Healthcare AI newswatch: California AI hoping, the jagged AI frontier, agents bumping off bots, more
Buzzworthy developments of the past few days.
- As California goes, so goes the nation. There’s more than a little truth in that maxim. The rub is that sometimes the Golden State leads the other 49 to a better way, other times not so much. In any case, California has moved assertively to legislate guardrails around AI in healthcare. The state’s Physicians Make Decisions Act prohibits payers from using AI to unilaterally deny coverage. Its rules around GenAI disclosure require providers to let patients know which communications were written by the technology. And a bill now under consideration would prevent AI sellers from fooling consumers into mistaking chatbots for healthcare workers. The author of the latter bill, Democrat Mia Bonta, tells a Politico tech publication her fondest wish is for healthcare AI to help lower medical bills. “I could imagine a world in which we are using the crazy amount of data that we have around why healthcare costs are so high to be able to generate some solutions that would be cost-saving to the individual,” Bonta says. “That is my pie-in-the-sky hope for how we’ll proceed with [healthcare AI] legislation moving forward.”
- AI in healthcare is a jagged frontier. Its settlers can see the contours of the unfolding landscape OK. But things seen from afar can look very different once you get up close. The challenge, as one keen observer sees it, is to not only blaze trails—i.e., to build sufficient tech infrastructure to scale AI—but also to train early end-users (call them the blazers) on how to use the tools you give them. “It’s about helping clinicians really understand and appreciate what that jagged frontier looks like, and what they can and should be using this technology for,” Tej Shah, managing director at Accenture, explains for MedCity News. “As often is the case with technology, it’s ‘people and process’ that truly determine the success of AI in healthcare.”
- Choose strategy over speed. That’s a bit of unsolicited but undeniably wise advice on AI adoption for healthcare providers. It comes from the World Economic Forum. “The winners in healthcare’s AI revolution won’t be those who rushed ahead but those who paused to build strong foundations,” the organization states in a March 20 post. The authors extol the advantages of deliberate pilot programs, iterative scaling projects and broad stakeholder engagement as key components to a sound healthcare AI strategy. “AI is not merely a tool for automation,” the post continues. “It is a catalyst for transformation—when implemented responsibly.” Read the rest.
- When will the difference between old-school healthcare bots and healthcare AI agents become clear? When the agents go beyond performing discrete tasks to taking over entire processes. An AI agent might educate patients on unfamiliar medical terms, generate code, speed up the process of reviewing a patient’s medical history or even scan swathes of medical papers to deliver insights, Deloitte healthcare AI specialist Tom Hittinger tells Silicon Angle. Hittinger and Google Cloud healthcare expert Schweta Maniar took a few questions from the outlet while at HIMSS.
- People tend to overestimate the short-term impact of new technologies while underestimating their long-term influence. Evidently this inclination is called Amara’s Law. It’s identified and cited in TechBullion by a Moscow-born and-based investor named Alexey Vladimirovich Bashkirov. “He compares AI’s trajectory to that of the internet, initially overhyped but ultimately transformative,” the outlet’s editor, Angela Scott-Briggs, reports. “He predicts that in the coming years, more AI startups in healthcare will fail than succeed, and the true impact of AI on medicine will become evident only after several years.” Interesting guy, this Alexey Bashkirov. (Google his name and see if you don’t agree.)
- We’re in a weird space with AI in healthcare. And that means all of us. Don’t take it from AIin.Healthcare. The observation is from Adam Rodman, MD, MPH, a Beth Israel Deaconess internist and Harvard Medical School instructor. “We say, ‘Wow, the technology is really powerful,’” Rodman says in conversation with the Harvard Gazette. “But what do we do with it to actually change things? My worry, as both a clinician and a researcher, is that if we don’t think big—if we don’t try to rethink how we’ve organized medicine—things might not change that much.” Rodman is just one of several well-credentialed sources who spoke about AI in healthcare with reporter Alvin Powell. Read the whole thing.
- AI may be incapable of willful evil, but sometimes it behaves in ways that might rightly be called duplicitous. Researchers have counted the ways how: at least five modes of seemingly deliberate deception. One of these, sandbagging, is what they labeled an algorithm’s penchant for making what looked like intentional mistakes when perfect scores might have drawn investigations from humans. No, really. The model at hand “strategically got more math questions wrong when it believed high performance would lead to unwanted modifications or retraining,” reports geologist Marc Defant, PhD, in Skeptic magazine. “Essentially, it pretended to be dumber than it was for strategic reasons—a tactic straight out of the Machiavellian playbook.” Read and ponder not only the veracity of the accounts but also the possible ramifications for AI in healthcare.
- Leave it to a 14-year-old to create an AI app that detects heart disease with 96% accuracy in less than 10 seconds. The kid is Siddharth Nandyala, originally of India and now of Texas. He calls his app CircadiaV. It’s been tested on more than 15,000 patients in the U.S. The algorithm behind the app analyzes heart sound patterns and detects irregularities that could indicate underlying cardiac conditions, eHealth reports. “The app’s rapid and precise diagnosis provides a crucial tool for both early intervention and preventive healthcare, making it an invaluable addition to global cardiovascular care.” More here.
- Recent research in the news:
- University of Colorado: Leveraging data science for disease prediction in the right against rheumatoid arthritis
- University of Virginia: Artificial intelligence tool to accelerate new treatments
- University of Helsinki: Innovative infant wearable uses artificial intelligence for at-home assessments of early motor development
- American College of Cardiology: AI-powered mammograms: A new window into heart health
- University of Colorado: Leveraging data science for disease prediction in the right against rheumatoid arthritis
- Funding news of note:
- From AIin.Healthcare’s news partners:
- Health Imaging: AI quantifies breast arterial calcifications on mammograms
- Cardiovascular Business: AI could help cardiologists predict bleeding, stroke risks in AFib patients on DOACs
- Radiology Business: Using robotics to automate X-ray, ultrasound workflows the goal of new GE HealthCare-Nvidia partnership
- Health Imaging: GE HealthCare launches new AI-enabled automated breast ultrasound system
- Health Imaging: AI quantifies breast arterial calcifications on mammograms
Dave P. has worked in journalism, marketing and public relations for more than 30 years, frequently concentrating on hospitals, healthcare technology and Catholic communications. He has also specialized in fundraising communications, ghostwriting for CEOs of local, national and global charities, nonprofits and foundations.