Industry Watcher’s Digest

Buzzworthy developments of the past few days.

• Good medicine is slow medicine. It’s a sensible point and one that kind of flies in the face of AI proponents stressing speed and efficiency gains. Safwan Halabi, MD, makes the statement in an article posted by Yahoo Finance July 2. Sane and sober people wouldn’t trust self-driving cars before they’d been put through the testing wringer in off-road settings, he suggests. So why in the world would anyone trust the safety of clinical AI before it’s been tested, tested and tested some more? Once a model clears that caveat, however, it’s ready for primetime. Healthcare AI is “here to stay, and we’re going to see more and more of it,” says Halabi, a radiologist who works at Lurie Children’s Hospital in Chicago as vice chair of medical imaging informatics. “We won't really notice it because it’s going to be happening in the background, or it’s going to be happening in the whole care process and not overtly announced or disclosed.”
   
• One aspect of healthcare AI seems to get scant attention: its potential for reducing hospital lengths of stay. Comments offered by Josh Fessel, MD, PhD, to Observer.com suggest a possible reason. Fessel, director of the NIH’s office of translational medicine, believes it’s wise to watch for AI solving one problem while worsening another. “What if it turns out that hospital length of stay is reduced, but readmission rates increase by some unacceptable factor?” he asks rhetorically, noting the complexity of healthcare compared with other sectors of the economy. “AI doesn’t necessarily know to look for that unless you train it to look for that.”
   
• AI will ‘take over’ in 10 realms of human activity next year. Healthcare is just one of them. That’s according to Jan Macarol, editor-in-chief of Ljubljana, Slovenia-based City Magazine. “AI systems will enable faster and more accurate diagnoses, which will reduce waiting times for treatment and improve treatment outcomes,” Macarol writes. “All of this will increase access to quality healthcare and enable better health outcomes for people around the world.” If this prediction sounds like hopeful hype from afar even before you’ve considered the other nine, well, so be it. This has been a slow news week in the U.S. Our nation’s birthday will do that. That doesn’t mean the piece is not worth a read.
   
• AI can help address healthcare disparities faced by many women. One woman is bullish on the proposition and well-situated to help nudge things toward fruition. She’s Priya Oberoi, general partner of Goddess Gaia Ventures, Europe’s first women-centric healthcare fund. Writing in Forbes, Oberoi says the technology can help boost women’s participation in clinical trials, increase the accuracy of their diagnoses, improve the effectiveness of fertility treatments and more. “AI is potentially able to address pertinent problems within the healthcare sector,” she writes, “whilst simultaneously allowing for the faster development of products and services catered specifically for women.”
   
• Healthcare loves GenAI. So observes David Talby, PhD, MBA, chief technology officer of John Snow Labs. Referencing a Snow survey released in April, Talby underscores four key findings at CIO.com: GenAI budgets are growing exponentially, task-specific language models reign supreme, use cases vary by technical experience and organization size, and human intervention remains necessary. “It’s clear technical leaders are spearheading GenAI in healthcare, as reflected by significant budget increases and a deep understanding of the technical advantages,” Talby writes. “However, challenges remain, particularly around accuracy, industry-specific requirements and ethical considerations for all groups and company sizes who have already deployed or are considering deploying GenAI.”
   
• We’re halfway through 2024. How’s your AI alignment? If you’re not sure, you know you could be doing better or you have the sense “AI alignment” is just another buzzword propagated by marketers, you should read a quick primer over at Pymnts.com. “At its core, AI alignment seeks to create AI systems that reliably pursue the objectives we want them to pursue rather than misinterpreting instructions or optimizing for unintended goals,” the piece helpfully clarifies. “The stakes are high—a misaligned AI system could cause significant harm if deployed in critical domains like healthcare, finance or national security.”
   
• Google is as high-profile a climate champion as corporate America has. For that reason, its recent admission that it’s falling far short of its green goals is garnering a lot of attention. And so is its naming of AI as a major electricity hog. Reaching its trumpeted goal of net zero by 2030 is now going to be really hard, chief sustainability officer Kate Brandt tells the Associated Press. “[O]ur approach will need to continue to evolve, and it will require us to navigate a lot of uncertainty, including this uncertainty around the future of AI’s environmental impacts.” There’s plenty more coverage online.
   
• Nvidia is making other AI players look like also-rans. But how? Whatever the secret of its success, the lopsided situation seems to be bugging multibillionaire Peter Thiel. Noting that as much as 85% of the money in AI is being made by that one company, Thiel—best known as the co-founder of PayPal and Palantir—spoke about Nvidia at last week’s Aspen Ideas Festival. According to Yahoo Finance, he described the market dominance of Jensen Huang’s high-flying brainchild as “very strange.” He wondered aloud about how eye-popping profits are being made “at the hardware layer—an area Silicon Valley doesn’t even know much about anymore.”
   
• Recent research roundup:
   
  • Karolinska Institute: Researchers in Sweden have developed nanorobots that kill cancer cells in mice
     
  • Ochsner Health: Using convolutional neural networks to improve the precision of nasal endoscopy
     
  • University of Alabama at Birmingham: Deep machine-learning speeds assessment of fruit fly heart aging and disease, a model for human disease
     
• Funding news of note:
   
  • Investors pour $27.1B into AI startups, defying a downturn
     
  • Nvidia rival Groq AI set to raise $300M in funding round
     
  • AI deals lift US venture capital funding to highest level in two years, data shows
     
• From AIin.Healthcare’s news partners:
   
  • Health Imaging: Skepticism and optimism: Radiologists are still divided on AI integration
     
  • Cardiovascular Business: Remote monitoring, AI to play key roles in the future of cardiology