The economics of artificial intelligence: health care challenges

In: National Bureau of Economic Research conference report

Abstract

"Amid sweeping conversations about the future of artificial intelligence and its impact on US industry and economy, one economic domain has remained relatively insulated from the discussion: health care. How is it possible that an industry so bemoaned for inefficiency and expense, an industry so large that it now makes up a quarter of the US economy, could escape the efficiency- and cost-driven disruptions of AI? How are doctor's offices still relying on fax machines in the age of driverless cars? Why is it the one industry where we'd like to see AI try some things the one that machines can't seem to infiltrate? The Economics of Artificial Intelligence: Health Care Challenges convenes contributions from health economists, physicians, philosophers, and legal scholars to identify the primary barriers to entry for AI in America's biggest industry. Across original papers and wide-ranging written responses, they find five domains of barriers: incentives; management; data availability; regulation. They also find evidence of real opportunity: AI has promise to improve outcomes and lower costs, and if paths to intervention are seized upon, improvements will follow"--