Priorities for the Japanese Economy in 2018 (January 2018)

What Differentiates the Companies that are Quick to Incorporate AI?

Consulting Fellow, RIETI

First-mover advantage and the business performance of companies

There are many companies that intend to incorporate artificial intelligence (AI) into their business. The relationship between the incorporation of innovations and the business performance of companies—namely what types of companies have been able to capture new markets with the incorporation of innovations and what types of companies have not been able to do so—has garnered academic interest again and again since the latter half of the 20th century. Steven Klepper, who left behind the significant achievement of developing economic theories regarding the evolution of industries and behavior of venture companies, proved that companies that quickly incorporate innovations have a higher probability of surviving than those that are slow to do so (Klepper & Simons, 2005). This is because the first-mover advantage obtained through the early incorporation of innovations can be used for the incorporation of subsequent innovations as well.

Empirical studies have shown that companies that have incorporated big data analysis into their value chain have a 5% to 6% higher productivity (McAfee, Brynjolfsson, Davenport, Patil, & Barton, 2012) and a 3% higher productivity growth (Tambe, 2014) than companies that have not.

Many Japanese companies have become aware of the value of such an innovation, AI, and are interested in incorporating this into their business ahead of others, but have not yet realized this. According to a survey by Morikawa (2016) on approximately 3,000 Japanese companies, 28% of companies believe that the utilization of AI has a positive impact on business, but in spite of this, only 3% of all respondent companies are "already using big data for business," and 40% responded that "we have no idea" how to utilize big data (Note 1).

As such, suppose a large number of companies want to incorporate AI in similar manner, what are the differences between the companies that will realize this earlier and those that will not? This article (Note 2) will look at organizational heterogeneity as to what kinds of organizations are the ones that incorporate AI before others.

Incorporation of AI and corporate organization

As AI's problem structures are highly complex and uncertain, it is inevitable to introduce heuristic inferential logic to solve such problems (Yashiro, 2016). For example, in the case of applying an inferential model using big data and explaining a target phenomenon, engineering intuition that enables the supposition of good initial values is required, and thus the organization needs to have the flexibility and adaptability to carry out this inference process.

In a study looking at the case of LinkedIn, one of the largest social networking services specializing in business, it is pointed out that an environment in which experiments with data can be carried out freely must be provided for AI scientists (Davenport & Patil, 2012). In experimental environments, it is important to repeatedly carry out inductive rather than deductive hypothesis verification using data (Constantiou & Kallinikos, 2015). This is because AI is created through processes with strong research characteristics in which experimentation and verification are repeated, rather than conventional IT product production processes in which the product is completed by taking the time to write the necessary code (Provost & Fawcett, 2013). If an organization allows this type of experimentation, this means that it will attract outstanding scientists who are brimming with curiosity.

Furthermore, for companies, incorporating AI may involve reforming decision-making methods (Bresnahan, Brynjolfsson, & Hitt, 2002) (Somers & Nelson, 2001). "Correct conclusions" that are derived from AI sometimes differ from the respected "intuitive conclusions" of executives at many companies thus far. In such cases, in principle, the former should be given precedence, and if data is used to support "intuitive conclusions," the incorporation of AI will fail (McAfee et al., 2012).

In addition to the above mentioned environment in which scientists who use AI can work easily, whether or not other personnel embrace AI in their work is an important factor for a company. One study (Lee, Kusbit, Metsky, & Dabbish, 2015) has analyzed the responses of drivers regarding AI analysis-based prices and vehicle dispatch proposals through conducting interviews with employees of Uber (Note 3) and Lyft (Note 4). The results showed that proposals that have clearly explainable details and match the feelings of the drivers are desirable. It seems difficult for drivers with many years of experience who rely on intuition to accept AI-based proposals. If this problem is left as it is, the incorporation of AI by companies will be delayed.

"Those who can imagine anything, can create the impossible"

Alan Turing, who developed the Turing machine, which was the starting point for today's computer science, stated, "Those who can imagine anything, can create the impossible."

As summarized in this article, organizations that possess flexibility that allows for the repetition of experiments ahead of other companies and can skillfully digest information obtained from AI will be able to incorporate AI more quickly than other companies. "Imagining anything" as mentioned by Alan Turing will likely serve as a first step in the development of such organizations.

December 27, 2017
  1. ^ These are among four possible responses to the question, "How does your firm think about big data?": "already using for business," "intend to use in future," "not related to our business," and "don't have any idea."
  2. ^ This article consists of the research results of a study group on the "effect that artificial intelligence has on economics" led by Kyushu University Professor Shunsuke Managi who is a faculty fellow at RIETI, and the research results of this study group are forthcoming under the title Jinko Chino no Keizai-gaku (The Economics of Artificial Intelligence).
  3. ^ This is a company offering vehicle dispatch services that was established in San Francisco in 2009. It currently offers a vehicle dispatch website and vehicle dispatch app in more than 450 cities in 70 countries and regions worldwide. Through Uber, the passenger makes a request to drivers for vehicle dispatch using current location information, a list of drivers that are available to pick up the passenger is sent to the passenger via Uber, the passenger selects a driver from the list, and that driver picks up the passenger. The drivers are sometimes taxi drivers and are sometimes persons driving private vehicles. There is a mechanism in which the fee is set and the passenger's credit card is charged in advance, and the rating of the driver by passengers is publicized, so there are advantages for passengers such as being able to avoid problems associated with unlicensed taxis and being able to find an alternate means of transportation if a taxi is not available. For drivers, there are advantages such as being able to earn money with their private vehicle whenever they have free time, and not having to give a margin to taxi companies.
  4. ^ This is a company offering vehicle dispatch services that was established in San Francisco in 2012. As with Uber, ordinary people take other people to their destination in private vehicles, rather than drivers hired by the company.
  • Bresnahan, T. F., Brynjolfsson, E., & Hitt, L. M. (2002). "Information technology, workplace organization, and the demand for skilled labor: Firm-level evidence." Quarterly Journal of Economics, 117(1), 339–376.
  • Constantiou, I. D., & Kallinikos, J. (2015). "New Games, New Rules: Big Data and The Changing Context of Strategy." Journal of Information Technology, 30(1), 44–57.
  • Davenport, T. H., & Patil, D. J. (2012). "Data scientist: the sexiest job of the 21st century: meet the people who can coax treasure out of messy, unstructured data." Harvard Business Review, 90 (October), 70–77.
  • Klepper, S. & Simons, K. L. (2005). "Industry Shakeouts and Technological Change." International Journal of Industrial Organization, 23(1), 23-43.
  • Lee, M. K., Kusbit, D., Metsky, E., & Dabbish, L. (2015). "Working with Machines : The Impact of Algorithmic and Data-Driven Management on Human Workers." In Proceedings of the 33rd Annual ACM Conference on Human Factors in Computing Systems (pp. 1603–1625).
  • McAfee, A., Brynjolfsson, E., Davenport, T. H., Patil, D., & Barton, D. (2012). "Big data: the management revolution." Harvard Business Review, 90 (10), 61–67.
  • Morikawa, M. (2016). "The Effects of Artificial Intelligence and Robotics on Business and Employment: Evidence from a survey on Japanese firms," RIETI Discussion Paper 16-E-066.
  • Provost, F., & Fawcett, T. (2013). Data Science for Business. O'Reilly Media Inc.
  • Somers, T. M., & Nelson, K. (2001). "The Impact of Critical Success Factors across the Stages of Enterprise Resource Planning Implementations." In Proceedings of the 34th Hawaii International Conference on System Sciences.
  • Tambe, P. (2014). "Big Data Investment, Skills, and Firm Value." Management Science, 60(6), 1452–1469.
  • Yashiro, Tomonari (2016), Innovation Management: Strategic Thinking for Process and Team Orchestration, University of Tokyo Press.

January 12, 2018

Article(s) by this author