Japanese telecommunications company Softbank will invest an additional 150 billion yen by next year to develop high-performance artificial intelligence (AI), local media reported on the 22nd.
According to the Nihon Keizai Shimbun (Nikkei), Softbank is planning to make world-class models similar to ChatGPT-4 made by OpenAI with this additional investment. Softbank invested 20 billion yen last year to build the base of calculating equipment needed for Generative AI development.
Nikkei said, “It is the largest for a Japanese company (with investments related to Generative AI),” and expected Softbank to purchase Nvidia semiconductors in the U.S. to improve its computing power by dozens of times.
Softbank is developing a large language model (LLM), a technology that is based on Generative AI. It plans to complete the 390 billion parameter model by March next year and then start developing a 1 trillion parameter model specialized in Japanese.
“Softbank is also pushing for the maintenance of AI data centers in line with the development of domestic Generative AI,” Nikkei said. “Recently, we decided to invest 65 billion yen to build Japan’s largest AI data center in Hokkaido.” He added, “Domestic Generative AI has also become important in terms of economic security, and the government and companies have begun to value ‘data sovereignty’ that manages their data in Korea.”
Earlier in October last year, Son Jeong-ui, chairman of SoftBank Group, predicted the expansion of the AI business, saying, “I want to make SoftBank Group the group that uses AI the most in the world.”
Around the world, the raw AI market is already aggressively progressing. It is questionable whether it will survive the competition with global companies that have already started.
SOPHIA KIM
US ASIA JOURNAL
