ABOUT THIS BOOK
Advances in artificial intelligence (AI) highlight the potential of this technology to affect productivity, growth, inequality, market power, innovation, and employment. This volume seeks to set the agenda for economic research on the impact of AI. It covers four broad themes: AI as a general purpose technology; the relationships between AI, growth, jobs, and inequality; regulatory responses to changes brought on by AI; and the effects of AI on the way economic research is conducted. It explores the economic influence of machine learning, the branch of computational statistics that has driven much of the recent excitement around AI, as well as the economic impact of robotics and automation and the potential economic consequences of a still-hypothetical artificial general intelligence. The volume provides frameworks for understanding the economic impact of AI and identifies a number of open research questions.
Contributors:
Daron Acemoglu, Massachusetts Institute of Technology
Philippe Aghion, Collège de France
Ajay Agrawal, University of Toronto
Susan Athey, Stanford University
James Bessen, Boston University School of Law
Erik Brynjolfsson, MIT Sloan School of Management
Colin F. Camerer, California Institute of Technology
Judith Chevalier, Yale School of Management
Iain M. Cockburn, Boston University
Tyler Cowen, George Mason University
Jason Furman, Harvard Kennedy School
Patrick Francois, University of British Columbia
Alberto Galasso, University of Toronto
Joshua Gans, University of Toronto
Avi Goldfarb, University of Toronto
Austan Goolsbee, University of Chicago Booth School of Business
Rebecca Henderson, Harvard Business School
Ginger Zhe Jin, University of Maryland
Benjamin F. Jones, Northwestern University
Charles I. Jones, Stanford University
Daniel Kahneman, Princeton University
Anton Korinek, Johns Hopkins University
Mara Lederman, University of Toronto
Hong Luo, Harvard Business School
John McHale, National University of Ireland
Paul R. Milgrom, Stanford University
Matthew Mitchell, University of Toronto
Alexander Oettl, Georgia Institute of Technology
Andrea Prat, Columbia Business School
Manav Raj, New York University
Pascual Restrepo, Boston University
Daniel Rock, MIT Sloan School of Management
Jeffrey D. Sachs, Columbia University
Robert Seamans, New York University
Scott Stern, MIT Sloan School of Management
Betsey Stevenson, University of Michigan
Joseph E. Stiglitz. Columbia University
Chad Syverson, University of Chicago Booth School of Business
Matt Taddy, University of Chicago Booth School of Business
Steven Tadelis, University of California, Berkeley
Manuel Trajtenberg, Tel Aviv University
Daniel Trefler, University of Toronto
Catherine Tucker, MIT Sloan School of Management
Hal Varian, University of California, Berkeley
TABLE OF CONTENTS
Introduction
DOI: 10.7208/chicago/9780226613475.003.0000
[artificial intelligence;general purpose technology;machine learning;innovation;growth]
In September 2017, the National Bureau of Economic Research held its first conference on the Economics of Artificial Intelligence in Toronto. The purpose of the conference and associated volume is to set the research agenda for economists working on AI. This introductory chapter organizes and summarizes the key ideas. We categorize the papers into four broad themes. First, several papers emphasize the role of AI as a general purpose technology, building on the existing literature on general purpose technologies from the steam engine to the internet. Second, many papers highlight the impact of AI on growth, jobs, and inequality, focusing on research and tools from macro and labor economics. Third, five chapters discuss machine learning and economic regulation, with an emphasis on microeconomic consequences and industrial organization. The final set of chapters explores how AI will affect research in economics. (pages 1 - 20)
This chapter is available at:
University of Chicago Press
Artificial Intelligence and the Modern Productivity Paradox: A Clash of Expectations and Statistics - Erik Brynjolfsson, Daniel Rock, Chad Syverson
DOI: 10.7208/chicago/9780226613475.003.0001
[artificial intelligence;general purpose technology;intangible assets;productivity;implementation lags]
We live in an age of paradox. Systems using artificial intelligence match or surpass human-level performance in more and more domains, leveraging rapid advances in other technologies and driving soaring stock prices. Yet measured productivity growth has declined by half over the past decade, and real income has stagnated since the late 1990s for most Americans. We describe four potential explanations for this clash of expectations and statistics: false hopes, mismeasurement, redistribution and implementation lags. While a case can be made for each explanation, we argue that lags have likely been the biggest contributor to the paradox. The most impressive capabilities of AI, particularly those based on machine learning, have not yet diffused widely. More importantly, like other general purpose technologies, their full effects won’t be realized until waves of complementary innovations are developed and implemented. The adjustment costs, organizational changes, and new skills needed for successful AI can be modeled as a kind of intangible capital. Some of the value of this intangible capital is already reflected in the market value of firms. However, going forward, national statistics could fail to measure the full benefits of the new technologies and some may even have the wrong sign. (pages 23 - 60)
This chapter is available at:
University of Chicago Press
https://academic.oup.com/chica...
The Technological Elements of Artificial Intelligence - Matt Taddy
DOI: 10.7208/chicago/9780226613475.003.0002
[artificial intelligence;machine learning;new technology;general purpose technology;automation]
In the past decade there has been a sharp increase in the extent that companies use data to optimize their businesses. Variously called the `Big Data' or `Data Science' revolution, this has been characterized by massive amounts of data, including unstructured and nontraditional data like text and images, and the use of fast and flexible Machine Learning (ML) algorithms in analysis. With recent improvements in Deep Neural Networks (DNNs) and related methods, application of high-performance ML algorithms has become more automatic and robust to different data scenarios. That has led to the rapid rise of an Artificial Intelligence (AI) that works by combining many ML algorithms together–each targeting a straightforward prediction task–to solve complex problems. We will define a framework for thinking about the ingredients of this new ML-driven AI. Understanding the components of these systems and how they fit together is important for those who will be building businesses around this technology. Those studying the economics of AI can use these definitions to clarify the conversation on AI's projected productivity impacts and data requirements. Finally, this framework should help clarify the role for AI in the practice of modern business analytics and economic measurement. (pages 61 - 88)
This chapter is available at:
University of Chicago Press
https://academic.oup.com/chica...
Prediction, Judgment, and Complexity: A Theory of Decision Making and Artificial Intelligence - Ajay Agrawal, Joshua Gans, Avi Goldfarb
DOI: 10.7208/chicago/9780226613475.003.0003
[artificial intelligence;machine learning;decision-making;complexity;automation;prediction;judgment]
We interpret recent developments in the field of artificial intelligence (AI) as improvements in prediction technology. In this paper, we explore the consequences of improved prediction in decision-making. To do so, we adapt existing models of decision-making under uncertainty to account for the process of determining payoffs. We label this process of determining the payoffs ‘judgment.’ There is a risky action, whose payoff depends on the state, and a safe action with the same payoff in every state. Judgment is costly; for each potential state, it requires thought on what the payoff might be. Prediction and judgment are complements as long as judgment is not too difficult. We show that in complex environments with a large number of potential states, the effect of improvements in prediction on the importance of judgment depend a great deal on whether the improvements in prediction enable automated decision-making. We discuss the implications of improved prediction in the face of complexity for automation, contracts, and firm boundaries. (pages 89 - 114)
This chapter is available at:
University of Chicago Press
https://academic.oup.com/chica...
The Impact of Artificial Intelligence on Innovation: An Exploratory Analysis - Iain M. Cockburn, Rebecca Henderson, Scott Stern
DOI: 10.7208/chicago/9780226613475.003.0004
[innovation;artificial intelligence;productivity]
Artificial intelligence (AI) may serve as a new general-purpose “method of invention” that can reshape the innovation process and R&D. We review the history of AI, including the distinction between robotics and the potential for “deep learning” to be a general-purpose method of invention. We assess evidence of this differential impact in the changing nature of measurable innovation outputs in AI, including papers and patents. We find evidence of a “shift” in the importance of application-oriented learning research since 2009 (relative to robotics and symbolic systems research), and that some of this shift began outside the United States. We consider implications of our findings, including changes in the innovation process, and policy and institutional responses to deep learning where it represents a general-purpose method of invention. There may be significant substitution away from routinized labor-intensive research towards research that exploits the interplay between passively generated datasets and enhanced prediction algorithms. The potential commercial reward of such research may trigger a period of racing, driven by powerful incentives for companies to acquire and control critical datasets and application-specific algorithms. We suggest that policies encouraging transparency and dataset sharing across public and private actors can stimulate more innovation-oriented competition and research productivity. (pages 115 - 148)
This chapter is available at:
University of Chicago Press
https://academic.oup.com/chica...
Finding Needles in Haystacks: Artificial Intelligence and Recombinant Growth - Ajay Agrawal, John McHale, Alexander Oettl
DOI: 10.7208/chicago/9780226613475.003.0005
[complexity;knowledge production function;economic growth;prediction;discovery rates;general purpose technologies;meta technologies;combinations]
There has been an explosion in data availability under the rubric of “big data” and computer-based advances in capabilities to discover and process those data. We can view these technologies in part as “meta technologies”—technologies for the production of new knowledge. Innovation is often predicated on discovering useful new combinations of existing knowledge in highly complex knowledge spaces. These needle-in-a-haystack type problems are pervasive in fields like genomics, drug discovery, materials science, and particle physics. We develop a combinatorial-based knowledge production function and embed it in the classic Jones growth model (1995) to explore how breakthroughs in artificial intelligence (AI) that dramatically improve prediction accuracy about which combinations have the highest potential could enhance discovery rates and consequently economic growth. This production function is a generalization (and reinterpretation) of the Romer/Jones knowledge production function. Separate parameters control the extent of individual-researcher knowledge access, the effects of fishing out/complexity, and the ease of forming research teams. (pages 149 - 174)
This chapter is available at:
University of Chicago Press
https://academic.oup.com/chica...
AI as the Next GPT: A Political-Economy Perspective - Manuel Trajtenberg
DOI: 10.7208/chicago/9780226613475.003.0006
[AI;GPT;skills;policy]
History suggests that dismal prophecies regarding the impact of technological advances rarely come to pass. Many occupations will indeed vanish with the advent of AI as the new General Purpose Technology (GPT), but there may be ways to ameliorate the detrimental effects of AI, and enhance its positive ones. One is education and skills development, revamping the centuries-old “factory model” of education, and developing skills relevant for an AI-based economy—analytical, creative, interpersonal, and emotional. Another is the professionalization of personal care occupations, particularly in healthcare and education. These will provide the bulk of future employment growth, yet currently involve little training and technology, and confer low wages. New, higher standards and academic requirements for these occupations would enable AI to benefit both providers and users. A third way is to affect the direction of technical advance. We distinguish between “human-enhancing innovations” (HEI), that magnify and enhance sensory, motoric, and other such human capabilities, and “human-replacing innovations” (HRI), which replace human intervention, and often leave for humans mostly “dumb” jobs. AI-based HEIs could unleash a new wave of creativity and productivity, particularly in services, whereas HRIs might decrease employment and give rise to unworthy jobs. (pages 175 - 186)
This chapter is available at:
University of Chicago Press
https://academic.oup.com/chica...
AI, Income, Employment, and Meaning
DOI: 10.7208/chicago/9780226613475.003.0007
[employment;income;productivity;automation]
The evolution of artificial intelligence and automation evokes strong emotions in people. Some imagine a dystopia in which people lose employment opportunities and are replaced by machines. Others envision a world where people will be able to enjoy their lives free from time and money constraints. Who is right? Most economists believe that automation promises a future of higher income that stems from the higher productivity that artificial intelligence will provide. This chapter discusses two aspects of such productivity gains: the effects of artificial intelligence and automation on employment and how people will spend their time if robots take their jobs; and the distribution of income gains and the role of social and political structure in that distribution. (pages 189 - 196)
This chapter is available at:
University of Chicago Press
https://academic.oup.com/chica...
Artificial Intelligence, Automation, and Work - Daron Acemoglu, Pascual Restrepo
DOI: 10.7208/chicago/9780226613475.003.0008
[AI;automation;displacement effect;labor demand;inequality;productivity;tasks;technology;wages]
We summarize a framework for studying the implications of automation and AI on the demand for labor, wages, and employment. Our framework emphasizes the displacement effect that automation creates as machines and AI replace labor. This displacement effect tends to reduce demand for labor and wages, but is counteracted by a productivity effect resulting from the cost savings generated by automation, which increases the demand for labor in non-automated tasks. The productivity effect is complemented by additional capital accumulation and the deepening of automation, both of which further increase demand for labor. These countervailing effects are incomplete. Even when they are strong, automation increases output per worker more than wages and reduces the share of labor in national income. The stronger countervailing force against automation is the creation of new labor-intensive tasks, which reinstates labor in new activities and increases the labor share to counterbalance automation’s impact. We highlight the constraints that decelerate the economy’s and labor market’s adjustment to automation and weaken the resulting productivity gains from this transformation: a mismatch between the skill requirements of new technologies, and the possibility that automation is introduced at an excessive rate, possibly at the expense of other productivity-enhancing technologies. (pages 197 - 236)
This chapter is available at:
University of Chicago Press
https://academic.oup.com/chica...
Artificial Intelligence and Economic Growth
DOI: 10.7208/chicago/9780226613475.003.0009
[artificial intelligence;economic growth;automation;singularity;superintelligence;Baumol's cost disease]
This chapter examines the potential impact of artificial intelligence (AI) on economic growth. We model AI as the latest form of automation, a broader process dating back more than 200 years. Electricity, internal combustion engines, and semiconductors facilitated automation in the last century, but AI now seems poised to automate many tasks once thought to be out of reach, from driving cars to making medical recommendations and beyond. How will this affect economic growth and the division of income between labor and capital? What about the potential emergence of “singularities” and “superintelligence,” concepts that animate many discussions in the machine-intelligence community? How will the linkages between AI and growth be mediated by firm-level considerations, including organization and market structure? The goal throughout is to refine a set of critical questions about AI and economic growth, and to contribute to shaping an agenda 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 what is essential and yet hard to improve. (pages 237 - 290)
This chapter is available at:
University of Chicago Press
https://academic.oup.com/chica...
AI and Jobs: The Role of Demand
DOI: 10.7208/chicago/9780226613475.003.0010
[automation;technical change;sectoral growth;labor demand;deindustrialization]
Artificial intelligence technologies will automate many jobs, but the effect on employment is not obvious. In manufacturing, technology has sharply reduced the number of jobs in recent decades. But before that, for over a century, employment grew, even in industries that experienced rapid technological change. What changed? Demand was highly elastic at first and then became inelastic. The effect of artificial intelligence on jobs will similarly depend critically on the nature of demand. This paper presents a simple model of demand that accurately predicts the rise and fall of employment in the textile, steel, and automotive industries. This model provides a useful framework for exploring how artificial intelligence is likely to affect jobs over the next 10 or 20 years. (pages 291 - 308)
This chapter is available at:
University of Chicago Press
https://academic.oup.com/chica...
Public Policy in an AI Economy - Austan Goolsbee
DOI: 10.7208/chicago/9780226613475.003.0011
[Artificial intelligence;Universal basic income;Policy;Antitrust;Future of work;Government;Regulation;privacy]
Much public discussion about an economy dominated by artificial intelligence has focused on robots and the future of work, particularly the destruction of jobs. On the other hand, economists have highlighted the historical record of job creation despite job displacement, and documented the way technological advances have eliminated jobs in some sectors but expanded jobs and increased wages in the economy overall. This chapter considers the role of policy in an artificial-intelligence-intensive economy (interpreting artificial intelligence broadly). It emphasizes the speed of adoption of the technology for the impact on the job market and the implications for inequality across people and across places. It also discusses the challenges of enacting a Universal Basic Income as a response to widespread adoption of artificial intelligence, and discusses pricing, privacy and competition policy, as well as the question of whether artificial intelligence could affect policy making itself. (pages 309 - 316)
This chapter is available at:
University of Chicago Press
https://academic.oup.com/chica...
Should We Be Reassured If Automation in the Future Looks Like Automation in the Past? - Jason Furman
DOI: 10.7208/chicago/9780226613475.003.0012
[artificial intelligence;productivity;inequality;labor force participation;automation;labor market institutions]
While artificial intelligence has not yet had a large macroeconomic impact, its potential economic impact is likely to grow in the coming years. Artificial intelligence offers both potential benefits, including higher productivity, and potential side effects, such as higher inequality or falling labor force participation. However, these outcomes are not inevitable. Instead, the ultimate impact of artificial intelligence will be shaped by public policy choices and other institutions. I argue that policies and institutions that help workers adapt to labor market changes brought on by artificial intelligence and which help to ensure that its benefits are widely distributed are likely to be more successful than large-scale changes to the social safety net, like establishing a universal basic income, at maximizing the potential benefits of artificial intelligence and minimizing its potentially disruptive side effects. (pages 317 - 328)
This chapter is available at:
University of Chicago Press
https://academic.oup.com/chica...
R&D, Structural Transformation, and the Distribution of Income - Jeffrey D. Sachs
DOI: 10.7208/chicago/9780226613475.003.0013
[aggregate factor income distribution;labor demand;labor productivity;innovation;technological change]
In Solow’s growth model, labor-augmenting technical change at a constant rate produces long-term growth in output per capita and wages at the same constant rate. The returns to capital are stable, as are the factor shares of national income going to labor and capital. These stylized facts have visibly broken down since around 2000. The US economy is characterized by science-led growth that is inherently unbalanced. There is a long-term shift in the share of national income from labor to capital, including physical, human, and intellectual capital. This shift was partially obscured in the past by including both labor income and human capital income within the traditional measure of labor income. The digital revolution will continue to shift national income from labor towards capital. I introduce a stylized computable general-equilibrium model to explore these long-term trends. (pages 329 - 348)
This chapter is available at:
University of Chicago Press
https://academic.oup.com/chica...
Artificial Intelligence and Its Implications for Income Distribution and Unemployment - Anton Korinek, Joseph E. Stiglitz
DOI: 10.7208/chicago/9780226613475.003.0014
[artificial intelligence;channels of inequality;technological unemployment;Malthusian dynamics]
Inequality is one of the main challenges posed by the proliferation of artificial intelligence (AI) and other forms of worker-replacing technological progress. This paper provides a taxonomy of the associated economic issues: First, we discuss the general conditions under which new technologies such as AI may lead to a Pareto improvement. Secondly, we delineate the two main channels through which inequality is affected – the surplus arising to innovators and redistributions arising from factor price changes. Third, we provide several simple economic models to describe how policy can counter these effects, even in the case of a “singularity” where machines come to dominate human labor. Under plausible conditions, non-distortionary taxation can be levied to compensate those who otherwise might lose. Fourth, we describe the two main channels through which technological progress may lead to technological unemployment – via efficiency wage effects and as a transitional phenomenon. Lastly, we speculate on how technologies to create super-human levels of intelligence may affect inequality and on how to save humanity from the Malthusian destiny that may ensue. (pages 349 - 390)
This chapter is available at:
University of Chicago Press
https://academic.oup.com/chica...
Neglected Open Questions in the Economics of Artificial Intelligence - Tyler Cowen
DOI: 10.7208/chicago/9780226613475.003.0015
[economics of technology;distribution;labor economics;public choice;international development]
The economics of artificial intelligence (AI) is commanding growing attention, but not all aspects of the problem have been considered adequately. For instance, in addition to its immediate effects on real wages, AI also has distributional effects through consumer surplus. Those effects may vary greatly, depending on whether there are constant or increasing returns to scale. AI also will have distributional consequences in the international sphere, for instance by making it harder for lower-wage nations to build their export capacities by offering lower wages to investors. Finally, AI will interact with political economy factors. Society may need to redistribute status as well as income, and furthermore the possibility of corruption may render large-scale redistribution impractical. (pages 391 - 396)
This chapter is available at:
University of Chicago Press
https://academic.oup.com/chica...
Artificial Intelligence, Economics, and Industrial Organization - Hal Varian
DOI: 10.7208/chicago/9780226613475.003.0016
[machine learning;cloud computing;algorithms]
Machine learning (ML) and artificial intelligence (AI) have been around for many years. However, in the last five years, remarkable progress has been made using multilayered neural networks in diverse areas such as image recognition, speech recognition, and machine translation. AI is a general purpose technology that is likely to impact many industries. In this chapter I consider how machine learning availability might affect the industrial organization of both firms that provide AI services and industries that adopt AI technology. My intent is not to provide an extensive overview of this rapidly evolving area, but instead to provide a short summary of some of the forces at work and to describe some possible areas for future research. (pages 399 - 422)
This chapter is available at:
University of Chicago Press
https://academic.oup.com/chica...
Privacy, Algorithms, and Artificial Intelligence
DOI: 10.7208/chicago/9780226613475.003.0017
[privacy;algorithms;digital data;artificial intelligence]
Artificial intelligence can use an individual’s data to make predictions about what they might desire, be influenced by, or do. The use of an individual’s data in this process raises privacy concerns. The desire (or lack of desire) for privacy will be a function of an individual’s anticipation of the consequences of their data being used in a predictive algorithm. In this way, the question of AI algorithms seems simply a continuation of the tension that has plagued earlier work in the economics of privacy. So, a natural question is whether AI presents new or different problems. This chapter focuses on what is novel about the world of artificial intelligence and privacy, arguing that the chief novelty lies in the potential for data persistence, data repurposing and data spillovers. (pages 423 - 438)
This chapter is available at:
University of Chicago Press
https://academic.oup.com/chica...
Artificial Intelligence and Consumer Privacy
DOI: 10.7208/chicago/9780226613475.003.0018
[privacy;consumer protection;data;security;artificial intelligence;law;technology]
Thanks to technological advances that have enabled a radical decline in the cost of collecting, storing, processing and using data in mass quantities, or “big data,” artificial intelligence (AI) has spurred exciting innovations. AI and big data are also reshaping the risk in consumer privacy and data security. Big data introduces three “new” problems for consumer privacy: (1) sellers initially have more information about future data use than buyers after the focal transaction; (2) sellers need not fully internalize potential harms to consumers because of the inability to trace harm back to a data collector; and (3) sellers may promise consumer-friendly data policy at the time of data collection but renege afterwards, as it is difficult to detect and penalize it ex post. All three affect data collection, data storage, and data use. In this chapter, I first define the nature of the problem and then present a few facts about the ongoing risk. The bulk of the chapter describes how the U.S. market copes with the risk in the current policy environment. It concludes with key challenges facing researchers and policy makers. (pages 439 - 462)
This chapter is available at:
University of Chicago Press
https://academic.oup.com/chica...
AI and International Trade
DOI: 10.7208/chicago/9780226613475.003.0019
[artificial intelligence;machine learning;trade;data;privacy;behind-the-border;agglomeration]
AI will likely raise average incomes and improve well-being, but it may also disrupt labor markets, raise inequality and drive non-inclusive growth. To the extent that progress has been made in understanding the impact of AI, we have been largely uninformed about its international dimensions. This chapter explores the international dimensions of the economics of artificial intelligence. Trade theory emphasizes the roles of scale, competition, and knowledge creation and knowledge diffusion as fundamental to comparative advantage. We explore key features of artificial intelligence with respect to these dimensions and describe the features of an appropriate model of international trade in the context of artificial intelligence. We then discuss policy implications with respect to investments in research, and behind-the-border regulations such as privacy, data localization, standards, and competition. We conclude by emphasizing that there is still much to learn before we have a comprehensive understanding of how artificial intelligence will affect trade. (pages 463 - 492)
This chapter is available at:
University of Chicago Press
https://academic.oup.com/chica...
Punishing Robots: Issues in the Economics of Tort Liability and Innovation in Artificial Intelligence - Alberto Galasso, Hong Luo
DOI: 10.7208/chicago/9780226613475.003.0020
[tort reforms;product liability;regulation;innovation;research and development;artificial intelligence;robotics;general-purpose technologies;risk-mitigating technologies]
Liability laws designed to compensate for harms that are caused by defective or dangerous products or that are the result of professional negligence may also affect innovation incentives. Advances in artificial intelligence and in robotics have generated lively debates over whether existing liability systems constrain technological progress and whether they present an opportunity to redesign liability rules. This chapter reviews empirical studies on the links between liability and innovation using a large sample of data.It aims to provide some insights into the potential impacts that liability laws and likely changes in the system may have on the rate and direction of innovation in robots and artificial intelligence,andto identify areas and questions for future research. (pages 493 - 504)
This chapter is available at:
University of Chicago Press
https://academic.oup.com/chica...
The Impact of Machine Learning on Economics - Susan Athey
DOI: 10.7208/chicago/9780226613475.003.0021
[machine learning;econometrics;causal inference;artificial intelligence;data science]
This chapter provides an assessment of the early contributions of machine learning to economics, as well as predictions about its future contributions. It begins by briefly overviewing some themes from the literature on machine learning, and then draws some contrasts with traditional approaches to estimating the impact of counterfactual policies in economics. Next, we review some of the initial “off-the-shelf” applications of machine learning to economics, including applications in analyzing text and images. We then describe new types of questions that have been posed surrounding the application of machine learning to policy problems, including “prediction policy problems,” as well as considerations of fairness and manipulability. We present some highlights from the emerging econometric literature combining machine learning and causal inference. Finally, we overview a set of broader predictions about the future impact of machine learning on economics, including its impacts on the nature of collaboration, funding, research tools, and research questions. (pages 507 - 552)
This chapter is available at:
University of Chicago Press
https://academic.oup.com/chica...
AI, Labor, Productivity, and the Need for Firm-Level Data - Manav Raj, Robert Seamans
DOI: 10.7208/chicago/9780226613475.003.0022
[artificial intelligence;AI;automation;labor;management;productivity]
We summarize existing empirical findings regarding the adoption of robotics and AI and its effects on aggregated labor and productivity, and argue for more systematic collection of the use of these technologies at the firm level. Existing empirical work primarily examines robotics rather than AI, and uses statistics aggregated by industry or country, which precludes in-depth studies regarding the conditions under which these technologies complement or substitute for labor. Further, firm-level data would also allow for studies of effects on firms of different sizes, the role of market structure in technology adoption, the impact on entrepreneurs and innovators, and the effect on regional economies amongst others. We highlight several ways that such firmlevel data could be collected and used by academics, policymakers and other researchers. (pages 553 - 566)
This chapter is available at:
University of Chicago Press
https://academic.oup.com/chica...
How Artificial Intelligence and Machine Learning Can Impact Market Design - Paul R. Milgrom, Steven Tadelis
DOI: 10.7208/chicago/9780226613475.003.0023
[Market design;incentive autions;online marketplaces;trust and reputation;online search]
In complex environments, it can be difficult to understand the underlying characteristics of transactions, and it is challenging to learn enough about them so as to design the best institutions to efficiently generate gains from trade. In recent years, artificial intelligence has emerged as an important tool that allows market designers to uncover important market fundamentals, and to better predict fluctuations that can cause friction in markets. Examples include retailers and marketplaces who mine vast amounts of data to identify patterns that help them increase the efficiency of their markets, and auction designers who train learning models to simplify auctions with complex sets of constraints. With better prediction tools, companies can better anticipate consumer demand and producer supply as well as identify risks to the integrity of transactions. This chapter offers some recent developments of how artificial intelligence helps market designers improve the operations of markets, and outlines directions in which it will continue to shape and influence market design. (pages 567 - 586)
This chapter is available at:
University of Chicago Press
https://academic.oup.com/chica...
Artificial Intelligence and Behavioral Economics
DOI: 10.7208/chicago/9780226613475.003.0024
[overfitting;clinical judgment;machine learning;judgment bias;behavioral economics;behavioral types;limited attention]
An important strand of judgment and decision research in the 1970s and 1980s, which influenced behavioral economics mostly indirectly, documented the fact that simple statistical models often predict better than experts do (beginning with Meehl 1954). This chapter revisits this phenomenon and connects it to modern machine learning (ML) debates. One view is that subjective judgment exhibits properties of “unregularized” ML which overfits, because human judgment does not naturally penalize overfitting.I also discuss how ML techniques, in large data sets, can help discover different “behavioral types” which correspond to, or extend, heterogeneous types hypothesized previously.ML applications in practice, such as recommender systems, also connect to behavioral concepts of how limited attention and preference assembly create opportunities to either benefit or harm consumers. (pages 587 - 610)
This chapter is available at:
University of Chicago Press
https://academic.oup.com/chica...
