Artificial Intelligence will Transform the Economy

MANAGI Shunsuke
Faculty Fellow, RIETI

The uncertain effects of innovative technology on society

Mankind has always developed new tools and technology to expand its habitat and to enrich its livelihood. In the history of mankind, social and economic activities have undergone drastic changes through various innovations such as new innovations such as food production, sanitation and transportation. Among them, the major technological innovation surrounding the invention of the steam engine, commonly referred to as the Industrial Revolution, dramatically improved our material well-being, with the ripple effect of enabling other various types of innovation that had not been possible earlier, which, in turn, exponentially transformed economic activities around the world.

However, the effects of such new technology innovation have not necessarily been positive and the ensuing drastic changes in the economy have also given rise to various social problems. Labor problem is one example. Through the invention and the widespread use of the steam engine, employers who had previously been major players in the manufacture of products were faced with the crisis of termination. As a result, in certain parts of England, laborers' fear of unemployment even gave rise to the radical Luddite movement, which advocated the destruction of machinery. Such crisis mentality became major issues in the subsequent technology innovation, and it is this crisis mentality that is attracting much attention in the IT revolution, which has been gaining momentum since the late-1990's as well as in the much-discussed development and dissemination of Artificial Intelligence (AI).

New technology innovation also has the potential of dramatically tipping the traditional economic balance of power between countries. Just as Japanese automobile manufacturers in the past achieved global technological superiority by successfully developing fuel-efficient cars and dominated the global automobile market, the development of self-driving cars has the potential of dramatically realigning each country's technological superiority in the automobile industry. Consequently, discussions on how to promote the development and the dissemination of AI and how to build a framework that can withstand economic changes on a global scale have become matters of the utmost urgency. To address this issue, we discussed how Japan should respond to this innovative technology from the perspective of international competition, and how society should respond to problems expected in the introduction of AI, from the standpoint of an economics study group on Artificial Intelligence (Reference: see book edited by Managi Shunsuke (2018)).

The present and the future of AI development and dissemination

Currently, the advances in Information and Communications Technology (ICT) have enabled corporations to utilize large volumes of big data and together with the technological progress made in machine learning and data mining, we have reached the stage where actual AI application to society has begun. Self-driving vehicles, albeit in their testing phases, are becoming a reality. Going forward the rapid development of AI is expected to enable its application to a wide range of fields.

However, there are also a number of major issues standing in the way of the dissemination of advanced AI in actual society, as currently envisioned. One major problem is the inordinate amount of time required for AI to learn elementary perception and motor skills. In other words, while we envision the widespread use of robots equipped with AI capable of acting and working like human beings in the near future, to realize this, advances in AI will not be enough.

It will require technology to control perception and motor skills, as well as engineering technology that will allow the robots to move flexibly and autonomously. Even if AI technology were to make rapid advances in the future, deeper application of AI to society will require the development of various related technologies, and myriads of uncertainties remain. In fact, industrial robots, which are currently in operation, can only perform extremely simplified tasks. Therefore, while the application of AI may not progress as we are currently envisioning, we anticipate the use of robots to steadily spread to simple tasks other than practical operations that contain layered processes, as well as to simple answering services capable of operating through mechanical pattern recognition in the near future. In this sense, the role of AI in economic activity is sure to expand.

Japan's delayed responses

In the history of AI development, there have been three periods of brisk activity, referred to as booms. The first AI boom was in the early 1960's, the second boom was in the 1980's and we are currently in the midst of the third AI boom. In past boom periods, Japan, with its advanced research, had managed to maintain an international competitive edge in its technology. However, in today's AI development, Japan has not only fallen behind China and India in the number of dissertations on AI by country, but also lags behind the US and the European countries, in terms of technology.

Our book explains why many companies have failed to utilize AI in their management from the standpoint of technology, human resources, management structure and external factors. Here, we will consider the problem of supplying the volumes of big data that are indispensable for AI. In the U.S., legislation pertaining to personal information has been passed allowing for the conditional use of personal information by companies. In Japan, on the other hand, under the Act on the Protection of Personal Information, third-party use of personal information is not allowed, which has resulted in the delay in the development of big data which is crucial for the development of AI. In May 2017, the amended Act on the Protection of Personal Information went into effect, allowing for conditional use of personal information (upon being processed so that individuals cannot be identified). The important point about this amendment of the Act on the Protection of Personal Information is that it provided a definition of personal information. Prior to this amendment, personal information had been insufficiently defined, which led to companies and universities utilizing personal information under relatively stringent conditions. This is at the root of the delay in the utilization of personal information in Japan. Given the limited amount of data held by Japanese companies, the foundations for collaboration between companies in the collection of information have yet to be laid.

How we should deal with AI going forward

Japan has already fallen behind the U.S. and the emerging countries such as India and China in the development of AI. This can be attributed not only to the lack of economic investments in the development and dissemination of AI but also to Japan's research framework itself, as well as delays in various fields including legislation and education. In other words, the accumulation of Japan's insufficient measures towards science and technology is at the core of this problem.

Consequently, Japan must rethink its previous stance toward science and technology, while at the same time think about how to specifically utilize AI, the innovative technology which is facing us. Rather than focusing on the basic technology, in which it has fallen behind, it would be worthwhile for Japan to formulate measures for catching up in the fields where the basic technology is to be applied. Currently, companies such as Google already exist with enormous volumes of big data on a global scale, and it would be difficult to compete against such companies in developing AI in terms of volumes of such data. Consequently, measures to enhance our competitiveness by collaborating with others in terms of basic AI technology, and by figuring out how to utilize AI in society are thought to be more effective.

It would then be possible to tap into new demand that would thoroughly utilize robot technology and other physical technology that has been nurtured in Japan. While AI in its current state of development is extremely useful in recognizing regularity in volumes of data, in order for it to be utilized in our lives in actual society, which is expected to begin in the 2020's, a new mechanism that will serve as the receptacle of AI, based on existing physical technology, will be required. However, as it stands, it is still extremely difficult for AI to carry out human movement, and greater developments in basic engineering technology to provide materials and controls for such movement will be required. To this end, it is believed that technology development in various new engineering fields centering on robots to serve as the receptacle for AI will also be required.

October 26, 2018
  • Managi, Shunsuke, The Economics of Artificial Intelligence: How Our Lives, Our Work and Our Society Will Change, Minerva Shobo Co., Ltd., 2018

November 19, 2018