【观点】中国工业经济学会会长、中国社科院大学教授江小涓：数据、算力与智能时代的创新范式

近日，中国工业经济学会会长、中国社科院大学教授江小涓在2023世界计算大会上表示，数字时代，数据和算力成为重要创新要素，科技与产业内在一体化程度增强，大型数字企业在创新中地位上升，开放开源成为创新资源配置的重要组织形态，推动科技创新的思路和组织科技项目的逻辑都要应时而变。

科技创新在不同时代有不同特点，江小涓称，数字时代的科技创新有重要变化，数据、算力等成为创新的重要资源，重新定义了创新各主体间的关系。观察分析这种变化，构建新的创新范式分析框架，既满足学术理论研究的需要，又具有重要的现实意义。 CXO UNION-CXO联盟（cxounion.cn）

长期以来，受市场环境封闭、通信技术受限、信息生成及传递成本高等因素影响，科学、技术与产业缺乏互动。随着通信技术和国际贸易发展，以电子信息为载体的创新资源和产业需求信息生成、传递、反馈能力提升，科学、技术与产业联动增强。江小涓称，互联网时代，信息搜索数量、广度和深度突破距离约束，双向信息流动频繁，科学、技术与产业间信息共享和交互能力极大增强。数字时代，原始信息伴随着科学技术和产业活动同步产生，科学、技术与产业呈现融合趋势。

江小涓强调，科技成果向产业应用的转化问题仍是一个突出问题。2022年，国家知识产权局发布的《中国专利调查报告》显示，我国有效发明专利产业化率为36.7%，其中，高校发明专利产业化率为3.9%，科研单位发明专利产业化率为13.3%，企业发明专利产业化率为48.1%。

江小涓指出，企业已经成为人工智能技术前沿创新的核心主体。2023年4月，美国斯坦福大学的研究人员以及谷歌、Anthropic和Hugging Face等人工智能公司编制的《2023人工智能指数报告》发布。报告指出，多年来，学术界在开发最先进的AI系统方面处于领先地位，但现在已被工业界牢牢接管。尖端人工智能研究需要大量资源，使其超出了学术界的能力范围。数据显示，从2002到2014年，大多数重要模型都是由学术界发布。而到2022年，32个重要的机器学习模型都诞生在工业界，而学术界仅有3个。 CXO UNION-CXO联盟（cxounion.cn）

在江小涓看来，大型数字企业生产和汇聚海量数据，能准确感知市场需求和应用场景。同时大型平台有能力大规模投入，组建研发簇群，研发内容可以覆盖产业生态所需的全链全域技术体系，并直接应用于自身庞大生态圈。这种模式下，创新、产业和用户一体化，不存在所谓的“科技成果转换”问题。

数字时代，大型数字企业具有巨量链接和海量数据获取、处理与迭代能力，是数字前沿技术的重要创新力量。在自动驾驶、云计算、渲染引擎、虚拟现实等前沿数字技术领域，国内发明专利企业占比超80%，国际发明专利排名前20位的均为企业，明显高出一些典型的传统领域，大型数字企业走向创新前沿。江小涓指出，国内平台企业不够强大不够活跃，调动它们的积极性是数智时代创新的关键。国家创新体系布局多年，高校汇聚较多高端人才和较多财源，原始创新力量较强；国家级实验室及各类新型科研机构各具优势；它们与企业的更好协同仍将发挥重要作用。

开源技术源于软件，指源代码向公众开放的软件技术。数字时代开源成为一种创新理念与文化形式，指共创共享的技术创新。江小涓认为，开源开放能汇聚更多数据和分布式创新投入，成为重要的创新组织形态。开源能够汇聚众智、多方协同，获得透明高效的海量数据、大量自动化协作工具、世界范围内智慧资源的分布式协作和接力式开发，推动技术持续迭代演进和大范围连结产品、企业和产业，构建大规模生产和应用场景。 CXO UNION-CXO联盟（cxounion.cn）

翻译：

Jiang Xiaojuan, President of the Chinese Society of Industrial Economics and Professor of the University of Chinese Academy of Social Sciences: Innovation paradigm in the era of data, computing power and intelligence

Recently, Jiang Xiaojuan, president of the Chinese Society of Industrial Economics and a university professor of the Chinese Academy of Social Sciences, said at the 2023 World Computing Conference that in the digital era, data and computing power have become important elements of innovation, the internal integration of science and technology and industry has been enhanced, the status of large digital enterprises has risen in innovation, and open source has become an important organizational form of innovation resource allocation. The idea of promoting scientific and technological innovation and the logic of organizing scientific and technological projects should change with the time.

Scientific and technological innovation has different characteristics in different times. Jiang Xiaojuan said that scientific and technological innovation in the digital era has undergone important changes, and data and computing power have become important resources for innovation, redefining the relationship between various subjects of innovation. To observe and analyze this change and construct a new analytical framework of innovation paradigm not only meets the needs of academic theoretical research, but also has important practical significance. CXO UNION-CXO联盟（cxounion.cn）

For a long time, due to the closed market environment, limited communication technology, and high cost of information generation and transmission, science, technology and industry lack of interaction. With the development of communication technology and international trade, electronic information as the carrier of innovative resources and industrial demand information generation, transmission, feedback ability has been improved, science, technology and industry linkage increased. Jiang Xiaojuan said that in the Internet era, the number, breadth and depth of information search break through the distance constraints, two-way information flow is frequent, and the information sharing and interaction ability between science, technology and industry is greatly enhanced. In the digital age, original information is produced simultaneously with science and technology and industrial activities, and science, technology and industry show a trend of integration. CXO UNION-CXO联盟（cxounion.cn）

Jiang Xiaojuan stressed that the transformation of scientific and technological achievements into industrial applications is still a prominent problem. In 2022, the “China Patent Survey Report” issued by the State Intellectual Property Office shows that the industrialization rate of effective invention patents in China is 36.7%, of which the industrialization rate of invention patents in universities is 3.9%, the industrialization rate of invention patents in scientific research institutions is 13.3%, and the industrialization rate of invention patents in enterprises is 48.1%.

Jiang Xiaojuan pointed out that enterprises have become the core subject of cutting-edge innovation in artificial intelligence technology. In April 2023, the “2023 Artificial Intelligence Index Report” compiled by researchers at Stanford University and AI companies such as Google, Anthropic and Hugging Face was released. For many years, academia led the way in developing state-of-the-art AI systems, but industry has now firmly taken over. Cutting-edge AI research requires significant resources, putting it beyond the reach of academia. The data shows that from 2002 to 2014, most of the important models were published by academia. By 2022, 32 important machine learning models will have been created in industry, compared to just three in academia. CXO UNION-CXO联盟（cxounion.cn）

In Jiang Xiaojuan’s view, large digital enterprises produce and gather massive data, and can accurately perceive market demand and application scenarios. At the same time, the large platform has the ability to invest on a large scale, set up research and development clusters, and the research and development content can cover the full-chain and full-domain technical system required by the industrial ecology, and directly apply to its own huge ecosystem. Under this model, innovation, industry and users are integrated, and there is no so-called “transformation of scientific and technological achievements” problem. CXO UNION-CXO联盟（cxounion.cn）

In the digital age, large digital enterprises have the ability of acquiring, processing and iterating massive links and massive data, which is an important innovation force of digital frontier technology. In the field of cutting-edge digital technology such as autonomous driving, cloud computing, rendering engines, and virtual reality, domestic invention patent enterprises accounted for more than 80%, and the top 20 international invention patents were all enterprises, significantly higher than some typical traditional fields, and large digital enterprises moved to the forefront of innovation. Jiang Xiaojuan pointed out that domestic platform enterprises are not strong enough and not active enough, and mobilizing their enthusiasm is the key to innovation in the era of digital intelligence. The national innovation system has been laid out for many years, and colleges and universities have gathered more high-end talents and more financial resources, and the original innovation force is strong. National laboratories and various new scientific research institutions have their own advantages; Better synergy between them and companies will still play an important role.

Open source technology is derived from software, which refers to software technology whose source code is open to the public. In the digital age, open source has become an innovative concept and cultural form, which refers to the technological innovation of co-creation and sharing. Jiang Xiaojuan believes that open source can gather more data and distributed innovation investment, and become an important innovation organization form. Open source can gather wisdom and multi-party collaboration, obtain transparent and efficient mass data, a large number of automated collaboration tools, distributed collaboration and relay development of worldwide intelligent resources, promote the continuous iterative evolution of technology and connect products, enterprises and industries in a wide range, and build large-scale production and application scenarios. CXO UNION-CXO联盟（cxounion.cn）

由CXO UNION-CXO联盟（cxounion.cn）转载而成，来源于中国计算机报，作者本报编辑部；编辑/翻译：CXO UNIONCXO联盟小U。

如需加入CXO UNION（CXO联盟）高管社群，请联系社群小伙伴哦~

免责声明: 本网站(http://www.cxounion.cn/)内容主要来自原创、合作媒体供稿和第三方投稿，凡在本网站出现的信息，均仅供参考。本网站将尽力确保所提供信息的准确性及可靠性，但不保证有关资料的准确性及可靠性，读者在使用前请进一步核实，并对任何自主决定的行为负责。本网站对有关资料所引致的错误、不确或遗漏，概不负任何法律责任。

本网站刊载的所有内容(包括但不仅限文字、图片、LOGO、音频、视频、软件、程序等) 版权归原作者所有。任何单位或个人认为本网站中的内容可能涉嫌侵犯其知识产权或存在不实内容时，请及时通知本站，予以删除。