Friday, December 27, 2024

OpenAI opens a hub in Tokyo, adds a GPT-4 model optimized for the Japanese language

Share

OpenAI expands its operations to Japan, with the opening of a new office in Tokyo and plans for a GPT-4 model optimized specifically for the Japanese language.

This move is vital for several reasons. This highlights the opportunity the company sees in attracting customers to business in the country, and underlines that as OpenAI grows, it will likely need to localize its technology into different languages. And from a more pragmatic point of view, as the advantages and disadvantages of artificial intelligence become increasingly clear to governments, their regulators and the entire public discourse, it becomes vital for OpenAI to have its feet on the ground to understand these currents and influence them to its advantage .

OpenAI offers huge name recognition, but it won’t come without competition. SoftBank also hopes to build on and capitalize on the current enterprise AI craze, launching SB Intuitions last November, which aims to build vast language models (LLM) and generative AI services natively in Japanese.

The Microsoft-backed developer of ChatGPT has just 1,200 employees worldwide, which is diminutive considering it is currently valued at around $80 billion and is struggling to handle high incoming demand. We are now looking demanding at how to scale to meet this demand. Last year, it opened its first international office in London, followed a few months later by the inaugural European Union (EU) office in Dublin. Tokyo will represent OpenAI’s first office in Asia and fourth in the world (including the headquarters in San Francisco).

CEO Sam Altman highlighted Japan’s “rich history of people and technology coming together to achieve more” as one of the reasons for the formal presence in the region. But from a more opportunistic standpoint, OpenAI’s global expansion efforts so far have been quite strategic.

The UK is a major hub for AI talent, while the EU is currently pursuing an AI regulatory agenda. Japan is one of the largest developers and users of humanoid robots and other artificial intelligence equipment, which played vital roles during Altman’s world tour last year, when he visited Japanese Prime Minister Fumio Kishida and I talked about OpenAI’s intentions opening an office in the country. Japan is the current chairman of the G7 and Chair of the G7 Hiroshima AI Processan initiative to promote AI security, including stronger AI governance.

The choice of OpenAI to lead the up-to-date hub is noteworthy. OpenAI Japan will be managed by Tadao Nagasaki, who joins the company after 12 years at Amazon Web Services (AWS), where he led Amazon’s regional cloud computing business. In other words, the main goal of the latest expansion is to develop OpenAI’s profile and cooperation with enterprises.

Enterprising

As president of OpenAI Japan, Nagasaki will be tasked with building a local team in the field to double OpenAI’s operations in the country. The company already counts Daikin (an industrial company best known for air conditioners), Rakuten and Toyota among customers using the enterprise version of OpenAI ChatGPT, which OpenAI says provides additional privacy, data analytics and customization options in addition to the standard consumer-class ChatGPT options.

OpenAI says ChatGPT is also being used by local governments to “improve the efficiency of public services in Japan.”

GPT-4 adapted for Japanese. Image credits: Open AI

ChatGPT has a long history of supporting multiple languages, including Japanese. However, optimizing the latest version of the GPT LLM core package for Japanese will enable a better understanding of the nuances of the Japanese language, including cultural understanding, which should augment its effectiveness, especially in business environments such as customer service and content creation.

OpenAI also claims that its custom model has better performance, which means it should run faster and be more cost-effective than its predecessor.

For now, OpenAI is providing early access to the custom GPT-4 model to some local companies, with access gradually opening up via the OpenAI API.in the coming months.”

Latest Posts

More News