Generative AI (artificial intelligence), the most well-known of which is ChatGPT, is attracting much attention around the world.^[i] This also applies to economics, which is my field of expertise. For example, Professor Acemoglu of the Massachusetts Institute of Technology, one of the leading economists of our time, has expressed the following concern:^[ii] "AI is robbing human beings of work and agency.^[iii] On the other hand, Professor Pissarides of the London School of Economics, winner of the Nobel Prize in Economics in 2010, expresses an optimistic view that the evolution of AI could help transition to a four-day workweek, and that generative AI could ultimately improve labor productivity.^[iv] Although these opinions are not necessarily contradictory, it is difficult to say that a clear consensus has been formed even among economists.

The rise of generative AI is reducing the functionality of the market economy

Recently, problems caused by generative AI are becoming increasingly common in the form of criminal damage. Lately, The New York Times and other newspapers have reported on the outbreak of a new type of publishing fraud. One example is the huge number of extremely low-quality travel guidebooks generated by AI listed on Amazon.^[v] Moreover, according to an article in The Tokyo Shimbun, identity fraud frequently occurs in China called deepfake, which uses AI to generate photos, videos, and audio and then processes them to resemble the actual subject. Furthermore, in the United States, a series of class-action lawsuits have been filed by artists alleging intellectual property infringement based on the claim that generative AI is using copyrighted text without permission.^[vi]

This kind of turmoil is likely to continue for the foreseeable future. Nevertheless, from a historical perspective, the modern economic turmoil surrounding ChatGPT may not be so unique.

Technological innovation temporarily disrupts the socioeconomic system: From a market quality perspective

According to the market quality theory^[vii] advocated by Makoto Yano (Specially Appointed Professor in the Institute of Economic Research, Kyoto University), who was my academic advisor when I was a college student and has been my long-time collaborator, at the time of past industrial revolutions experienced by humankind at least three times, there was a temporary decline in market quality which resulted in low growth and economic stagnation.

The view that we are currently facing the Fourth Industrial Revolution is widely shared not only by some researchers but also by policymakers in Japan. AI is a core technology of the Fourth Industrial Revolution, and it continues to develop at a speed that is just as fast^[viii] or even faster than envisioned several years ago. According to market quality theory, it is a natural phenomenon in line with economic laws that market quality declines and society disrupts temporarily due to major innovations such as ChatGPT and other forms of AI technology.

Stagnation can be overcome

Then, in order to avoid economic stagnation due to a decline in market quality, should we stop the development and utilization of AI technology? From an economic perspective, we should avoid overly simplistic binary questions such as, "Are you for or against using generative AI in your work?" Instead, with a focus on the near future when AI technology will advance considerably, we should deepen the discussion on how to socially implement AI (utilize AI in the real world) in a way that is desirable for society. We should also discuss topics such as the framework of a safety net to deal with unemployment and low wages caused by the introduction of AI in the future.

Let's look back at history. As shown above in the diagram of industrial revolution cycles, in the past, the world economy stagnated after the occurrence of an industrial revolution. However, high-quality markets were recovered through efforts such as the development of legal systems (Labor Act during the First Industrial Revolution, and Antitrust Act and Securities Act during the Second Industrial Revolution). After escaping from the temporary turmoil, further technological progress and economic development continued. Together with Professor Yano, I theoretically clarified the mechanism behind this historical observation in a study published in the Proceedings of the National Academy of Sciences (Yano and Furukawa, 2023).^[ix] Based on these considerations, even if generative AI were to disrupt modern society, it should be possible to overcome the turmoil by designing appropriate legal systems and policies.

Theory of directed technical change

As a key to consider such a design, I would like to explain the concept of directed technical change by Professor Acemoglu, whom I introduced above. I will refrain from discussing the details of the research in this article, but the core idea is that "technology progresses in a specific direction." So, what are the specific directions in which technology is changing? The original research focused on the (specialized) skills and abilities possessed by workers. However, if I were to explain the research in my way based on the context up to this point, I would say that it is possible to consider two different directions. The first type is technology that advances as a substitute for work (labor) which had previously been performed by humans (labor substitution). In contrast, the second type is the technology which advances in a way that helps improve the quality of labor and increases labor productivity (labor complimentary).^[x]

If I were to give an example of each type in connection with generative AI (I will focus on ease of understanding and give extreme examples in the assumption that I will not be misconstrued), the first type would be using ChatGPT to create in 10 minutes a report that used to take an hour to create, while maintaining the same level of quality. The second type would be using ChatGPT to improve the quality of the report while keeping the time required for creation at one hour. In the case of the former type of technological progress that is a substitute for labor, AI simply replaces existing labor,^[xi] so there is the possibility that the evolution of AI will cause unemployment or deprive humans of their agency.

Based on this concept and the market quality theory, the following message can be gleaned from the views of Professor Acemoglu and Professor Pissarides as introduced at the beginning of this article: "It is important to aim for both human agency and improved labor productivity by fostering a shared awareness for steering the evolution of AI technology in the direction of complimenting labor, and by designing desirable rules for supporting such complimentary behavior."

Conclusion: Economics of cultural customs

Designing rules and legal systems is easier said than done. One of the basic propositions of market quality theory is that appropriate coordination of market infrastructure is a necessary condition for a high-quality market (Yano, 2009). Here, the term of market infrastructure is a complex concept that includes various factors surrounding people's market activities, such as culture, customs, and values, with the legal system at the core. This basic proposition does not mean that simply creating laws is enough to reach a resolution; instead, it suggests the importance of thinking in advance about how to ensure the proper functioning of laws. Although market infrastructure includes various elements, I focus on culture and national character in particular.

In a joint study with Associate Professor Tat-kei Lai (IESEG School of Management Paris) and Associate Professor Kenji Sato (Osaka Metropolitan University) that was recently published in Macroeconomic Dynamics (Furukawa, Lai, and Sato, 2023), we clarified both theoretically and through data analysis that the culture and national character of people who love new things play an important role in the speed and direction of technical change. These findings indicate that even if systems and policy designs that have been successful in foreign countries are introduced directly into Japan, they may not work well due to differences in cultures and customs. When discussing the ideal direction of development for AI technology and how that technology should be socially implemented for people, it is important to pay attention to coordination with national and regional cultures and customs.

---

[i] The same is true in the university-industry to which I belong. Each university is formulating and publishing provisional rules for the use of AI. For example, on May 26, 2023, The University of Tokyo published provisional rules regarding the use of generative AI tools. As expressed by the term of provisional, at the individual level, there is a wide range of ideas, from the position that generative AI should be prohibited in principle to the position that the use of AI should be encouraged (with certain restrictions). There is ongoing debate regarding these ideas.
[ii] From an interview article by Ayako Hirono (Deputy Editor-in-Chief of Nikkei Business) entitled Warning from MIT Professor Acemoglu: We Must Restrict AI Development That Robs Human Beings of Agency. Nikkei Business, Global Intelligence (May 12, 2023).
[iii] Similar opinions have also been submitted by researchers in the natural sciences. For example, internationally known physicist Stephen Hawking (1942-2018) mentioned the possibility of AI completely replacing humans in an interview with WIRED in 2017, the year before his death.
"Stephen Hawking: I fear AI may replace humans altogether," João Medeiros, WIRED, (November 28, 2017).
[iv] "ChatGPT could finally turn workers' 4-day week dream into reality, says Nobel Prize-winning economist," Fortune online edition (April 5, 2023).
[v] "A New Frontier for Travel Scammers: A.I.-Generated Guidebooks," Seth Kugel and Stephen Hiltner, New York Times online edition (August 5, 2023).
[vi] "Michael Chabon, Other Literary Giants Sue OpenAI Over Alleged Copyright Infringement," Cyrus Farivar, Forbes online edition (September 11, 2023).
[vii] Yano (2009). General explanations of market quality theory include Makoto Yano's book (Yano, 2005) and articles, which I wrote while serving as a Faculty Fellow at the Research Institute of Economy, Trade and Industry.
[viii] In 2016, Japanese Economy 2016-2017 published by the Cabinet Office included a detailed examination of the growth of AI as part of the industrial changes that are currently underway.
[ix] Yano and Furukawa (2023). The same authors also wrote an explanatory text intended for the general public: Market Quality That Creates the Industrial Revolution Cycle: Policies That Create a Virtuous Cycle. Nikkei Business, Global Intelligence (May 19, 2023).
[x] For more information on this theory of directed technical change, see Acemoglu (1998, 2023). Furthermore, this field is also my area of expertise. For example, in a joint project with Professor Angus Chu of the University of Macau and Professor Guido Cozzi of the University of St. Gallen, we have published research on directed technical change in U.S.-China relations (Chu, Cozzi, and Furukawa, 2015).
[xi] Of course, even in the case of labor substitution raised in this example, the 50 minutes created by saving time can be used for other pursuits; for example, studying English, attending graduate school for working adults, or generating ideas for new products and projects. The ability to make investments leading to improved productivity in the future should not cause unemployment or loss of agency. In this case, in a broad sense, it may be possible to say that AI technology is being implemented in society as a supplement to labor. Furthermore, Acemoglu and Restrepo (2018) theoretically examined the possibility that the advancement of AI which replaces labor in a narrow sense does not create long-term unemployment. In this context, the aforementioned paper which I wrote together with Professor Yano (Yano and Furukawa, 2020) introduces a new perspective on AI that expands in a self-propagating manner without human input.

---

[Reference Literature]

・ Acemoglu, D. (1998) "Why do new technologies complement skills? Directed technical change and wage inequality," Quarterly Journal of Economics 113, 1055-1089.
・ Acemoglu, D. (2023) "Distorted innovation: Does the market get the direction of technology right?" American Economic Review Papers and Proceedings 113, 1-28.
・ Acemoglu, D. and P. Restrepo (2018) "The race between man and machine: Implications of technology for growth, factor shares and employment," American Economic Review 108, 1488-1542.
・ Chu, A., G. Cozzi, and Furukawa, Y. (2015) "Effects of Economic Development in China on Skill-Biased Technical Change in the US," Review of Economic Dynamics 18, 227-242.
・ Furukawa, Y., Tat-kei Lai, and Sato, K. (2023) "Love of Novelty: A Source of Innovation-Based Growth... or Underdevelopment Traps?" Macroeconomic Dynamics, forthcoming.
・ Yano, M. (2009) "The Foundation of Market Quality Economics," Japanese Economic Review 60, 1-32.
・ Yano, M. and Furukawa, Y. (2020) "Economic Black Holes and Labor Singularities in the Presence of Self-replicating Artificial Intelligence," RIETI Discussion Paper Series 20-E-009, February 2020.
・ Yano, M. and Furukawa, Y. (2023) "Two-Dimensional Constrained Chaos and Industrial Revolution Cycles," Proceedings of the National Academy of Sciences 120 (5), e2117497120.
・ Yano, M. (2005) "System Reform in the Era of Quality: What is a Good Market?," Iwanami Shoten.

Yuichi Furukawa／Professor, Faculty of Economics, Chuo University
Area of Specialization: Theoretical Economics

Yuichi Furukawa was born in 1977. He graduated from the Faculty of Economics, Keio University in 2000. He received his PhD (economics) from the Graduate School of International Social Sciences, Yokohama National University in 2005. He held positions such as Professor in the School of Economics, Chukyo University, and Professor in the Faculty of Economics, Aichi University before assuming his current position in 2022.

He served as a Faculty Fellow at the Research Institute of Economy, Trade and Industry until June 2022. In 2019, he received the Special Field Research Encouragement Prize (the Oda Prize) from the Japan Society of International Economics. Since 2019, he has been involved in the editing of international academic journals such as Economic Modelling and International Journal of Economic Theory.

His area of expertise is macroeconomics. Recently, he has been particularly interested in the influence of culture on innovation and ultra-long-term economic changes.

His major written works include Market Quality and Modern Economies (author and editor, Keiso Shobo, 2016).

• Share articles
• 
• 
•