Several Reflections on the Development of Artificial Intelligence

In recent years, artificial intelligence, represented by large models, has ushered in a new wave of development. It is penetrating every aspect of economic and social life at an unprecedented speed, drawing close to each person’s life and work, and profoundly reshaping the productive forces and civilizational form of human society. General Secretary Xi Jinping has pointed out that AI is a key driving force for a new round of scientific and technological revolution and industrial transformation, and will have a far-reaching impact on global economic and social development and the progress of human civilization. At present, major countries around the world are all seizing opportunities for AI development. Recently, the State Council issued the “Opinions on Deeply Implementing the ‘AI+’ Action,” which clarifies the overall requirements, development goals, and key directions for implementing the “AI+” action, providing directional guidance for further promoting the extensive and deep integration of AI with various industries and fields of the economy and society. We must deeply comprehend General Secretary Xi Jinping’s important expositions on AI, properly grasp the dialectical relationships between innovation and openness, development and security, government and market, and application and governance, base ourselves on industrial development needs, conform to the trends of governance reform, and forge a path for AI development with Chinese characteristics.

近年来，以大模型为代表的人工智能迎来新的发展浪潮，正以前所未有的速度深入渗透经济社会发展的方方面面，贴近每个人的生活和工作，深刻重构着人类社会生产力和文明形态。习近平总书记指出，人工智能是新一轮科技革命和产业变革的重要驱动力量，将对全球经济社会发展和人类文明进步产生深远影响。当前，全球主要国家都在抢抓人工智能发展机遇。近日，国务院印发《关于深入实施“人工智能+”行动的意见》，明确实施“人工智能+”行动的总体要求、发展目标和重点方向，为进一步推动人工智能与经济社会各行业各领域广泛深度融合提供了方向指引。我们要深刻领会习近平总书记关于人工智能的重要论述，把握好创新与开放、发展与安全、政府与市场、应用与治理的辩证关系，立足产业发展需求，顺应治理变革趋势，走出一条具有中国特色的人工智能发展之路。

The dialectical unity of independent innovation and open cooperation. General Secretary Xi Jinping has emphasized that “independent innovation is innovation in an open environment.” This important thesis provides the fundamental guidance for handling the relationship between AI technology autonomy and globalized collaboration.

自主创新与开放合作的辩证统一。习近平总书记强调，“自主创新是开放环境下的创新”。这一重要论断为处理好人工智能技术自主与全球化协作提供了根本遵循。

Independent innovation is the “root” of AI development. Without the autonomy of core technologies, open cooperation may degenerate into “dependent development,” leaving key areas vulnerable to control by others. Since 2018, in the face of the severe blockade and frenzied suppression of high-tech industries such as AI and chips by the United States, China has consistently adhered to independent innovation, achieving breakthroughs in underlying technology fields such as chip architecture and core algorithms, and initially building an autonomous industrial ecosystem. For example, DeepSeek, through independently optimizing its algorithm architecture, achieved performance levels comparable to international products at a cost far lower than similar products abroad. Another example is that the independent innovation of Chinese enterprises in key links such as LiDAR and high-precision maps has propelled China into the ranks of leading countries in autonomous driving. These technological breakthroughs are inseparable from China’s favorable conditions for independent innovation. The domestic super-large-scale market is a “testing ground” for the implementation and iteration of AI technology. China possesses the world’s largest digital economy market, providing rich scenarios for AI applications. The complete domestic industrial chain system provides effective support for the autonomous development of AI technology.

自主创新是人工智能发展的“根”，没有核心技术的自主化，开放合作就可能沦为“依附式发展”，在关键领域受制于人。2018年以来，面对美国对人工智能和芯片等高技术产业的严密封锁和疯狂打压，我国始终坚持自主创新，在芯片结构、核心算法等底层技术领域实现突破，初步构建了自主化的产业生态。例如，深度求索公司通过自主优化算法架构，用远低于国际同类产品的成本，达到了相近的性能水平。再如，我国企业在激光雷达、高精度地图等关键环节的自主创新，让我国跻身自动驾驶领域领先国家。这些技术的突破离不开我国良好的自主创新条件。国内超大规模市场是人工智能技术落地与迭代的“试验场”，我国拥有全球最大的数字经济市场，为人工智能应用提供了丰富场景；国内完整的产业链体系，为人工智能技术的自主发展提供了有效支撑。

Independent innovation is not “working behind closed doors.” Adhering to open cooperation is necessary to keep pace with the direction of technological development and avoid stagnation or becoming a “technological island.” In the field of basic research, deep learning frameworks contributed by the global open-source community, such as TensorFlow and PyTorch, provide underlying support for China’s AI innovation. China has also relied on the open-source ecosystem of large models to break the monopoly of the closed systems of American technology giants, providing Chinese solutions for global AI development. In terms of high-end talent, the distribution of top global AI researchers is dispersed, requiring the gathering of wisdom through international exchange. AI governance is a global challenge that requires joint efforts by all countries. Against the backdrop of intensified technological competition among major powers, international AI governance presents a complex situation where multipolar competition coexists with fragmented cooperation, and the struggle over technical standards is intertwined with the game of ethical values. The United States uses its technological hegemony to shape global AI leadership, the European Union takes legislative regulation as its guide to consolidate its leading position in global AI governance, while China adheres to an open and cooperative posture, advocating for narrowing the digital divide through “technology sharing and capacity co-building,” and proposes AI governance strategies through international platforms such as the World Artificial Intelligence Conference, providing useful references for global AI governance.

自主创新并非“闭门造车”，坚持开放合作才能紧跟科技发展方向，避免裹足不前或沦为“技术孤岛”。在基础研究领域，全球开源社区贡献的TensorFlow、PyTorch等深度学习框架，为我国人工智能创新提供了底层支撑，我国也依靠大模型的开源生态，打破了美国科技巨头的封闭系统垄断，为全球人工智能发展提供了中国解决方案。在高端人才方面，全球顶尖人工智能研究者分布具有分散性，需要通过国际交流集聚智慧。人工智能治理是全球性的挑战，需要各国携手共进。在大国科技博弈加剧的背景下，国际人工智能治理呈现多极化竞争与碎片化合作并存、技术标准争夺与伦理价值博弈交织的复杂态势。美国利用技术霸权塑造全球人工智能领导力，欧盟以立法监管为导向巩固在人工智能全球治理中的领导地位，我国则坚持以开放合作的姿态，主张通过“技术共享、能力共建”缩小数字鸿沟，并通过世界人工智能大会等国际平台提出人工智能治理方略，为全球人工智能治理提供有益参考。

Practice shows that independent innovation is the foundation of open cooperation, and open cooperation is the driving force for independent innovation. Only by organically combining the two can AI release its maximum value under the premise of being safe and controllable. China’s AI development must adhere to the dialectical unity of independent innovation and open cooperation, supporting the “depth” of technological innovation through the “thickness” of the domestic market, expanding the “height” of technological value through the “breadth” of international cooperation, ensuring that AI not only serves domestic high-quality development but also helps China take the initiative in global competition, striving to form a new pattern of “centered on ourselves, open and win-win.”

实践表明，自主创新是开放合作的根基，开放合作是自主创新的助力，唯有将二者有机结合，才能让人工智能在安全可控的前提下释放最大价值。我国人工智能发展，必须坚持自主创新与开放合作的辩证统一，通过国内市场的“厚度”支撑技术创新的“深度”，通过国际合作的“广度”拓展技术价值的“高度”，确保人工智能既服务于国内高质量发展，又助力我国在全球竞争中占据主动，努力形成“以我为主、开放共赢”的新格局。

The dynamic balance between development and security. Many technologies are “double-edged swords.” A breakthrough in a new technology, while bringing a leap in productive forces, is often accompanied by potential risks. Security is the prerequisite for development, and development is the guarantee of security. Developing AI must follow the laws of technological evolution and firmly guard the bottom line of national security.

发展与安全的动态平衡。很多技术都是“双刃剑”，一项新技术的突破，在带来生产力跃升的同时，往往也伴随着潜在风险。安全是发展的前提，发展是安全的保障。发展人工智能必须遵循技术演进规律，守牢国家安全底线。

In recent years, the widespread application of generative AI has changed production models in many industries, but it has also brought many problems and risks. For example, the explosive innovation of generative AI in the field of content creation may lead to security threats such as deepfakes and information manipulation. In terms of ideology, large models are susceptible to manipulation in links such as data feeding and algorithm training, leading to issues like data poisoning and algorithmic bias. In education and scientific research, the quality of various AI-generated content is uneven. Coursework and academic research completed with the assistance of AI reduce the processes of critical screening and analytical research, potentially producing more false information and academic garbage. In production and operation, the deep application of AI is prone to issues such as data compliance risks, copyright infringement, and leakage of trade secrets.

近年来，生成式人工智能的广泛应用改变了多个行业的生产模式，同时也带来不少问题和风险。例如，生成式人工智能在内容创作领域的爆发式创新，可能引发深度伪造、信息操纵等安全威胁。在意识形态方面，大模型在数据投喂、算法训练等环节易受到操纵，出现污染投毒和算法偏见等问题；在教育科研方面，各种人工智能生成内容质量良莠不齐，人工智能辅助完成的课业和学术研究减少了批判甄别和分析研究过程，会产出更多虚假信息和学术垃圾；在生产经营方面，人工智能的深度应用易出现数据合规风险、版权侵权、商业秘密泄露等问题。

Therefore, AI development must achieve a dynamic balance between safety, controllability, and innovative development, properly handling the relationship between the speed of technological progress and society’s risk-bearing capacity. We must neither blindly “hit the brakes” out of fear of risk, missing the opportunity for AI to revolutionize productive forces, nor blindly “step on the accelerator” in pursuit of innovation, plunging society into the predicament of “technological runaway.” Only by taking safety and controllability as the anchor and innovative development as the engine can we, while safeguarding national security, industrial competitiveness, and social welfare, promote the leap of human civilization to a higher form.

因此，人工智能发展必须做好安全可控与创新发展的动态平衡，把握好技术进步速度与社会风险承受能力的关系，既不能因恐惧风险而一味“踩刹车”，错失人工智能对生产力革命性提升的机遇，也不能因追求创新而只顾“踩油门”，让社会陷入“技术失控”的困境。唯有以安全可控为锚点，以创新发展为引擎，才能在保障国家安全、产业竞争力和社会福祉的同时，推动人类文明向更高阶形态跃迁。

The coordinated synergy of government guidance and market drive. General Secretary Xi Jinping has emphasized that “we must neither use the decisive role of the market in resource allocation to replace or negate the role of the government, nor use the better role of the government to replace or negate the decisive role of the market in resource allocation.” The effective synergy between the government and the market is not only the core mechanism for the efficient operation of a modern economic system but also an important principle that must be followed in the development of cutting-edge fields such as AI.

政府引导与市场驱动的协同发力。习近平总书记强调，“既不能用市场在资源配置中的决定性作用取代甚至否定政府作用，也不能用更好发挥政府作用取代甚至否定使市场在资源配置中起决定性作用”。政府与市场的有效协同，既是现代化经济体系高效运转的核心机制，也是人工智能等前沿领域发展必须遵循的重要原则。

The early funding by the U.S. Defense Advanced Research Projects Agency (DARPA) for the predecessor of the internet, “ARPANET,” and for disruptive technologies like “autonomous driving,” as well as the support from China’s “Science and Technology Innovation 2030 – ‘New Generation Artificial Intelligence’” major project for basic research on brain-computer interfaces and neuromorphic computing, both reflect the government’s complementary role in areas where “the market is unwilling or dares not to act.” When technology enters the stage of application transformation and commercialization, the market can achieve rapid iteration through enterprise competition and user feedback, realizing industrial development, commercial value, and the improvement of productive forces. For example, OpenAI’s ChatGPT large model, from basic research to commercialization, benefited both from early funding by the U.S. National Science Foundation and from Microsoft’s market-oriented investment and scenario implementation, facilitating the formation of a closed loop of “basic research – application transformation – commercial monetization.”

美国国防高级研究计划局对互联网的前身“阿帕网”、“自动驾驶”等颠覆性技术的早期资助，我国“科技创新2030—‘新一代人工智能’”重大项目对脑机接口、类脑计算等基础研究的支持，均体现了政府在“市场不愿做、不敢做”领域的补位作用。当技术进入应用转化和商业化阶段，市场则能通过企业竞争和用户反馈快速迭代，实现产业的发展、商业价值的实现和生产力的提升。例如，OpenAI的ChatGPT大模型从基础研究到商业化，既受益于美国国家科学基金会的早期资助，又依赖微软的市场化投入与场景落地，有利于形成“基础研究—应用转化—商业变现”的闭环。

At the same time, the development of the AI industry requires support from computing power infrastructure, data element resources, standard systems, cross-industry coordination, etc. Relying solely on free market development can easily lead to disorderly competition. This requires the government to play the role of a “coordinator,” guiding industrial development rationally, solving problems such as disorderly development and unfair competition, and maximizing comprehensive benefits at the national level. For example, through the deployment of the “East Data West Computing” project [1]1 The “East Data West Computing” project is a Chinese national strategy initiated around 2022 to channel data processing and computing power demands from the economically developed eastern regions to the resource-rich western regions, aiming to optimize resource allocation and promote regional development., China encourages and guides enterprises to build computing power centers in the western region, which not only solves the problem of tight computing power in the east but also drives the development of the digital economy in the west, achieving a win-win situation for both efficiency and fairness.

同时，人工智能产业发展需要算力基础设施、数据要素资源、标准体系、跨行业协同等支持，单靠市场自由发展易陷入无序竞争。这就需要政府发挥“协调者”的作用，通过合理引导产业发展，解决无序发展和不正当竞争等问题，实现国家层面的综合效益最大化。例如，我国通过部署“东数西算”工程，鼓励和引导企业在西部建设算力中心，既解决了东部算力紧张问题，又带动了西部数字经济发展，实现效率与公平的双赢。

In the era of artificial intelligence, the complexity of technological innovation, the systemic nature of industrial upgrading, and the intensity of global competition all highlight the irreplaceability of this synergy. China must better leverage the synergistic role of government and the market under the socialist market economy system, allowing the “visible hand” and the “invisible hand” to each display their strengths, so as to achieve the safe, orderly, and efficient development of artificial intelligence.

在人工智能时代，技术创新的复杂性、产业升级的系统性、全球竞争的激烈性，更凸显了这种协同的不可替代性。我国更要发挥好社会主义市场经济体制下政府与市场的协同作用，让“看得见的手”和“看不见的手”各展其长，实现人工智能的安全有序高效发展。

The value integration of industrial application and social governance. The rapid advancement of AI industrialization and its extensive empowerment of myriad industries, while demonstrating enormous transformative potential, also bring multiple social challenges, and social governance has increasingly become a focus of concern for many parties. Promoting the value integration of AI social governance and industrial application is a key pathway to ensuring that technological development both serves economic growth and safeguards social equity.

产业应用与社会治理的价值融合。人工智能产业化的快速推进与广泛赋能千行百业，在展现巨大变革潜力的同时，也带来多重社会挑战，社会治理日益成为多方关注的焦点问题。推动人工智能社会治理与产业应用的价值融合，是确保技术发展既服务于经济增长，又保障社会公平的关键路径。

As AI deeply intervenes in human decision-making and production and daily life, issues such as algorithmic bias, data abuse, and job displacement are becoming increasingly prominent. For example, some internet platform enterprises, while using their dominant position to precisely “profile” users, also infringe upon consumers’ legitimate rights and interests through practices such as “price discrimination against regular customers” [2]2 “Price discrimination against regular customers” (杀熟, shā shú): a practice where platforms charge higher prices to long-term or loyal customers compared to new users. and “information cocoons” [3]3 “Information cocoons” (信息茧房, xìnxī jiǎnfáng): a term describing the phenomenon where algorithms feed users only content that aligns with their existing preferences, isolating them from diverse information.. For another example, while AI creates new employment positions, it also replaces many traditional ones. In 2024, the International Monetary Fund pointed out that approximately 40% of global employment positions will be affected by AI technology. Currently, a large number of positions such as telemarketing and customer service have already been replaced by AI. Some autonomous driving platforms have triggered employment anxiety among ride-hailing and taxi drivers, as well as public concerns about driving safety, bringing about public opinion and social risks.

随着人工智能深度介入人类决策及生产生活，算法偏见、数据滥用、就业替代等问题也日益凸显。例如，有些互联网平台企业利用其主导地位，对用户精准“画像”的同时，通过“杀熟”、“信息茧房”等方式侵犯消费者正当权益。又如，人工智能在创造新就业岗位的同时也会取代许多传统岗位。2024年，国际货币基金组织指出，全球约40%的就业岗位将受人工智能技术的影响。目前，大量电话销售、客服等岗位已经被人工智能替代。一些自动驾驶平台曾引发网约车、出租车司机就业焦虑与公众对行驶安全的担忧，带来舆论和社会风险。

The essence of AI is a “tool for enhancing human capabilities.” If it pursues only efficiency maximization without constraints, technological development will deviate from the original intention of “serving humanity,” and may even trigger social conflict. Therefore, it is necessary to ensure that AI industrial application is built upon reasonable social governance norms, promoting AI to find a dynamic balance between efficiency and fairness, innovation and safety, commerce and people’s livelihoods, so that AI can better benefit society.

人工智能的本质是“增强人类能力的工具”，若只追求效率最大化而不加约束，会导致技术发展背离“服务人类”的初衷，甚至引发社会冲突。因此，必须确保人工智能产业应用建立在合理的社会治理规范之上，推动人工智能在效率与公平、创新与安全、商业与民生之间找到动态平衡点，让人工智能更好造福社会。

(Author: Chairman of the Board and Party Secretary of China International Engineering Consulting Corporation)

（作者：中国国际工程咨询有限公司党委书记、董事长）

Online Editor: Tongxin

网络编辑：同心