Chapter 3. Country case studies
3.4 South Korea
國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
freedom, autonomy, and rights” (Knight, 2019). This is in line with the assumption that China is interested in international commercialization facilitated by complying with the global common sense based ethical dimension of data-derived applications.
3.4 South Korea
3.4.1 Conditions for data-derived value
Korea holds a strong position in the global semiconductor industry with a diversified manufacturing base. In 2018, Samsung overtook U.S. chipmaker Intel in total global revenue and henceforth led global sales in semiconductors with US$52 billion revenue in 2019 and, on the purchasing side, a more than two-fold increase, all of which depicts the dynamic development of the domestic tech sector (Holst, 2020).
Korea was also identified by the Center for a New American Security as a prime strategic partner for mutually beneficial cooperation regarding semiconductor manufacturing, AI hardware, and research collaboration. As of 2019, Korea hosts 8 out of the 500 most powerful supercomputers in the world, with supercomputer Nurion ranking 14th (Strohmaier et al., 2019).
The general characteristics of South Korea’s industrial structure can be largely traced back to Korean business conglomerates called chaebols which have been exercising a corporate influence on governmental policymaking since becoming the dominant force in South Korea’s economy in the 1960s and 1970s (Kalinowski, 2009). Their traditionally good relationships with the government and banking sector as well as the high functionality of their corporate structures provided the chaebols competitive advantages on the way to the center of the world economy, a way that was marked by the small capacity of their internal market and predestined for export-led growth strategies (Pohlmann, 2005). Although the South Korean government has been trying to reduce the influence of the chaebols since the 1997 Asian crisis, they remain a determining factor in the South Korean political economy that can be
considered a function of path-dependent forces. Their intertwining with policymaking may be conducive to establishing an innovation-driven digital economy as knowledge and know-how of experts from the private sector provide valuable input on how to best leverage data for corporate purposes. However, the chaebol system is prone to corruption and collusion that contrarily has the potential to impede innovation by eliminating true competition, as well as hampering the social agenda of South Korea
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
if vested corporate interest is prioritized over social issues. As a potential
countermeasure, the Presidential Committee on the Fourth Industrial Revolution (PCFIR), as discussed below, is a heterogeneous committee of public and private stakeholders to deliberate on 4.0-related issued from multiple perspectives (PCFIR, 2020). As for data-related technologies, the chaebol structure may have a significant advantage, nonetheless. Since chaebol operations are characterized by fast decision-making and the ability to change and adapt faster than other bureaucratic large corporations (Pohlmann, 2005), it is not only easier for conglomerates to gather and merge data from a broad spectrum of their respective portfolios, they may also be efficient in steering the formation of technocratic ecosystems according to identified application fields and further develop those into a comparative advantage among global digital production networks. Considering the socio-economic transition to and uncertainty of the digital economy, path-dependence comes into play in that social insecurity through “volatile destructive neoliberalism” impacting the country’s
“developmental liberal order” (Chang, 2019) and susceptibility to political promises of economic miracles have led the South Korean people and economic stakeholders to rely on a developmental state perspective and growth-oriented solutions through conglomerates that historically have showcased their capability to do so, thus, fostered their position in the public psyche to generate wealth (Kalinowski, 2009;
Kang, 2003). Kang (2003) denotes the relationship between the private chaebols and the government’s public stakeholder role as a reciprocal mechanism that has
prevented to give excessive power to either the public or private sector and, thus, kept corruption from growing beyond control, calling this a ‘mutual hostage situation’
giving rise to peculiar South Korean institutional complementarities comparable to a system of checks and balances.11
As for big data technologies, the private sector plays a major role in
investments, with about three quarters coming from large enterprises and the chaebols like Samsung and LG (Chang, 2019). Therefore, market-oriented R&D and
11 Kang (2003) puts forth the idea that if bribes were to be equated with a transfer of wealth from corporate chaebols to the political realm resulting in productive investments, this could actually lead to overall gains for the economy and society in overcoming pressing societal issues. Throughout the democratic transition, this “mutual hostage situation” had nurtured favorable conditions for a sociopolitical fabric with, for example, a social security network and labor unions, now at stake through destructive neoliberal forces since the Asian financial crisis (Chang, 2019).
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
oriented application feasibility are at the core of technology investments. State proximity comes in handy because favorable conditions are created for these large conglomerates to pursue and integrate basic research on AI as discussed in the next section on institutional frameworks. Support comes in form of public investments to strengthen basic research, education, and open-access databases for public use and AI development, all of which aims to nurture the talent pool that thereupon, however, is most likely to be absorbed by the conglomerates with the financial capacity to buy experts for their own R&D purposes in corporate research facilities such as the
Samsung AI Center, LG AI Lab, or Hyundai Motors AI established between 2016 and 2017 (Kim, 2019).
3.4.2 Institutional framework
In the wake of the 2017 impeachment of former president Park Geun-hye, current president Moon Jae-in pledges a more liberal course and declared Korea’s “potential to emerge as an AI powerhouse … [as] the government’s duty to turn that potential into reality” (Moon, 2020). He set up the Presidential Committee on the Fourth Industrial Revolution (PCFIR), with the chair held by former startup entrepreneur and investor Chang Byung-gyu, and a committee composed of 24 representatives from various ministries12, academics13, and private sector practitioners14. The multi-layered composition of the committee can be considered a first guarantor for holistic
perspective-taking as opposed to policy approaches singularly serving corporate chaebol interest in the face of the socio-economic transition toward the digital economy. With input from the private and public sector, the Ministry of Science and ICT (MSIT) finds key tasks regarding 4.0 areas to be analyzed thoroughly by the ministries and two sub-committees, respectively: the special sub-committee for issues regarding smart city infrastructure or healthcare, and the innovation sub-committee with the three fields of expertise in science and technology, industry and economy, and the social system; the sub-committees report back to the ministry representatives,
12 E.g. Ministry of Science and ICT, Ministry of Trade, Industry and Energy, Ministry of Employment and Labor, Ministry of SMEs and Startups, Ministry of Land, Infrastructure and Transport (PCFIR, 2020)
13 E.g. Department of Biomedical Systems Informatics at Yonsei University, CK SMART Aged friendly Service at Hallym University (PCFIR, 2020)
14 E.g. Korea Mobile Internet, Hyundai Motor Company, Korea Startup Forum (PCFIR, 2020)
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
share policy directions and modify innovation policies to be carried out by the ministries (PCFIR, 2020). Key fields proposed to the innovation sub-committee include not only further developing ICT networks for data flows but also the
“strengthening of data production and utilization technologies” to provide for
“ecosystems for innovative tech startups … and developing new services” (PCFIR, 2020). These are identified as a key contribution to The People-Centered Response Plan released by the PCFIR that can be integrated within the smart city and healthcare vision to alleviate chronic social problems by putting forth intelligent technologies to prepare the workforce for change and create high-quality jobs based around emerging industries that “secure data and networks accessible to all” (PCFIR, 2020).
The MSIT set out the I-Korea 4.0 roadmap to define industries and sectors that the government targets with their innovation growth engine policy (MSIT, 2018) including measures promoting smart factories at the manufacturing base, digital growth incubators and emerging fields for economic diversity, and enhanced support strategies for a knowledge-based digital economy to transform Korea incrementally with targets set for 2023 and 2030. The blueprint lays out a holistic approach to transforming, merging, and connecting various socio-economic domains of the digital economy. As for artificial intelligence, the MSIT (2018, pp. 3-4) classified these domains and highlighted artificial intelligence as a core technology to be developed in each single one: a) the industrial base, for example, acquiring core technologies for AI semiconductor devices (intelligent semiconductors) with a global market share of 7 percent by 2022; b) the intelligent infrastructure with a large high-quality data pool for open data utilization, analysis, and prediction, as well as kick-starting companies specialized in AI and boost their number from 34 as of 2017 to 100 by 2022 in order to disseminate AI and secure proprietary algorithm-based technologies of the future;
c) smart moving objects relying on the industrial base for production and the
intelligent infrastructure for application, including drones and autonomous vehicles;
and d) convergence services relying on big data such as smart city solutions and medical healthcare safety-robots, and virtual / augmented reality platforms and devices to induce industry fusion, namely education and manufacturing. The innovation growth engine policy (MSIT, 2018) shows governmental efforts in leveraging Korea’s comparative advantages such as innovating on the strong
manufacturing base by transforming it into the smart supplier of hardware needed for effectively applying data-based core technologies in tandem with creating competitive
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
ecosystems and jobs for innovative thinking and entrepreneurial spirit outside the chaebol network. For instance, one of the objectives in the domain of convergence services is to foster and establish more than 100 enterprises with $US10 million sales revenue in the fusion field of virtual and augmented reality by 2022 in order to diversify the development potential outside of the leading companies Samsung and LG Electronics (MSIT, 2018, p. 4). For AI development specifically and its
commercialization, a public-private joint project, the AI Hub program as elaborated in the following section, was established in 2019.
3.4.3 Regulation and commercialization
The holistic development approach, similar to Japan, not only considers the revitalization of Korea’s strength in manufacturing but first and foremost its
reconfiguration through industrial AI utilization for adding value in manufacturing by exploiting advanced ICT. With the government strengthening and steering through industrial incentives towards convergence of AI and algorithm-based hardware and software, South Korean corporations are holding strong positions in global value chains and have great potential to monetize upon digital technologies. Despite the structural dominance of conglomerates, competition exists among the big players as stipulated by the law, incorporating suppliers along the value chain, and driving product innovation coupled with IoT applications: smart speakers were first launched by SK Telecom in 2016, followed by devices by South Korea's largest online
platforms Naver and market-listed Kakao in 2017 (Won & Hye, 2019). As internet-based platforms connecting a variety of services such as search engines, streaming and media outlets, and e-commerce portals, their access to data on the Korean market enables them to develop IoT applications of such kind and innovate upon demand. In 2020, Samsung Electronics will also introduce an advanced IoT smart speaker. In fact, the Samsung conglomerate plans on pushing the boundaries towards state-of-the-art consumer electronics with pstate-of-the-articular regard to smstate-of-the-art home devices and
connectivity. A visit to Samsung D’light, the company’s exhibition center in Seoul, in 2019, provided me with an immersive experience of Samsung Electronics’ ambitions to integrate IoT as the new digital utility in people’s lives with AI at its core. Besides an exhibition on the new generation of semiconductors which the IoT is based on in its physical dimension, a showroom exhibiting existing and currently being planned
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
and designed gadgets by the company for the smart home of the future: from smart beds that regulate room temperatures according to our body and behavioral data collected during our sleep, then transmitted to intelligent mirrors in the bathroom that scan our skin condition and process all the gathered data to give suggestions on healthy habits and foods, with the latter being stored in Samsung’s smart fridge with AI-powered food recognition, temperature control, supply notifications, and
integrated purchase suggestions. The list is long but interestingly epitomizes the overall convergence of Korea’s strength in hardware production and 4.0 ambitions in software development.
As for data-driven growth potential, President Moon sees creativity and the media industry as driving forces behind prospective growth and is eager for Korea “to stand tall as a media powerhouse in the global media market that is seeing seismic changes” (Moon, 2020). For example, the government invested in and nurtured the creative industry over years and promoted the emergence of an entirely new genre of Korean pop-music to an extent that gave rise to a new sector dedicated to digital content distribution that gave rise to new marketing strategies based on entertainment:
In 2018, South Korea had the fifth-largest sales of digital media in the world and the sector continues to grow and gain popularity abroad (Holroyd, 2019, p. 13). For instance, the Korean boyband BTS was the first Korean music group to enter the U.S.
album charts at number one with a foreign-language music album on the English-dominated U.S. market, gaining acknowledgment by President Moon (Chiu, 2018).
Domestically, the Korean entertainment industry blends in with formerly separated sectors such as the financial sector and finds its place in the 2st-century fintech sector (financial technology): through data-enabled targeted marketing, domestic online platforms such as aforementioned Naver and Kakao are better able to track users’
behavior and aspirations through their extended network and access to user data.
Kakao, for example, combines services such as ride-hailing and shopping outlets and has launched the first Korean digital-only bank Kakao Bank (Wilson, 2018), with a digital payment wallet system for convenience that allows the provider to place targeted loan and credit options with the user’s favorite Korean pop-music group or entertainer, making the product more interesting through customized advertising. To develop new business models and forge innovative digital ecosystems around media, Seoul’s newly erected Digital Media City, Korea's first creative cluster, houses broadcasting channels and was set up to connect small businesses with big players
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
through subsidizes office rooms, tax breaks, etc. (Cohen, 2013). The Digital Media City is an entirely new city district with attractive housing and leisure options
strategically located between the city center and the industrial cluster around Incheon airport. Such creative hubs have become a target point in creative policy formulation with regard to open up new channels of enhancement for the digital economy and explore new areas for growth with particular regard to AI-based applications as outlined in the innovation growth engine policy (MSIT, 2018) under I-Korea 4.0.
Within the framework of I-Korea 4.0 and emphasis on enhancing the “DNA industry” (data, networks and AI) as put by president Moon (2020), the AI Hub program was jointly launched by the MSIT, the National Information Society Agency, the National IT Industry Promotion Agency, and the Electronics and Telecommunications Research Institute in order to advance the nation’s
competitiveness in AI and other algorithm-based technologies. AI Hub is an online platform providing AI infrastructures15 for businesses, students, researchers, and developers that integrates public and private data and open-sources it for everyone to use in an attempt to spark innovative and smart service models for successful
commercialization (AI Hub, 2019). Data as an open resource provides high quality AI learning datasets that would otherwise be difficult for small and medium companies, research institutes, and individuals to obtain by themselves due to high cost and input time. Overall, this approach strikes a balance between providing fundamental
researchers with the resource needed at low cost and the market-oriented R&D ambitions of corporate ventures in that the program allows everyone to use it for research or monetizing on public-private data, respectively. However, it pertains to core objectives in two domains of the I-Korea 4.0 roadmap regarding leveraging AI with particular regard to its commercialization potential (MSIT, 2018): establishing the intelligent infrastructure with high-quality databases for open data utilization, analysis, and prediction, as well as kick-starting companies specialized in AI and boost their number from 34 as of 2017 to 100 by 2022 in order to disseminate AI and secure proprietary algorithm-based technologies of the future; and establishing convergence services relying on big data such as smart city solutions and medical healthcare. For instance, to work towards the (MSIT, 2018) objective of developing thirty medical devices with export potential and twelve with export sales exceeding
15 infrastructure services include AI data, AI software, AI computing, AI Easy Builder (AI Hub, 2019)
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
US$100 million, AI Hub (2019) was provided with medical images for further dissemination, including disease diagnosis image AI data such as 30,000
mammography images and 4,500 fundus images for eye disease detection (AI Hub, 2019). There is a need to build an efficient and systematic medical knowledge base not only to activate AI startups for the medical industry but also to advance these technologies for the sake of finding solutions to pressing issues, such as an aging society. Other data include published CCTV data showing anomalies in people’s behavior, such as assaults, fighting, theft, vandalism, fainting, or intoxication. The AI Hub program is relatively new, at the time of writing this thesis, was announced to be complemented with 60 million new data items in 2020 (Yonhap, 2019). However, these data are sensitive in that medical conditions and individuals’ behavior are captured and exposed on the internet. This impediment was addressed by the ministry, saying that in Korea, it is still difficult to open and use high-quality data due to
various issues such as personal information infringement, copyright, and portrait rights. Thus, there is a need for strategies to ensure corporate utilization of de-identified information while safeguarding privacy.
Data privacy has in fact been addressed through the Personal Information Protection Act (PIPA) by the Ministry of Interior and Safety (MOIS). In fact, Korean regulations are very much on par with GDPR provision, mostly addressing the anonymization of personal data as well as data subjects’ rights requiring consent to collect, process, retain information for corporate utilization (Kang & Lee, 2019;
Simmons, 2019). The equalization of personal data protection laws with EU
regulations bears many advantages for future trade and corporate investments between digital economies pledging data protection. Through the anonymization of personal data, that is data that cannot be used to identify an entity, their legitimate distribution enables businesses to leverage these with regard to developing competitive big data applications in a multinational context. For instance, the PIPA states various rights of persons, called data subjects, such as:
“the right to be informed of the processing of such personal information … the right to confirm the processing of such personal information, and to request access (including the provision of copies …) to such personal information … the right to suspend the processing of … and destruction of such personal
“the right to be informed of the processing of such personal information … the right to confirm the processing of such personal information, and to request access (including the provision of copies …) to such personal information … the right to suspend the processing of … and destruction of such personal