Japanese tech conglomerate Rakuten is poised to venture into the domain of artificial intelligence language models (LLM), as CEO Hiroshi “Mickey” Mikitani disclosed in an interview with CNBC.
Developing a Unique LLM
Mikitani highlighted that Rakuten, known for its diverse portfolio encompassing banking, e-commerce, and telecommunications, possesses a vast reservoir of “very unique” data. This data will serve as the foundation for teaching and refining its proprietary Large Language Model.
Utilization and Expansion Plans
The primary objective for Rakuten’s LLM deployment lies in enhancing operational efficiency and marketing strategies, with an ambitious target of a 20% improvement, according to Mikitani. Moreover, the company aims to extend the model beyond its internal use, offering it to third-party entities for further development and utilization.
Mikitani teased an imminent development, stating that Rakuten “will have something coming in a few months.” However, a Rakuten representative clarified that this alluded not to the timing of the launch but rather a potential announcement regarding a significant language model, possibly within the next few months.
The Landscape of Language Models
In the realm of large language models, US and Chinese tech giants dominate the scene, notes NIX Solutions. Industry behemoths like OpenAI, Amazon, Google, Baidu, Alibaba, and Tencent have already unveiled their respective models. While Japanese companies have historically lagged, recent initiatives by entities like NTT and now Rakuten signify an accelerated drive to bridge the gap.
Closing the Gap
NTT, a prominent telecommunications group, revealed plans for its LLM slated for a March release, setting the stage for Rakuten’s announcement of its own entry into the AI language model sphere.
In summary, Rakuten’s venture into the realm of AI language models marks a significant shift in the landscape of tech innovation, with Japanese firms intensifying efforts to catch up with their American and Chinese counterparts.