The landscape of healthcare in the United States has been characterized by a gradual but often hesitant embrace of technological innovations. From the slow implementation of electronic health records (EHR) to the cautious exploration of artificial intelligence (AI) tools, the sector has faced challenges in adopting general-purpose technologies that promise to revolutionize the nature of healthcare work. However, on the one-year anniversary of the public launch of ChatGPT, there is growing optimism, particularly from healthcare thought leaders like Dr. Robert Wachter, Chair of UC San Francisco’s School of Medicine, about the potential of generative artificial intelligence (genAI) to overcome historical barriers and usher in transformative changes.

The U.S. healthcare system has been notorious for its resistance to change, grappling with misaligned incentives, regulatory complexities, and an overall reluctance to embrace digital transformation. In a commentary released in the Journal of the American Medical Association (JAMA), Dr. Wachter and co-author Erik Brynjolfsson, Director of the Digital Economy Lab at Stanford University, argue that genAI possesses unique properties that could break the historical pattern of slow adoption and yield accelerated productivity gains.

Historically, the healthcare industry has faced the “productivity paradox of information technology,” a phenomenon described by Brynjolfsson in 1993. Despite the hype surrounding the adoption of general-purpose technologies, including the digitization of health records, significant gains in productivity have often taken years to materialize. In the late 2000s, fewer than one in ten U.S. hospitals had implemented EHR, highlighting the sector’s lag in embracing transformative technologies.

The authors identify several factors that make genAI different and potentially more conducive to rapid adoption in the U.S. healthcare environment. Unlike previous technologies, genAI is relatively easy to use and does not necessitate wholesale changes in hardware or workflow. Healthcare professionals, including doctors and nurses, are already accustomed to utilizing digital tools, setting the stage for seamless integration.

Importantly, the healthcare ecosystem has become better prepared for genAI. With increased use of digital data and systems, the integration of third-party software tools has become more straightforward. The current pressures on healthcare systems to deliver high-quality, safe, and cost-effective care, coupled with shortages of clinical and non-clinical personnel, create a conducive environment for genAI to address clinical and business needs.

The early applications of genAI in healthcare are expected to target administrative frictions, addressing patient needs such as appointment scheduling, medication refills, and finding healthcare providers. For healthcare professionals, genAI holds the potential to streamline tasks like creating clinical notes, handling prior authorization requests, composing letters to patients and colleagues, and summarizing complex patient records.

However, the road to widespread genAI adoption in U.S. healthcare is not without challenges. Continuous improvement in genAI technology is crucial, especially as it takes on more significant roles in clinical decision-making. The integration of AI into existing electronic health record systems, while easier than before, requires further refinement for seamless functionality.

Financial considerations pose another challenge, as healthcare systems must find resources to invest in AI technology. Potential labor-management tensions, witnessed in other industries where AI adoption led to strikes, may also arise in the healthcare sector. Despite these challenges, the pressing need for improved efficiency, quality, and equitable healthcare provision, coupled with leadership awareness and governance structures in institutions like UCSF Health, positions the sector for successful genAI implementation.

In conclusion, the optimism surrounding generative artificial intelligence in U.S. healthcare stems from its unique attributes, potential for seamless integration, and a more receptive healthcare ecosystem. While challenges exist, the promise of transformative gains in productivity, quality of care, and patient experience could mark a significant turning point in the healthcare industry’s relationship with cutting-edge technologies. As regulatory frameworks evolve to establish effective guardrails for genAI, the United States may witness a healthcare revolution powered by artificial intelligence.