On Thursday, South Korean tech giant Naver Corp. announced an updated version of their hyperscale AI model HyperCLOVA X and its conversation chatbot service designed for the Korean language.
The move is part of Naver’s effort to compete in the global race for linguistic AI, which has been heating up since the launch of ChatGPT late last year. The model is supported by Naver’s supercomputing and data-analyzing capabilities, and the experiences they have gained from their all-in-one online platform, which covers their largest search engine, cloud services, email, shopping, and other internet products.
Naver CEO Choi Soo-yeon said, “Our technologies and experiences that Naver has learned from various sectors, including searching, shopping and booking for decades, support technological competitiveness of our generative AI model HyperCLOVA X.” She added that the company has invested 1 trillion won (US$754 billion) in AI over the last five years.
To boost their data-analyzing abilities, Naver will be opening a new large data center in Sejong in November, which will house more than 600,000 servers. This will be the brain center for Naver’s hyperscale AI services.
Naver is also working with Samsung Electronics Co. to develop semiconductor solutions tailored for hyperscale AI models to speed up the handling of massive AI workloads. They are also in talks with global partners to develop local models of their generative AI service for global expansion.
The Korean tech company also revealed CLOVA X, a large language model-based AI chatbot, and CUE:, a generative AI search service, both of which are powered by HyperCLOVA X. CLOVA X will go on beta testing later in the day, while CUE: will be launched next month.
Choi said, “Naver is ready to face the new change opened by generative AI. We will support our sellers, creators and partners to encourage their diversity and grow further with our technology.”