Not a day goes by when the talk of technology revolves around ChatGPT. Thousands of articles are driving perspectives around the use of ChatGPT, as well as cautionary tales noting how the use of AI and solutions like ChatGPT can go wrong or even harm businesses and individuals.

Like any new innovation, there is always great skepticism and a sense of excitement and opportunity as well. The good news is that AI innovation is clearly alive and kicking and already proving itself with various healthcare tech use cases. Furthermore, when competitors like Google and Microsoft catch a wave of innovations, not necessarily created by themselves, they immediately jump on the bandwagon and push hard to introduce innovations around AI and ChatGPT in the context of their business and solution offerings for fear of being left behind. In healthcare, we see the same thing happening: optimism, skepticism, and a hard push to embrace ChatGPT and other AI solutions that may help the healthcare space where it is practical and valuable to the patient journey and engagement.

Consumers have had many years of experience looking for healthcare content relating to their own healthcare experiences using search engines, reading content on healthcare-specific content sites, watching YouTube videos, and reading through blog posts and forums. With ChatGPT, patients now have access to yet another avenue to help them in their healthcare journey. However,  like search engines and other sources of healthcare information, patients should always tread carefully when it comes to ChatGPT and always consult with a physician or medical professional as they would normally do. As far as healthcare and ChatGPT is concerned, it is important to pay attention to the following perspective:  

ChatGPT is a very broad language model trained mostly on large amounts of publicly available internet content; however, ChatGPT does not possess the highly specialized medical knowledge and training of experienced healthcare professionals. It may provide inaccurate or misleading information, leading to incorrect diagnoses, treatment recommendations, or advice.

With a limited understanding of a particular context: The ChatGPT language model may struggle to understand the specific context of a particular medical scenario. It may misinterpret or fail to capture critical details the interested patient provides, leading to flawed responses or recommendations.

Healthcare data is highly sensitive and subject to strict privacy regulations such as HIPAA. When using ChatGPT or any other language model solutions, there is a risk of data breaches or unauthorized access to patient information. Robust security measures must be in place to protect patient privacy. Patients should always be careful about asking specific and detailed healthcare questions with personal information as where that data ends up is only partially understood at this point.

If a patient’s healthcare decisions or actions are based solely on ChatGPT advice or information, there can be potential legal and ethical implications. Patients must ensure that their healthcare decisions are made by highly qualified professionals who are held accountable for their actions.

Language models can reflect and even amplify the biases present in the data they are trained on. ChatGPT may inadvertently provide biased or discriminatory responses if the training data contains biased information sources from biased content found on the internet, potentially exacerbating existing healthcare disparities.

Very importantly, ChatGPT simply lacks empathy and the ever-so-important human touch. While language models can provide potential helpful information and assistance, they cannot replace the human touch and empathy that healthcare professionals offer. Patient care involves emotional support and nuanced understanding, which may be lacking when relying solely on solutions such as ChatGPT and other AI solutions.

To mitigate these concerns, it is essential to consider ChatGPT, AI solutions, and similar language models as tools to support healthcare professionals rather than replace them. When integrating these models into healthcare systems, critical validation, verification, and supervision processes must be implemented. Furthermore, ongoing, robust research and development efforts are necessary to address these challenges and enhance the safety and effectiveness of AI applications and ChatGPT in healthcare.

Learning about your healthcare situation through ChatGPT might help you drive your healthcare literacy so that you can help yourself lead a healthier life and be more informed like other sources of information. However, it is important to note that while ChatGPT can support patient engagement, it should not replace human healthcare professionals. ChatGPT can complement healthcare professionals’ work by providing accessible information and support, enhancing the overall patient experience. We are early on this journey with solutions like ChatGPT and hope that it continues to evolve to help us all in a meaningful and helpful way in the context of our healthcare journey.