South Korean ‘Google’ unveils HyperCLOVA X, to drive AI services & personalised experiences

Naver, a prominent internet company in South Korea, has introduced its latest innovation, HyperCLOVA X, a suite of generative artificial intelligence (AI) tools. This new offering compliments their existing lineup, which includes ChatGPT from OpenAI, Bard from Alphabet, Llama from Meta, and Ernie Bot from Baidu. Naver envisions its AI solution as a means to provide AI-driven search capabilities for users while also delivering personalised services for enterprise clients.

Choi Soo-yeon, CEO of Naver, remarked, "We are entering a new phase defined by generative AI. In this era, a fierce competition is underway to be adopted by services and business partners."

The flagship component of the HyperCLOVA X suite is the chatbot application CLOVA X, designed to enhance web searches, online shopping experiences, and navigation services, much like the AI integration in Google Search and Bing. The chat and generative AI functionalities within CLOVA X are managed by a component known as Cue.

Collaboration with Samsung

Referred to as the 'Google of South Korea,' Naver has already initiated beta services for CLOVA X and plans to commence the beta phase for Cue in September. In a collaborative effort with Samsung, Naver is also actively working on the development of new chip solutions.

These chips are intended to be smaller and more efficient, supporting the advancement of Naver's AI technology. Additionally, the company has scheduled the opening of a data centre dedicated to HyperCLOVA X services in November.

Naver stands apart from giants like Google and Microsoft by focusing on untapped markets that have yet to be adequately addressed by American and Chinese tech conglomerates. This strategic direction targets specialised regions like the Middle East, characterised by political sensitivities, as well as non-English-speaking territories such as Japan and Southeast Asia.

GPT-3.5 Turbo and Code Llama streamline workflows

In the broader landscape, recent developments include OpenAI's announcement of GPT-3.5 Turbo's fine-tuning capability, enabling businesses and customers to refine this potent and cost-effective model using their own data.

In a similar way, Meta has introduced Code Llama, a sizable language model (LLM) engineered to generate and discuss code through text prompts. This innovation holds the promise of streamlining and enhancing workflows for developers, potentially ushering in increased efficiency.