?vid=9&sid=d5824ec7-e32d-41b2-bcb5-e7d777973a13%40redis&bdata=JkF1dGhUeXBlPXNoaWImc2l0ZT1laG9zdC1saXZl#AN=36733854&db=mdl
Karami, A., Qiao, Z., Zhang, X., Hadi Kharrazi, Parisa Bozorgi, & Bozorgi, A. (2024a). Health Use Cases of AI Chatbots: Identification and Analysis of ChatGPT Prompts in Social Media Discourses. Big Data and Cognitive Computing, 8(10), 130–130. https://doi.org/10.3390/bdcc8100130
Guo, P., Liu, G., Xiang, X., & An, R. (2025). From AI to the Table: A Systematic Review of ChatGPT’s Potential and Perfor...: EBSCOhost. Ebscohost.com. https://web.p.ebscohost.com/ehost/detail/detail?vid=23&sid=d5824ec7-e32d-41b2-bcb5-e7d777973a13%40redis&bdata=JkF1dGhUeXBlPXNoaWImc2l0ZT1laG9zdC1saXZl#AN=184204507&db=cul
Theodore , A., Poupi, T., Nfor, K. A., Kim, J.-I., & Kim, H.-C. (2024). Applications of Artificial Intelligence, Machine Learning, and Deep Learnin...: EBSCOhost. Ebscohost.com. https://web.p.ebscohost.com/ehost/pdfviewer/pdfviewer?vid=26&sid=d5824ec7-e32d-41b2-bcb5-e7d777973a13%40redis