Never unveils a large language model (LLM).. determined to win the AI race against Google and Microsoft

폴 리

hoondork1977@alphabiz.co.kr | 2023-08-25 00:52:22

 

 

[Apha Biz=(Chicago) Reporter Paul Lee] Choi Soo-yeon, CEO of Naver, unveiled the next-generation LLM (Large Language Model) 'HyperCLOVA X' at the 'DAN 23', a conference to share its own technology direction and business strategy.

The LLM, which best understands Korean culture and context, has challenged the global generated AI (Artificial Intelligence) competition occupied by Google, Microsoft, and Meta. To this end, Naver has invested more than 1 trillion won in AI over the past five years.

HyperCLOVA X advanced the world's third LLM 'HyperCLOVA' (parameter 204 billion) developed by Naver in 2021.

CEO Choi said, "We will support sellers, creators, and partners in Naver with strong technologies and services so that they can be more competitive than anyone else in the generational AI era."

 


 

Based on this, various B2C·B2B services are poured out. In particular, attention is focused on the interactive AI service 'CLOVAX' and the search service 'CUE:' If CLOVAX, which was released at 4 pm on this day, is 'Chat GPT' of open AI, CUE, which will be released in September, is close to Microsoft's 'Bing Chat' and Google's 'Bard'.


CLOVAX provides answers to various questions such as creation, summary, reasoning, translation, and coding. Multimodal AI is applied to understand and create photos, images, and sounds in addition to text. Like the plug-in of ChatGPT, we have introduced a 'skill' system that links internal and external services of Naver.

The generative AI market plans to improve profitability with the B2B business as it costs large-scale operating costs.

We are also promoting the overseas expansion of HyperCLOVA X. HyperCLOVA X will be applied to LineWorks in Japan by the end of this year and corporate services will be released.

 


[ⓒ AlphaBIZ. 무단전재-재배포 금지]

많이 본 기사