Pattern Recognition
A monthly digest of key trends and developments in healthcare AI
About “Pattern Recognition”
Healthcare’s Transformation Imperative
People frequently tell me that healthcare has been transformed by technology. My response: “Are you kidding?”
A century ago, America’s most valuable companies were the likes of US Steel, Standard Oil, and American Harvester. Today, of course, they’re companies like Apple, Nvidia, Microsoft, and Tesla – none of which existed when I was in high school in the 1970s.
How about in healthcare? Well, a list of America’s top ten hospitals in 1925 would be nearly identical to a top ten list today. The unemployment rate among doctors and nurses is essentially zero. With the possible exception of education, no industry has been as un-disrupted by technology as mine.
This stability might be OK, except for the fact that our healthcare system is failing to deliver in countless ways. In the U.S., we spend nearly 20 percent of our gross domestic product, about $5 trillion each year, on healthcare. (The second-place spender is Germany, which devotes about 13 percent of its GDP.) Warren Buffett calls healthcare the “tapeworm” of the U.S. economy, literally starving our society of resources it would otherwise invest in infrastructure, housing, or public education.
Our profligacy might be acceptable if it resulted in better health, but it does not. Approximately one out of three of these dollars is spent on administrative paper-pushing or other activities that add next to nothing to a patient’s welfare. Our average life expectancy is four years lower than the average in a peer group of other wealthy countries that includes Canada, France, Germany, and Australia; we also have the highest rate of avoidable deaths from causes like diabetes, hypertension, and several cancers, and the highest infant and maternal mortality rates.
Don’t get me wrong – U.S. healthcare delivers miracles every day, particularly when it comes to cutting-edge and intensive care, things like transplants, cancer immunotherapy, and gene editing. And the people who work in healthcare – both clinicians and non-clinicians – are generally caring, dedicated, and well trained. But the healthcare system itself is a headache wrapped in red tape inside the nightmare that Franz Kafka himself might have dreamed up while on hold with the insurance company. It’s a system in desperate need of transformation.
Is This Our Hemingway Moment?
Our traditional response to growing patient needs and demands, new drugs and procedures, and mounting bureaucratic and regulatory hoops has been to hire more humans: medical specialists, nurse practitioners, care coordinators, billing and prior authorization request clerks. But we’ve hit a wall – not only can’t we afford these staff; we increasingly can’t find enough humans to fill these jobs.
When one of Hemingway’s characters in The Sun Also Rises went bankrupt, another asked him how it happened. “Two ways,” he replied. “Gradually, and then suddenly.” This combination of a failing healthcare system and a breathtaking new technology – generative AI – makes me believe that we’re approaching our “suddenly” moment.
I’ve spent the past 18 months studying all of this in the process of writing a book: “A Giant Leap: How AI is Transforming Healthcare, and What That Means for our Future,” which will be published by Portfolio/Penguin in February 2026. Researching (including more than 100 interviews with leaders across tech, medicine, business, and policy) and writing the book convinced me that we are on the cusp of real transformation – a story that’s consequential, fascinating, and unpredictable. While the new AI is the catalyst, this isn’t really a tech story. As I write in the book:
The story of whether AI will finally transform healthcare will mostly be about whether the healthcare system can implement AI in ways that produce better outcomes for patients, lower costs, and some relief for beleaguered doctors and nurses. And that, in turn, will be determined as much by history, politics, economics, pride, regulations, leadership, lawsuits, guilds, culture, workflows, inertia, greed, hubris, vibes, and zeitgeist as by graphical processing units, diffusion models, and neural networks.
While I hope you’ll read A Giant Leap, there is too much happening to wait for it to come out. This Substack will be my attempt to synthesize emerging issues in AI and healthcare – major studies, key business developments, big policy swings, tech breakthroughs, focusing more on the forest than the trees.
About Me
My perspective is that of a practicing physician, researcher, teacher, and chair of the biggest department in a large and well-regarded academic healthcare system. I also have the advantage of living in San Francisco, AI’s epicenter, and advising several health tech companies, most of them integrating AI into their businesses (I’ll let you know when I mention any of them). Since I wrote my 2015 book, The Digital Doctor: Hope, Hype, and Harm at the Dawn of Medicine’s Computer Age, I’ve been fascinated and frustrated by our rocky digital transformation. The advent of generative AI has only raised the stakes – both the upside potential and the risks.
I plan to post each month. My first post, on the primacy of platforms and the commoditization of AI scribes, will go up in a few days. Thanks for joining me for what’s sure to be a wild ride!