Talent Drain in the US and China... No One to Lead AI

No 'Korean' Companies in Top 20 Global AI Firms
Salary and Work Environment Gaps Large... "Startup Development Is More Effective"

[Asia Economy Reporter Jin-gyu Lee] Kim Byung-hak, Vice President of Kakao Enterprise, lamented at last year's 'if kakao Developer Conference' that "there is a lot to be done in the field of artificial intelligence (AI), but there is a severe shortage of talent to work together." Since 2017, Kakao has been holding recruitment events in the United States and has been working to attract global AI talent, but this year it is facing difficulties due to the novel coronavirus disease (COVID-19). Although it has secured 500 to 600 AI-related personnel so far, internal evaluations indicate that it falls short of the original goal. Other companies are in the same situation. As American and Chinese companies sweep up global AI talent, domestic companies are complaining about a talent shortage.

◆Risk of Declining AI Global Competitiveness = According to the 'AI Research Ranking 2019' by the globally prestigious AI conferences, the International Conference on Artificial Neural Networks and the International Conference on Machine Learning, no domestic companies are ranked within the top 20 global companies leading AI research. In contrast, U.S. companies such as Google (1st), Microsoft (2nd), Facebook (3rd), and Chinese companies Tencent (6th), Alibaba (7th) occupied the top ranks. Japan's Toyota was also ranked 11th. Among the top 40 global organizations leading AI research, including companies and academia, only KAIST (28th) and Seoul National University (38th) from Korea were ranked, with no companies making the list. Experts analyze that one reason for the decline in AI competitiveness is the global talent's avoidance of domestic companies. Lee Kyung-jun, an AI expert and professor of Business Administration at Kyung Hee University, explained, "There is a large salary gap between domestic companies and global companies, and the lack of English working environments in Korean companies makes it difficult for domestic companies to recruit core AI talent from English-speaking or Chinese-speaking regions. While domestic companies sometimes hire Korean-American talent active in the U.S., it is rare for them to scout core American or Chinese talent."

Even in the case of Go AI, the personnel gap between domestic and overseas companies is clear. The startup DeepMind, which developed Google's AI Go 'AlphaGo,' had more than 100 global AI personnel at the time, and Tencent's 'Jueyi' development team reportedly had more than 10 pure AI personnel and dozens of big data experts. In contrast, NHN's Go AI 'Handol' had only about five AI personnel involved. Lee Chang-yul, head of NHN's AI development team, said, "Reinforcement learning used in Handol still has fewer personnel inflows domestically compared to other machine learning fields, making it difficult to recruit personnel. Chinese and American companies offer better treatment to secure AI talent, so unless global talent applies, we are not considering active recruitment."

◆Need for Measures to Foster Startups = Although AI talent from the U.S. and China is still evaluated as more competitive than domestic talent, they tend to avoid domestic companies. According to a survey by the Korea Economic Research Institute under the Federation of Korean Industries, when the AI talent competitiveness of the U.S. is rated as 10, China is rated 8.1, but Korea's talent competitiveness is only 5.2.

Some domestic companies judge that AI does not lead to profit generation and withdraw from the scouting competition. While the annual salary of senior AI personnel at U.S. Silicon Valley companies like Google or Facebook easily exceeds 400 million KRW, it is difficult to surpass 100 million KRW at domestic large companies. An industry insider said, "A few years ago, domestic companies were eager to secure AI talent and tried to hire AI-related personnel with good treatment, but recently, as they face difficulties in generating profits through AI, they are reducing AI investment and AI personnel salaries." Professor Wi Jung-hyun of Chung-Ang University, who serves as president of the Korea Game Society, explained, "Domestic IT companies, including game companies, cannot offer global AI talent salaries more attractive than domestic AI personnel. If the salary gap widens in the seniority-based culture, existing domestic personnel may resist." He advised, "Since domestic companies have limits in attracting global AI talent with huge salaries, fostering startups and providing attractions such as stock option exercises when unicorn companies grow would be more effective."

This content was produced with the assistance of AI translation services.