A lot of monster tech companies are doing a lot of AI development. Start the list with Amazon, Google, Microsoft and OpenAI. What they and a host of others have in common is a voracious appetite for GPU chips capable of processing lofty AI business visions into blockbuster AI products.

Demand is outstripping supply, and GPU superstar Nvidia is feasting on the math.

This week news broke that the Silicon Valley “overnight sensation” had its valuation upped to nearly $2 trillion. More than one business outlet pointed out this makes Nvidia worth more than the entire Chinese stock market. Here’s what people have been saying about Nvidia’s gradual-then-sudden rise to riches—and what it says about expectations around AI in 2024.  

‘This last year, we’ve seen generative AI really becoming a whole new application space, a whole new way of doing computing,’ Jensen Huang, Nvidia’s co-founder and chief executive, said in a Wednesday earnings call. “A whole new industry is being formed, and that’s driving our growth.” (Source: multiple)

Jensen Huang isn’t just banking on selling AI chips to companies. He’s also looking at selling to entire countries as part of a concept known as sovereign AI. Sovereign AI boils down to a single country or region setting up its own AI capabilities, allowing its AI to meet its own needs. (Source: Yahoo! Finance)

Five years ago Nvidia was mostly known as a graphics and video-game chip maker. Its roughly $100 billion market valuation was less than half as much as integrated chip-maker Intel’s. Few investors foresaw Nvidia’s transformation into the world’s leading chip firm—or the way artificial intelligence would emerge.” (Source: The Wall Street Journal)

Demand for downgraded chips risks being reduced amid ferocious competition between companies in China to get ahead on AI. That’s especially so as domestic players seek to offer chips that can rival the very best Nvidia has to offer.” (Source: Business Insider)

Public fascination with ChatGPT’s ability to generate humanlike language [has] led to a flood of investment in OpenAI and spurred on its competitors. Improving sophisticated AI systems like ChatGPT requires increasingly large numbers of the chips Nvidia and its competitors make. (Source: WSJ.com)

Nvidia has established its dominance due to the use of its GPUs for training AI systems. However, demand for chips to support inference—the process of generating answers or results from AI models—instead of training is expected to rise as a proportion of the market. That could potentially create some space for rivals, but Nvidia looks to be establishing a strong position in inference too. (Source: Barron’s)

Investors like AI because, like a lot of technology, it transforms labor into a thing that can be sold at a profit … and this has the potential to radically reorder the priorities of thousands of companies. Calling Nvidia “the most important stock on planet earth” is, of course, outrageous hyperbole. But if AI is as important as Wall Street thinks it is, it might not be wrong. (Source: New York magazine)

Big Tech companies account for nearly 40% of Nvidia revenues, but its customers have diversified as more industries rush to invest in hardware for AI computing. Huang said industries including automotive, financial services and healthcare were now spending on its chips at a multibillion-dollar level. (Source: Financial Times)

‘Accelerated computing and generative AI have hit the tipping point,’ says Nvidia founder and chief executive Jensen Huang. “Demand is surging worldwide across companies, industries and nations.” (Source: multiple)

Buzzworthy developments of the past few days.

• A recent survey showed strong levels of hope and optimism among doctors that AI will relieve them of administrative burdens. However, there’s also “significant skepticism—and significant fear—that AI will be yet just another thing that physicians have to deal with that’s super-imposed upon them without any input and without the appropriate implementation.” The quote is from Nele Jessel, chief medical officer of Athenahealth, who spoke about physician burnout with ChiefHealthcareExecutive.com. More here.
   
• It was largely end-users who drove the successes of desktop computers and Internet connections into businesses, organizations and enterprises of every kind. Given those experiences, an AI plan is “just exactly what we shouldn’t waste our time on” now. The thinking is from IT consultant Bob Lewis, who makes his case in a tongue-in-cheek “dialogue” published at CIO.com. The fictional interlocutors are a bunch of supposed experts and one wise if prickly “Dr. Yeahbut.” (Expert No. 12: “Ensure there’s value in each anticipated use case.” Dr. Yeahbut: “No. Don’t do this. … It’s a cure that would be far worse than the disease.” Read the rest.
   
• AHIMA has built what it calls a ‘pioneering centralized knowledge space’ for health IT professionals interested in AI. The group, whose acronym stands for the American Health Information Management Association, has released a white paper to complement the resource hub. Check out both here.
   
• ‘As AI’s tentacles reach deeper into medical practices every day, the fast-evolving technology faces almost no regulation. … Hospitals aren’t even required to tell patients when they’re using AI.” That’s from a feature article in the Kansas City Beacon. Health reporter Suzanne King looks at the promise and risk of healthcare AI in her neck of the nation and beyond. The piece is posted here.
   
• Alert fatigue is the phenomenon in which nurses get so desensitized to sounds constantly coming from patient monitors that they just tune it all out. Researchers at Vanderbilt University Medical Center are using AI to “transform alert improvement into a continuous learning process”—for machines and clinicians alike. Details here.
   
• At Penn State, they’re streamlining the way AI summarizes patient info. The aim is to make the outputs faster to arrive yet easier to trust. In testing, the technique yielded a “consistent improvement in faithfulness,” which was confirmed by physicians who checked the work. University coverage here.
   
• Also of note from the AI research beat:
   
  • Mass General Brigham: Generative AI used to create translatable pediatric care educational videos for hospitals in resource-poor countries
     
  • UC-San Francisco Health: How AI Can Help Spot Early Risk Factors for Alzheimer’s Disease
     
  • Tufts Medicine: Tufts Medicine Announces Virtual-First Approach to Improve Access to Care
     
  • Raboud University Medical Center: Help from AI significantly improves colonoscopy—40% more polyps detected when the computer assists
     
  • University Hospital Bonn: Artificial intelligence recognizes patterns in behavior
     
• From AIin.Healthcare’s news partners:
   
  • Radiology Business: Startup that creates avatars via CT, MR images raises $5.4M from radiologists and investors
  • Health Imaging: Light-powered 3D imaging system guides vascular surgeries without an X-ray