人工智能和经济增长研究报告（英文）

Artificial Intelligence and Economic GrowthPhilippe AghionCollege de France and LSEBenjamin F. JonesNorthwestern University and NBERCharles I. JonesStanford GSB and NBER ∗October 10, 2017 – Version 1.0AbstractThis paper examines the potential impact of artificial intelligence (A.I.) on eco-nomic growth. We model A.I. as the latest form of automation, a broader pro-cess dating back more than 200 years. Electricity, internal combustion engines,and semiconductors facilitated automation in the last century, but A.I. now seemspoised to automate many tasks once thought to be out of reach, from driving carsto making medical recommendations and beyond. How will this affect economicgrowth and the division of income between labor and capital? What about thepotential emergence of “singularities” and “superintelligence,” concepts that an-imate many discussions in the machine intelligence community? How will thelinkages between A.I. and growth be mediated by firm-level considerations, includ-ing organization and market structure? The goal throughout is to refine a set ofcritical questions about A.I. and economic growth and to contribute to shaping anagenda for the field. One theme that emerges is based on Baumol’s “cost disease”insight: growth may be constrained not by what we are good at but rather by whatis essential and yet hard to improve.∗We are grateful to Ajay Agrawal, Mohammad Ahmadpoor, Adrien Auclert, Sebastian Di Tella, PatrickFrancois, Joshua Gans, Avi Goldfarb, Pete Klenow, Hannes Mahlmberg, Pascual Restrepo, Chris Tonetti,Michael Webb, and participants at the NBER Conference on Artificial Intelligence for helpful discussionand comments.2P. AGHION, B. JONES, AND C. JONES1.IntroductionThis paper considers the implications of artificial intelligence for economic growth.Artificial intelligence (A.I.) can be defined as “the capability of a machine to imitateintelligent human behavior” or “an agent’s ability to achieve goals in a wide range ofenvironments.”1 These definitions immediately evoke fundamental economic issues.For example, what happens if A.I. allows an ever-increasing number of tasks previouslyperformed by human labor to become automated? A.I. may be deployed in the ordinaryproduction of goods and services, potentially impacting economic growth and incomeshares. But A.I. may also change the process by which we create new ideas and tech-nologies, helping to solve complex problems and scaling creative effort. In extreme ver-sions, some observers have argued that A.I. can become rapidly self-improving, lead-ing to “singularities” that feature unbounded machine intelligence and/or unboundedeconomic growth in finite time (Good (1965), Vinge (1993), Kurzweil (2005)). Nordhaus(2015) provides a detailed overview and discussion of the prospects for a singularityfrom the standpoint of economics.In this paper, we speculate on how A.I. may affect the growth process. Our primarygoal is to help shape an agenda for future research. To do so, we focus on the followi