People's Daily Online, Beijing, December 18 - On December 16, the National Key Laboratory of Communication Content Cognition, which is in charge of the People's Daily Online, and the People's Daily Science and Technology Company jointly hosted the "Opportunities and Challenges in the Era of Intelligent Internet" large-scale model exchange meeting in Beijing. The conference brought together many artificial intelligence industry experts, founders of investment institutions and representatives of more than 30 large-scale model enterprises, making full use of large-scale model technology to promote the high-quality development of the industry and industry, while ensuring the sustainable and safe development of large-scale model technology.
The relevant person in charge of the National Key Laboratory of Communication Content Cognition introduced the AI capabilities of People's Daily Online in the fields of government services, spiritual education, safety communication, and health science popularization, and shared a series of artificial intelligence scientific and technological achievements such as the government-related intelligent audit platform "Zhixiaozhu" and the deep synthetic content detection platform "Tianmu" launched by the laboratory. Daniel Zhang, chief scientist of the laboratory, said that People's Daily has mobilized more than 1,000 editors and reporters since April to "work hard" and initially built a mainstream value corpus of professional artificial Q&A, which can help the large model greatly improve the level of value alignment after integration and verification with the mainstream large model.
Huang Minlie, deputy director of the Basic Model Center of Tsinghua University, and Che Wanxiang, deputy director of the Institute of Artificial Intelligence of Harbin Institute of Technology, expounded their unique insights and views on the ability evaluation of large language models, AI security in the era of large language models, and the development opportunities and risks and challenges of artificial intelligence. You Peng, President of HUAWEI CLOUD Big Data and Artificial Intelligence, Zhang Ji, Vice President of Alibaba Cloud Tongyi Lab, Zhang Fan, Chief Operating Officer of Zhipu AI, Yang Xinhui, Vice President of Thunderda, Wu Shihong, Partner of Meihua Venture Capital, and many representatives of enterprises in the field of large models shared their latest thoughts on artificial intelligence, research results, and development prospects for large models in 2024.
During the meeting, the "Large Model Responsibility Alliance" was announced under the guidance of the Science and Technology Ethics Committee of the National Key Laboratory of Communication Content Cognition, and jointly initiated by People's Daily Science and Technology Company and large model enterprises. There are 30 members of the alliance in the first batch, including Huawei Cloud, Alibaba Cloud, Kuaishou, Zhipu AI, China Unicom, China Telecom AI, Thunderda, Daguan Data, Kingsoft Office, Zhongke Wenge, Kunlun Wanwei, Yunzhisheng, TravelSky Mobile Technology, iSoftStone, Hang Seng Juyuan, Facing Wall Intelligence, Transwarp Technology, New One Technology, ArcherMind Technology, Yuncong Technology, Chayi Technology, China Industrial Internet, Turing Robot, Fourth Paradigm, Xiling Force, Zhongke Vision, Beike Housing, Hualong Technology, Shuhui Space, Xingxing AI.
The alliance will cooperate in the implementation of scientific research projects, promote industry exchanges, promote the construction of standards, organize the cultivation of talents, safeguard public interests and moral standards in the process of R&D, application and promotion of large models, and build a responsible and reliable healthy development ecology of large models.
At present, People's Daily Online Technology Co., Ltd. has reached in-depth cooperation with HUAWEI CLOUD and Zhipu AI on mainstream value corpus and co-construction of large models. In 2024, we will carry out diversified cooperation with more outstanding upstream and downstream enterprises in the field of large models.