Google has been conducting tests of its medical AI chatbot, Med-PaLM 2, in various hospitals, including the renowned Mayo Clinic research hospital. The AI chatbot has been undergoing testing since April, as reported by WSJ.

Med-PaLM 2 is an advanced AI tool specifically designed to provide answers to medical inquiries. It is an updated version of PaLM 2, which was introduced at Google I/O in May. The chatbot is trained using a curated collection of medical expert demonstrations, enhancing its conversational capabilities in the healthcare domain.

While Med-PaLM 2 shows potential in aiding countries with limited access to doctors, there have been concerns about its accuracy. A research paper published by Google in May highlighted inaccuracies and irrelevant information in the chatbot's responses, as identified by physicians, compared to answers provided by human doctors.

However, Med-PaLM 2 has demonstrated strength in other aspects, such as displaying evidence of reasoning, providing consensus-supported answers, and exhibiting correct comprehension. Similar accuracy issues have been encountered with other widely used chatbots, including OpenAI's ChatGPT and Google's own Bard.

Furthermore, the report suggests that customers testing Med-PaLM 2 will have control over their data, with encryption in place to ensure that Google does not have access to it.

According to Greg Corrado, Google's senior research director mentioned in the report, Med-PaLM 2 is still in its early stages. While he personally wouldn't rely solely on it for his family's healthcare, he believes that Med-PaLM 2 significantly expands the potential benefits of AI in the healthcare sector.