Towards an AI Economy That Works for All

Stephen Herzenberg and John Alic |

Executive Summary

This is the first report of a Keystone Research Center project on the “Future of Work.” The aim is to identify public policies that could help ensure that the application and diffusion of artificial intelligence (AI) over the next several decades fosters an economy in which Americans generally thrive. The project is motivated, in part, by concern that the opposite could occur: that AI will exacerbate the already high levels of income and wealth inequality in the United States. Our most important conclusion is that AI need not make our inequalities more severe. Creative public policies could lead to an AI economy “that works for the many, not just the few.”

The study design has been informed by the two principal authors’ experience at the one-time Office of Technology Assessment (OTA) of the US Congress. To guide the undertaking and provide feedback on its products, we recruited an advisory panel of nationally recognized academics and representatives of think tanks and the corporate, labor, and non-profit sectors. The project methodology combines interviews with technology experts, policy analysis, synthesis of research literature and, still to come, sectoral studies.

This first report contains three main parts. (1) Following an introduction, Sections II-IV contain an analysis of AI’s likely impacts through the lens of technology. Section II reviews past impacts of innovations including robotics and information technology on the economy and jobs. Section III looks at AI itself, how it does and does not go beyond previous technologies and substitute for human capacities and intelligence. Section IV explores the difficulties of predicting AI’s job impacts. (2) Section V, “The Plight of the American Worker,” considers the labor market context in which AI systems will spread and the roots of the economic inequality from which the nation suffers. (3) Section VI surveys policies that could influence inequality and the distribution of the benefits of productivity growth as AI spreads.