Google’s medical AI chatbot is already being tested in hospitals

AI News


Google’s Med-PaLM 2, an AI tool designed to answer questions about medical information, has been in testing since April at the Mayo Clinic Research Hospital and others. wall street journal Reported this morning. Med-PaLM 2 is a variant of PaLM 2 announced at Google I/O in May of this year. PaLM 2 is the language model behind Google’s Bard.

The paper also touches on a study (pdf) published by Google in May, in which Med-PaLM 2 still has some of the accuracy problems we are already accustomed to seeing in large language models. is shown. In this study, doctors found more inaccuracies and irrelevant information in the answers provided by his Med-PaLM and Med-PalM 2 on Google than other doctors’ answers.

Nevertheless, Med-PaLM 2 performed more or less as well as real physicians on nearly every other measure, whether showing evidence of reasoning, giving a consensus-supported answer, or showing no signs of misunderstanding. showed.

WSJMore Customers testing Med-PaLM 2 control their data and report that it is encrypted and that Google can no longer access it.

Greg Collard, senior research director at Google, said: WSJMore He said Med-PaLM 2 is still in its early stages. Corrado doesn’t want to be part of his family’s “healthcare journey,” but Med-PaLM 2 “occupies a place where AI can be beneficial in health care, magnifying it tenfold.” said he believed.

We contacted Google and the Mayo Clinic for more information.



Source link

Leave a Reply

Your email address will not be published. Required fields are marked *