As described in this issue, we are undergoing two major transitions in healthcare: a demographic transition and a technology transition. On the demographic side, aging populations will put enormous strain on health are systems by increasing demand at a time when we might become supply constrained in terms of personnel. One major healthcare system in the United States is touting its ability to maintain its provider level despite the rapid pace of retirements as a competitive advantage.
It’s no secret that technology is moving forward at an amazing pace. One of my colleagues at Stanford, Fei-Fei Li, has just written a memoir about the evolution of AI (The Worlds I See). She describes Geoffrey Hinton’s first successful use of a neural network for computer vision research in 2012, little more than a decade ago. It’s important to reflect on how early we are in the use of this technology. One great metaphor I have heard is that we’re back at the dawn of electricity – we’ve invented the copper wire, but not yet the light bulb, and so we’re constantly getting shocked as we explore use cases and applications. The withdrawal of Cruise from San Francisco and the recall of Tesla Autopilot are examples of this process of discovery in real time. Thus, the appropriate concern about safety in healthcare applications – applications that are, frankly, being developed with teams that have far fewer resources than these well-funded efforts.
But AI is a technology. In the Clayton Christensen world, to have true cost and quality improvements in a market, we need technology innovation and business model innovation. For all of the excitement of AI, we’re ignoring the business model and business process challenges. Building from the work of Gerry Anderson and Ge Bai at Johns Hopkins, my coauthors and I recently described how dysfunctional the healthcare market has become in terms of complexity (https://jamanetwork.com/journals/jama/fullarticle/2812255).
We are going to have to wake up to this unfortunate reality if we are going to realize the full potential of this technology as a solution to our demographic and work force challenges. Dissecting and solving the business process challenge is a great role to play for all of us studying healthcare management and healthcare systems.
Kevin Schulman, MD, MBA
BAHM President & HMPI Editor-in-Chief
Professor of Medicine, Stanford University