Japan embraces ChatGPT in controversy and is unwilling to miss out on the AI craze again! From Government to Enterprise AI | Japan | Enterprise
After 41 days of trial operation, Yokosuka City in Kanagawa Prefecture, Japan decided to be the first person to eat crabs.
According to Japanese media reports, the Kanagawa Prefectural Government in Japan took the lead in using ChatGPT in administrative work, becoming the first government department in Japan to use ChatGPT.
In the previous trial operation, about 4000 civil servants of the Yokosuka City Government attempted to use ChatGPT for official duties such as "meeting minutes" and "policy formulation" to verify ChatGPT's ability in administrative work.
Yokosuka Mayor Kamei Ueda said, "It is the first time in Japan that ChatGPT has been fully introduced to local governments for official duties... The introduction of artificial intelligence in civil work is very important and can effectively reduce labor costs. Now we finally have this excellent tool."
Not only Kanagawa Prefecture, but also Hota City in Saitama Prefecture established a research group this month to study the potential use of ChatGPT in official business. Unlike the strong regulation advocated in Europe, in Japan, from the central government to local governments and businesses, there has been a strong interest in ChatGPT.
Regarding this, Zhu Qiaochu, an assistant researcher at the Japan Research Center at Shanghai Jiao Tong University, who has been paying attention to the development of artificial intelligence in Japan and related legal protection issues, told First Financial that there is a significant difference in attitude towards generative AI between Japan, the United States, and Europe. "Unlike the strong regulatory attitude in Europe, Japan and the United States hope to guide AI technology to develop along a people-centered path through government guidance and other 'soft law governance' methods."
"The ChatGPT party has started"
I asked ChatGPT, who is Taro Kono? The answer is actually the Prime Minister of Japan. Previously, Japan's Minister of Digital Transformation, Taro Kono, shared this awkward moment in a media interview, saying, "It gave the wrong answer, so you need to be careful." Despite admitting the mistake to the Japanese Prime Minister, Japan's enthusiasm for artificial intelligence has not diminished.
"We are very optimistic about ChatGPT," SoftBank CEO Runichi Miyagawa said earlier. "These days, most of our meetings have been related to ChatGPT topics. The ChatGPT party has begun."
SoftBank, a telecommunications subsidiary of SoftBank Group, announced that it will join the global competition to create the ChatGPT version. According to Japanese media reports, the company established a new entity in March and selected approximately 1000 people to develop the Japanese version of the artificial intelligence chatbot ChatGPT under OpenAI.
Not only SoftBank, but Japanese manufacturing giant Hitachi also announced in mid May that it will develop its own generative AI and combine its technology with existing overseas generative AI such as ChatGPT for use in its own business. Hitachi's General Manager of Data and Design, Junichi Yoshida, stated at a briefing in mid May that "we hope to accelerate the utilization of generated AI by combining various knowledge while avoiding risks." In addition, Japanese communication companies such as NTT are also entering the field of artificial intelligence.
Japanese Prime Minister Fumio Kishida himself has a strong interest in ChatGPT. Recently, he has ordered the establishment of a new artificial intelligence team within the Japanese government to investigate the economic potential of artificial intelligence technology and how to address its risks. In mid April, OpenAI founder Altman visited Japan and was met by Kishida to discuss establishing a research base for ChatGPT in Japan and correcting the systematic bias of ChatGPT based on English world learning.
Among the civil servants who participated in the test in Kanagawa Prefecture, 10.9% reported a significant improvement in work efficiency, and 71.6% reported an improvement. In addition, 5.8% reported using it every day. This test has accumulated over 26000 cases of ChatGPT usage, mainly applied in scenarios such as information search, copywriting, summarization, proofreading, and recommendation of new ideas.
Oligogenesis at an advanced age is key
Zhu Qiaochu is not surprised by Japan's top-down embrace of generative AI. She told First Financial that in fact, this is closely related to the long-standing problem of Japanese society - the aging population with fewer children.
She gave an example that as early as 2016, the Abe Cabinet of Japan proposed the Social 5.0 Development Plan and made the development of people-oriented artificial intelligence technology an important part of the plan, hoping to solve Japan's labor shortage and healthcare challenges in an aging society through artificial intelligence technology. Subsequently, under the leadership of the cabinet, various departments such as the Ministry of General Affairs, the Ministry of Economy and Industry, and the Ministry of Health, Labour, and Welfare carried out a step-by-step policy promotion along the direction of social 5.0 development.
The latest data from the Ministry of General Affairs of Japan shows that as of April 1, 2023, the number of children under the age of 15 in Japan was 14.35 million, a decrease of 300000 from the previous year and a 42 year decline in a row. Faced with the severe reality of Japan's population, Kishida once said, "The next 6 to 7 years are the last opportunity to reverse the declining birth rate." He warned that a rapid decline in the proportion of young people would lead to economic contraction and affect social security.
Zhu Qiaochu said that after the outbreak of the COVID-19, Japan, from the government to the people, was deeply aware that its conservative attitude towards information technology had led to its backwardness and shortcomings in this field, thus accelerating the pace of the Japanese government in promoting a digital society. In 2021, Japan introduced the Basic Law on the Formation of a Digital Society and established a Digital Department specifically responsible for coordinating the work of various government departments, promoting digital reform at the administrative level in Japan, and formulating key areas for promoting the digital society.
"In addition, Japan also sees the digital industry as an opportunity to help it regain its position as a world power and regain its voice in the international community," said Zhu Qiaochu.
"The development of generative AI based on big data training and the free flow of cross-border data advocated by DFFT have logical coherence. Therefore, Japan has sufficient motivation to support the development of generative AI to further promote consensus among countries on establishing DFFT mechanisms." She said, "At the previous G7 meeting, Kishida directly expressed his willingness to contribute funds to promote the cross-border flow of data, in order to support data training as the cornerstone of AI technology development."
Can slow paced Japan keep up with this wave of AI?
Of course, while Japan's public and private sectors have shown a positive attitude towards the introduction of ChatGPT, Zhu Qiaochu stated that there are concerns or doubts from various sectors in Japan. For example, generative AI generated content can easily breed risks of copyright infringement and personal privacy, and may be used to create false information to disrupt public opinion. For example, Shinichi Maki, a representative of the Japanese Reform Party, previously stated that "ChatGPT is still under development and has security risks. There are still many issues that need to be solved, and it is not mature enough to be used by government agencies.".
In Zhu Qiaochu's view, "Europe has always adhered to a strong regulatory posture in the regulation of personal information processing, strictly regulating the cross-border flow of data. In order to meet the sufficiency recognition conditions of the General Data Protection Regulations, Japan has made several revisions to its domestic Personal Information Protection Law in order to obtain an evaluation of 'having the same level of protection as the European Union in personal information protection'. Finally, in 2019, it became one of the 14 sufficiency recognition jurisdictions of the European Union. Therefore, even if Japan intends to relax the regulatory intensity of data circulation to support the research and application of ChatGPT, it will not overly weaken the regulatory intensity of personal information processing activities."
Faced with various doubts, Japanese Minister of General Affairs Masayoshi Matsumoto stated, "Although ChatGPT has rumors of insufficient privacy protection and inaccurate answers to questions, these issues should not be used as reasons to stop using ChatGPT. The administrative department must make appropriate judgments and rule making based on the situation."
Zhu Qiaochu stated that Japan began using computers to carry out national administrative processing work as early as 1959. In the 1960s, local regulations were adopted to regulate computer data processing activities. In 1970, the academic community launched a large-scale discussion on the institutionalization of personal information protection. "However, when discussing personal information, the context often focused on how to strengthen the protection of personal information to limit the government's improper use of personal information. In the 1980s, Japan was generally sensitive and conservative towards issues such as personal information protection and privacy leakage. Around 2000, there were constant reports of improper handling of personal information by private enterprises," she said.
"In such an atmosphere where both the private sector and the government emphasize the strong supervision of personal information, it is naturally difficult to develop AI technology based on data and information, which also led to Japan missing the Internet development boom around the millennium to some extent." Zhu Qiaochu said.
Nowadays, the Japanese government is also trying to change and promote the use of personal information at the public level. Taking the Japan Digital Agency as an example, the first task that has been vigorously promoted since its establishment in 2021 is to popularize and promote the implementation and implementation of personal data systems.
The latest news shows that Japanese researchers will use the local "Fuyue" supercomputer to develop a chatGPT style generative artificial intelligence big language model centered on Japanese, and plan to provide this technology to domestic Japanese companies free of charge in the next fiscal year.