Healthcare AI newswatch: Fail-fast AI skunk works, writerly doctors avoiding AI scribes, more

Buzzworthy developments of the past few days.

• If only AI could help healthcare providers get paid faster, more accurately and with less hassle. Actually, it can. It just has to be prodded along by people working together at predictable sticking points in the revenue cycle. At times that means goosing cooperation between providers and payers. By working together, these stakeholders can ensure that AI tools are designed and implemented in a way that supports cash flow on both ends. Several subject matter experts with the nonprofit Council for Affordable Quality Healthcare flesh out the best-case scenario in the American Journal of Managed Care. “AI can be used to proactively identify services that require prior authorization, retrieve necessary documentation from electronic health records and submit it to health plans for adjudication,” Erin Weber and colleagues explain. “Such collaborative efforts can help providers realize the full potential of AI in improving efficiency and reducing costs.” 
   
• Is trial and error any way for a sprawling health system to roll out AI? Duke Health’s CMIO, Eric Poon, MD, MPH, believes so. In fact he’s convinced that, looking back later, 150-site Duke will see that dollars won’t have been wasted on the errors—as long as people learned from them. And that’s whether even costly stumbles can be blamed on the AI products or the users themselves. “We know that not everything that we put our hands on or that comes our way is going to pan out,” Poon says in remarks made to HealthLeaders. “Our governance philosophy is that we need to fail often and fail fast so we [can] provide just-in-time advice.” An especially bold approach seeing as Duke Health is already using 80 or so separate AI tools. 
   
• The dean of MIT’s computing college seconds that motion. “Machine learning is going to reinvent many areas of work,” Daniel Huttenlocher, PhD, told attendees of an MIT conference on AI this week. “If you don’t have skunk works projects, you’re going to be left behind.” 
   
• Not every physician is thrilled to let AI scribes take the lead on note-taking duties. One who loves to write describes her misgivings over letting go of the pen in an opinion piece published by Stat. “As a doctor, an important way I sustain and deepen my relationships with my patients is by writing about them,” explains Christine Henneberg, MD, whose nonclinical writings include both fiction and nonfiction. “This is the very process that the AI scribe asks me—or forces me—to give up.” Read the rest. 
   
• It’s not immediately apparent how AI would affect work for midwives and OB nurses. That doesn’t mean they have no cause for enthusiasm—or concern. Down Under, an officer with the Australian Nursing & Midwifery Federation lays out the pros and cons. On the plus side of the ledger are things like education and work augmentation. On the minus (or mixed-blessing) side: workplace surveillance, work intensification and liability. “AI can never truly replicate the complex and often physical aspects of care-based work, so nurses and midwives will not be replaced by AI,” the ANMF officer, Paul Yiallouros, reassures readers of Australian Nursing & Midwifery Journal. “Decision-making in healthcare requires accountability, which only registered professionals can provide.” 
   
• When this teenager talks healthcare AI and consumer protection, her state’s legislature listens. The young person is Paisley Tuel, a student at Case Western Reserve University. The governing body is the Ohio House of Representatives. After writing a 50-page paper on the need for AI-specific laws to guardrail healthcare in the Buckeye State, Tuel was asked by a key House rep to help draft a bill covering much the same material. “We want AI tools to be tested and monitored for widespread use in Ohio hospitals because that has not happened yet,” the undergrad tells the Cincinnati Enquirer. “We want patients to know when AI in involved in their healthcare decisions.” 
   
• Here’s another teen making news of interest to healthcare AI watchers. This one is only 18 and already wealthy after developing an AI app for tracking calorie intake. Zach Yadegari blasted into the headlines after he posted an essay on X describing his rejection by 15 prestigious colleges despite his real-world accomplishments. The post spurred heated bickering over the elusive “real reason” so many admissions boards gave the young man the cold shoulder. The kerfuffle went viral. And now many millions of people have heard the name Zack Yadegari. TechCrunch recounts the turn of events here, and there’s a good deal more coverage elsewhere. 
   
• Recent research in the news:
   
  • Columbia University: Nurses and AI collaborate to save lives, reduce hospital stays 
     
  • American College of Cardiology: AI-powered assessment for fractional flow reserve proves comparable to conventional assessment
     
  • Boston University: ‘Artificial intelligence is the future of peer review’
     
  • University of Technology Sydney: AI-driven smart devices to transform healthcare 
     
  • George Mason University: AI can recommend if you need to be screened for cancer 
     
  • UC-Irvine: AI effectively predicts dementia risk in American Indian/Alaska Native elders
     
• Funding news of note:
   
  • Alphabet-backed Isomorphic Labs raises $600M for AI drug development
     
  • AI-based RCM startup RapidClaims raises $8M in Series A funding
     
  • Sohar Health lands $3.8M to ‘open the front door to faster, fairer care’
     
  • Ex-Meta alum’s Parallel lands $3.5M to cut hospitals’ administrative burden with AI agents 
     
  • FlocCare secures $500K to drive healthcare awareness and patient engagement with Gen AI
     
  • Lord Abbett: Generative AI investing: Tracking the transformation of healthcare
     
• From AIin.Healthcare’s news partners:
   
  • Health Imaging: New AI-enabled platform could be ‘game changer’ for opportunistic screening in women
     
  • Radiology Business: American College of Radiology partners with AI developer, seeking to improve Alzheimer’s diagnoses