Google's Medical AI Chatbot 'Med-PaLM 2' Enters Hospital Testing Phase
Google's medical AI chatbot, Med-PaLM 2, has entered the testing phase at prominent research hospitals such as the Mayo Clinic, according to a report by The Wall Street Journal. Med-PaLM 2 is a modified version of PaLM 2, which was introduced at the Google I/O conference in May. PaLM 2 serves as the language model behind Google's Bard.
The report states that an internal email indicated Google's belief that the updated model could be particularly valuable in regions with limited access to healthcare professionals. Med-PaLM 2 underwent training using a carefully curated collection of medical expert demonstrations, which Google believes will enhance its performance in healthcare conversations compared to general chatbots like Bard, Bing, and ChatGPT.
The article also references a research paper released by Google in May, which revealed that Med-PaLM 2 still exhibits some of the accuracy issues commonly associated with large language models. Physicians participating in the study identified more inaccuracies and irrelevant information in the answers provided by Google's Med-PaLM and Med-PaLM 2 compared to responses from other doctors.
Nevertheless, Med-PaLM 2 performed reasonably well in various other metrics, including demonstrating evidence of reasoning, providing consensus-supported answers, and displaying correct comprehension.
WSJ reports that customers involved in testing Med-PaLM 2 will retain control over their encrypted data, ensuring that Google does not have access to it.
Greg Corrado, Google's senior research director, mentioned that Med-PaLM 2 is still in the early stages of development. While Corrado admitted that he would not want it to be part of his family's healthcare journey at this point, he believes that Med-PaLM 2 has the potential to significantly expand the areas where AI can be beneficial in healthcare.