Japanese technology group appears on the AI ​​"stage"

In an interview published on CNBC on December 11, Rakuten CEO Hiroshi “Mickey” Mikitani said the group is researching and developing AI language models called LLMs. These are algorithms trained on large-scale data sets, which are the foundation for current AI applications such as ChatGPT by OpenAI (USA).

Mr. Mikitani added that Rakuten has a huge amount of data across many fields, from banking and finance to e-commerce and telecommunications, to train LLM's algorithms.

According to Rakuten's CEO, the group plans to use the AI ​​model internally to improve operational efficiency by 20%. In addition, Rakuten also aims to provide this model to third-party businesses, similar to the way e-commerce group Amazon or US software group Microsoft do. Japanese companies can create Japanese language LLM, giving them a greater advantage over AI products from US and Chinese technology groups.

Mikitani believes that promoting AI models will bring Rakuten “huge profit growth”.

To date, American and Chinese tech giants such as OpenAI, Amazon, Google, Baidu, Alibaba, Tencent, etc. have all launched or are developing their own AI language models. Japanese companies are somewhat behind their American and Chinese rivals in the AI ​​race, but they are also trying to catch up quickly.

Japanese telecoms giant NTT announced that its proprietary LLM will launch in March 2024. Last month, SoftBank Group's telecom arm also said it was using the company's AI computing platform.