ANN/JAPAN TIMES – With generative artificial intelligence (AI) rapidly gaining traction around the world, Japanese firms ranging from SoftBank to Hitachi are developing or incorporating the technology into their businesses. At the same time, the government is working toward crafting a national AI strategy.
Following the public debut of Microsoft-backed OpenAI’s ChatGPT last November, global tech giants such as California-based Google and China’s Baidu have rolled out their own AI-powered chatbots, but Japanese firms have been conspicuous in their absence.
Still, the nation’s companies are beginning to make their presence felt, with SoftBank’s mobile unit declaring earlier this month that it will develop a Japanese equivalent of ChatGPT.
Line, which is Japan’s most popular messaging app provider and is co-owned by SoftBank and South Korea-based Naver, has been developing AI, with the parent company intending to invest more in this technology, SoftBank CEO Junichi Miyakawa said during a news conference last week.
“The SoftBank and Line alliance has been discussing what we need to do to catch up with OpenAI,” he said.
Miyakawa said SoftBank launched a company in March that will be developing and seeking ways to tap into generative AI applications, with about 1,000 engineers already selected for this task.
Given that Line already has its own large language model (LLM) AI, “we have no choice but to challenge. It’s not a matter of whether we can beat Chinese or United States (US)firms”, Miyakawa said.
Chatbot tools such as ChatGPT are powered by LLMs that are capable of generating text and handling a variety of language tasks, such as summarising and translating.
Line and Yahoo Japan operator SoftBank-backed Z Holdings Corp agreed to merge in 2019 with the aim of challenging US and Chinese online behemoths by becoming a “leading AI technology-based” information technology company.
Japanese tech giant NTT also said last week that it plans to develop its own LLM this fiscal year and provide it to other businesses.
Meanwhile, CyberAgent announced on Wednesday that it has released its own LLM with which companies can create AI chatbot tools. “Since most of the existing LLMs are trained with the English language, there are not many models specialised in the Japanese language and culture,” CyberAgent said in a statement.
On top of the speed up in development of generative AI tools by Japanese firms, more firms are also seeking to utilise the technology.
Industrial conglomerate Hitachi said Monday that it has set up an internal body called the Generative AI Center that consists of employees from across various fields such as AI, data science, security and law.
The centre aims to facilitate the use of generative AI to improve the productivity of Hitachi’s employees. The firm also plans to provide consulting services on AI for other companies.
It’s not just the private sector that is ramping up efforts to ride the AI wave, as the government is also taking action.
Last week, the administration of Prime Minister Fumio Kishida launched a new panel of experts to draft a national AI strategy. The panel, chaired by the University of Tokyo’s renowned AI expert Yutaka Matsuo, will discuss the use and development of AI, as well as
its risks. Kishida himself seems to be keen to boost Japan’s AI competitiveness, with the prime minister inviting several leading AI experts and young business operators to his office last week to talk about the technology.
The prime minister asked them about Japan’s strengths and weaknesses in the field, as well as how they think AI will impact society, according to a government official who attended the meeting and briefed reporters.
The participants told Kishida that because Japan’s population is expected to keep declining, the country has a strong incentive to take advantage of AI to enhance productivity.
They also said that in order to speed up the development and use of the new technology, the public, private and academic sectors all need to work closely together.
Bolstering the development and use of AI among Japanese firms may be major priorities, but the government will also have to deal with risks, such as AI potentially eliminating jobs and spreading fake news.
It is still unclear how Japan will get a handle of the risks and whether it will take a strict or soft stance on regulations. Other countries and regions will likely take different approaches, with a recent move by the European Union (EU) to regulate AI coming under the spotlight.
The EU is taking steps to introduce an AI bill to rein in generative AI tools using a risk-based approach. The proposed law categorises AI into four levels of risk, with those that are deemed an “unacceptable level of risk” -including subliminal or manipulative systems as well as social-scoring platforms based on people’s behaviour and social and economic backgrounds – being prohibited.
How to govern the use of AI is a key agenda item at the leaders’ summit of Group of Seven nations in Hiroshima that kicked off yesterday, with Japan looking to lead the discussion.