by Ajay Agrawal
York: Harper & Row.
236 Daron Acemoglu and Pascual Restrepo
Syverson, Chad. 2017. “Challenges to Mismeasurement Explanations for the US
Productivity Slowdown.” Journal of Economic Perspectives 31 (2): 165– 86.
Tuzel, Selale, and Miao Ben Zhang. 2017. “Economic Stimulus at the Expense of
Routine- Task Jobs.” Unpublished manuscript, Marshall School of Business, Uni-
versity of Southern California.
Zeira, Joseph. 1998. “Workers, Machines, and Economic Growth.” Quarterly Journal
of Economics 113 (4): 1091– 117.
9
Artifi cial Intelligence
and Economic Growth
Philippe Aghion, Benjamin F. Jones, and Charles I. Jones
9.1 Introduction
This chapter considers the implications of artifi cial intelligence for eco-
nomic growth. Artifi cial intelligence (AI) can be defi ned as “the capability
of a machine to imitate intelligent human behavior” or “an agent’s ability to
achieve goals in a wide range of environments.”1 These defi nitions immedi-
ately evoke fundamental economic issues. For example, what happens if AI
allows an ever- increasing number of tasks previously performed by human
labor to become automated? Artifi cial intelligence may be deployed in the
ordinary production of goods and services, potentially impacting economic
growth and income shares. But AI may also change the process by which we
create new ideas and technologies, helping to solve complex problems and
scaling creative eff ort. In extreme versions, some observers have argued that
AI can become rapidly self- improving, leading to “singularities” that feature
unbounded machine intelligence and/or unbounded economic growth in
Philippe Aghion is a professor at the Collège de France and at the London School of Economics. Benjamin F. Jones is the Gordon and Llura Gund Family Professor of Entrepreneurship, professor of strategy, and faculty director of the Kellogg Innovation and Entrepreneurship Initiative at Northwestern University, and a research associate of the National Bureau of Economic Research. Charles I. Jones is the STANCO 25 Professor of Economics at the Graduate School of Business at Stanford University and a research associate of the National Bureau of Economic Research.
We are grateful to Ajay Agrawal, Mohammad Ahmadpoor, Adrien Auclert, Sebastian Di
Tella, Patrick Francois, Joshua Gans, Avi Goldfarb, Pete Klenow, Hannes Mahlmberg, Pascual Restrepo, Chris Tonetti, Michael Webb, and participants at the NBER Conference on Artifi cial Intelligence for helpful discussion and comments. For acknowledgments, sources of research support, and disclosure of the authors’ material fi nancial relationships, if any, please see http://
www .nber .org/ chapters/ c14015.ack.
1. The former defi nition comes from the Merriam- Webster dictionary, while the latter is from Legg and Hutter (2007).
237
238 Philippe Aghion, Benjamin F. Jones, and Charles I. Jones fi nite time (Good 1965; Vinge 1993; Kurzweil 2005). Nordhaus (2015) provides a detailed overview and discussion of the prospects for a singularity
from the standpoint of economics.
In this chapter, we speculate on how AI may aff ect the growth process.
Our primary goal is to help shape an agenda for future research. To do so,
we focus on the following questions:
• If AI increases automation in the production of goods and services,
how will it impact economic growth?
• Can we reconcile the advent of AI with the observed constancy in
growth rates and capital share over most of the twentieth century?
Should we expect such constancy to persist in the twenty- fi rst century?
• Do these answers change when AI and automation are applied to the
production of new ideas?
• Can AI drive massive increases in growth rates, or even a singularity, as
some observers predict? Under what conditions, and are these condi-
tions plausible?
• How are the links between AI and economic growth modulated by
fi rm- level considerations, including market structure and innovation
incentives? How does AI aff ect the internal organization of fi rms, and
with what implications?
In thinking about these questions, we develop two main themes. First,
we model AI as the latest form in a process of automation that has been
ongoing for at least 200 years. From the spinning jenny to the steam engine
to electricity to computer chips, the automation of aspects of production
has been a key feature of economic growth since the Industrial Revolution.
This perspective is taken explicitly in two key papers that we build upon:
Zeira (1998) and Acemoglu and Restrepo (2016). We view AI as a new form
of automation that may allow additional tasks to be automated that previ-
ously were thought to be out of reach from automation. These tasks may
be nonroutine (to use the language of Autor, Levy, and Murnane [2003]),
like self- driving cars, or they may involve high levels of skill, such as legal
services, radiology, and some forms of scientifi c lab- based research. An
advantage of this approach is that it allows us to use historical experience
on economic growth and automation to discipline our modeling of AI.
A second theme that emerges in our chapter is that the growth conse-
quences of automation and AI may be constrained by Baumol’s “cost dis-
ease.” Baumol (1967) observed that sectors with rapid productivity growth,
such as agriculture and even manufacturing today, often see their share of
gross domestic product (GDP) decline while those sectors with relatively
slow productivity growth—perhaps including many services—experience
increases. As a consequence, economic growth may be constrained not by
what we do well but rather by what is essential and yet hard to improve. We
suggest that combining this feature of growth with automation can yield a
Artifi cial Intelligence and Economic Growth 239
rich description of the growth process, including consequences for future
growth and income distribution. When applied to a model in which AI
automates the production of goods and services, Baumol’s insight gener-
ates suffi
cient conditions under which one can get overall balanced growth
with a constant capital share that stays well below 100 percent, even with
near- complete automation. When applied to a model in which AI automates
the production of ideas, these same considerations can prevent explosive
growth.2
The chapter proceeds as follows. Section 9.2 begins by studying the role
of AI in automating the production of goods and services. In section 9.3,
we extend AI and automation to the production of new ideas. Section 9.4
then discusses the possibility that AI could lead to superintelligence or even
a singularity. In section 9.5, we look at AI and fi rms, with particular atten-
tion to market structure, organization, reallocation, and wage inequality. In
section 9.6, we examine sectoral evidence on the evolution of capital shares
in tandem with automation. Finally, section 9.7 concludes.
9.2 Artifi cial Intelligence and Automation of Production
One way of looking at the last 150 years of economic progress is that it
is driven by automation. The Industrial Revolution used steam and then
electricity to automate many pr
oduction processes. Relays, transistors, and
semiconductors continued this trend. Perhaps artifi cial intelligence is the
next phase of this process rather than a discrete break. It may be a natural
progression from autopilots, computer- controlled automobile engines,
and MRI machines to self- driving cars and AI radiology reports. While up
until recently automation has mainly aff ected routine or low- skilled tasks,
it appears that AI may increasingly automate nonroutine, cognitive tasks
performed by high- skill workers.3 An advantage of this perspective is that it
allows us to use historical experience to inform us about the possible future
eff ects of AI.
9.2.1 The Zeira (1998) Model of Automation and Growth
A clear and elegant model of automation is provided by Zeira (1998). In
its simplest form, Zeira considers a production function like
n
(1)
Y = AX 1 X 2 . . . X n where
= 1.
1
2
n
i
i=1
2. In the appendix we show that if some steps in the innovation process require human R&D, AI could possibly slow or even end growth by exacerbating business stealing, which in turn discourages human investments in innovation.
3. Autor, Levy, and Murnane (2003) discuss the eff ects of traditional software automating routine tasks. Webb et al. (2017) use the text of patent fi lings to study the diff erent tasks that AI, software, and robotics are best positioned to automate.
240 Philippe Aghion, Benjamin F. Jones, and Charles I. Jones While Zeira thought of the X s as intermediate goods, we follow Acemoglu
i
and Autor (2011) and refer to these as tasks; both interpretations have merit,
and we will go back and forth between these interpretations. Tasks that have
not yet been automated can be produced one- for- one by labor. Once a task
is automated, one unit of capital can be used instead:
L if not automated
(2)
X =
i
.
i
K if automated
i
If the aggregate capital K and labor L are assigned to these tasks optimally, the production function can be expressed (up to an unimportant constant) as
(3)
Y = A K L 1 ,
t
t
t
t
where it is now understood that the exponent refl ects the overall share and
importance of tasks that have been automated. For the moment, we treat
as a constant and consider comparative statics that increase the share of
tasks that get automated.
Next, embed this setup into a standard neoclassical growth model with
a constant investment rate; in fact, for the remainder of the chapter this is
how we will close the capital/ investment side of all our models. The share
of factor payments going to capital is given by and the long- run growth
rate of y ≡ Y/ L is
(4)
g =
g ,
y
1
where g is the growth rate of A. An increase in automation will therefore increase the capital share and, because of the multiplier eff ect associated
with capital accumulation, increase the long- run growth rate.
Zeira emphasizes that automation has been going on at least since the
Industrial Revolution, and his elegant model helps us to understand that.
However, its strong predictions that growth rates and capital shares should
be rising with automation go against the famous Kaldor (1961) stylized
facts that growth rates and capital shares are relatively stable over time. In
particular, this stability is a good characterization of the US economy for
the bulk of the twenieth century, for example, see Jones (2016). The Zeira
framework, then, needs to be improved so that it is consistent with historical
evidence.
Acemoglu and Restrepo (2016) provide one approach to solving this prob-
lem. Their rich environment allows for a constant elasticity of substitution
(CES) production function and endogenizes the number of tasks as well as
automation. In particular, they suppose that research can take two diff erent
directions: discovering how to automate an existing task or discovering new
tasks that can be used in production. In their setting, a refl ects the fraction of tasks that have been automated. This leads them to emphasize one possible
Artifi cial Intelligence and Economic Growth 241
resolution to the empirical shortcoming of Zeira: perhaps we are inventing
new tasks just as quickly as we are automating old tasks. The fraction of
tasks that are automated could be constant, leading to a stable capital share
and a stable growth rate.
Several other important contributions to this rapidly expanding literature
should also be noted. Peretto and Seater (2013) explicitly consider a research
technology that allows fi rms to change the exponent in a Cobb- Douglas
production function. While they do not emphasize the link to the Zeira
model, with hindsight the connections to that approach to automation are
interesting. The model of Hemous and Olsen (2016) is closely related to
what follows in the next subsection. They focus on CES production instead
of Cobb- Douglas, as we do below, but emphasize the implications of their
framework for wage inequality between high- skill and low- skill workers.
Agrawal, McHale, and Oettl (2017) incorporate artifi cial intelligence and
the “recombinant growth” of Weitzman (1998) into an innovation- based
growth model to show how AI can speed up growth along a transition path.
The next section takes a complementary approach, building on this lit-
erature and using the insights of Zeira and automation to understand the
structural change associated with Baumol’s cost disease.
9.2.2 Automation and Baumol’s Cost Disease
The share of agriculture in GDP or employment is falling toward zero.
The same is true for manufacturing in many countries of the world. Maybe
automation increases the capital share in these sectors and also interacts
with nonhomotheticities in production or consumption to drive the GDP
shares toward zero. The aggregate capital share is then a balance of a rising
capital share in agriculture/ manufacturing/ automated goods with a declin-
ing GDP share of these goods in the economy.
Looking toward the future, 3D printing techniques and nanotechnology
that allow production to start at the molecular or even atomic level could
someday automate all manufacturing. Could AI do the same thing in many
service sectors? What would economic growth look like in such a world?
This section expands on the Zeira (1998) and Acemoglu and Restrepo
(2016) models to develop a framework that is consistent with the large struc-
tural changes in the economy. Baumol (1967) observed that rapid productiv-
ity growth in some sectors relative to others could result in a “cost disease”
in which the slow- growing sectors become increasingly important in the
economy. We explore the possibility that automation is the force behind
these changes.4
4. The growth literature on this structural transformation emphasizes a range of possible mechanisms, see Kongsamut, Rebelo, and Xie (2001), Ngai and Pissarides (20
07), Herrendorf, Rogerson, and Valentinyi (2014), Boppart (2014), and Comin, Lashkari, and Mestieri (2015).
The approach we take next has a reduced form that is similar to one of the special cases in Alvarez- Cuadrado, Long, and Poschke (2017).
242 Philippe Aghion, Benjamin F. Jones, and Charles I. Jones Model
Gross domestic product is a CES combination of goods with an elasticity
of substitution less than one:
1/
1
(5)
Y = A
X di
where > 0,
t
t
it
0
where A = A egt captures standard technological change, which we take to t
0
be exogenous for now. Having the elasticity of substitution less than one
means that tasks are gross complements. Intuitively, this is a “weak link”
production function, where GDP is in some sense limited by the output of
the weakest links. Here, these will be the tasks performed by labor, and this
structure is the source of the Baumol eff ect.
As in Zeira, another part of technical change is the automation of produc-
tion. Goods that have not yet been automated can be produced one- for- one
by labor. When a good has been automated, one unit of capital can be used
instead:
L if not automated
(6)
X =
it
.
it
K if automated
it
This division is stark to keep the model simple. An alternative would be to
say that goods are produced with a Cobb- Douglas combination of capital
and labor, and when a good is automated, it is produced with a higher expo-
nent on capital.5
The remainder of the model is neoclassical:
(7)
Y = C + I ,
t
t
t
(8)
K = I
K ,
t
t
t
1
(9)
K di = K ,
it
t
0
1
(10)
L di = L.
it
0
We assume a fi xed endowment of labor for simplicity.
Let be the fraction of goods that that have been automated as of date
t
t. Here, and throughout the chapter, we assume that capital and labor are
allocated symmetrically across tasks. Therefore, K / units of capital are