Artificial Intelligence, Economics, and Industrial Organization

Article Source: in The Economics of Artificial Intelligence: An Agenda, Ajay K. Agrawal, Joshua Gans, and Avi Goldfarb, eds., University of Chicago Press, 2019
Publication Date:
Time to Read: 2 minute read
Written By:



Deployment of artificial Intelligence (AI) and machine learning (ML) systems will affect the size and choices of firms that provide or use AI services. Economists consider the effects of AI on pricing, firm size, competition, privacy, and security.


Policy Relevance:

The overall effect of AI and ML on firm behavior and choices is hard to predict.


Key Takeaways:
  • Computers can now outperform humans in many tasks that involve image recognition, voice recognition, language translation, and similar tasks.
  • Kaggle sets up competitions in which an organization provides some data and a problem statement, challenging data scientists to use the data to solve the problem for prize money.
    • One contest winner improved the accuracy of Homeland Security threat recognition.
    • Another predicted the effects of genetic variants to enable the development of personalized medicine.
  • Because one person's use of data does not diminish another person's use, data is different from a good like oil; one should think about access to data, rather than ownership of data.
  • Data exhibits decreasing returns to scale, like other factors of production; the accuracy of AI increases as the amount of training data increases, but at a decreasing rate, so that learning proceeds faster with the first set of images than the last set.
  • Key questions for economists include whether firms that adopt ML later will imitate early adopters; the role of patents, copyright, and trade secrets, and whether early adopters will enjoy a competitive advantage.
  • Researchers who study industrial organization will discover whether ML tools and data will be combined to create value within or across corporate boundaries, affecting vertical integration.
    • Cloud vendors now compete intensely to provide tools for data analysis.
    • Providers seek to standardize data systems, but also seek to differentiate their offerings from competitors.
  • AI might enable some firms to adopt price discrimination, which would tend to mean that poorer households would pay less for products than more wealthy households.
  • ML also might enable what is known as "algorithmic collusion," in which firms that interact with one another repeatedly begin to cooperate (for example, to fix prices) rather than to compete.
  • The effect of ML on competition in sectors that provide ML services is hard to predict; some firms that use a lot of data processing power can sell excess capacity to smaller firms, which would tend to increase competition in provision of ML.



Hal Varian

About Hal R. Varian

Hal R. Varian is an emeritus professor in the School of Information, the Haas School of Business, and the Department of Economics at the University of California at Berkeley. He is the Chief Economist at Google. He started in May 2002 as a consultant and has been involved in many aspects of the company, including auction design, econometric analysis, finance, corporate strategy and public policy.