hankyoreh

(Pixabay)

Artificial intelligence (AI), a major component of the wave of technological innovation known as the Fourth Industrial Revolution, made rapid progress in the 2010s.

According to a report released by the Institute for Human-Centered Artificial Intelligence at Stanford University, AI performance increased at seven times the speed of Moore’s law in the 2010s.

Moore’s law states that the processing ability of computer chips doubles every two years.

The improvement of AI capabilities picked up speed after the emergence of groundbreaking “deep learning” technology, developed by Geoffrey Hinton in 2012. Those capabilities doubled every 3.4 months in the second half of the decade.

With AI’s potential confirmed, the world’s most advanced countries began adopting a national strategy for AI in the late 2010s, led by China. An intensifying rivalry over AI technology has heated up tensions between the US, long the leader in the field, and China, which seeks to unseat it.

Countries are devoting so much effort to developing AI technology because of the huge effect it will have on human lives. AI will have a sweeping impact not only on human occupations and working hours and methods but on all areas of our daily lives.

Technologies that affect all aspects of human society are called general-purpose technologies. Previous examples include electricity, computers, and the internal combustion engine.

But what about South Korea? Did it make substantial progress on AI technology while innovation was sweeping the field in the 2010s?

Clarivate Analytics, a multinational database analysis firm, and the KAIST Center for Innovation Strategy and Policy published a report Tuesday titled “The Global Race for AI Innovation: Present and Future.” The report compared progress in AI innovation in ten technology-leading countries, including South Korea, in the 2010s.

The report’s assessment of South Korea can be summed up as follows: Korea has achieved quantitative growth but is lacking in qualitative achievements.

The report, which analyzed the number and content of technical AI patents around the world between 2010 and 2019, found that 147,000 patents were registered during that period. The number of patents increased by an average of 31% each year.

The ten countries in the report accounted for 136,000 of those patents, or 92% of the total. In effect, those ten countries are leading the development of AI technology around the world.

In terms of patent registration numbers, the top ten countries break down into three groupings: two on the top, two in the middle, and six in the bottom. China was far and away the leader in number of patents, holding 91,236, or 60% of the total.

That’s the result of a massive material and workforce investment by the Chinese government, which hopes to become the global leader in AI technology by 2030. Although the US is the traditional leader in the field, its researchers only registered 24,708 patents in the same period.

Further behind were Japan in third place, with 6,754 patents, and South Korea in fourth place, with 6,317 patents.

Japan and South Korea had a considerable lead over the third grouping, ranked from fifth to tenth place. Germany, Taiwan, the UK, Canada, France, and India had all registered between 2,280 and 529 AI patents.

While South Korea occupied fourth place in the number of patents, a quantitative measurement, it placed much lower in a qualitative assessment. Only 8% of Korea’s patents were among the top 10% of patents in terms of influence, ranked by their combined patent impact (CPI).

CPI measures the number of times a patent is cited in other technology, indicating the influence of the patented technology.

On average, 14% of registered patents in the ten countries were in the top 10% by CPI. That means that South Korea may have more patents than other countries, but its patents are less likely to make an impact.

In qualitative terms, 43% of American patents were in the top 10% in terms of CPI, by far the highest percentage among the ten countries.

“Since the US has a large number of registered patients and its patents have the most influence, it can be regarded as leading innovation in global AI technology,” the report said.

Another indicator of the level of AI technology is the number of patents registered overseas. China is an outlier in this regard since a full 96% of its patents are registered domestically. That’s much higher than the 60% share of domestic patent registration in South Korea, Taiwan, and Japan, as well as the US.

But that figure is no reason to disparage Chinese technology. The report said that this is largely due to the immense size of China’s domestic market. Since South Korea has a smaller market, it should work even harder to register patents overseas than the US or Japan, with their larger domestic markets.

“Technical influence in the global market is ultimately determined not by the number of technological innovations but by the qualitative superiority of technological ability. The time has come to develop a strategy that can lead not to quantitative growth but to qualitative growth based on superb technological ability,” the report said.

“When AI patents are categorized by their researchers, South Korea is second only to China in the percentage of patents obtained by universities, but universities rank very low in the impact of their patents. We need to set up an AI ecosystem in which R&D experts at universities and industrial experts on the ground can work together organically to create competitive technology,” said Kim Won-joon, director of KAIST Center for Innovation Strategy and Policy.

By Kwak No-pil, senior staff writer

Please direct comments or questions to [english@hani.co.kr]