Of the countless applications of AI, the one for hospitals and medical solutions is certainly among the most fascinating. He knows it well Google which he presented to the world Med-PaLM 2, an artificial intelligence developed to operate in the healthcare sectorwhich has been in testing at the Mayo Clinic research hospital since April.
As reported by the Wall Street Journal (via Ansa) Google’s goal would be to offer a useful tool for countries where doctors are in short supply.
How Med-PaLM 2, Google’s artificial intelligence in the medical field, works
Med-PaLM 2 takes advantage of technology LLM (Large Language Model) mentioned in the name for incorporate all a medical database verified by experts. This is how Google provides a reliable conversational chatbot that they should guarantee a higher quality of healthcare conversations compared to other generic and less reliable chatbots (such as Bard, Bing and ChatGpt).
Despite this ambitious goal, Google’s AI is anything but perfect. The Wall Street Journal also reports that Med-PaLM 2 still has some accuracy problems, typical of large language models. However in some reasoning tests the chatobot proved to be at the level of human doctors.
Greg Corrado, senior director of research at Google, said the software is still in its infancy. The Wall Street Journal added that customers who test Med-PaLM 2 are in control of their data, that they are encrypted, and that Google cannot access them. An important piece, given that one of the most heated discussions on artificial intelligence concerns the privacy issue.