| | | | |  | | By Evan Peng, Daniel Payne, Erin Schumaker and Ruth Reader | | | | | 
Google is honing its health AI offerings. | Getty Images | Google is fine-tuning its family of artificial intelligence models for the health care industry, and they’re now available for Google Cloud customers in the U.S. The family of models, called MedLM, builds on Google’s existing products, including Med-PaLM 2, the large language model that’s the basis of the MedLM models. MedLM consists of two models: a large one that’s designed for complex tasks and a medium-sized model that can be modified for specific use cases, an attribute Google says is best for scaling across tasks. What’s it for? The blog post announcing MedLM presented several use cases for the AI models from companies Google has piloted the technology with. — HCA Healthcare, a health care facility operator, is testing out a product from health tech company Augmedix. The product uses MedLM and natural language processing to help doctors convert their conversations with patients into medical notes that can be inserted into their electronic health records. — BenchSci, a health tech company whose ASCEND platform is an AI-powered tool to expedite drug development, is integrating MedLM into ASCEND to improve the speed and quality of preclinical research and development. — Deloitte, a consulting firm, is working with Google Cloud to build a chatbot to help corporate clients better understand the provider options that their insurance covers. What’s next? Google continues to test Med-PaLM 2 with health care organizations and, in the coming months, intends to add MedLM models built with Gemini, Google’s largest and strongest AI model.
| | | | 
New York City | Shawn Zeller/POLITICO | | This is where we explore the ideas and innovators shaping health care. Fertility drugs' high cost is pushing some patients to the black market, Bloomberg reports. After spending thousands of dollars on drugs, IVF and egg freezing, patients are often left with excess medication, and some choose to sell or trade it illegally using listervs, Facebook and Discord groups. Share any thoughts, news, tips and feedback with Carmen Paun at cpaun@politico.com, Daniel Payne at dpayne@politico.com, Evan Peng at epeng@politico.com, Ruth Reader at rreader@politico.com or Erin Schumaker at eschumaker@politico.com. Send tips securely through SecureDrop, Signal, Telegram or WhatsApp. Today on our Pulse Check podcast, host Chelsea Cirruzzo talks with POLITICO health care reporter Robert King, who breaks down the Biden administration's efforts to curb high prescription drug prices by requiring drugmakers receiving research funding from one HHS agency to commit to a fair price, or no higher than they charge in other countries.
| | | | | | | 
AI is built on a long history of tech advancement. | Getty Images | AI tech advancing at “hypersonic speed” offers novel ways to face health systems’ financial and workforce problems with near equal force — at least that’s how Sunil Dadlani, chief information and digital transformation officer at Atlantic Health System, sees it. At a Healthcare Information and Management Systems Society AI summit in San Diego this week, Dadlani laid out the biggest challenges for health care in the 21st century, suggesting they’re also the avenues for AI’s biggest gains: labor shortages, clinician burnout and declining profitability. Whether by reducing paperwork, stretching the abilities of clinicians or making care more cost-effective, AI could help solve health executives’ problems. “Generative AI is playing a dominant role,” he said, adding: “I don’t think there is any facet in the health care industry that will not be touched by AI.” How did we get here? Other predictive AI technologies have paved the way — and empowered — the current generative and language-focused AI that mimics human intellect and communication, he said. Hardware advances shouldn’t be discounted either, he said. Establishing 5G networks has allowed devices to work together in new ways, and advances in processing power have made the systems more practical. Even so: Some industry leaders caution that the current iteration of AI is limited in what it can do and shouldn’t be counted on to move beyond its proven scope.
| | | | 
Cooperation between tech companies and health care organizations aims to speed development of AI tools. | Justin Sullivan/Getty Images | Health care organizations are standing up partnerships with tech companies — and each other — as they consider AI’s applications in their work. One reason: to speed development of tools that have proven they’re worth further investment. No matter how flashy the new tech is, health executives and makers of AI products told Daniel, they must have a solid business case behind them. “We’ve got to deliver results,” Albert Martinez, the Cleveland Clinic’s chief analytics officer, said at a HIMSS panel about strategically bringing artificial intelligence into health organizations. “In many cases, that’s really hard for us to do in the short term as an internal team.” Cue the partnerships: The Cleveland Clinic is at the forefront of that effort, joining the new AI Alliance with a number of companies and nonprofits across sectors. They hope to share resources and lessons learned to make better AI tools faster. And those in health organizations championing the new tech hope to parlay those advancements into further internal investments. That investment includes hiring a new kind of worker for these organizations: people with roles centered around AI’s use and development. “We’ve got to develop new talent internally,” Martinez said, highlighting the interplay of internal and external work to develop AI for health systems. | | | | Follow us on Twitter | | | | Follow us | | | | |
