Japanese Tech Giant Rakuten Plans to Launch Its Own AI Large Language Model Within the Next Two Months
-
Japanese tech giant Rakuten plans to launch its own large language model (LLM) within the next two months. CEO Hiroshi 'Mickey' Mikitani revealed in an interview with CNBC that due to Rakuten's involvement in various business sectors such as banking, e-commerce, and telecommunications, the company has access to a significant amount of 'unique' data that will be utilized to train its large language model.
Image credit: AI-generated image, licensed by Midjourney
Mikitani stated, 'No other company has a dataset like ours.' He plans to use the AI model internally to enhance operational efficiency and marketing effectiveness, with an expected improvement of 20%. Additionally, he aims to make the model available to third-party businesses, similar to the approaches of Amazon or Microsoft.
Mikitani disclosed that the company will roll out this AI model in the coming months. This move is part of Rakuten's effort to compete with other tech companies, such as OpenAI, Amazon, and Google in the U.S., as well as Baidu, Alibaba, and Tencent in China, all of which have launched their own large language models.
Compared to tech giants in the U.S. and China, Japanese companies have been somewhat lagging in the field of large language models. However, they are striving to catch up. Earlier this month, telecom group NTT announced that its proprietary large language model will be launched in March. Additionally, SoftBank's telecom division announced in November that its generative AI computing platform is now operational.
Japanese companies have the opportunity to create large language models specifically tailored for Japanese, which could give them a competitive advantage. Mikitani stated that entering the AI field would bring 'significant profit growth' for Rakuten.