Economics and Artificial Intelligence

Paper Session

Friday, Jan. 6, 2017 3:15 PM – 5:15 PM

Hyatt Regency Chicago, Grand Ballroom CD North
Hosted By: American Economic Association
  • Chair: Avi Goldfarb, University of Toronto

Are We Approaching an Economic Singularity?

William Nordhaus
,
Yale University

Abstract

What are the prospects for long-run economic growth? The present study looks at a more recently launched hypothesis, which I label Singularity. The idea here is that rapid growth in information technology and artificial intelligence will cross some boundary, after which economic growth will increase sharply as an ever-accelerating pace of improvements cascade through the economy. The paper develops a growth model that features Singularity and presents several tests of whether we are rapidly approaching Singularity. The key question for Singularity is the substitutability between information and conventional inputs. The tests suggest that the Singularity is not near.

Exploring the Impact of Artificial Intelligence: Prediction versus Judgment

Ajay K. Agrawal
,
University of Toronto
Joshua S. Gans
,
University of Toronto
Avi Goldfarb
,
University of Toronto

Abstract

Based on recent developments in the field of artificial intelligence (AI), we examine what type of human labour will be a substitute versus a complement to emerging technologies. We argue that these recent developments reduce the costs of providing a particular set of tasks – prediction tasks. Prediction about uncertain states of the world is an input into decision-making. We show that prediction allows riskier decisions to be taken and this is its impact on observed productivity although it could also increase the variance of outcomes as well. We consider the role of human judgment in decision-making as prediction technology improves. Judgment is exercised when the objective function for a particular set of decisions cannot be described (i.e., coded). However, we demonstrate that better prediction impacts the returns to different types of judgment in opposite ways. Hence, not all human judgment will be a complement to AI. Finally, we explore what will happen when AI prediction learns to predict the judgment of humans.

The Race Between Machine and Man: Implications of Technology for Growth, Factor Shares and Employment

Daron Acemoglu
,
Massachusetts Institute of Technology and NBER
Pascual Restrepo
,
Massachusetts Institute of Technology

Abstract

The advent of automation and the simultaneous decline in the labor share and employment among advanced economies raises concerns that labor will be marginalized and made redundant by new technologies. This paper examines this proposition in a task-based framework wherein not only are tasks previously performed by labor automated, but also more complex versions of existing tasks, in which labor has a comparative advantage, can be created. We fully characterize the structure of equilibrium in this model, establishing how the allocation of factors to tasks and factor prices are determined by the available technology and the endogenous choices of firms between capital and labor. We then demonstrate that although automation tends to reduce employment and the share of labor in national income, the creation of more complex tasks has the opposite effect. Our full model endogenizes the direction of research and development towards automation and the creation of new complex tasks. We show that, under reasonable conditions, there exists a stable balanced growth path in which the two types of innovations go hand-in-hand. Consequently, an increase in automation reduces the wage to rental rate ratio, and thus discourages further automation, encourages greater creation of more labor-intensive tasks, and restores the share of labor in national income and the employment to population ratio back towards their initial values. Though the economy is self-correcting, the equilibrium allocation of research effort is not optimal: to the extent that wages reflect quasi-rents for workers, firms will engage in too much automation. Finally, we extend the model to include workers of different skills. We find that inequality increases during transitions, but the self-correcting forces also serve to limit the increase in inequality over longer periods.
Discussant(s)
Benjamin F. Jones
,
Northwestern University
Erik Brynjolfsson
,
Massachusetts Institute of Technology
Orie Shelef
,
Stanford University
JEL Classifications
  • O3 - Innovation; Research and Development; Technological Change; Intellectual Property Rights
  • O4 - Economic Growth and Aggregate Productivity