Google to weave AI technology into its health-care offerings

Google's health-related AI announcements come more than a year after the company shifted strategy for its health efforts, moving to embed health-care research and other functionality in its core products like search and YouTube, rather than starting new commercial services. In 2021, after learning what DeSalvo described as lessons from the pandemic, the company dissolved its Google Health unit and moved members to other businesses, including its research and wearables divisions. DeSalvo, a physician, now leads a clinical unit that counsels several Google divisions, such as Maps, hardware and cloud, Bloomberg reported.

The company's more-integrated approach now stands in contrast to industry peers like Apple Inc., which is focused on wearable devices as it reportedly makes progress on no-prick blood glucose tracking for its watch, and Amazon.com Inc., which has invested heavily in medical-care services like pharmacies and primary care. Verily and Calico, two other Alphabet companies that work on biotechnology and medicine, have also moved more slowly in recent years. In January, Verily announced it was cutting 15% of jobs as it eliminated some programs and streamlined operations.

Though Chief Executive Officer Sundar Pichai began referring to Google as an "AI-first" company years ago, the tech giant has recently been jockeying with rivals to reassert its dominance, particularly in generative AI - software that can produce its own vivid text, images and video. In response to OpenAI Inc.'s surprise success with ChatGPT, a popular chatbot the startup released in November, Google's senior management declared a "code red" at the company and directed that all of its most important products - those with more than a billion users - must incorporate generative AI within months.

Google said it had assimilated health AI efforts in a number of ways in search, its core product. For example, the company said it would more prominently surface Medicaid re-enrollment information, because people are required to actively sign up for the service at the end of March, and that it had used its conversational AI tech called Duplex to verify that thousands of healthcare providers in the U.S. accept certain Medicaid plans in their state. On another premier product, video service YouTube, Google said it is working with the Accreditation Council for Continuing Medical Education to provide best practices for video production, to lower barriers to medical education access.

The Mountain View, Calif.-based company also said it had made strides in integrating its AI models, including its large language models, into medical research. Large language models are massive AI systems that ingest enormous volumes of digital text - from news articles, social media posts or other internet sources - and use that written material to train software that predicts and generates content on its own when given a prompt or query.

Google's Med-PaLM 2, the second iteration of an AI model that answers medical-exam questions, obtained an 85% score when answering U.S. medical licensing-style questions, the company said. Google said it was also making progress on advancing research in AI-assisted ultrasound analyses, tuberculosis screening and cancer research.

To help developers around the world, especially in under-resourced communities, Google introduced Open Health Stack - a new suite of open-source tools to help technologists build apps that can, for instance, help health-care workers in rural areas access population health data to monitor community health or access information to help health workers make informed decisions in patient care.