Lin Huizhi Shenzhen Special Commissioner
Scholars of interviewees believe that the official strengthening supervision is expected, but the content of the draft is similar to that of China's existing regulations on other Internet services, so as not to curb the overall development of the field.
The Chinese Internet giant Alibaba launched the same day similar to the ChatGPT chat robot.The core values of socialism do not contain information that subverts the state power.
Interviewed scholars believe that the official strengthening supervision is expected, but the content of the draft is similar to that of China's existing regulations on other Internet services, so as not to curb the overall development of the field.
ChatGPT developed by OPENAI in the United States has set off a huge wave in the Chinese technology circle after the market in November last year. Many technology companies have launched services to target chat robots since March this year.In order to strengthen the supervision of such technologies, the China National Internet Information Office issued a draft draft of the management of artificial intelligence services on Tuesday (April 11) to publicly solicit feedback from the public.
The draft shows that China supports the independent innovation of basic technologies such as artificial intelligence algorithms and frameworks.However, before providing the people with genetic artificial intelligence services, enterprises must declare a security assessment to the online information department; the content of the production should reflect the core values of socialism.Promote terrorist extremism, national hatred, may disrupt economic and social order.
The draft also stipulates that enterprises must also ensure that the content is true and accurate, and take measures to prevent false information; prohibit illegal acquisition, disclosure, and use personal information and privacy and business secrets.Users must provide real identity information in accordance with the law.
In addition to China, including the United States, Germany, Canada, Japan, South Korea and other countries, it has also stated that it is necessary to strengthen the supervision of artificial intelligence tools such as ChatGPT.In early April, Italy banned ChatGPT on the grounds of suspected violation of privacy protection. It was the first Western country to ban this chat robot.
Just a few hours before the release of the above draft, Alibaba officially launched a large -scale language model at the 2023 Alibaba Cloud Summit in Beijing in Beijing.In a recorded video, Tongyi Qianwen showed how it responded to human instructions, writing emails, writing code, generating recipes, and answering mathematical questions.
Zhang Yong, chairman and CEO of Alibaba Group, announced at the event that all products under the company will access the Tongyi Qianqi model in the future, including Taobao, Tmall, Gaode Map, Youku, etc.He said that generating artificial intelligence and cloud computing are a technical watershed, and in the era of intelligence, all products are worthy of redemption with large models.
Alibaba Hong Kong stocks closed at HK $ 99.30 on Tuesday, with 1.59%.
Lin Dawei, a senior researcher at the East Asian Research Institute of the National University of Singapore, said in an interview with Lianhe Zaobao that the relevant regulation was not surprising given by China's central management and centralized system.However, these political characteristics also mean that when China wants to access the actual application of chat robots, the execution speed will be much faster than that of Western countries.
Zhu Feida, an associate professor at the School of Information Systems of Singapore's University of Management, also agreed that supervision was expected, but considering that China has always responded to data supervision, the speed and efficiency of this control measures were surprising.He judged that this may be related to the fierce competition between China and the United States and the phenomenon of ChatGPT in China.
Zhu Feida explained that China felt the leading advantage of the United States in large language models and promoted the vigorous development of the field and the investment in a large number of resources.He said that the Chinese government has formulated corresponding control measures in a short time to reflect that officials are also paying attention to the development of this area. These norms are actually pointing for enterprises and guiding them to invest in the direction of resources.
As for whether supervision will hinder the development of artificial intelligence, Zhu Feida disagreed that the content of the draft is similar to the existing requirements of the official Internet industry and online media.For disadvantages.
