Mayo Clinic Trials Google’s AI-Powered Medical Chatbot

According to The Wall Street Journal, Google is currently conducting trials of its Med-PaLM 2 AI chat technology at the Mayo Clinic and various other hospitals. This technology is built upon Google’s PaLM 2 large language model (LLM), which serves as the foundation for Bard, Google’s competitor to ChatGPT. Bard was recently unveiled at Google I/O a few months ago.

Unlike the base model, Med-PaLM-2 is trained on questions and answers from medical licensing exams and presentations from a curated series of medical experts. It provides the expertise to answer health-related questions, and it can also perform labor-intensive tasks such as summarizing documents and organizing research data, according to the report.

During I/O, Google released a paper detailing its work on Med-PaLM2. On the positive side, it demonstrated features such as “consistency with medical consensus,” reasoning ability, and even the ability to generate answers that respondents preferred over doctor-generated answers. On a more negative note, it showed the same accuracy issues we’ve seen in other Chat AI models.

Microsoft is also developing medical AI chat technology based on OpenAI’s ChatGPT in collaboration with healthcare software company Epic. Google is also working to use its AI in ultrasound diagnosis and cancer treatment, it revealed in March. Both companies have promised to keep patient data confidential, saying they do not train their models on patient data. Last month, Microsoft expressed concern that doctors are using its ChatGPT technology to improve communication with patients.

In an internal email seen by the WSJ, Google said it believes the updated model “could have tremendous value in countries where access to doctors is more limited.” Still, Google has admitted that the technology is still in its infancy. “I don’t feel like this kind of technology is at a point yet where I would want it in my family’s healthcare journey,” said Greg Corrado, Google’s senior director of research. However, he added that the technology “takes the places in healthcare where AI can be useful and expands them 10-fold.”