OpenAI, which developed ChatGPT, is hard at work developing healthcare applications. Providence's CEO Rod Hochman, MD, and president and COO Erik Wexler, joined a group of health system executives on a visit to OpenAI's headquarters in San Francisco earlier this year.
"[They showed us] all the cool stuff they're doing, and I would say we were totally blown away by what we saw," said Dr. Hochman during an interview at the Becker's CEO+CFO Roundtable, Nov. 11-14 in Chicago. "They took a case from the New England Journal of Medicine, and I can tell you it was the most complicated case you could ever come up with. [ChatGPT] solved the case to a hundred percent, and not only that, it came up with a treatment plan."
Dr. Hochman, an immunologist and rheumatologist by training, was impressed by the speed and accuracy of AI diagnosis. Physicians traditionally trained to memorize facts and information about their specialty; Dr. Hochman recalled carrying the Washington Manual in the pocket of his white coat with the "sum total of medical knowledge" for reference.
"Today, if I'm an intern on the ward somewhere, I've got the sum total of all medical knowledge that has ever been produced from Hippocrates on, and I am able to sort through that," said Dr. Hochman. "What people have to say, it's not AI making the decision, it's AI aiding the physician in their decision-making. That's a really big difference in terms of what's going on. From my standpoint, I think it's tremendously exciting. It's going to transform healthcare."
But not all CEOs feel the same way. Some, Dr. Hochman said, have decided to ignore AI and move forward with business as usual. But AI likely is already within the hospital's system as digital technology integrates within operations.
"It's incumbent upon every health system to put the guardrails in place," said Dr. Hochman. "Where's the ethics committee? Who's making the decisions on what should get implemented? That takes a lot of work. How do you engage your board in understanding AI? What are the principles around what we implement? Who do we partner with? There's now thousands of applications. How do you prioritize which are the most important?"
Providence went to every department and asked: If you could use AI to fix two things, what would it be? And then the departments started prioritizing which areas they wanted to focus on to make the biggest difference for health system operations.
Dr. Hochman also sees the big ethical challenges upcoming. During the visit, OpenAI leaders had one of the other CEOs say a few sentences about healthcare and then the AI wrote a talk in the CEO's voice that Dr. Hochman thought "was pretty good."
"If I closed my eyes, I could not tell the difference between [the CEO] and the production," said Dr. Hochman. "It raises lots of issues, and I think what's incumbent upon the people in the room and in healthcare is how do we use AI for good, not evil, because the good stuff is just going to be fabulous."