Understand questions and provide quick answers; Train your voice and cover classic songs; According to the description, create interesting paintings... Recently, generative artificial intelligence based on large models has demonstrated excellent abilities in language comprehension and content generation, attracting social attention. Empowered by large models, generative artificial intelligence is triggering a new wave of intelligence. Thanks to its massive data, parameters, and good learning ability, large models enhance the universality of artificial intelligence. From chatting smoothly with people to writing contracts and scripts, from detecting program security vulnerabilities to assisting in game creation and even movies... the evolution of generative artificial intelligence skills is accelerating. With technological iteration, more efficient and "smart" large models will penetrate into more and more fields, and are expected to become a new foundation for artificial intelligence technology and applications, a fundamental tool for people's production and life, and thus bring about profound changes in economic and social development and industry. The strong innovative potential of artificial intelligence models has made them one of the focal points of global competition. After multiple efforts, China's artificial intelligence large-scale model has laid a certain foundation. The "Research Report on China's Artificial Intelligence Big Model Map" released at the 2023 Zhongguancun Forum shows that China's artificial intelligence big models are showing a thriving development trend. According to incomplete statistics, as of now, 79 large models with parameter scales above 1 billion have been released nationwide. China has established a systematic research and development capability covering theoretical methods and software and hardware technologies in the field of large models. It should also be noted that the large model of artificial intelligence cannot be separated from the integration and innovation of multiple technologies. There is still a gap between China and the international advanced level in cutting-edge basic theories and algorithms. To firmly establish the foundation of the intelligent era, it is necessary to focus on the shortcomings and promote original technological breakthroughs in generative algorithms, frameworks, and other fields in the field of large models. At the same time, we should also leverage the advantages of our country's application scenarios, further deepen our efforts in vertical fields, and build professional large models in fields such as finance, healthcare, and electricity based on industry-specific training datasets. We need to optimize with high-quality applications and data feedback technology to help iterate and upgrade large models. The quality of data affects the intelligence of large models. The reason why some big models are leading in the world is closely related to the training of a large number of open high-quality data. China has a vast amount of data and rich application scenarios. We should gradually open up and share high-quality data, and promote the orderly opening of data classification and classification through the development of sharing directories and rules, so as to stimulate innovation vitality through mobile data. For example, the recently issued Action Plan for Accelerating the High Level Application of Artificial Intelligence in Shenzhen (2023-2024) proposes to "establish a multimodal public dataset and create high-quality Chinese corpus data". It is expected that all industries around the country will strengthen the supply of high-quality data from the reality to provide sufficient "food" for the growth of the big model. The research and development cycle of artificial intelligence large models is long, with high investment and high risk. After several years of continuous research and development, the internationally leading large model has gathered good resources and talents. At present, many colleges and universities and enterprises in China are doing R&D work, focusing on big models, Big data, big computing power and other aspects, and the R&D force is relatively decentralized. As a catcher, it is necessary to further strengthen the dominant position of enterprises in technological innovation,