Med-PaLM 2 is a medical AI chatbot that Google has been testing in hospitals, including the Mayo Clinic research hospital. The AI chatbot has reportedly been tested since April, according to a WSJ story. An AI tool called Med-PaLM 2 is made to answer medical-related queries. PaLM 2, which was introduced at Google I/O in May, has been updated. To enhance its capabilities for healthcare interaction, the chatbot is trained on a selected collection of demos from medical experts.
According to an internal email cited in the paper, countries with poor access to physicians can benefit most from Google’s Med-PaLM 2. However, as noted in a study released by Google in May, there are some accuracy problems with the chatbot. When compared to other doctors’ responses, doctors found Med-PaLM 2’s to contain more errors and unrelated material.
Despite the accuracy problems, Med-PaLM 2 did well on other metrics like offering responses that were supported by agreement and displaying accurate comprehension. Other well-known chatbots that are accessible to the general public, such as OpenAI’s ChatGPT and Google’s own Bard, have also experienced accuracy problems.
Additionally, according to the article, users testing Med-PaLM 2 will have control over their encrypted data and Google won’t have access to it.
Med-PaLM 2 is still in its early phases, according to Greg Corrado, Google’s senior research director (quoted in the story). Although he personally wouldn’t rely on it for his family’s health, he thinks that Med-PaLM 2 considerably increases the potential advantages of AI in healthcare.
