by Ajay Agrawal
matrix factorization, 51, 511
judgment and decision- making (JDM)
meta technologies, 153
research, 596– 98
ML. See machine learning (ML)
model averaging, 511
knowledge creation, 477– 79
MOOCS (massive open online courses),
knowledge externalities, and artifi cial intel-
181
ligence, 470– 71
knowledge spillovers, 479
neural networks, 72– 74, 123, 124– 25, 510,
511
labor: comparative advantage of, and new
new economic geography (NEG), 479– 80
tasks, 217– 18; model of demand for,
211– 14
online marketplaces, using artifi cial intelli-
labor demand, technology and, 214– 21
gence to promote trust in, 576– 81
labor productivity growth, technologies and,
optical lenses, invention of, 121
26– 28
optimism, sources of technological, 24– 26
learning by doing, 412– 13
liability, innovation and, 494; empirical evi-
Pareto improvement, 363
dence on, 496– 98; theoretical model of,
patent systems, deep learning and, 142
494– 96
Pen Register Act (1986), 456
liability, tort, development of artifi cial intel-
policy analysis, using methods of prediction
ligence technologies and, 498– 502
for, 516– 19
political economy: of artifi cial intelligence,
machine learning (ML), 3, 24– 25; applica-
11; of income redistribution, 394– 95;
tions of, 401– 2; defi ned, 509– 10;
of technological disruptions, 176– 79
double, 523– 24; early use cases of, 510–
prediction: in absence of judgment, 101– 2;
15; for fi nding behavioral variables,
artifi cial intelligence as tool for, 16– 17;
587– 96; general purpose, 67– 71; human
as complements/ substitutes to judg-
prediction as imperfect, 496– 603; im-
ment, 102– 3; costs of, and AI, 92– 93;
pact of, 12– 13, 15– 17, 507– 9; incentive
judgment and, 91– 92; using methods
auctions and, 569– 76; new literature
of, in policy analysis, 516– 19
on, 519– 34; overview, 399– 406; predic-
premature deindustrialization, 393– 94
tions about impact of on economics,
price discrimination, artifi cial intelligence
534– 42; regulation and, 12– 15; super-
and, 604
vised, 511– 12; unsupervised, 510– 11;
principal components analysis, 74, 92,
vertical integration and, 408– 9. See also
510
artifi cial intelligence (AI)
privacy, 13– 14; artifi cial intelligence and,
Subject Index 629
425– 26; current models of economics
search frictions, artifi cial intelligence for
and, 424– 25; data spillovers and, 431.
reducing, 581– 83
See also consumer privacy
security policy, 416
privacy policy, 416– 17
singularities, 253– 61; examples of techno-
privacy regulation, trade policies and,
logical, 254– 58; objections to, 258– 61
482– 85
skills: mismatch of technologies and, 221–
privileged access to government data, trade
23; technologies and, 209
policies and, 486– 87
Solow paradox, 24; potential explanations
production, automation of, and artifi cial
for, 28– 31
intelligence, 239– 50
source code, trade policies and, 487– 88
productivity: automation and, 210– 11; miss-
spectrum reallocation, 569– 76, 572f
ing growth of, and new technologies,
spillovers, data, 431
223– 26
spreadsheet software, invention and impact
productivity eff ects, 198, 203– 4, 214– 16
of, 90
productivity growth: low current, reasons
stochastic gradient descent optimization,
why it is consistent with future techno-
77– 81
logical growth, 41– 44; rates of, tech-
strategic trade policy, 473– 74
nologies and, 26– 28; slow, and future
structural change, 293– 96
productivity growth, 31– 36
structural models, 532– 34
productivity optimism, technology- driven
superstar scientists, role of, 474– 76
case for, 36– 39
supervised machine learning, 511; methods
for, 511– 12
radiology, case of, 94– 95
supplementary analysis, 530
random forest, 511
support vector machines, 511
R&D. See research and development (R&D)
symbolic processing hypothesis, 123
recommender systems, 603
symbolic systems, 123
regional clusters, theory of, 479
regression trees, 511
tasks, new, 205– 7; comparative advantage
regularization on norm of matrix, 510
of labor and, 217– 18; creation of, 198;
regularized regression, 511
model of, 211– 14
regulation, 12– 15; machine learning and,
technological changes: factor- biased, 376–
12– 15
77; and levels of employment, 220– 21;
reinforcement learning, 25, 66, 81– 84, 400–
types of, 212– 13
401
technological disruptions, political economy
reinstatement eff ect, 8, 198, 206
of, 176– 79
research and development (R&D), 336; pro-
technological growth, reasons why it is con-
ductivity, eff ects of rise in, 341– 43
sistent with low current productivity
research tools, economics of new, 118– 22
growth, 41– 44
robotics, 123– 24; tort law and, 493– 94.
technological optimism, sources of, 24– 26
See also robots
technological progress: channels of inequal-
robots, studies on, 554– 55. See also robotics
ity and, 365– 70; determining scenarios
Romer/ Jones knowledge production func-
that best describe economy, 363– 64;
tion, 151– 52
endogenous, 364– 65; fi rst- best scenario,
353– 56; imperfect markets scenario,
scale, economies of, and artifi cial intelli-
361– 62; perfect markets but costly
gence, 470
redistribution scenario, 358– 61; perfect
Schumpeterian model with artifi cial intel-
markets ex post and no costs of redis-
ligence, 276– 79
tribution scenario, 356– 58; welfare and,
scientifi c discovery, rate of, 6
353– 65; worker- replacing, redistribu-
scientists, role of, 472– 73; superstar, 474– 76
tion and, 370– 77
scope, economies of, and artifi cial intel-
technological singularities, examples of,
ligence, 470
254– 58
630 Subject Index
technological unemployment, 377– 81
universal basic income (UBI), 312– 14; cost
of replacing current safety net with,
current productivity growth, 41– 44;
325– 26
labor demand and, 214– 21; mismatch
unsupervised machine learning, 510– 11
of skills and, 221– 23; skills and, 209
USA Patriot Act (2001), 456
technology- driven economic growth, pros-
pects for, 149– 53
vertical integration, machine learning and,
tort law, robotics and, 493
408– 9
tort liability, 14
vertical research spillovers, 122
total factor productivity growth, 32
trade models, basic, 476– 77
wages, automation and, 200– 211
trade policies: data localization rules and,
welfare, technological progress and, 353– 65
485– 86; industrial and strategic, case
work, automation and, 200– 211
for, 471– 81; industrial regulation and,
worker- replacing technological progress:
487; privacy regulation and, 482– 85;
dynamic implications of, 373– 74; redis-
privileged access to government data
tributing innovators’ surplus and, 374–
and, 486– 87; role of university- related
76; redistribution and, 370– 77; static
talent, 472– 76; source code and, 487–
pecuniary externalities of, 371– 72
88; strategic, 473– 74
Zeira model of automation and growth,
UBI. See universal basic income (UBI)
239– 41
unemployment, technological, 377– 81
“zero- shot” learning systems, 66
Document Outline
Contents
Acknowledgments
Introduction / Ajay Agrawal, Joshua Gans, and Avi Goldfarb
I. AI as a GPT 1. Artificial Intelligence and the Modern Productivity Paradox: A Clash of Expectations and Statistics / Erik Brynjolfsson, Daniel Rock, and Chad Syverson, Comment: Rebecca Henderson
2. The Technological Elements of Artificial Intelligence / Matt Taddy
3. Prediction, Judgment, and Complexity: A Theory of Decision-Making and Artificial Intelligence / Ajay Agrawal, Joshua Gans, and Avi Goldfarb, Comment: Andrea Prat
4. The Impact of Artificial Intelligence on Innovation: An Exploratory Analysis / Iain M. Cockburn, Rebecca Henderson, and Scott Stern, Comment: Matthew Mitchell
5. Finding Needles in Haystacks: Artificial Intelligence and Recombinant Growth / Ajay Agrawal, John McHale, and Alexander Oettl
6. Artificial Intelligence as the Next GPT: A Political-Economy Perspective / Manuel Trajtenberg
II. Growth, Jobs, and Inequality 7. Artificial Intelligence, Income, Employment, and Meaning / Betsey Stevenson
8. Artificial Intelligence, Automation, and Work / Daron Acemoglu and Pascual Restrepo
9. Artificial Intelligence and Economic Growth / Philippe Aghion, Benjamin F. Jones, and Charles I. Jones, Comment: Patrick Francois
10. Artificial Intelligence and Jobs: The Role of Demand / James Bessen
11. Public Policy in an AI Economy / Austan Goolsbee
12. Should We Be Reassured If Automation in the Future Looks Like Automation in the Past? / Jason Furman
13. R&D, Structural Transformation, and the Distribution of Income / Jeffrey D. Sachs
14. Artificial Intelligence and Its Implications for Income Distribution and Unemployment / Anton Korinek and Joseph E. Stiglitz
15. Neglected Open Questions in the Economics of Artificial Intelligence / Tyler Cowen
III. Machine Learning and Regulation 16. Artificial Intelligence, Economics, and Industrial Organization / Hal Varian, Comment: Judith Chevalier
17. Privacy, Algorithms, and Artifi cial Intelligence / Catherine Tucker
18. Artificial Intelligence and Consumer Privacy / Ginger Zhe Jin
19. Artificial Intelligence and International Trade / Avi Goldfarb and Daniel Trefler
20. Punishing Robots: Issues in the Economics of Tort Liability and Innovation in Artificial Intelligence / Alberto Galasso and Hong Luo
IV. Machine Learning and Economics 21. The Impact of Machine Learning on Economics / Susan Athey, Comment: Mara Lederman
22. Artificial Intelligence, Labor, Productivity, and the Need for Firm-Level Data / Manav Raj and Robert Seamans
23. How Artificial Intelligence and Machine Learning Can Impact Market Design / Paul R. Milgrom and Steven Tadelis
24. Artificial Intelligence and Behavioral Economics / Colin F. Camerer, Comment: Daniel Kahneman
Contributors
Author Index
Subject Index