AI and future of S. Korea

The release of ChatGPT last November kicked off a race for artificial intelligence (AI) supremacy among global tech giants. The importance of this race is not lost on the world’s great powers, which increasingly put tech supremacy at the center of national industrial policies. The question on the minds of many is which state or company may reign supreme in the end.

Currently, South Korea seems poorly positioned to take advantage of AI technologies. Firstly, research and development (R&D) expenditures in the country are too concentrated to make an early impact in the field of AI. As such, the country will most likely play catch-up for years to come. Secondly, the positive regulatory regime in the country will stifle the creation of new sectors and enterprises that can absorb workers when they are replaced by computerization.

Currently, the South Korean government does a commendable job of contributing a share of GDP to R&D that is on par with its competitors. According to 2019 data from the U.S. National Science Foundation, the government’s share of total R&D funding in South Korea was 20.7 percent. This compared to 21 percent, 20.5 percent and 14.7 percent, in the United States, China and Japan respectively. Of course, due to the size of South Korea’s economy, total funds pale in comparison. The South Korean government spends almost $70 billion less than Japan, even though the Japanese government spends considerably less in percentage terms.

Adding to this lack of aggregate government funding is the concentration of R&D expenditure by sector and enterprise. According to the Analytical Business Enterprise R&D database of the OECD, nearly 86 percent of R&D in South Korea was spent in the manufacturing sector in 2021. This amounts to 57.1 trillion won ($4.32 billion) of 80.8 trillion won in business enterprise R&D spending. A full 49.3 percent of total enterprise R&D spending, about 39.9 trillion won, is spent on the hardware category of computer, electronic and optical products, while only 2.87 trillion won or 3.5 percent of spending is allocated to software publishing. According to Stanford University’s AI Index, the top fields for private investment in AI in 2022 were medical and healthcare, management, processing, cloud and fintech. South Korea’s economic staple of semiconductors was far behind in 13th place.

In country comparison terms, spending by business enterprises on computer, electronic and optical products is about 12 percent in Germany, 17 percent in Japan, 18 percent in the United States and 16 percent in China. The only close competitor more concentrated than South Korea in this sector of the economy is Taiwan at 78 percent. This means most of South Korea’s competitors have distributed their R&D spending allowing more economic sectors to potentially explore and implement the adaptation and application of AI technologies at a faster and more cost-effective pace.

R&D concentration aside, the addition of AI technologies will be disruptive to the labor force and will increasingly replace workers in a number of occupations. This is not debatable. In their groundbreaking 2013 paper “The Future of Employment: How Susceptible Are Jobs to Computerization?” Oxford University scholars Carl Frey and Michael Osborne concluded that 47 percent of total U.S. employment is at risk of computerization. As the South Korean economy is similar in many ways to that of the United States, there is little reason to conclude that total South Korean employment does not face a similar risk. Keep in mind, Frey and Osborne drew their conclusions nine years before ChatGPT opened to the public.

Nonetheless, many policymakers and proponents of technology around the world claim that more new jobs will be created as technologies like AI create new needs. This assumption should be reexamined in the case of South Korea.

South Korea’s concentrated R&D spending and potential risks to employment clash when we consider that South Korea operates a positive regulatory system — a system by which all new economic endeavors must be approved by government regulators before implementation. The country needs to spend more on R&D and distribute that money more widely to innovate and create more new economic sectors and enterprises.

These new creations would presumably offer employment for those displaced by AI disruptions. However, the bureaucracy operating in a positive regulatory environment will likely block the creation of new economic sectors due to the potential disruption to entrenched interest groups. Therefore, the country will have limited methods by which to absorb the increasing number of laborers potentially displaced by AI technologies.

The renowned economist John Maynard Keynes often predicted that economizing labor will outpace finding new uses for it. This may happen with AI. When this occurs in a positive regulatory environment, it could be catastrophic for employment. South Korean policymakers and enterprise leaders must make significant changes soon, if South Korea is to weather the AI storm.

Sean O’Malley (seanmo@dongseo.ac.kr) is a tenured professor of international studies at Dongseo University, where he teaches classes on free trade, regionalism and international relations.