SK Telecom (CEO Youngsang Yoo) participated within the presentation of the ‘M360’ session held on the Westin Chosun Seoul on the 2nd with Softbank, NTT Docomo, etc., saying that artificial intelligence (AI) technology will improve the performance of communication and networks and produce latest business possibilities. revealed.
M360 APAC is a mobile conference held annually by GSMA since 2013. Following last yr, it was held in Korea again this yr, and KT, which is lively as a GSMA board member, served because the host sponsor for the second consecutive yr on behalf of domestic telecommunications firms.
This yr’s agenda was ‘Digital National Development through AI’, and communication-related technologies akin to AI innovation and next-generation networks (5G Advanced-6G) were mainly mentioned.
SKT, which has been actively advancing AI, announced that it would develop an AI agent optimized for communication based on the Large Language Model (LLM) ‘A.X’.
It was announced that continuous LLM learning and advancement is in progress with its own network-based documents, and that it would play an auxiliary role in resolving network and communication problems.
Specifically, Jennifer MS Park, head of the AI-DT division of SKT Infrastructure, said, “If we use AI, we are going to find a way to transcend easy abnormality detection to discover early signs and progress to self-healing (auto-healing).”
He went on to say, “We are going to find a way to grasp the context, akin to customer movement routes and transportation methods, and use it for marketing,” adding that business expansion through AI can also be possible.
NTT Docomo, Japan’s largest mobile communications operator, which recently signed a “joint research agreement on AI technology within the next-generation communications field” with Samsung Electronics, also emphasized the “potential of AI.” Takehiro Nakamura, CSO of NTT Docomo, said, “With a purpose to increase network operation efficiency, the bottom line is to quickly receive and analyze fresh data and quickly reply to network problems based on this.”
Nonetheless, limitations of AI were also identified. “With a purpose to use AI in actual workplaces, integration into existing back-end systems can be essential,” said Choi Jin-seong, senior fellow at Softbank’s Advanced Technology Research Center. “Many efficiency tools akin to ML Ops and AI Ops are emerging, but the event environment might be simplified. “We want next tools that might be integrated,” he said.
“If the back-end system is integrated, many problems akin to retraining of existing engineers, high-quality training data, and talent development might be solved,” he said. “In the long run, latest varieties of AI agents can be possible beyond chatbot, inference, and agent stages.” said.
SKT cited the large infrastructure issues required for AI learning. SKT also said that it has accomplished a lot of GPU contracts, and explained that extensive systems and latest equipment are essential for AI to process communication and network information, and are also essential from the attitude of a ‘network vendor’.
Moreover, although generative AI technology continues to be evolving and its effectiveness can’t be accurately predicted, the common opinion was that it should give attention to network architecture and equipment architecture.
Meanwhile, Linda Sham, Vice President of Hong Kong Telecommunications (HKT) Mobility and Telecommunications Business Division, participated within the ‘Diversity of AI Introduction’ session and explained the potential of ‘AI Satellite Connectivity Service’.
Through AI, ▲Network optimization akin to traffic pattern evaluation and bandwidth evaluation ▲Supporting service stability akin to predicting device defects ▲Predicting and defending against cyber attacks ▲Predicting disruptive aspects akin to emergencies and natural disasters ▲Providing customized plans based on evaluation of consumer behavior patterns and lifestyles Which means it is feasible.
He quoted Microsoft founder Bill Gates and emphasized, “AI is the primary technology with no limits.”
Reporter Jang Se-min semim99@aitimes.com
