Key Takeaways
- AI‑driven automation of specific clinical tasks is unlikely to shrink the overall health‑care workforce; instead, it may expand demand for clinicians over the long run.
- Historical patterns (e.g., cataract surgery, joint replacement) illustrate the Jevons paradox, where gains in efficiency lead to higher utilization of a service.
- The “lump of labor” fallacy—assuming a fixed amount of work—misguides fears of job loss; technology reshapes the type and volume of work needed.
- AI can create new treatments, care models, and specialization opportunities, increasing the need for human expertise in areas that remain unautomated.
- The O‑ring theory underscores that medicine is a chain of interdependent steps; automating isolated tasks does not eliminate the need for clinician oversight, especially in high‑stakes settings.
- When AI tools are priced near marginal cost and lower overall care expenses, they can stimulate greater service volume, further supporting workforce growth.
Introduction: Reassessing AI’s Impact on Clinician Employment
The rapid diffusion of artificial intelligence in medicine has sparked concerns that algorithms will supplant physicians, radiologists, pathologists, and psychiatrists, leading to a net loss of jobs. In a Perspective published Aug. 29 in The New England Journal of Medicine, Dr. Dhruv Khullar—associate professor of population health sciences at Weill Cornell Medicine and a hospitalist at NewYork‑Presbyterian/Weill Cornell Medical Center—challenges this narrative. He argues that, contrary to pessimistic forecasts, AI agents could ultimately increase the number of health‑care professionals working in the United States. Drawing on economic theory and historical precedents, Khullar outlines why automation of discrete tasks may expand, rather than contract, the clinical workforce.
Economic Theory Predicts Expansion, Not Contraction
Khullar begins by acknowledging that “there’s little doubt that clinical roles will change in the AI era, and that some types of health‑care jobs could be replaced.” Yet he quickly pivots to the counterintuitive claim rooted in economic reasoning: “But economic theory and history offer distinct reasons to believe that in the long run, adoption of AI agents—even models capable of performing some forms of cognitive work—could lead to an expansion, rather than a contraction, of the clinical workforce.” This assertion rests on two well‑established concepts: the Jevons paradox and the rejection of the lump‑of‑labor fallacy. By framing AI as a productivity‑enhancing technology, he sets the stage for a discussion of how efficiency gains can stimulate greater demand for services—and thus for the humans who deliver them.
The Jevons Paradox in Medicine: Lessons from Cataract Surgery and Joint Replacement
To illustrate his point, Khullar cites cataract surgery and joint replacement as classic examples of the Jevons paradox. He notes that these procedures “have become more efficient by reducing clinical effort, shortening recovery time, and improving safety, which have enabled more patients to receive them.” The paradox predicts that when a technological advance makes a resource cheaper or more effective, overall consumption of that resource rises rather than falls. Applying this logic to AI, Khullar reasons that if AI systems lower the marginal cost of delivering care—by, say, automating image analysis or streamlining diagnostic workflows—then the volume of services such as imaging, pathology reads, or preventive consultations could rise, creating additional workload for clinicians who oversee, interpret, and act on AI‑generated outputs.
Rejecting the Lump‑of‑Labor Fallacy
A second pillar of Khullar’s argument tackles the “lump of labor” fallacy, the belief that there is a fixed amount of work to be done in an economy. He writes that “the type and amount of work changes over time in response to technological advances, and that AI may allow clinicians to prevent or treat conditions in ways that aren’t currently possible or imaginable.” By emphasizing that labor demand is elastic, he contends that AI will not simply replace existing tasks but will unlock new therapeutic possibilities—such as personalized treatment algorithms, real‑time risk stratification, or AI‑augmented surgical planning—that require fresh skill sets and, consequently, more clinicians to implement and refine them.
Creating New Professional Capabilities and Specializations
Expanding on the idea of novel work, Khullar states: “AI may increase the demand for new professional capabilities by creating new treatments, care modes, and opportunities for specialization.” In practice, this could mean the emergence of roles focused on AI model validation, ethical oversight of algorithmic decisions, or integration of predictive analytics into chronic disease management. As AI handles routine pattern‑recognition tasks, clinicians may shift toward higher‑order functions—complex decision‑making, patient communication, and care coordination—thereby expanding the scope and appeal of medical professions rather than diminishing them.
The O‑Ring Theory: Why Human Oversight Remains Essential
Khullar invokes the O‑ring theory, named after the faulty component that doomed the Space Shuttle Challenger, to stress that medicine is a series of interdependent steps where a single failure can jeopardize outcomes. He explains: “Delivering high-quality care involves much more than successfully executing a given task. Rather, the practice of medicine depends on numerous skills and many interconnected steps, from interpreting test results to developing and negotiating treatment plans.” Because errors at any link can cascade, high‑stakes contexts will continue to necessitate clinician supervision for safety and trust. Thus, even if AI automates isolated components—such as flagging a suspicious lesion on a mammogram—the value of the remaining human tasks, like discussing biopsy options with a patient or tailoring therapy based on comorbidities, may actually increase.
Policy and Pricing Considerations: AI Near Marginal Cost
The author adds a pragmatic note: AI’s impact on workforce size will also hinge on how the technology is priced and deployed. If AI tools are made available at or near marginal cost—akin to a utility—then the overall expense of delivering care drops, potentially prompting health systems to expand service offerings to meet previously unmet demand. This scenario mirrors historical trends where cheaper diagnostics (e.g., rapid blood tests) led to more frequent testing and greater need for staff to manage the increased workload. Consequently, thoughtful reimbursement models that encourage broad, affordable AI adoption could amplify the workforce‑expanding effect Khullar describes.
Conclusion: A Balanced Outlook for the Clinician Workforce
In sum, Khullar’s Perspective reframes the AI‑jobs debate by emphasizing that technological change in health care is unlikely to be a zero‑sum game. By leveraging economic theory, historical analogies, and the intrinsic complexity of medical practice, he presents a compelling case that AI agents could stimulate growth in the number of health‑care professionals rather than diminish it. As AI continues to evolve, the challenge for policymakers, educators, and health‑care leaders will be to harness these opportunities—ensuring that clinicians are equipped with the skills to thrive alongside intelligent machines, while safeguarding the patient‑centered, safety‑first ethos that remains at the heart of medicine.
Quoted material from the original article is presented verbatim where indicated, adhering to fair‑use standards for journalistic reporting.
https://news.weill.cornell.edu/news/2026/08/how-will-ai-impact-the-future-of-the-clinical-workforce