Chapter 5. Conclusion
5.1 Case study summary and implications
國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
Chapter 5. Conclusion
5.1 Case study summary and implications
Economies in the Asia Pacific have great potentials for incorporating data as a new resource in their growth and development models, especially towards more
inclusiveness through digital connectivity and incorporating smaller businesses in regional and global production networks. Therefore, policies must be formulated adequately to leverage this potential and implement national strategies towards digital economy models matching up with domestic conditions. If poorly steered, the digital divide can grow and give rise to new forms of social and spatial exclusion
domestically and region-wide. Policy approaches vary. However, economies in the Asia Pacific certainly reflect a set of institutional configurations that derive from path-dependent forces and can be deemed conducive to emerging digital ecosystems. The export-led catch-up experience of Asian economies in the past has led to a sound manufacturing base of global importance, with Japan leading the regional catch-up phase and followed by the NIEs. Additionally, China rose to the colloquial status of the ‘world factory’ and has now become a global player in big tech. Most recently, labor offshoring occurred to the benefit of emerging Southeast Asian markets with rising incomes and increasing wealth. Asia is predicted to host two-thirds of the global middle-class by 2030. Growing incomes give rise to consumption aspirations of individual and personalized goods and services, and post-materialist aspirations including the “new customer experience”. Big data enables business models to monetize on these trends through exploitation and fusion of data feeding business intelligence and prediction mechanisms like AI. Secondly, the amount of data is crucial to the creation of such systems. Machine learning, in particular, as applied in AI systems, feeds on these data. Generally speaking, the more data, the better the predictions. A spatial characteristic of the Asia Pacific includes that large parts of its population live in dense metropolitan areas that provide the public and private sector with aggregate data. This is conducive to developing and implementing AI
technologies at scales, such as smart city systems or autonomous vehicles.
To effectively leverage the potential of AI technologies and generate data-derived value, a national manufacturing base remains a necessary condition to catch up or innovate. The manufacturing sector has been contributing largely to Asian economies’ GDP growth and, as opposed to other developed countries, continues to
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
do so whilst maturing and shifting towards knowledge-intensive industries. The case studies show that economies target this potential. China’s Made in China 2025 industrial upgrading strategy aims to achieve a large-scale improvement of general automation and notably its competitiveness in the production and manufacturing sector. The Three-Year Guidance for Internet Plus Artificial Intelligence Plan and backed by the Three-Year Action Plan for Promoting Development of a New Generation Artificial Intelligence Industry puts AI at the core of the upgrading process in manufacturing sectors, laying out timely guidelines and prompting the industry to propel R&D for breakthroughs in AI to create a next-generation internet infrastructure for smart factories and devices. In Japan, premier Shinzo Abe strives for economic revitalization to realize Society 5.0 that goes beyond industry 4.0.
Singapore’s manufacturing industry accounts for about 20 percent of its GDP and upgrading the semiconductor industry and high-tech computing capacity remains an important governmental target. Also, Korea holds a strong position in the global semiconductor industry with a diversified manufacturing base. Similar to China’s 2025 strategy, the I-Korea 4.0 roadmap defines industries and sectors that the government targets with their innovation growth engine policy, promoting smart factories and enhance production through AI in order to massively increase the value-added ratio of the manufacturing sector. As global production networks are reshuffled and China also starts to offshore labor, ASEAN’s emerging economies can leverage data-driven off-the-shelf AI technologies to dramatically boost their own
manufacturing sectors’ productivity and further expedite the catch-up process.
However, structural imbalances and the digital divide in ASEAN and within individual ASEAN states themselves can be tremendously large and there still is a need for enhancing digital connectivity through ICT infrastructure improvements.
What all policy initiatives pronounce is to not simply focus on developing AI
software but to particularly complement it with the hardware, such as South Korea’s smart semiconductor ambitions, or generally achieve higher automation levels
through AI and IIoT system applications in the manufacturing process. Prioritizing to upgrade the domestic manufacturing sector epitomizes the ambitions to maintain the
“factory Asia” model in tandem with shifting toward “value-added factory Asia” as proposed by (Kam, 2017). The policy objectives of AI implementation in
manufacturing combines a holistic and inclusive approach of boosting productivity, incorporating small and medium enterprises in the production network through ICT
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
enabled low transaction costs particularly in the ASEAN region, expanding and reshuffling production networks for greater consumer welfare, and nurturing nascent digital ecosystems through value-adding digitized products.
While this upgrading process lifts the economy up the value chain, disrupting technologies like AI that automate core businesses are associated with changing labor environments and job displacement. To ensure an inclusive development trajectory throughout the transition to the digital economy, arising labor issues have been addressed, respectively. Chinese policymakers realized that innovation in AI does not come from within its SOEs and delegates digital governance in R&D to private companies while exercising its influence over the direction of research through incentives for picked national heroes in an attempt to create more competitive digital ecosystems for high-quality jobs, thus, binding financial support to the overarching objective of becoming an AI superpower. Korea and Japan similarly aim to strengthen fundamental AI research through cooperation between universities and the private sector in order to nurture their domestic talent pool. Both countries lack in innovative startup ecosystems around AI, partly due to yet again path-dependent forces that have formed institutional complementarities such as conglomerates steering innovation and remuneration systems based on seniority rather performance and creativity. Singapore already possesses a vivid startup ecosystem due to investment-friendly environments and a strong knowledge-based economy with tech-savvy and highly educated human resources. A fifth of full-time jobs in services such as banking and insurance could be displaced, more than anywhere in ASEAN. Thus, the AI Singapore initiative is a top political priority trying to enhance its digital ecosystem with programs such as AI for Everyone to provide up to 100,000 Singaporeans with knowledge on how to
incorporate AI in their businesses, or the 100Experiments program connecting research and businesses to jointly develop and commercialize AI solutions.
Generally, all the policies under scrutiny take an application-oriented and commercial approach to AI, however, this does not exclude the social implementation of AI technologies. The policies in the case study emphasize and promote the added value for society derived from the technology’s marketability. In Korea and Japan, deregulation has been implemented geographically in strategic special zones to try out new business models and generate necessary data for further research,
commercialization, and nationwide usage upon success. Both countries stress AI-solutions as a remedy to social issues, namely their fast aging societies. Japan’s
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
Society 5.0 concept is considered as the next step in human evolution and job displacement through AI is not perceived as a threat against the backdrop that automation and smart devices can help solve issues around Japan’s aging society.
China faces similar demographic issues. Therefore, all countries identified strategic application fields for social AI and set out incentives to spark consumer demand in these sectors, including public services, healthcare, and transportation. This entails emerging patterns of digital platformization, coined as platform capitalism, that takes over the states’ role as infrastructure provider for the digital utility. Governments reinforce this trend through policies to facilitate the fusion of public and private sector data as they perceive it as a driver of innovation. Korea’s AI Hub program provides access to ready-to-use high-quality datasets of CCTV footage and medical images for academic or business research and AI training; Japan’s Basic Act for the Advancement of Public and Private Sector Data Utilization involves the entire administration to be digitized and public and private data will be made available to the private sector;
Singapore’s Smart Nation Initiative merges data from the public and private sector to match people’s profile with jobs and career recommendations to nurture the AI talent pool; for the rest of ASEAN, no such mechanism was identified within the e-ASEAN program, but may exist in individual member states. The Chinese government
partnered up with partly state-owned voice recognition developer iFlyTek providing it with biometric data and also supports the AI platform City Brain by Alibaba Cloud to realize truly smart cities, with complementing data sourced from Apollo by Baidu for autonomous driving, or Tencent’s AI platform for a smart public health system.
The fusion of public and private data raises questions of data ownership, which has been addressed by governments, respectively. Singapore and Japan
amended their copyright regulations to allow duplication of copyrighted data for third party information analysis such as training AI algorithms for commercial applications.
The distinction in the Korean and Chinese copyright is not yet clear and is set to be amended in 2020. The ASEAN Digital Integration Framework Action Plan 2019-2025 aims to propel policy streamlining and transparency of domestic laws with particular regard to intellectual property rights mainly regarding e-commerce platforms.
Concerning personal information and data protection, the European Union’s GDPR was used as a benchmark for data privacy protection in this study. Singapore’s
Personal Data Protection Act has no definition of ‘sensitive personal data’ and is less strict than the GDPR on matters of consent provision for data collection and handling
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
for it was enacted for purely economic purposes as opposed to the GDPR's notion of privacy as a human right. Japanese and Korean regulations are on par with GDPR provision to a considerable extent, mostly addressing the anonymization of personal data as well as data subjects’ rights requiring consent to collect, process, retain information for corporate utilization. Japan amended its Personal Information Protection Act in 2017 not least due to the 2019 EU-Japan Economic Partnership agreement to facilitate cross-border data flows. There are no unifying regulations regarding data protection across ASEAN countries and provisions differ territorially.
The APEC Cross-Border Rules System, however, requires business activities to comply with the APEC Privacy Framework (2017) setting forth basic guiding
principles of privacy protection to ensure a certain extent of synchronized information laws. These principles have been incorporated in most APEC members' jurisdictions, namely Japan, South Korea, Singapore, and China, and in ASEAN states, such as Indonesia, the Philippines, Vietnam, and Malaysia.
Data protection has gained distinguished relevance, not least because of recent incorporation in free trade agreements, namely with regard to e-commerce, digital products and cross-border flows of personal data and information. However, unconcise definitions for digital products and services hamper enhanced policy streamlining in the Asia-Pacific. The WTO definition of digital products and
electronic transmissions is ambiguous and intersecting. Thus, several countries have resorted to bilateral preferential trade agreements to clarify the context, leading to a growing heterogeneity in definitions. This is problematic in the context of economies entering into free trade agreements such as CPTPP and RCEP, which are seen as important drivers of market integration in the Asia Pacific. If definitional
heterogeneity remains, these inconsistencies may be reflected by individual countries’
approach to digital policymaking even on a domestic level and, internationally, result in costly dispute settlement procedures. RCEP falls short of significance in digital trade as it does not promote cross-border data flow for businesses nor prevent customs duties on digital products. CPTPP does give clear directions. Through RCEP, China could fill this gap and set regulatory standards if definitional heterogeneity sustains.
E-commerce is thriving and China is committed to enhancing cross-border e-commerce cooperation as the ASEAN–China Free Trade Area (ACFTA) was supplemented by respective provisions. In this case, personal data security could not be fully ensured due to existing legislation that the Chinese government could force
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
companies to hand over data to comply with specific regulations. However, digital policymaking is market-oriented and adapts to industrial needs. Inter-governmental cooperation beyond the scope of deal-making merely exists. ASEAN shows the most efforts in policy streamlining, most notably due to Singapore’s special role and aspirations for an integration of the CMLV. However, AI-related content was not found in any of the guiding principles among ASEAN frameworks and is yet to be integrated.