China News Network Guangzhou, May 5 (Reporter Cai Minjie) Relying on digital human technology, voice generation technology, and plot script creation technology, the production of digital human live broadcast room can be completed in 30 minutes; It can support free input from users, and AI can make text into beautiful paintings in seconds with just one sentence...

At present, the development of artificial intelligence has made it possible for general intelligence to be empowered by all walks of life and thousands of households. In the recently held 2023 Zhongguancun Forum, 2023 Vientiane Conference, 2023 Xiaomanwaist Science and Technology Conference and other forums and conferences, the big model has become a hot word for discussion. As the hottest AI subdivision field in recent years, large models have many advantages such as multi-scenario commonality and large-scale replication, and are regarded as an important research direction for the realization of general artificial intelligence.

"The big language model will further have an impact on natural science and technology, economic form and operation, social structure and other fields." Qiao Weibing, president and editor-in-chief of China Translation Publishing House, recently said that previously, although the machine can efficiently capture and reorganize information from the physical environment, it is difficult to analyze information autonomously, and the big language model has achieved a breakthrough in this point, solving complex natural language processing, computer vision and language recognition and other tasks, "In the future, the development of the big language model will emerge more new functions closely related to its technical characteristics, which will be able to better understand and process different types of data, so as to achieve a wider range of applications."

According to the "Research Report on the Map of Chinese Intelligent Large Model", domestic general-purpose large models are continuing to expand their application fields and create a cross-industry generalized artificial intelligence capability platform. At the same time, professional large models in vertical fields are also deepening, and a number of large models for vertical fields such as biopharmaceuticals, remote sensing, and meteorology give full play to their in-depth advantages to provide high-quality professional solutions for specific business scenarios.

Since the end of last year, the domestic general-purpose language model represented by Wen Xin Yiyan and MOSS has led a new round of changes in the field of information technology. He Junjie, senior vice president of Baidu Group, recently proposed to reconstruct the mobile ecology with AI native thinking, "refreshing" the user experience of search, library, e-commerce and other products and services, he said that the previous big language model is to make AI "big", and this reconstruction of the mobile ecology is to strive to make AI "small", more specific and available to everyone.

"We must let AI go down the 'altar' of technology and go deep into the 'soil' of application." He Junjie said that this is the only way for all information technology to eventually become inclusive. The reconstruction of Baidu's mobile ecosystem by AI native thinking gives AI the opportunity to become a general productivity that is available to everyone and different for everyone, from "AI sees the same" to "AI sees differently", bringing personalized experience to everyone. For example, in terms of search experience, "AI partners" can realize intelligent dialogue and multi-round interaction, reducing the threshold and cost of user input expression.

In view of the future development of large models, Zhao Zhiyun, director of the China Institute of Science and Technology Information, said that first of all, it is necessary to strengthen the coordination of resources and R&D forces to promote the orderly development of large models; Secondly, we should accelerate basic research and technological innovation, and enhance the influence of academia and open source; Thirdly, it is necessary to strengthen the scene traction role in the development of large models and create large model benchmarking projects; Finally, we must strengthen international cooperation and actively participate in global AI governance. (End)