Bae Gyeong-hun, nominated as the Minister of Science and ICT, is recognized for fostering a horizontal atmosphere within the organization that allows young talents to showcase their skills freely. He is known as a leader who sets stringent standards for the expertise and performance of his members.

According to industry sources on the 25th, Bae focused on creating a culture at the LG AI Research Institute where members can freely exchange opinions and achieve results based on autonomy. In fact, he is known to have established a 'nim' culture, which has transcended titles, since the institute was founded in 2020. In the business world, LG Group's corporate culture is regarded as conservative. The LG AI Research Institute is noted to be the only place where the 'nim culture' has taken root among LG Group's affiliates and subordinate organizations.

An insider familiar with the circumstances at the LG AI Research Institute noted, 'It's an atmosphere where Bae is called 'Gyeong-hun' instead of 'Director Bae.' It's not just formal horizontal communication, but an actual environment where anyone can freely express their ideas.'

Bae is known to have publicly stated since the early days of the LG AI Research Institute's establishment in 2020 that he would 'secure our own research space.' With the expansion and transfer of the LG AI Research Institute to a new building in Magok, Seoul, in May last year, he has upheld his promise to his researchers. He expressed hope that 'this space (the new building) will serve as another starting point.'

Bae's leadership suited the characteristics of an organization with many young researchers. The institute, which started with 70 members, has grown to over 300 in just 4 to 5 years. He is said to have enjoyed using the expression 'playground' internally, indicating his desire to create a research environment where AI talents can autonomously and joyfully demonstrate their capabilities.

Bae Gyeong-hun, the head of LG AI Research Institute, is conducting a Q&A session at a networking event held during 'ACL (Association for Computational Linguistics) 2024' in Thailand last year./Courtesy of LG Group

An industry source explained, 'Instead of simply filling the LG AI Research Institute's office space with desks and seats, we created a space for naturally gathering, exchanging ideas, and freely discussing, adding that this environment seems to have led to results.'

Bae is known to have set strict standards for his team regarding job performance. The LG AI Research Institute won the overall championship at the Continual Learning competition hosted by CVPR in 2020, the year it was established, and achieved first place in the Stanford SQuAD competition in 2021. In 2023, it also reached the top in KorQuAD 2.0.

Bae, a native of Seoul, graduated from the Department of Electronic Physics at Kwangwoon University and obtained a master's and doctorate in electronic engineering from the same school. He later earned an MBA from Columbia Southern University in the United States and completed the executive project management program at Stanford University. During his doctoral studies, he joined 3R as a technology leader in the fields of image recognition and 3D imaging. He also conducted AI-based research on unmanned autonomous robots at Samsung Thales and worked on image recognition and big data-based AI research at SK Telecom.

Bae Gyeong-hun, the head of LG AI Research Institute, is delivering a keynote speech at an online event commemorating the launch of LG AI Research Institute in 2020./Courtesy of LG Group

Bae, who joined the LG Group in 2016, has led the LG AI Research Institute since 2020 after serving as the head of the LG Economic Research Institute, LG Uplus, and the AI Promotion Division at LG Science Park. He has led the development and commercialization of the super-large AI 'EXAONE' and has expanded the global collaboration network. In 2023, he was awarded the Silver Tower Industrial Medal by the Ministry of Science and ICT.

The LG AI Research Institute released EXAONE 3.0 as open source in June last year, followed by EXAONE 3.5 in December, which included the 32B model. This year, it unveiled the data risk automatic analysis AI 'NEXUS' and the 'EXAONE Deep' model specialized in mathematics, science, and coding inference. Notably, the 32B model was named on the 'notable AI models' list by the U.S. private research organization EpochAI.

※ This article has been translated by AI. Share your feedback here.