AI newswatch: OpenAI goes to the UAE, patients get cyberchondria, more

Buzzworthy developments of the past few days.

• Like remoras hitching a ride on a shark, private businesses are capitalizing on President Trump’s AI-aggressive trip to the Middle East. Exhibit A: OpenAI. On Thursday the company best known for ChatGPT and Sam Altman firmed up its plan to construct a physical complex in the United Arab Emirates. OpenAI says the move is tied to Stargate, the company’s AI infrastructure platform. Stargate, the May 22 announcement reminds, “represents our long-term vision for building frontier-scale compute capacity around the world in service of safe, secure and broadly beneficial AGI.” That’s AGI as in artificial general intelligence, as in “human level” AI. OpenAI says the project will feature two primary ventures. One is a 1-gigawatt “Stargate UAE” computer cluster in Abu Dhabi, immediately slated for a 2026 go-live. The other is an investment of undisclosed magnitude by G42, the UAE firm focused on accelerating AI advancement and uptake, in Stargate infrastructure in the United States.  
   
  • The collaboration “reinforces OpenAI’s commitment to strengthening U.S. infrastructure while helping allies gain access to transformative AI responsibly and securely,” OpenAI says. “The new partnership will support the UAE in leveraging OpenAI’s tools across critical sectors like government, energy, healthcare, education and transportation to help accelerate innovation and economic growth while creating lasting benefits for its people. … Stargate UAE has the potential to provide AI infrastructure and compute capacity within a 2,000-mile radius, reaching up to half the world’s population” (emphasis added). 
     
  • OpenAI announcement here, mass media coverage here. 
     
• Picture a doctor tag-teaming with a chatbot to see patients. The AI bot goes first, then sends in the human. That’s where CMS Administrator Mehmet Oz, MD, MBA, sees things going. Medical professionals get bored reciting the same old song and dance about things like diabetes, Oz explained at a May 20 conference in Washington hosted by the Duke-Margolis Institute for Health Policy. “What if some of that work was done by an avatar that could actually communicate with the patient in a way that may be better, more patient than a clinician?” Oz suggested, as noted in coverage by MedPage Today. “And then the clinician comes in toward the end of the discussion, the last 25 minutes or 10 minutes of the discussion, and actually connects on the real issue by looking the patient in the eye and making sure that these stories are sticking.” Wrestled by human-machine combos like that, he added, the physician shortage “becomes less of a concern.” 
   
• Come to think of it, the Dr. Oz scenario is already reality. It’s just not yet unfolding in doctors’ offices. Instead, healthcare consumers get informed by “Dr. Google” and “Dr. AI” before going in to be seen in person. Organizational psychologist Amy Bucher, PhD, believes this dynamic is a good thing. Especially when it’s leveraged by primary care providers to head off patients’ health problems now rather than react to them later. “By thoughtfully integrating AI into clinical workflows and patient routines, we can unburden providers, empower patients and put prevention at the center of care,” Bucher enthuses in Physicians Practice. “That’s not just a technological upgrade—it’s a behavioral one.”
   
• AI has passed another medical board test. This time it’s the Chinese LLM DeepSeek R1 satisfactorily completing the 455-question gastroenterology board exam administered by the American College of Gastroenterology. The base model R1 correctly answered 77.1% of the problems. And a search-augmented edition of R1 notched a score of 81.5%, surpassing earlier performances of legacy ChatGPT models. DeepSeek “exhibited passing performance on the gastroenterology board examination, but gaps in niche topics and image exclusion limit utility,” report researchers at Texas Tech University and Baylor College of Medicine. “DeepSeek R1 may supplement [gastroenterologist] education if validated by specialists.”
   
• Odds are you haven’t heard of Twofold, Twill Health or TytoCare. But there’s a pretty good chance you will—and that’s apart from this little news item. Those three are among six AI vendors that are “quietly reshaping healthcare delivery,” according to a piece published May 21 by International Business Times UK. Get the rest of the list and a summary description of each solution here. 
   
• Between 20% and 30% of Floridians self-report symptoms commonly associated with ‘cyberchondria.’ This has been defined in the literature as “a clinical phenomenon in which repeated Internet searches regarding medical information result in excessive concerns about physical health.” In other words, it’s hypochondria for the always online among us. The figures may well reflect national averages, although they come from a sample set of 500 Sunshine Staters surveyed by the University of South Florida and Florida Atlantic University. The researchers also found respondents “cautiously optimistic—but uncertain—about AI’s long-term impact on healthcare.” Many believe AI has the potential to improve health outcomes and reduce medical errors, but more than a few “remain unconvinced or unsure.” More findings here. 
   
• By the end of this year, AI will be using more power than France, the U.K. and a whole host of other not-small countries. So forecasts data scientist Alex de Vries-Gao in the sustainable energy journal Joule. “A significant portion of [the needed] hardware may end up in the United States, as OpenAI has partnered with several others in a joint venture called Stargate”—sound familiar?—“to invest up to $500 billion over four years in new data center infrastructure across the country,” he writes. “[W]hile a growing reliance on fossil fuels threatens to undermine climate goals, effective policy responses first require urgent transparency.” 
   
• Investments in AI will take a Big Tech player to a valuation of $5 trillion over the next five years. Hint: The company is already a household name. The prediction is from Adam Spatacco, a former financial analyst and current close watcher over the money side of healthcare, technology and consumer goods for The Motley Fool. Interestingly, the AI-happy juggernaut he expects to top the $5T mark by 2030 isn’t one of the 10 best stocks for investors to buy right now, according to a separate Fool analysis. (If you guessed Amazon, you’re of like mind on the markets with Adam Spatacco.) 
   
• Some 8 of 10 Gen Z-ers would consider marrying a sweetheart who happens to be an AI bot. The survey behind the finding may be a tad tainted by selection bias. It was conducted by Joi AI, formerly Eva AI, an online service that matches subscribers with highly attentive—albeit corporeally nonexistent—companions. Mulling the naiveté of the presumably lonely 80%, TechRadar editor-at-large Lance Ulanoff doesn’t conceal his shock. “I don’t understand why Joi AI’s respondents, even Generation Z—who are much more deeply immersed in technology, social media and AI than any generation before it—would accept an AI as a life mate,” he writes. “How long before some dude is marrying his AI bot in Vegas?”