According to a report from The Wall Street Journal, Google is currently conducting tests of its Med-PaLM 2 AI chat technology at the Mayo Clinic and other medical facilities. Med-PaLM 2 is based on Google’s PaLM 2 large language model (LLM), which is also the foundation for Bard, Google’s competitor to ChatGPT. The technology was recently introduced at Google I/O.
Med-PaLM 2 is different from the base model as it has been specifically trained on questions and answers from medical licensing exams, as well as a curated collection of demonstrations by medical experts. This specialized training equips it with the ability to provide knowledgeable responses to health-related queries. Additionally, Med-PaLM 2 can perform labor-intensive tasks such as document summarization and research data organization.
During Google I/O, the company published a paper outlining the development of Med-PaLM 2. The paper highlighted positive features such as “alignment with medical consensus,” reasoning capabilities, and generating answers preferred by respondents over those generated by physicians. However, it also acknowledged accuracy issues similar to those encountered with other AI chat models.
Microsoft is also working on its own medical AI chat technology, leveraging OpenAI’s ChatGPT and partnering with healthcare software company Epic. In March, Google revealed its efforts in utilizing AI for ultrasound diagnosis and cancer therapy. Both companies have assured the confidentiality of patient data and emphasized that their models are not trained on such information. Microsoft recently expressed concerns about doctors using ChatGPT for improving patient communications.
In an internal email obtained by the WSJ, Google expressed its belief that the updated model could have significant value in regions with limited access to doctors. However, Google acknowledges that the technology is still in its early stages. Greg Corrado, Google’s senior research director, stated that he does not consider this technology suitable for his family’s healthcare at present but believes it has the potential to greatly enhance healthcare practices.
Please note that the content provided is for informational purposes only and does not constitute endorsement or recommendation by BuyTechBlog. The publication’s editorial team independently selects products, and some articles may contain affiliate links.
Frequently Asked Questions (FAQs) about AI chatbot
What is Med-PaLM 2?
Med-PaLM 2 is an AI chatbot developed by Google based on a large language model. It is designed to provide expertise in answering medical queries and assisting with tasks like document organization.
Where is Med-PaLM 2 being tested?
Med-PaLM 2 is currently being tested at the Mayo Clinic, along with other hospitals.
What sets Med-PaLM 2 apart from the base model?
Unlike the base model, Med-PaLM 2 has been specifically trained on medical licensing exam questions, along with expert demonstrations. This specialized training equips it with knowledge in health-related topics.
Can Med-PaLM 2 perform labor-intensive tasks?
Yes, Med-PaLM 2 is capable of performing labor-intensive tasks such as summarizing documents and organizing research data.
Are there any concerns about accuracy with Med-PaLM 2?
Yes, similar to other AI chat models, Med-PaLM 2 has shown accuracy issues. However, Google is actively working on addressing and improving this aspect.
Is patient data used to train Med-PaLM 2?
No, both Google and Microsoft, the developers of the AI chatbot technology, have assured that patient data is not used to train their models. They prioritize patient confidentiality.
How does Google envision the potential of Med-PaLM 2?
Google sees potential for Med-PaLM 2 to be of significant value, particularly in countries with limited access to doctors. However, they also acknowledge that the technology is still in its early stages of development.
More about AI chatbot
- The Wall Street Journal: Google Tests Its AI Chatbot at Mayo Clinic
- Google I/O: Med-PaLM2: A Large, Multilingual, Domain-Specific Language Model for Medical Conversations
- Microsoft News: Microsoft and Epic Collaborate to Empower Clinicians with AI-Powered Virtual Health Assistants
- Google Blog: Exploring the Use of AI in Healthcare