China Mil Watch Mandarin-source monitoring · Chinese military & security reporting
Independent monitor
Official PRC military media, read in the original
← Daily Brief
Modernization Doctrine Cyber & Info

Walking the Path of AI Innovation and Development with Chinese Characteristics — A Dialogue with Zhang Shuangxi, Researcher at an Academy of the Academy of Military Sciences

走好中国特色人工智能创新发展之路——对话军事科学院某院研究员张双喜
PLA Daily (解放军报) 29 July 2026
View original source ↗
Zhang Shuangxi, a researcher at an academy of the Academy of Military Sciences, gave a PLA Daily interview assessing China's AI development as having reached a level of "running in parallel with the world globally, and leading in certain domains" (全球并跑、部分领域领跑), while candidly acknowledging persistent gaps in foundational theory, original algorithms, high-end chips, and industry-academia-research conversion pipelines. The interview cites the U.S. DoD's FY2026 unmanned systems budget of over $13 billion as a reference point for international competition, framing China's trajectory against active great-power rivals. This is standard PLA science-and-technology advocacy content; its value is as a record of how the Academy of Military Sciences publicly frames AI self-reliance, talent cultivation, and the "strangled at the throat" (卡脖子) narrative for a military audience, not as evidence of a specific capability or programmatic shift.

Walking the Path of AI Innovation and Development with Chinese Characteristics — A Dialogue with Zhang Shuangxi, Researcher at an Academy of the Academy of Military Sciences

■ Li Jiaqi, Li Long, PLA Daily Correspondent Cheng Xue

[Image: Conceptual illustration of artificial intelligence technology. AI-generated image.]

At the National Science and Technology Awards Conference, the Conference of Academicians of the Chinese Academy of Sciences and the Chinese Academy of Engineering, and the 11th National Congress of the China Association for Science and Technology, artificial intelligence was identified as a frontier domain in which China must advance research and development of key core technologies.

China's development of artificial intelligence has gone from starting with nothing and being subject to others' control, to achieving breakthroughs on multiple fronts and leading in certain areas. Behind this lies the quiet dedication of generation after generation of science and technology workers, who used perseverance to break monopolies and innovation to forge a new path.

The course of scientific and technological development has proven to us time and again that key core technologies cannot be begged for, bought, or borrowed. Only by persisting in self-reliance and self-strengthening in science and technology can we thoroughly break foreign technology monopolies and standards blockades, and firmly grasp the initiative in future development.

In this installment of the Science and Technology Interview Series, we speak with Zhang Shuangxi, a researcher at an academy of the Academy of Military Sciences, to hear him discuss the substance of artificial intelligence, the current state of its development, and future directions — and to hear the story of how China's science and technology workers, through independent research and independent innovation, have continuously produced original and pioneering results in the field of artificial intelligence and have sustained their deep cultivation at the frontier of AI technology.

[Image: Conceptual illustration of artificial intelligence technology. AI-generated image.]

A New Engine Driving the Development of the Age

Reporter: At this conference, the key elements and era-defining value of artificial intelligence were elaborated upon. Could you explain in plain language what artificial intelligence is? Compared with traditional informatized technology, where does its disruptive innovation lie?

Zhang Shuangxi: At present, artificial intelligence is surging forward. It takes data, computing power, and algorithms as its key elements; is grounded in theoretical breakthroughs in neuroscience, cognitive science, computational science, and related fields; and is characterized by the collective breakthroughs and broad permeation and enablement of cutting-edge intelligent technologies. It exhibits prominent features such as data-driven operation, interconnection of all things, human-machine collaboration, and cross-domain integration.

Compared with traditional informatized technology, the disruptive innovation of artificial intelligence technology is manifested primarily in the leap from "rapid transmission" to "autonomous cognition (自主认知)."

In the 1980s and 1990s, the core software and hardware of domestic large-scale computer rooms and enterprise data platforms were almost entirely dependent on imports. At that time, systems had no capacity for independent analysis; they could only passively store and transmit data. All data organization, information assessment, and pattern summarization had to be completed manually.

To give the most intuitive example: when processing early meteorological data, researchers could only analyze and calculate by hand, and computing a single forecast took an extremely long time. The machines of that era could only "store data and transmit data" — they could not think, screen, or anticipate. This was the limitation of traditional informatized technology at the time: transmission only, no cognitive capability.

With the development of artificial intelligence technology, this situation has been thoroughly transformed. Today's intelligent systems can autonomously complete the full closed-loop cycle of "perception — decision — execution (感知—决策—执行)," integrate and analyze multi-source information, intelligently mine complex patterns, and accurately anticipate future trends.

To summarize briefly: traditional informatized technology addressed the problem of information "mutual exchange and rapid transmission (互通有无、快速传递)," whereas intelligentization (智能化) achieves "autonomous cognition and intelligent creation (自主认知、智能创造)" of information — and this is precisely the embodiment of artificial intelligence technology's "disruptive innovation." From informatization (信息化) to intelligentization (智能化): this is an important technological leap in the development of China's artificial intelligence technology over the past several decades.

Reporter: In today's world, a new round of scientific and technological revolution is accelerating, competition over frontier technologies is fierce, and artificial intelligence has become the main battlefield of international science and technology competition. From a global perspective, what is the competitive landscape in artificial intelligence? At what overall level does China's development stand?

Zhang Shuangxi: At present, countries are continuously increasing their research investment in the field of artificial intelligence, concentrating innovative resources, and competing to seize future tracks.

According to the U.S. Department of Defense's FY2026 unmanned systems budget documents, its annual research, development, and procurement budget for unmanned intelligent systems is as high as more than 13 billion dollars, with funding focused on humanoid combat robots, all-domain swarms across land, sea, and air, and extreme-environment adaptive systems. The Russian military, grounded in operational requirements, is rapidly improving the anti-jamming and battlefield survivability of intelligent equipment and advancing its large-scale application.

Looking at China, in terms of AI development achievements, our overall development stands at the advanced level of "running in parallel with the world globally, and leading in certain domains (全球并跑、部分领域领跑)." From early complete dependence on foreign computing equipment, to a total computing power scale that now steadily ranks among the world's top, to continuous breakthroughs in technologies such as computer vision and natural language processing — it can be said that China has thoroughly escaped the predicament of the past, in which it was "subject to others' control and passively following (受制于人、被动跟随)."

Boldly Venturing into "Uncharted Territory (无人区)": From Following to Running in Parallel to Leading

Reporter: In the scientific pursuit of continuously achieving breakthroughs in artificial intelligence technology, what spirit of innovation has been accumulated?

Zhang Shuangxi: Today, we enjoy the development dividends of the digital age and the intelligent age, but very few people know that today's development was ground out step by step from a zero baseline under total blockade.

In the past, faced with comprehensive technology blockades, the older generation of scientists, under conditions of having nothing, used paper and pen to derive formulas, used abacuses to iterate data, and used simple equipment to build experimental platforms. Over the course of decades, they laid a solid foundation for China's computer science.

It is precisely because of the perseverance and dedication of generation after generation of science and technology workers that the history of "China has no core algorithms, China has no indigenous computing power" was thoroughly ended, and the spark was planted for today's development of artificial intelligence.

After reform and opening up, China's information technology gradually caught up, but core chips, underlying systems, and foundational algorithms remained long monopolized by foreign entities. Around the year 2000, the domestic internet and digital industries developed rapidly, but almost all underlying architectures and core components were entirely dependent on imports. Once foreign suppliers restricted supply or tightened licensing, large numbers of domestic platforms and systems would face the risk of shutdown at any moment.

Faced with this long-standing passive situation of being "strangled at the throat (卡脖子)," a new generation of researchers persisted in the path of independent innovation, deeply cultivating algorithms and breaking through computing power bottlenecks.

This scientific research character — unafraid of blockades, willing to sit on the cold bench (甘坐冷板凳), daring to venture into uncharted territory (敢闯无人区), and pressing forward in sustained assault across generations — is the core secret of the rapid development of China's artificial intelligence technology.

Deeply Enabling High-Quality Development of Human Society

Reporter: China's artificial intelligence technology has now been widely deployed across many fields including public services, industrial upgrading, and social governance. Could you introduce several examples of applied AI achievements based on real cases?

Zhang Shuangxi: After years of iterative development, China's artificial intelligence has moved out of the laboratory and moved comprehensively toward industrialization, integration into daily life, and universal accessibility, deeply penetrating every aspect of human social development and becoming a powerful driver for improving people's wellbeing and driving industrial upgrading.

In the field of public services, intelligent government affairs, smart healthcare, and intelligent education are gradually becoming widespread. Relying on AI intelligent analysis systems, government approval processes can achieve second-level review and intelligent verification, allowing the public to make fewer trips while data does more of the running; medical AI-assisted diagnostic systems can rapidly screen imaging lesions, helping to improve the quality and efficiency of grassroots healthcare and bringing high-quality medical resources down to thousands of households.

In the field of industrial upgrading, intelligent manufacturing and intelligent quality inspection have been widely deployed. Traditional manufacturing industries are using artificial intelligence to achieve automated production, intelligent inspection, and intelligent scheduling, leaving behind the old model of manual assembly lines and experience-based production. Many established industrial bases have achieved transformation and upgrading through AI digital transformation, radiating entirely new vitality.

In the field of public governance, intelligent meteorological forecasting, smart traffic dispatching, and intelligent monitoring of natural resources have substantially improved the precision and scientific level of social governance. Relying on big data and AI computing power, the accuracy of extreme weather early warnings and disaster risk assessments has been substantially improved.

In the field of basic scientific research, artificial intelligence is enabling original innovation. In areas such as new materials research and development, biopharmaceutical breakthroughs, astronomical and geographical research, and foundational mathematical simulation, artificial intelligence technology is helping researchers shorten research cycles, reduce experimental costs, and improve innovation efficiency, and has become an indispensable "intelligent assistant."

It can be said that today, artificial intelligence technology is no longer a distant frontier concept; it has become a key technology serving society, benefiting people's livelihoods, and supporting innovation, injecting continuous scientific and technological momentum into advancing Chinese-style modernization (中国式现代化).

Squarely Facing Shortcomings and Weaknesses, Cultivating Innovative Talent for the New Era

Reporter: This conference objectively pointed out that China's scientific and technological development still has problems such as insufficient capacity for original innovation and low efficiency of science and technology investment. Measured against this scientific assessment, what shortcomings remain in China's current AI development?

Zhang Shuangxi: First, the accumulation of foundational innovation is still insufficient. China's artificial intelligence technology iterates rapidly, deploys broadly, and achieves good results at the application level, but in source innovation domains such as underlying foundational theory, original algorithm systems, and foundational mathematical models, the accumulation is relatively weak.

Second, there are still shortcomings in high-end core foundational components. The performance of certain ultra-high-precision computing power chips, extreme-environment-adapted components, and high-end intelligent sensors still needs improvement, and key foundational materials and core processes still require sustained breakthroughs.

Third, the system for deep integration of industry, academia, and research still needs improvement. At present, some frontier research results remain at the laboratory stage; the transformation chain from theoretical breakthrough and technology prototype to large-scale industrial application is not sufficiently smooth, and innovative value has not been fully released.

Nevertheless, looking back at the history of China's AI technology development, it has itself been a process of continuously discovering problems and solving problems. Squarely facing gaps and deficiencies is the foundation of our confidence in continuing to move forward. We will anchor on the goal of building a strong science and technology nation, fill in the shortcomings in artificial intelligence, and continuously activate innovative value.

Reporter: The conference emphasized the need to treat the building of a young science and technology talent workforce as a strategic project, and to carry out cultivation, recruitment, and utilization work in an all-around manner. To support the long-term development of artificial intelligence, how should we strengthen the talent workforce and carry forward the spirit of scientists?

Zhang Shuangxi: Looking back at the history of China's scientific and technological development, the most precious wealth has never been equipment and technology, but rather the scientific research founding aspiration (科研初心) and spirit of serving the nation (报国精神) passed down from generation to generation.

From the 1950s and 1960s, when countless top scientists gave up comfortable conditions abroad and resolutely returned to China to start from scratch and work hard under difficult conditions; to the young scientific researchers of the new era who bravely shoulder heavy responsibilities and dare to forge ahead and innovate — the spirit of patriotism, truth-seeking, innovation, and dedication (爱国、求实、创新、奉献) has never changed.

The scientific research founding aspiration and spirit of serving the nation have always been passed on. During the development of a certain large domestic model, many young research teams were rooted in the laboratory year-round, iterating models through the night, repeatedly refining algorithms, and continuously optimizing performance. With a tenacity that refused to concede defeat, they rapidly caught up with the international advanced level. These young people took the baton from their predecessors and shouldered the banner of scientific and technological self-reliance and self-strengthening in the new era.

Looking to the future, we hope to temper young talent through major scientific research projects and key innovation platforms, while at the same time carrying forward the spirit of scientists, guiding the broad community of science and technology workers to hold their country in their hearts, take root at the frontier of technology, and work persistently over the long term — so that through a lifetime of perseverance and deep cultivation, they may lift up the bright future of China's intelligent science and technology.

Layout design: Wu Huaijiang

Original Chinese
走好中国特色人工智能创新发展之路 ——对话军事科学院某院研究员张双喜 ■李佳奇 李 龙 解放军报记者 程 雪 图为人工智能技术概念图。AI制图 在国家科学技术奖励大会、两院院士大会、中国科协第十一次全国代表大会上,人工智能被列为我国需要推进关键核心技术攻关的前沿领域。 我国人工智能的发展,从最初一穷二白、受制于人,到如今实现多点突破、部分领域领跑,这背后是一代代科技工作者默默付出,用坚守打破垄断、用创新走出新路。 科技发展的道路一次次向我们证明,关键核心技术求不来、买不来、讨不来,我们唯有坚持科技自立自强,才能彻底打破外来技术垄断和标准封锁,牢牢掌握未来发展的主动权。 本期科技访谈录,我们对话军事科学院某院研究员张双喜,一同听他讲述人工智能的内涵、发展现状与未来动向,讲述我国科技工作者自主攻坚、自主创新,不断产出人工智能领域原创性、引领性成果,持续深耕人工智能技术前沿的故事。 图为人工智能技术概念图。AI制图 驱动时代发展的新引擎 记者:此次大会,阐释了人工智能的关键要素与时代价值。请您用通俗的语言,为大家解读什么是人工智能?对比传统信息化技术,它的颠覆性创新体现在哪里? 张双喜:当前,人工智能蓬勃兴起,它以数据、算力、算法等为关键要素,以神经科学、认知科学、计算科学等理论突破为基础,以智能前沿技术群体性突破和广泛渗透赋能为标志,呈现数据驱动、万物互联、人机协同、跨界融合等显著特征。 对比传统信息化技术,人工智能技术的颠覆性创新,主要体现在从“快速传递”到“自主认知”的跨越。 上世纪八九十年代,国内大型机房、企业数据平台等,核心软硬件大都依赖进口。那时候,系统没有自主分析能力,只能被动存储、传输数据。所有数据整理、信息研判、规律总结工作,全部依靠人工完成。 举个最直观的例子:处理早期气象信息数据时,科研人员只能手动分析计算,算一次预报耗时极长。而当时的机器只能“存数据、传数据”,不会思考、不会筛选、不会预判。这就是当时传统信息化技术的局限:只能传输,没有认知能力。 而随着人工智能技术的发展,这一局面被彻底打破。今天的智能系统,可以自主完成“感知—决策—执行”全流程闭环,融合分析多源信息,智能挖掘复杂规律,精准预判未来趋势。 简要概括,传统信息化技术解决的是信息“互通有无、快速传递”的问题,而智能化实现的是信息的“自主认知、智能创造”,而这也正是人工智能技术“颠覆性创新”的体现。从信息化到智能化,这是数十年来,中国人工智能技术发展的一次重要的技术跨越。 记者:当今世界,新一轮科技革命加速演进,前沿技术竞争激烈,人工智能已经成为国际科技竞争的主战场。从全球范围看,人工智能竞争呈现怎样的态势?我国整体发展处于什么水平? 张双喜:当前,各国持续加大对人工智能领域的科研投入,集聚创新资源,抢占未来赛道。 据美国防部2026财年无人系统预算文件披露,其年度无人智能系统研发采购预算高达130多亿美元,资金重点投向人形作战机器人、陆海空全域集群、极限环境自适应系统等领域。俄军则立足实战需求,快速提升智能装备抗干扰与战场生存适应能力,推动其规模化应用。 反观我国,在人工智能发展成效上,整体发展处于“全球并跑、部分领域领跑”的先进水平。从早期完全依赖国外计算设备,到如今算力总规模稳居世界前列,再到计算机视觉、自然语言处理等技术不断突破,可以这样说,我国已彻底摆脱了过去“受制于人、被动跟随”的困境。 勇闯“无人区”,从跟跑到并跑再到领跑 记者:在实现人工智能技术不断突破的科技求索中,沉淀了怎样的创新精神? 张双喜:今天,我们享受着数字时代、智能时代的发展红利,但很少有人知道,如今的发展,都是从零基础、全封锁中一步步熬出来的。 曾经,面对全面技术封锁,老一辈科学家在一无所有的条件下,用纸笔推演公式、用算盘迭代数据、用简易设备搭建试验平台。他们用几十年时间,为中国计算机科学打下了坚实的基础。 正是一代代科技工作者的坚守和付出,才彻底终结了“中国无核心算法、中国无自主算力”的历史,为今天人工智能的发展埋下了火种。 改革开放后,我国信息技术逐步向前追赶,但核心芯片、底层系统、基础算法依然长期被国外垄断。2000年前后,国内互联网、数字产业快速发展,但几乎所有底层架构、核心元器件全部依赖进口。一旦国外限制供货、收紧授权,国内大量平台、系统将随时面临停摆风险。 面对这种长期“卡脖子”的被动局面,新一代科研人坚持走自主创新之路,深耕算法、突破算力瓶颈。 这种不畏封锁、甘坐冷板凳、敢闯无人区、代代接续攻坚的科研品格,就是中国人工智能技术快速发展的核心密码。 深度赋能人类社会高质量发展 记者:当前我国人工智能技术已广泛落地民生服务、产业升级、社会治理等诸多领域,请您结合实际,介绍几个人工智能技术的应用成果? 张双喜:经过多年迭代发展,我国人工智能已经从实验室走出,全面走向产业化、生活化、普惠化,深度融入人类社会发展的方方面面,成为提升民生福祉、驱动产业升级的强大动力。 在民生服务领域,智能政务、智慧医疗、智能教育逐渐普及。依托AI智能分析系统,政务审批可实现秒级审核、智能核验,让群众少跑腿、数据多跑路;医疗AI辅助诊断系统可快速筛查影像病灶,助力基层医疗提质增效,让优质医疗资源下沉到千家万户。 在产业升级领域,智能制造、智能质检等广泛落地。传统制造业依靠人工智能实现自动化生产、智能检测、智能调度,告别了过去人工流水线、经验化生产的老旧模式。很多老牌工业基地,通过AI数字化改造实现转型升级,焕发全新活力。 在公共治理领域,智能气象预判、智慧交通调度、自然资源智能监测,大幅提升社会治理精细化、科学化水平。依托大数据与AI算力,极端天气预警、灾害风险研判准确率得到大幅提升。 在基础科研领域,人工智能正在赋能原始创新。在新材料研发、生物医药攻关、天文地理研究、基础数学模拟等领域,人工智能技术帮助科研工作者缩短科研周期、降低试验成本、提升创新效率,已经成为他们离不开的“智能助手”。 可以说,如今,人工智能技术不再是遥远的前沿概念,已经成为服务社会、惠及民生、支撑创新的关键技术,为推进中国式现代化注入源源不断的科技动能。 正视短板弱项,培育新时代创新人才 记者:此次大会客观指出,我国科技发展中还存在原始创新能力不足、科技投入效能不高等问题。对标这一科学判断,当前我国人工智能发展还存在哪些短板? 张双喜:一是基础创新积淀仍有不足。我国人工智能技术在应用层面迭代快、落地广、成效好,但在底层基础理论、原创算法体系、基础数学模型等源头创新领域,积累较为薄弱。 二是高端核心基础器件仍有短板。部分超高精密度算力芯片、极端环境适配元器件、高端智能传感器等,性能仍有待提升,关键基础材料、核心工艺仍需持续突破。 三是产学研深度融合体系仍需完善。目前,还有部分前沿科研成果停留在实验室阶段,从理论突破、技术样机到产业规模化应用,转化链条不够顺畅,创新价值未能完全释放。 不过,回望我国人工智能技术发展的历史,本身就是一个不断发现问题、解决问题的过程。正视差距和不足,是我们继续前进的底气。我们将锚定建成科技强国目标,补齐人工智能短板,不断激活创新价值。 记者:大会强调,要把青年科技人才队伍建设作为一项战略工程,全方位做好培养、引进、使用工作。支撑人工智能长远发展,我们如何建强人才队伍、传承科学家精神? 张双喜:回望中国科技发展史,最宝贵的财富,从来不是设备和技术,而是代代相传的科研初心与报国精神。 从上世纪五六十年代,无数顶尖科学家放弃海外优渥条件,毅然回国、白手起家、艰苦奋斗;到新时代青年科研工作者勇挑重担、敢闯敢创,爱国、求实、创新、奉献的精神始终未变。 科研初心与报国精神一直在传递。国内某大模型攻关过程中,许多青年科研团队常年扎根实验室,通宵迭代模型、反复打磨算法、持续优化性能,凭着一股不服输、不认输的韧劲,快速追赶国际先进水平。这群年轻人,从前辈手中接过接力棒,扛起了新时代科技自立自强的大旗。 面向未来,我们希望依托重大科研项目、重点创新平台历练青年人才,同时赓续科学家精神,引导广大科技工作者心怀家国、扎根技术前沿久久为功,以一生坚守、一世深耕,托举起中国智能科技的美好未来。 版式设计:吴淮江