Japan's culture in technology and innovation appears to be that of a follower. The country's patience in following over decades has been the key to grasping and mastering targeted areas of technology and innovation. The strong and meticulous fascination with material-centric perfection has been the key to Japanese success in the industrial economy over the last 100 years. For this reason, often carpenters and maintenance specialists as opposed to scientists and engineers were behind the success of Japan's industrial economy. For example, the carpentry skill of Sakichi Toyoda led to the success of Toyota. Similarly, a camera repairman named Goro Yoshida started the journey of Canon.
As opposed to conceiving an alternative to existing cameras in 1936, his interest was primarily to figure out how to shape metal and rubber to replicate an existing German camera. But in the age of AI (artificial intelligence), the main competence is in the software part. It is about theorisation of reality in the abstract world, as opposed to acquiring skill through apprenticeship in perfectly shaping physical objects. Abstraction and zero cost in replication of software-centric AI innovation appear to be a misfit for Japanese culture of innovation. Such observation raises a vital question: Will Japanese industry keep succeeding by being the follower in the era of artificially intelligent machines?
AI, of late, has become a buzzword -- often filled with hypes. The increasing processing power of microchips and the growing capability of sensors are opening up opportunities for creations in the abstract domain of software to empower machines to perform very complex tasks with fewer errors. Upon succeeding with Google's AlphaGo capability of computing and deep learning algorithms to develop its strategy for defeating the best human players in the board game 'Go', the possibility of adding such ability to physical machines has become a real possibility. For example, driving a car is a complex task. Despite the growth of different aspects of automobiles, the driving is still left to the human. But human drivers have many limitations, starting from cognitive delay and fatigue to even intoxication.
Although cars are more comfortable than before, the rate of road accidents has not been declining. More than 1.0 million people are being killed every year due to road accidents, and such accidents cost as high as 3.0 per cent of gross domestic product (GDP) in many countries. The progression of human capability in building machine intelligence to delegate this driving role to automobiles is opening up a new dimension of value creation in automobile service consumption. Japan's success in the automobile industry appears to be blended within the nation's culture of perfection in making mechanical parts. It has been found that culturally Japanese are very much patient in understanding mechanical objects and in mastering the skill to keep upgrading them. But will such skill be sufficient to create the AI ability in automobiles? It appears that there is a discontinuity in the skill-set for enabling Japanese industry to upgrade their products by adding the AI layer.
The success of AI innovation in improving existing products largely depends on software and network externality. It has been found that innovation and entrepreneurship culture in the domain of software and network externality is entirely different from Japan's core competence in the material-centric world. As opposed to incremental and efficiency innovations, the success in the AI domain is primarily dominated by disruptive thinking. Instead of steady progress in shaping the material in the conventional world of innovation, often successes in the domain of software and network end up in creating discontinuity. Often such discontinuity creates a decision-making dilemma for the managers of highly successful prevalent products.
Prof. Clayton Christensen has defined it as the innovators' dilemma faced by existing managers. Strong loyalty and hierarchically long decision-making loops are often detrimental to leverage such innovation opportunities. As a result, the same culture that made Japan's corporations successful during most of the 20th century, could be a barrier to upgrading their products in the AI age. In the software and network externality-intensive intelligent machine age, the nucleus of success has moved from making individual components to adding a new layer of value on top of them.
For example, Apple has emerged as the most valuable company from mere ashes. Apple's successes primarily hinge on the iPhone. But, Apple does not produce a single physical component of iPhone. By capitalising on the strengths of software, artificially intelligent multi-touch interface and network externality, Apple has succeeded in making the iPhone by sourcing components from over 200 suppliers. Against Apple's such capability, Japan's Sony or Canon lost many of their iconic innovations, such as portable video players and compact digital cameras. Although Sony makes just a couple of dollars from each image sensor chips supplied to Apple, Apple pays less than 50 per cent of the retail price to component makers and assemblers while recording a staggering profit.
A similar possibility has arisen in other areas of Japan's significant core industrial capabilities. For example, one of the notable core capabilities of Japan's industrial economy is automobile. It appears that within the couple of decades, self-driving or autonomous cars will be a reality. The autonomous ability will be the key success criteria. As most of the automobile parts are no longer made by known brands like Toyota or Honda, a start-up having robust AI module with autonomous driving capability may emerge for assembling components sourced from numerous suppliers to offer autonomous cars in the market. Like the way Apple's iPhone caused disruption to Nokia within a span of just two years, this autonomous car poses the risk of causing disruption to the major Japanese automobile makers-changing the game of its industrial economy. Similar threats are emerging in many other industrial products, starting from medical imaging to manufacturing machineries. It is to be noted that due to this inherent weakness in Japan's technology and innovation capability, despite showing remarkable mechanical aptitude in developing humanoid robot ASIMO, Honda could not succeed in making it a useful smart machine.
It is understood that Japan has been working to address this core limitation of innovation in the age of artificially intelligent machines. Not surprisingly, Japan's industry is looking overseas to fill this gap. For example, Toyota started its own billion dollar AI Research Lab in Silicon Valley in 2015, and other companies are following suit with investments of hundreds of millions of dollars each year, even by establishing their own venture capital arm. Such investments in AI start-ups may add some strength, but such an alternative does not replicate Japanese corporations' core capability of in-house or on-shore innovation.
It is also understood that there has been significant focus on university-based research, industrial innovation, and start-up ecosystem formation. But, Japan may have lost decades before focusing on such vital capability of university-based start-up ecosystems to complement its industrial conglomerates' efforts to upgrade their products with AI capabilities. Although research is still going on to unearth the fact behind Japan's "lost decade," but we should not be surprised if it is found that Japan's weakness in innovation in the software-intensive AI machine capability has been creating a barrier to the revival of economic power Japan enjoyed during the 1970s and 1980s.
M. Rokonuzzaman, PhD, is Academic, Researcher and Activist on Technology, Innovation and Policy.
