광주과학기술원 웹진

GIST, a mecca for fostering

unprecedented AI professionals

President Kiseon Kim

In March 2016, a five-game Go match between Lee Sedol, a 9 dan rank, and Google’s AI (commonly known as AlphaGo Lee) that resulted in a 4:1 victory by AlphaGo sent shockwaves. An article published in the renowned science magazine, Nature in October 2017, less than a year after the match, introduced scientific evidence that AlphaGo Zero, an improvised version of AlphaGo Lee would be able to beat AlphaGo Lee by 100:0. As it can be seen in the continuous evolution of AlphaGo, artificial intelligence is exhibiting superiority over human in various areas along with image processing, which was believed to be distinct area of human intelligence, and people who used to feel that AI was something ambiguous that could only be seen in novels or movies are starting to feel awe for the first time about this new world. Starting from 2015, AI is developing continuously by reinforcing learning results through utilization of general and regional information simultaneously based on temporal and spatial experience (data) by using a learning structure that replicates artificial nerve network that resembles a person's brain. Just like how a child grows into an adult by growing independently when given the chance to learn, the era of post-human AI that can think beyond a person and can become super-intelligent has begun.

If so, we need to focus on how we can prepare for this era of AI and how to gain an upper hand by utilizing new opportunities. The government of the Republic of Korea announced a 5-year National Operation Plan in July 2016, revealing details on the creation of ‘AI-centric industrial convergence complex’ in preparation for transforming old, industrialized cities into AI centric cities through the 4th Industrial Revolution. Subsequently, Gwangju City announced itself as ‘Gwangju, an AI-centric city’ in January of this year and has started to implement the creation of world-class industrial convergence complex and to train professionals. It plans to build key AI infrastructure such as Data Center, Validation Block, and Startup Block over 46,200 ㎡ in the High Tech District 3 by 2024, and it is expected to accommodate the AI Cloud Data Center, Validation Test Bed in three fields: energy, automotive, and health care which are the mainstream industries for the AI+ region. The best performance driving simulator and artificial intelligence living lab in Korea will also be constructed, too. In addition, Health Care Start-up Valley is going to be housed in Dong-gu, Gwangju, an old downtown regeneration region. Investment promotion by outstanding companies at both home and abroad as well as investment funds are also expected.

For the national project, ‘AI-centric industrial convergence complex’, roles have been assigned to various AI and software specialists as well as institutes specializing in innovation nationwide for participation, and GIST, as a research-centric educational institute, will contribute through research, education, and start-up support. Training specialists is the most fundamental core of GIST in order to expand AI technology and to apply it effectively in all industrial fields. GIST considers all members of society as potential specialists. 1) Training a graduate school-level global AI research workforce based on the research competitiveness of GIST, 2) training college-level industrial specialists for applied science technicians in each field of AI field, and 3) training general specialists capable of handling and supporting data for AI are the three areas for training specialists.

First, in March of this year, GIST established the AI Graduate School. It is the first step towards creating global standards for a high-quality workforce that is being supported by the nation’s program to train specialists. The GIST AI Graduate School is centered on an integrated masters and Ph.D. course, receiving up to KRW 19 billion over 10 years. The curriculum of the AI Graduate School is comprised of the entire life-cycle of industries and key details include 9 basic subjects of AI, 22 core subjects, 13-start-ups, site application and expansion with the objective of nurturing specialized AI knowledge and convergence skills. Graduate school students would learn basic key theories and practical skills, experience internships and global cooperation in the 1st-3rd year after admission, and in the 4th-5th year, they will undergo specialized education, conducting educational research closely related to industry centered on validation in preparation for start-ups in the related specialized fields. In order to reinforce education and research on 'AI+ X' (Artificial Intelligence Convergence), research on convergence related to autonomous driving, digital health and energy have already commenced in collaboration with electrical, electronics, computers, machinery, convergence specialist technology centers and has been gaining attention from domestic and overseas institutions from its initial stage because of its differentiated infrastructure and research environment.

Second, GIST would be playing the role of a mediator, providing consistent education direction and content related to AI and data science with the objective of training as workforce that could contribute to the industrialization of AI by collaborating with the natural science and engineering colleges within the region to train a college-level workforce in order to supply technicians to each domain of the AI field. In particular, education using AI cloud data and large-scale validation testing facilities that would be installed in the region will contribute directly to training an industrial workforce for reinforcing not just regional but national AI competitiveness.

Third, training general specialists capable of handling and supporting data for AI is an area of education that could contribute to the generation of jobs regardless of the major. It is similar to the workforce in call centers, providing specialized education for industrial call centers over a short period of time. A manager who can conveniently handle big data for AI and data connection, which is the first step to applying new AI tools in conventional industries, is required. Third, the future workforce involves supporting the use of big data and training managers a socially conscientious workforce through cooperation with social education specialists in college is an extremely crucial part.

The paradigm shift of AI technology is a crisis for South Korea, but it is also an opportunity for dominating a new market. The revenue expected to be gained through the creation of an AI ecology amounts to KRW 295 billion in Gwangju alone. The project for creating the AI industrial and R&D convergence complex in Gwangju is expected to turn the tables for domestic AI related fields and address the absence of AI specialists. This is a direct expected effect of having a graduate school workforce with natural science and engineering expertise, and, considering the indirect ripple effects such as fostering big data workforce managers and creating of jobs, Gwangju is expected to become a base as a future growth engine for the economy with revitalized technologies, companies, and jobs within the region. GIST’s AI specialists will be standing at the center of AI.

South Korea and the entire world are looking at the course and performance of GIST's approach. GIST is going to contribute to the reinforcement of national AI competitiveness by utilizing its 25 years of research infrastructure, results, and educational abilities. As demonstrated over the past year, GIST will continuously cooperate with the regional government to establish a national model for pioneering AI centered in the region. Now, GIST will endeavor to train various AI specialists and world-class AI research competence, headed by research centers including the AI industrial convergence complex, AI education center, and AI convergence program. Collaboration with renowned domestic and foreign AI research institutes such as NVIDIA is actively taking place, too. GIST is going to play a pivotal role in creating a future industrial ecosystem based on AI by adding new life through AI to all industrial fields, including conventional manufacturing industries that have been responsible for maintaining jobs in the region. South Korea, transforming together with AI, has started from GIST. We seek for your interest and support.