[Asia Future Business Forum] Baek Gyeonghun, LG AI Director: "Level of AI Utilization Determines Corporate Competitiveness"

11th Asia Future Business Forum
On the Theme of 'Strategies for Companies in the Era of Supermassive AI'

"There is no business sector unrelated to artificial intelligence (AI). The level of AI utilization will vary depending on the level of data possession and investment scale, and this will soon become a company's competitive advantage."

On the 17th, Baek Gyeong-hoon, head of LG AI Research Institute, attended the 11th Asia Future Business Forum held at Lotte Hotel in Sogong-dong, Jung-gu, Seoul, and made these remarks under the theme "The Era of Massive AI, Corporate Response Directions."

At the "2023 Asia Future Business Forum" held on the 17th at Lotte Hotel in Jung-gu, Seoul, hosted by Asia Economy, Baek Kyung-hoon, head of LG AI Research Institute, is giving a special lecture on the theme of "The Future of AI." Photo by Yoon Dong-joo doso7@

Director Ba summarized the stages of AI development into three phases: ▲ ANI (Artificial Narrow Intelligence), focused on executing specific tasks ▲ AGI (Artificial General Intelligence), which performs a wide range of tasks based on reasoning abilities comparable to humans ▲ ASI (Artificial Super Intelligence), possessing intelligence surpassing human capabilities.

Director Ba explained, "When AI reaches ASI, it will have a significant impact on various industries," adding, "There will be major changes beyond prediction and imagination, such as reversing aging and extending lifespan to achieve human immortality, or the disappearance of distinctions between virtual and real worlds."

The current AI market is expected to grow from 87 trillion won in 2021 to 478 trillion won in 2026, with an average annual growth rate of 37%. Director Ba emphasized, "Generative AI will occupy a large portion," and added, "The next decade will be the era of generative AI, where AI creates content, so focusing on generative AI is essential to advance toward ASI."

However, Director Ba noted, "Currently, generative AI suffers from severe hallucination issues, especially in specialized fields," and stated, "For generative AI to be used in actual corporate settings, it is important to secure high-quality data reliability, implement industry-specific specialized functions, and achieve a balance between performance and cost-effectiveness."

The "2023 Asia Future Business Forum" was held on the 17th at Lotte Hotel in Jung-gu, Seoul, hosted by Asia Economy. AI-related businesspeople attending the event are listening to the lecture with serious expressions. Photo by Dongju Yoon doso7@

To this end, LG aims for 'Expert AI' that provides higher expert capabilities to professionals and is nurturing 'EXAONE.' Director Ba said, "EXAONE learns Korean and English simultaneously to effectively access more knowledge and focuses on specialized knowledge such as academic papers and patent data," adding, "It learns text and images concurrently and develops models based on reliability, possessing differentiated competitiveness that other massive AI models do not have." In particular, he stated, "In July, we will launch a reliable generative AI technology, an expert-type model of EXAONE," and added, "We also plan to build a system platform for commercial services."

Director Ba believes that if companies utilize massive AI, it will lead to productivity improvements and provide significant assistance not only to researchers and creators but also to corporate decision-makers. However, he lamented, "Unfortunately, there are no Korean companies among Forbes' '50 Promising AI Companies,'" emphasizing, "I hope many Korean companies will develop as promising enterprises."

Finally, he said, "AI is not a technology that replaces humans but evolves together with humans. It is necessary to overcome indiscriminate fear of AI," and added, "Socially responsible leaders should understand the reliability issues of AI well and proactively respond to AI innovation through discovering specialized functions and accumulating data."