To use this site, please disable the ad blocking feature and reload the page.
This website uses cookies to collect information about your visit for purposes such as showing you personalized ads and content, and analyzing our website traffic. By clicking “Accept all,” you will allow the use of these cookies.
Users accessing this site from EEA countries and UK are unable to view this site without your consent. We apologize for any inconvenience caused.
Yomiuri Shimbun file photo NTT Corp. President Akira Shimada is seen in Chiyoda Ward, Tokyo, in June.
The Yomiuri Shimbun
13:52 JST, November 2, 2023
NTT Corp. plans to launch a business-focused generative artificial intelligence service in March, the company has announced.
The system’s Japanese language-processing capability is on a par with such overseas generative AI models as ChatGPT, developed by U.S.-based OpenAI Inc., NTT said Wednesday.
NTT has developed an in-house large language model (LLM) named “tsuzumi” to serve as the basic technology. Tsuzumi was named after a hand drum used in traditional Japanese music.
By having tsuzumi consume high-quality data relating to each company’s operations, the system size can be kept to a minimum, allowing firms to implement it cheaply.
Tsuzumi also is capable of reading business documents such as invoices containing charts and diagrams. NTT expects the service to be put to such uses as creating medical record databases at hospitals and recording operator responses in call centers.
The company plans for its AI model — which will also be available in English — to achieve annual sales of ¥100 billion by 2027.
“[Tsuzumi] can achieve high performance with low power consumption,” NTT President Akira Shimada said at a press conference.
NTT plans to design tsuzumi to cope with challenges relating to potential copyright issues that could arise from the data used to train it, the firm said.
Meanwhile, SoftBank Corp. announced plans Tuesday to develop an LLM specializing in the Japanese language by the end of 2024. SoftBank aims to create one of the largest generative AI models in the country, using a supercomputer that was developed with an investment of about ¥20 billion that includes support from the government.
