LG to unveil latest EXAONE AI platform in August

LG AI Research President Bae Kyung-hoon speaks on the future of AI during the 2024 CEO Summer Forum hosted by the Federation of Korean Industries (FKI) at Lotte Hotel Jeju, Thursday. Courtesy of FKI

SEOGWIPO, Jeju Island — LG Group is set to introduce the latest version of its hyper-scale artificial intelligence (AI) platform — EXAONE — in August, hoping to enhance efficiency and accuracy in research and business activities across the group's key affiliates, LG's AI chief said Thursday.

"Companies should not remain hesitant in adopting AI systems," LG AI Research President Bae Kyung-hoon said on the sidelines of this year's CEO Summer Forum held by the Federation of Korean Industries (FKI).

"All LG affiliates have worked hard for the AI transformation over the past four years. Each CEO of the group's subsidiaries studies AI on their own. We will unveil our newest generative AI next month."

LG unveiled EXAONE for the first time in 2021. EXAONE is an expert-level AI model using images and text. It can understand academic documents and is widely adopted by LG's major affiliates, such as LG Chem and LG Display. EXAONE 2.0 was unveiled in 2023.

The head of LG AI Research also advised top executives from any company to remain agile in adopting AI systems for their businesses.

"LG Group also faced challenges in adopting AI at its infancy, as most executives from affiliates showed little interest in AI for financial issues," he said. "But our top management started to have a firm belief in AI and established the AI-dedicated research facility. Leaders from LG's subsidiaries are now taking the lead in embracing AI."

He showed confidence in creating tangible outcomes by introducing AI systems.

For instance, LG Group's chemical affiliates, such as LG Chem and LG Household and Health Care, created much faster output in molecular design tests with EXAONE. Bae said it typically took three years for chemists to do so, but EXAONE resolved it in about a month.

LG's affiliates ask LG AI Research for assistance in conducting a number of their research projects in areas such as pharmaceutical development, electrolytes and light-emitting materials due to a limited pool of scientists and researchers, he said.

"The function of AI models in specific professional fields, such as medical sectors, has been well-proven," he said. If AI systems get more sophisticated, this will help drive innovation in more promising growth areas, such as chemistry and biology, according to him.

Bae also stressed the importance of catching up with the rapid growth of AI.

"The most important thing is speed in the era of AI," he said.

"More companies should be willing to quickly understand and introduce AI systems fit for their own businesses."

Even if AI is very skillful in predicting the future with previous data, it still comes with a shortcoming in that it cannot forecast pandemics or wars, he added.