Science & Technology

Four Chinese AI Models Pass Official Assessment

In China, the first group of four large generative artificial intelligence models has passed an official assessment of compliance with standards.

Four Chinese AI Models Pass Official Assessment

The media reports that the mentioned procedure, which was handled by machine intelligence configurations, paves the way for faster and better regulation of the development of the industry of content created using digital thinking systems in the space of the world’s second-largest economy.

Also, the assessment of compliance with standards will allow the creation of a catalog of requirements for the artificial intelligence technology sector.

Large models of machine intelligence that have passed the state assessment have been developed by 360 Group, Baidu, Tencent, and Alibaba. The assessment was initiated by the China Electronics Standardization Institute under the Ministry of Industry and Information Technology. In this case, the emphasis is on the study of multi-domain and multidimensional evaluation of artificial intelligence models. Factors such as generality, intelligence, and safety are used as evaluation criteria. Language, speech, and visual content are also taken into account.

Baidu reported that its chatbot is based on artificial intelligence called Ernie as of November this year had a user base of more than 70 million people. The interaction of consumers with this digital product takes place based on more than 4,000 scenarios.

The current year was a period of active development for Ernie. In May, the digital product was updated to version 3.5. In October, the development’s version 4.0 appeared. The chatbot has achieved significant success in understanding information, generating original content based on consumer commands, and the ability to carry out mental activity in the paradigm of human logic and memory.

Alibaba said its Tongyi Qianwen artificial intelligence model is the only open-source digital product among the AI configurations assested in December. The company notes that its system of machine intelligence has a wide user base among developers and corporate clients around the world. Alibaba noted that Tongyi Qianwen has been subjected to close public scrutiny in terms of productivity and security.

Pan Helin, a professor at Zhejiang University’s International Business School, said in a media commentary that the national assessment has raised the threshold for large artificial intelligence models. The expert believes that the earlier intervention of government regulation will contribute to the growth of the AI industry. Pan Helin says that it is better to define standards for sphere activities at the initial stage than to destroy and reshap it later.

Currently, Chinese companies are striving to launch large artificial intelligence models as quickly as possible to compete with OpenAI ChatGPT and GPT-4. Baidu, SenseTime, and startups such as Baichuan Intelligent Technology and Zhipu AI have introduced their chatbots with digital intelligence.

Chinese media reports that currently there are more than 130 large language models in this country, which is a 40 percent share of the global volume of these developments.

Zhou Hongyi, founder and chairman of the board of 360 Security Technology, said in a comment to the media that in the next stage of global competition in the sphere of artificial intelligence, the use of technology in real-world scenarios will be a particularly important task. He also noted that one of the goals of his company is to increase the level of integration of AI into the online search system. According to him, the artificial intelligence model developed by 360 Security Technology, after training, will formulate a search query based on keywords and interactive questions sent by the consumer. Zhuo Hongyi says that ChatGPT is still 12-18 months ahead of similar Chinese digital products in terms of intelligence development.

The largest artificial intelligence models developed by companies from an Asian country have equaled GPT-3.5 in terms of abilities, and the technical gap with GPT-4 is narrowing. This was stated by Dou Dejing, associate professor at the Faculty of Electronic Engineering at Tsinghua University.

As we have reported earlier, China Softens Stance on Gaming.

Serhii Mikhailov

2212 Posts 0 Comments

Serhii’s track record of study and work spans six years at the Faculty of Philology and eight years in the media, during which he has developed a deep understanding of various aspects of the industry and honed his writing skills; his areas of expertise include fintech, payments, cryptocurrency, and financial services, and he is constantly keeping a close eye on the latest developments and innovations in these fields, as he believes that they will have a significant impact on the future direction of the economy as a whole.