Deep Dive into Qiushi's Latest Edition: How to Build Future Industries? What are China's Fault-lines in Basic Research? How are LLMs Impacting Knowledge? What's Driving Japan's 'Neo-Militarism'?
Hi folks,
In today’s edition, since I found little interesting in the paper, I am offering breakdowns of interesting articles in the latest edition of the Qiushi journal.
I am covering four articles below:
The first one is by the journal’s editorial department on China’s approach to future industries.
The second is by Zhang Jun on the importance of basic/foundational research for cultivating future industries. He is quite critical in this regard.
Third, there’s a piece by Sun Maosong talking about the impact of LLMs and generative AI on knowledge production.
Finally, I am briefly discussing the foreign affairs article in the journal, which deals with Japan’s so-called “neo-militarism”.
I am also offering some takeaways from my perspective on some these articles. I hope you enjoy this format. If it works, I might continue this for future editions of the journal.
Cheers,
Manoj
In the latest edition of Qiushi, the lead article was an excerpt of a speech that Xi Jinping delivered during the 24th collective study session of the Politburo in January 2026. The speech focussed on cultivating future industries. Sinocism has the full translation of Xi’s article.
The journal’s editorial department also has an interesting take on future industries, which I am sharing below. Instead of the full text, I am covering some of the key points.
“Developing future industries cannot mean trying to cover everything and exerting force equally in all directions; it is imperative to clarify the direction and accurately identify the priorities. The Fourth Plenary Session of the 20th CPC Central Committee proposed that quantum technology, biomanufacturing, hydrogen energy and nuclear fusion energy, brain-computer interfaces, embodied intelligence, and sixth-generation mobile communications be promoted as new points of economic growth. In this important speech, General Secretary Xi Jinping explicitly identified these fields as the main areas of focus for the development of future industries for China during the 15th FYP period. Only by gaining a deep understanding of the development dynamics, industrial ecosystems, and prospective pathways of these fields can we better grasp the priorities, apply precisely targeted policies, and drive significant progress in the development of future industries during the 15th FYP. 发展未来产业，不能面面俱到、平均用力，必须明确方向、找准重点。党的二十届四中全会提出，要推动量子科技、生物制造、氢能和核聚变能、脑机接口、具身智能、第六代移动通信等成为新的经济增长点。在这篇重要讲话中，习近平总书记明确把这些领域作为“十五五”时期我国未来产业发展的主攻方向。只有深入认识这些领域的发展态势、产业生态与前景路径，才能更好把握重点、精准施策，推动未来产业发展在“十五五”时期取得明显进展.
Why have these six major future industries been chosen as the main areas of focus? This was not done arbitrarily, but was a scientific and prudent choice made on the basis of an accurate assessment of factors such as global trends in science and technology, national strategic needs, and the laws governing industrial transformation. In terms of global trends in science and technology, all six of these industries represent frontier directions of the technological revolution and industrial transformation, and stand at critical junctures for technological breakthroughs. According to relevant research, nuclear fusion energy, sixth-generation mobile communications, and brain-computer interfaces are currently in the embryonic stage of development; hydrogen energy, quantum computing and precision measurement, and humanoid robots are in the growth stage; and quantum communications, new types of pharmaceutical manufacturing, and the like have already entered the expansion stage. The tiered distribution of these six industries chosen by China balances short-term breakthroughs with medium- and long-term development. 发展未来产业，不能面面俱到、平均用力，必须明确方向、找准重点。党的二十届四中全会提出，要推动量子科技、生物制造、氢能和核聚变能、脑机接口、具身智能、第六代移动通信等成为新的经济增长点。在这篇重要讲话中，习近平总书记明确把这些领域作为“十五五”时期我国未来产业发展的主攻方向。只有深入认识这些领域的发展态势、产业生态与前景路径，才能更好把握重点、精准施策，推动未来产业发展在“十五五”时期取得明显进展.
In terms of national strategic needs, these six industries span the domains of future manufacturing, future information, future energy, and future health, and are characterised by high technological content, high added value, and low resource consumption. They align closely with China’s strategic objectives of developing new-quality productive forces, ensuring energy security, and achieving a green, low-carbon transition, and they carry the mission of safeguarding national security and seizing the commanding heights of future development. In terms of the laws governing industrial transformation, the development of these six industries will not only enable them to form clusters of hundreds of billions and trillions in the future, shaping new drivers of growth, but will also exert powerful technological spillover and industrial linkage effects, effectively driving the transformation and upgrading of traditional industries and achieving an overall leap forward in the industrial system. 为什么把这六大未来产业作为主攻方向？这并非随意为之，而是在准确研判全球科技发展趋势、国家战略需求和产业变革规律等因素的基础上，作出的科学而慎重的抉择。从全球科技发展趋势看，这六大产业均代表着科技革命和产业变革的前沿方向，处于技术突破的关键节点。据相关研究，目前核聚变能、第六代移动通信、脑机接口处于发展萌芽期，氢能、量子计算与精密测量、人形机器人处于成长期，量子通信、新型医药制造等已进入扩展期。我国选择的这六大产业梯次分布，兼顾了短期突破与中长期发展。从国家战略需求看，这六大产业涵盖未来制造、未来信息、未来能源、未来健康等领域，具有科技含量高、附加值高、资源消耗低等特点，同我国发展新质生产力、保证能源安全、实现绿色低碳转型等战略目标高度契合，承载着维护国家安全、抢占未来发展制高点的使命。从产业变革规律看，这六大产业的发展不仅自身在未来能够形成一批千亿级、万亿级产业集群，塑造发展新动能，而且具有强大的技术溢出和产业关联效应，能够有效推动传统产业改造升级、实现产业体系整体跃升.
What major problems can the development of the six major future industries solve? These six industries point directly at some of the most critical bottlenecks and chokepoints in China’s current and long-term development, each corresponding to real difficulties in future development such as technological bottlenecks, energy dependence, and industrial security. Their development and maturation will reshape, from the ground up, modes of production, energy structures, and patterns of information exchange, thereby providing the foundation and conditions for fundamentally resolving ‘chokepoint’ problems and the like.
Quantum technology is a frontier technology that processes, measures, and transmits information based on the principles of quantum mechanics, comprising three major fields, i.e., quantum computing, quantum communications, and quantum measurement, and holds immeasurable strategic significance for safeguarding national information security, enhancing basic research capabilities and efficiency, and driving leapfrog development in industry.
Biomanufacturing is a green manufacturing model that uses biological tissues or organisms (such as enzymes, microorganisms, and animal and plant cells) as ‘factories’ to produce chemicals, materials, pharmaceuticals, and other products at scale through biotransformation processes; it holds important value for meeting challenges such as resource scarcity, environmental pollution, and food security, and for reshaping the landscape of the materials, chemical, and pharmaceutical industries.
Hydrogen energy and nuclear fusion energy represent two important directions for transformation in the energy sector. Hydrogen energy is a key lever for the present-day energy transition and the decarbonisation of industry, while nuclear fusion energy is the ‘ultimate energy solution’ of the future; once a breakthrough in controlled nuclear fusion is achieved, it will completely transform the world’s energy landscape and even its geopolitical configuration.
Brain-computer interfaces integrate multiple disciplines including neuroscience, biomedical engineering, and AI, with application prospects spanning medical rehabilitation, thought-based control, immersive gaming, and specialised operations, and hold great significance for improving human health and quality of life. The typical embodiment of embodied intelligence is the humanoid robot, which aims to allow AI to break through the boundaries of the virtual world and truly enter the physical world to perform complex tasks, bringing profound transformation to manufacturing, services, home life, and even space exploration.
6G mobile communications is the next generation of mobile communications technology aimed at commercial use around 2030 and beyond. Compared with 5G, it will achieve order-of-magnitude improvements in key indicators such as speed, latency, and connection density, and will introduce new capabilities such as integrated communications and sensing, integrated space-air-ground networks, and endogenous intelligence, empowering thousands of industries.
发展六大未来产业能够解决哪些重大问题？这六大产业直指我国当前和长远发展中一些至为关键的卡点、堵点，无一不对应着未来发展的技术瓶颈、能源依赖、产业安全等现实难题。其发展和成熟，将从底层重塑生产方式、能源结构、信息交互模式等，从而为我们从根本上解决“卡脖子”等问题提供基础和条件。量子科技是基于量子力学原理进行信息处理、测量和传输的前沿技术，主要包括量子计算、量子通信和量子测量三大领域，对于维护国家信息安全、提升基础科研能力和效率、推动产业跨越式发展具有不可估量的战略意义。生物制造是以生物组织或生物体（如酶、微生物、动植物细胞）为“工厂”，通过生物转化过程，规模化生产化学品、材料和药物等产品的绿色制造模式，对于应对资源短缺、环境污染、粮食安全等挑战，重塑材料、化工和医药产业格局具有重要价值。氢能和核聚变能代表了能源领域转型的两个重要方向，氢能是当下能源转型与工业脱碳的重要抓手，核聚变能则是未来的“终极能源解决方案”，可控核聚变一旦突破，将彻底改变世界能源乃至地缘政治格局。脑机接口融合了神经科学、生物医学工程、人工智能等多个学科，应用前景涉及医疗康复、意念控制、沉浸式游戏、特种作业等，对改善人类健康、提升生活质量具有重要意义。具身智能的典型载体是人形机器人，旨在让AI突破虚拟世界的边界，真正进入物理世界执行复杂任务，将给制造业、服务业、家庭生活乃至太空探索等领域带来深刻变革。第六代移动通信（6G）是面向2030年及以后商用的下一代移动通信技术，与5G相比，将在速率、时延、连接密度等关键指标上实现数量级的提升，并引入通信感知一体化、空天地一体化、内生智能等新能力，赋能千行百业.
Ecosystems are the soil that gives rise to innovation and concentrates industry. Forming a sound innovation ecosystem and industrial ecosystem is crucial to the development of the six major future industries. In some areas of these six future industries, China has already begun to build relatively complete industrial and supply chains. In embodied intelligence, for example, China has seen the emergence of robotics industry clusters, holds an advantage in the hardware manufacturing of humanoid robots, possesses a complete manufacturing supply chain and abundant industrial and consumer application scenarios, and is therefore better positioned to achieve economies of scale. Another example is the field of hydrogen energy. China’s hydrogen-energy industry has already begun to take shape. The construction of fuel-cell vehicle demonstration city clusters is advancing steadily. In 2025, annual hydrogen production exceeded 37 million tons, ranking first in the world for many consecutive years. At the same time, it must be recognised that the industrial ecosystem for future industries is not yet well developed, and these six industries in China still face problems such as the lack of fundamental breakthroughs in key core technologies, shortages of talent, and insufficient long-term capital investment. The core test lies in whether the original breakthroughs of ‘from 0 to 1’, the engineering scale-up of ‘from 1 to 10’, and the large-scale application of ‘from 10 to 100’ can be forged into a sustainable ecosystem. On this basis, it is necessary to adhere to the combination of an effective market and a capable government, innovate institutional systems and mechanisms, and advance basic research, pilot-scale validation, scenario implementation, financial support, talent cultivation, and policy safeguards in an integrated manner, striving to be the first to build a relatively complete innovation and industrial ecosystem and seize the initiative and discourse power in the competition. 生态是催生创新、集聚产业的土壤，形成良好的创新和产业生态对六大未来产业的发展至关重要。在六大未来产业的一些领域，我国已初步构建起较为完整的产业链供应链。比如，在具身智能领域，我国已出现机器人产业集群，在人形机器人硬件制造上占优，拥有完备的制造供应链和丰富的工业及消费应用场景，更易形成规模效应。又比如，在氢能领域，我国氢能产业已初具规模，燃料电池汽车示范城市群建设稳步推进，2025年氢气年产量超过3700万吨，连续多年居世界第一。同时也要看到，未来产业的产业生态尚不完善，我国这六大产业仍面临关键核心技术尚未取得根本突破、人才短缺、长期资本投入不足等问题。其核心考验在于，能否把“从0到1”的原创突破、“从1到10”的工程化放大、“从10到100”的规模化应用打造成可持续的生态体系。基于此，必须坚持有效市场和有为政府相结合，创新体制机制，把基础研究、中试验证、场景落地、金融支撑、人才培育、政策保障等贯通起来推进，力争率先建成比较完整的创新和产业生态，掌握竞争的主动权、话语权.
In terms of what must be done going ahead, the piece says:
“In strengthening overall planning, we must do a good job in top-level design and scientifically plan the overall layout. Future industries are marked by pronounced uncertainty; the directions of their development are numerous and their technological pathways are not yet clear. Without accurate assessment and coordinated planning, there is a risk both of missing development opportunities through poor choices of route or improper layout, and of falling into duplicated construction as localities rush in all at once and blindly follow trends. It is necessary to comprehensively implement General Secretary Xi Jinping’s important requirements regarding accurately grasping development directions, scientifically evaluating technological routes, controlling the pace of development, and strengthening industrial coordination, so as to ensure that the development of China’s future industries always advances in the right direction. It is necessary to uphold the principle that the whole country is one chessboard, improve technological foresight mechanisms, formulate specialized plans by category, clearly define roadmaps and timetables for each subfield, and advance work actively yet prudently. It is necessary to rely on local industrial foundations and innovation resources, advance the construction of future-industry pilot zones in an orderly manner, and guide localities to develop according to local conditions and pursue differentiated development. 在加强统筹谋划上，要做好顶层设计、科学谋篇布局。未来产业具有明显的不确定性特征，发展方向众多，技术路径尚未明晰，若无精准研判和统筹谋划，既有可能因路线选择或布局不当而错失发展机遇，也容易因各地一哄而上、盲目跟风而陷入重复建设。要全面贯彻落实习近平总书记关于把准发展方向、科学论证技术路线、把握发展节奏、强化产业协同等重要要求，保证我国未来产业发展始终沿着正确方向前进。要坚持全国一盘棋，健全技术预见机制，分门别类编制好专项规划，明确各细分领域的路线图和时间表，积极稳妥推进。要依托地方产业基础和创新资源，有序推进未来产业先导区建设，引导各地因地制宜、错位发展.
With regard to scientific and technological innovation, it is necessary to leverage the advantages of the new nationwide system and enhance the capability of science and technology to provide support and leadership. Future industries originate from breakthroughs in frontier technologies. The capability for scientific and technological innovation determines the speed, breadth, and depth of future-industry development. At present, our country still has obvious shortcomings in areas such as basic research and key core technologies. For example, bottleneck constraints exist in critical fields such as advanced chips; the intensity of investment in basic research remains relatively low; original and leading achievements are insufficient; the mechanisms for transforming scientific and technological achievements are not smooth; and the innovation chain is not tightly connected to the industrial chain. In accordance with the requirements of General Secretary Xi Jinping’s important speech, it is necessary to strengthen the role of the nation’s strategic scientific and technological forces, adhere to the principle that ‘industry sets the questions, science and technology provides the answers’, and use mechanisms such as ‘revealing the list and appointing the commander’ to achieve breakthroughs in ‘chokepoint’ problems as quickly as possible. It is necessary to fully bring into play the role of innovation platforms such as national laboratories and national key laboratories, strengthen the supply of foundational generic technologies, and strive to fundamentally solve theoretical and foundational problems. It is necessary to accelerate the transformation and industrialization of scientific and technological achievements, plan and establish concept-verification centers and pilot-testing platforms, and open up the ‘last mile’ from the laboratory to the production line. 在科技创新上，要发挥新型举国体制优势，提升科技支撑引领能力。未来产业源于前沿技术的突破，科技创新能力决定了未来产业发展的速度、广度和深度。当前，我国在基础研究、关键核心技术等方面仍存在明显短板。比如，在高端芯片等关键领域存在瓶颈制约；基础研究投入强度偏低，原创性、引领性成果不足；科技成果转化机制不畅，创新链与产业链衔接不紧，等等。要按照习近平总书记这篇重要讲话的要求，强化国家战略科技力量作用，坚持“产业出题、科技答题”，采用“揭榜挂帅”等机制尽快突破“卡脖子”难题。充分发挥国家实验室、全国重点实验室等创新载体作用，加强基础共性技术供给，努力从根本上解决原理性、基础性问题。加快科技成果转化和产业化，布局建设概念验证中心和中试平台，打通从实验室到生产线的“最后一公里”.
With regard to giving full play to the principal role of enterprises, it is necessary to promote the concentration of various innovation resources toward enterprises. Enterprises are the principal actors of innovation and are also the carriers through which industry is transformed from a blueprint into reality. The rise of many future industries has been driven by enterprises achieving breakthroughs step by step. For some time now, the growth momentum of China’s technology enterprises has been strong. Hard-tech enterprises such as Unitree Robotics and DeepSeek have accelerated their breakthroughs and emergence. At the same time, it must also be recognized that many technology enterprises still face problems such as scattered resources, talent shortages, and poorly functioning mechanisms for industry-academia-research collaboration. In response to this, General Secretary Xi Jinping has made targeted arrangements from multiple points of view. First, through policy guidance, institutional innovation, and ecosystem optimization, it is necessary to vigorously cultivate technology-leading enterprises and high-tech enterprises that possess leading core technologies and strong innovation capabilities. Second, it is necessary to support central SoEs in developing future industries in conjunction with their primary responsibilities and principal businesses, enabling them to serve both as the national team for scientific and technological innovation and as the main force in future industries. Third, it is necessary to strengthen the supply of public services and cultivate a large number of technology-based small and medium-sized enterprises, ‘specialised, refined, distinctive, and innovative’ enterprises, single-product champion enterprises, and unicorn enterprises. 在发挥企业主体作用上，要推动各类创新资源向企业集聚。企业是创新的主体，也是产业从蓝图变为现实的承载者，很多未来产业的兴起是靠企业一步步突破带动的。一段时间以来，我国科技企业成长势头强劲，宇树科技、深度求索等硬核企业加速突围。同时也要看到，仍有不少科技企业面临资源分散、人才短缺、产学研协同机制不畅等问题。对此，习近平总书记从多个方面作出针对性部署。一是要通过政策引导、机制创新、生态优化，大力培育核心技术领先、创新能力强的科技领军企业和高新技术企业。二是要支持中央企业结合主责主业发展未来产业，既当科技创新的国家队，又成为未来产业的主力军。三是要强化公共服务供给，培育一大批科技型中小企业、专精特新企业、单项冠军企业、独角兽企业.
With regard to the policy environment, it is necessary to provide good support and services, and create fertile soil for the vigorous growth of future industries. Future industries generally possess characteristics such as long cultivation cycles, rapid technological iteration, and high market risks. From basic research to industrialization generally requires 10 years or even longer. Relying solely on the spontaneous forces of the market is often insufficient to cross the ‘valley of death’. For this reason, General Secretary Xi Jinping has explicitly required that ‘there must be vigorous policy support, and the government must provide good services’. Through means such as optimising and refining policies and doing a good job on talent, it is necessary to support and encourage the growth of future industries with greater intensity and more precisely targeted measures. In terms of financial support, it is necessary to guide patient capital to invest in future industries, promote investment models such as ‘combining grants with investment’ and ‘invest first, take equity later’, build a financial-services system suited to the financing needs of future industries across their entire life cycle, and support long-term capital in ‘investing early, investing small, investing for the long term, and investing in hard technology.’ In terms of application scenarios, it is necessary to build demonstration scenarios around key fields such as raw materials and consumer goods, support the promotion and application of first-set and first-batch products, and drive the large-scale implementation of new technologies and new products. In terms of talent, it is necessary to pursue cultivation, recruitment, and utilisation in tandem, bring forth a cohort of strategic scientists, outstanding engineers, and highly skilled personnel with forward-looking vision, and let more ‘innovation pioneers’ become the core code for activating future industries. 在政策环境上，要做好支持和服务，营造未来产业拔节生长的沃土。未来产业普遍具有培育周期长、技术迭代快、市场风险高等特征，从基础研究到产业化一般需要十年乃至更长时间，单靠市场自发力量往往难以跨越“死亡之谷”。为此，习近平总书记明确要求，“政策上要大力支持，政府要做好服务”。要通过优化和完善政策、做好人才工作等方式，以更大力度、更精准举措为未来产业成长撑腰鼓劲。在金融支持方面，要引导耐心资本投入未来产业，推广“拨投结合”、“先投后股”等投资模式，构建与未来产业全生命周期融资需求相适应的金融服务体系，支持长期资本“投早、投小、投长期、投硬科技”。在应用场景方面，要围绕原材料、消费品等重点领域打造示范场景，支持首台（套）、首批次商品的推广应用，推动新技术、新产品规模化落地。在人才方面，要坚持培养、引进、使用并举，造就一批具有前瞻视野的战略科学家、卓越工程师和高技能人才，让更多“创新先锋”成为激活未来产业的核心密码.
With regard to industrial governance, it is necessary to strengthen policy coordination and improve the governance system. The development of future industries involves many actors and spans frontier fields such as AI, gene editing, and brain-computer interfaces. Without a sound governance system, industrial development could descend into disorder. It is necessary to uphold and strengthen the centralised and unified leadership of the Party Central Committee, improve mechanisms for inter-ministerial coordination and central-local collaboration, and prevent situations such as fragmented policymaking and dispersed efforts. It is necessary to coordinate development and security well, explore scientific and effective regulatory approaches, precisely delineate ‘red lines’ and ‘bottom lines’, and allow ‘trial and error’, guarding against risks such as technology spinning out of control while leaving ample space for innovation to explore. Future industries require innovative cooperation and joint progress on the part of the international community. It is necessary to actively participate in global governance, strive for discourse power in the formulation of relevant international rules and standards, and work hard to promote the joint building of standards, joint consultation on rules, and joint advancement of industry by all parties. 在产业治理上，要加强政策协同，健全治理体系。未来产业发展涉及诸多主体，涵盖人工智能、基因编辑、脑机接口等前沿领域。没有健全的治理体系，产业发展就可能失序。要坚持和加强党中央集中统一领导，健全部际协同和央地协作机制，防止出现政出多门、力量分散等情况。统筹好发展和安全，探索科学有效的监管方式，精准划定“红线”和“底线”，允许“试错”，既防范技术失控等风险，又为创新留足探索空间。未来产业需要国际社会创新合作、携手前行。要积极参与全球治理，在相关国际规则和标准制定中争取话语权，努力推动各方标准共建、规则共商、产业共促.
Another interesting piece on this same issue is by Zhang Jun, who is the Party Secretary of the Beijing Institute of Technology. Zhang offers biting criticism of China’s approach towards future industries.
My Note: The focus of his critique isn’t the government or broad policy, although one can read policy criticism into what he is arguing. The core point that he is making is that Chinese research institutes and enterprises must focus far more on foundational research breakthroughs if it wants stable and sustainable advancement of future industries.
This paragraph below, early on in the article, makes the core point as clearly as anyone could have: “Whether a country can seize the initiative and win an advantage in global scientific and technological competition does not depend on how many application scenarios it rolls out or how many industrial concepts it forms in the short term, but on whether it can continuously pose new questions, discover new principles, and open up new directions, using original breakthroughs to open new space for industrial development.” 一个国家能否在全球科技竞争中占据主动、赢得优势，不取决于短期内铺开了多少应用场景、形成了多少产业概念，而取决于能不能持续提出新问题、发现新原理、开辟新方向，以原创性突破打开产业发展新空间.
To be honest, I think Zhang’s argument is conditionally true depending on whether you’re playing the great-power game. One needs to, of course, achieve the 0-1 breakthroughs, if one aims to remain at the frontier of international competition. Such breakthroughs are also critical if one desires to reap sustained economic gains, spin off new technologies and new business models. But is it essential for countries to be doing so in order to become prosperous? Even in the current world disorder, where we are witnessing controls on export of technology capital, knowledge, products and talent, I don’t think this is the case. The semiconductor example is a case in point. Countries around the world where there are companies that use chips do not need to focus on foundational breakthroughs. Building application scenarios, in fact, might be the most efficient pathway for them to achieve prosperity. Of course, the ones who invest in foundational research can leverage their dominance as a tool of coercion. One has to, therefore, identify a toolkit of carrots and sticks to navigate such situations.
My argument is that resource and other constraints will mean that an overwhelming majority of countries will choose the path of prosperity riding on someone else’s foundational knowledge. And that’s okay. The right strategy will require you to assess your exposure to coercion, and then build resilience or bandwagon or do a bit of both if you can. Does this bring challenges? Yes. But it is a faster path to prosperity for much of the developing world. Now, if you believe you are a great power in a world of great power competition for the future, then indeed Zhang’s argument is critical. In other words, one must read this in the context of the geopolitical scenario that Zhang believes China is in.
Anyway, here’s the piece:
Future industries are driven by frontier technologies and have notable characteristics of being forward-looking, strategic, and disruptive. Unlike mature industries, their difficulty lies not in simply expanding scale, but in the fact that key core technologies are still evolving and the underlying support for innovation is not yet firmly established. Precisely for this reason, future industries are most prone to the problem of ‘being able to see the track, but not clearly seeing the source’; industrial plans can take shape relatively quickly, application scenarios can be rolled out at an accelerated pace, and capital and projects can rapidly gather, yet the underlying scientific questions, key technological routes, and engineering pathways that sustain the continued development of an industry often require a longer time to explore and validate. Without the original breakthroughs of ‘from 0 to 1’, the technological iteration of ‘from 1 to 10’ is impossible, and the industrial leap of ‘from 10 to 100’ is even harder to achieve. It can be said that the level of basic research determines the height, breadth, and resilience of future-industry development, and also profoundly influences the depth and staying power of a country’s innovation capability. 未来产业由前沿技术驱动，具有显著前瞻性、战略性、颠覆性等特点，不同于成熟产业，其难点不在于简单扩大规模，而在于关键核心技术仍在演进、底层创新支撑尚不稳固。正因如此，未来产业最容易出现“看得见赛道、看不清源头”的问题：产业规划可以较快成形，应用场景可以加快铺开，资本和项目也可以迅速集聚，但支撑产业持续发展的底层科学问题、关键技术路线和工程化路径，却往往需要更长时间加以探索和验证。没有“从0到1”的原创性突破，就不可能有“从1到10”的技术迭代，更难形成“从10到100”的产业跃迁。可以说，基础研究的水平决定未来产业发展的高度、广度和韧性，也深刻影响一个国家创新能力的厚度和后劲.
At present, the world’s major countries are accelerating their deployment of national laboratories, major scientific and technological infrastructure, and frontier interdisciplinary research platforms around fields such as quantum technology, AI, life sciences, and advanced materials, and are continuously increasing the intensity of their investment in basic research, with the fundamental aim of seizing the commanding heights of future industrial development. Looking at China’s own development practice, the reason fields such as manned spaceflight, advanced nuclear energy, high-speed rail, and new-generation information and communications technology have been able to achieve leapfrog development lies in fully bringing into play the advantages of the new type of nationwide mobilisation system, tackling basic scientific problems through sustained effort, and forming a relatively strong capacity for independent innovation. At the same time, it must be recognised that China still faces problems such as insufficient original theoretical support and weak key core technologies in frontier fields such as embodied intelligence, commercial spaceflight, and biomanufacturing; and although some industrial fields generate considerable enthusiasm and have active application scenarios, original achievements with genuine global influence and industrial dominance remain relatively few. This requires us to strengthen basic research with greater intensity and more concrete measures, to continuously enhance China’s capacity for original innovation, and to further lay a solid foundation for building a country strong in science and technology. 当前，全球主要国家纷纷围绕量子科技、人工智能、生命科学、先进材料等领域加快布局国家实验室、重大科技基础设施和前沿交叉研究平台，持续提升基础研究投入强度，其根本目的就在于抢占未来产业发展的制高点。从我国发展实践看，载人航天、先进核能、高速铁路、新一代信息通信技术等领域之所以能够实现跨越发展，关键在于充分发挥新型举国体制优势，围绕基础科学问题持续攻关，形成了较强自主创新能力。同时也要看到，我国在具身智能、商业航天、生物制造等前沿领域仍面临原创性理论支撑不足、关键核心技术薄弱等问题；一些产业领域虽然热度较高、应用场景活跃，但真正具有全球影响力和产业主导力的原创性成果仍然偏少。这就要求我们必须以更大力度、更实举措加强基础研究，不断提升我国原始创新能力，进一步打牢科技强国建设根基.
Zhang adds:
As future industries accelerate their development, some fields are gradually exhibiting the practical problem of ‘hot industry, weak source’. Examining the root causes of this, it is evident that the issue lies not only in poor linkage between the industrial side and the scientific research side, but also in shortcomings such as insufficient forward-looking planning in relevant basic research, inadequate accumulation of original innovations, and weak long-term stable support. An effective innovation chain connecting basic research, engineering validation, and industrial evolution has not yet been formed, and the resulting structural contradictions are becoming increasingly prominent. 随着未来产业加快发展，一些领域逐渐呈现出“产业热、源头弱”的现实问题。究其根源，不只是产业端与科研端衔接不畅，更在于相关基础研究仍存在前瞻性布局不足、原创性积累不够、长期稳定支持不强等问题短板，基础研究、工程验证、产业演进之间尚未形成有效贯通的创新链条，由此带来的结构性矛盾日益凸显.
Insufficient support from scientific discovery for industrial development The source of future industries often comes from scientific discoveries and technological explorations that have not yet fully matured. However, in actual work, the things most easily seen, assessed, and showcased are often the construction of industrial parks, the implementation of projects, the introduction of enterprises, and the opening of application scenarios. When some localities plan future industries, their first considerations are often whether ‘the track is generating enough buzz’, ‘whether projects can be implemented’, and ‘whether the industrial park can take shape as quickly as possible’, while paying too little attention to, and conducting too little deep research on, the basic scientific questions, the key technological routes, and the underlying capabilities that lie behind them. For example, some localities have intensively issued plans, established platforms, and laid out industrial parks around such directions as the low-altitude economy, humanoid robots, and synthetic biology. Within a short period of time, these efforts have indeed generated a high degree of attention. However, the problem of ‘industry running fast while science lags behind’ has also gradually become apparent. Underlying capabilities such as environmental perception and dexterous manipulation in humanoid robots still pose many difficulties, and key scientific questions in brain-computer interfaces, such as the mechanisms of brain cognition and the parsing of neural signals, have not yet truly been broken through. Industrial concepts can quickly become ‘hot’, but scientific breakthroughs require a long period of ‘sitting on a cold bench’. If research on scientific questions lags behind industrial evolution, what seems like seizing the track may in reality amount only to remaining at the level of following along at the application level. What appears to be the creation of momentum and visibility may in reality lack the roots needed to support the sustained and healthy growth of an industry. 科学发现支撑产业发展不足。未来产业的源头往往来自尚未完全成熟的科学发现和技术探索，但在实际工作中，最容易被看见、被考核、被呈现的，往往是园区建设、项目落地、企业引进和场景开放。一些地方布局未来产业，首先考虑的是“赛道有没有热度”、“项目能不能落地”、“产业园能否尽快成形”，对于背后的基础科学问题、关键技术路线、底层能力支撑，关注得不够多、研究得不够深。比如，一些地方围绕低空经济、人形机器人、合成生物等方向密集出台规划、挂牌平台、布局园区，短时间内确实形成了较高关注度，但“产业跑得快、科学跟得慢”的问题也逐渐显现。人形机器人的环境感知、灵巧操作等底层能力仍存在不少难题，脑机接口的大脑认知机理、神经信号解析等关键科学问题也尚未真正突破。产业概念可以很快“热”起来，科学突破却需要长期坐“冷板凳”。如果科学问题研究滞后于产业演进，看似抢占了赛道，实则可能只是停留在应用跟随层面；看似形成了声势，实则缺乏支撑产业持续健康生长的根系。
Insufficient long-term accumulation supporting technological application. For many key technologies in future industries, moving from an experimental breakthrough to stable application often requires a prolonged process of theoretical deepening, process optimization, reliability verification, and the improvement of industrial support systems. It is not the case that an industry naturally forms once laboratory results exist, nor that market competitiveness is acquired once a prototype or sample product exists. In fact, many frontier achievements are not stuck at the question of ‘whether they can be made’, but rather at ‘whether they can be made stably, at scale, and sustainably’. For example, some high-end chips can already complete the trial production of sample wafers, but reaching large-scale mass production still requires continuous improvement in areas such as process maturity and yield control; behind the rapid development of the low-altitude economy, related research on complex-environment perception, autonomous flight control, and airspace coordination requires long-term deepening; and in some advanced materials and biomanufacturing fields, new technological bottlenecks are also encountered at the stages of pilot-scale scale-up, process stabilisation, and large-scale preparation. An important reason these problems arise is that some fields pay more attention to phased results and rapid translation while attaching insufficient importance to long-term stable accumulation and the building of generic foundational capabilities; some research deployments are prone to frequent adjustment in step with market heat and the direction of capital, and the phenomenon of research resources ‘chasing hot spots’ still exists, so that some fields appear to progress very quickly while the core capabilities that truly sustain continued industrial development have not formed in parallel. At the same time, between ‘paper achievements’ and ‘industrial capabilities’ there remains a relatively long translation cycle and a relatively high translation threshold. This causes many achievements to have visible prospects yet be unable to leave the laboratory; some technologies can produce samples but cannot become industries. As a result, the development of future industries faces the most immediate and most difficult commercialization chasm. 长期积淀支撑技术应用不足。未来产业很多关键技术从实验突破走向稳定应用，往往需要经历长期的理论深化、工艺优化、可靠性验证和产业配套完善，而不是有了实验室成果就能自然形成产业，也不是有了样机样品就具备了市场竞争力。事实上，许多前沿成果并非卡在“能不能做出来”，而是卡在“能不能做得稳、做得大、做得可持续”。比如，一些高端芯片已经能够完成样片试制，但距离规模化量产，仍需在工艺成熟度、良率控制等方面不断提升；低空经济快速发展背后，复杂环境感知、自主飞行控制、空域协同等相关研究需长期深化；一些先进材料、生物制造领域，在中试放大、工艺稳定、规模化制备阶段也面临新的技术瓶颈。出现这些问题的重要原因在于，一些领域更关注阶段性成果和快速转化，对长期稳定积累和共性基础能力建设重视不够；一些研究布局容易随着市场热度和资本风向频繁调整，科研资源“追热点”现象仍然存在，致使有的领域看似进展很快，但真正支撑产业持续发展的核心能力并未同步形成。同时，从“论文成果”到“产业能力”之间，还存在较长的转化周期和较高的转化门槛，致使很多成果看得见前景、走不出实验室，一些技术做得出样品、做不成产业，导致未来产业发展面临最现实、最艰难的转化鸿沟.
Insufficient support for system innovation through cross-disciplinary collaboration. Breakthroughs in many frontier directions of future industries increasingly no longer result from the linear advancement of a single discipline or a single technology. Rather, they depend on the formation of systemic innovation capabilities through deep multidisciplinary integration and coordinated efforts by multiple actors. At present, however, basic research in some fields still tends toward fighting one’s own battles in isolation: there is much ‘physical splicing’ between disciplines and little deep integration, and some interdisciplinary research remains more at the level of stacking directions, combining projects, and short-term cooperation, without truly forming an organisational model for conducting collaborative effort around major scientific questions. For example, quantum technology requires not only breakthroughs in basic theory but also the coordinated advancement of materials, devices, algorithms, precision manufacturing, and more; the development of synthetic biology likewise depends heavily on interdisciplinary support from computational science, chemical engineering, intelligent manufacturing, and other fields. At the same time, some research platforms and innovation resources are insufficiently open and shared, the mechanisms for interdisciplinary talent cultivation, joint research, and the evaluation of results are not well developed, and the approaches to research evaluation and resource allocation still largely follow traditional disciplinary logic, providing insufficient support for long-term interdisciplinary collaborative research and making it difficult to form a sustained and stable capacity for systemic innovation. Competition in future industries is, to a large extent, a contest of complex-systems innovation capability and cross-domain collaboration capability; without interdisciplinary integration in the true sense, it is difficult to achieve major breakthroughs that are original and pioneering. 交叉协同支撑系统创新不足。未来产业很多前沿方向的突破，越来越不是单一学科、单项技术的线性推进，而是多学科深度交叉、多主体协同攻关形成系统性创新能力。但当前部分领域基础研究仍存在各自为战的倾向，学科之间“物理拼接”多、深度融合少，一些交叉研究更多停留在方向叠加、项目组合和短期合作层面，尚未真正形成围绕重大科学问题开展协同攻关的组织模式。例如，量子科技不仅需要基础理论突破，还需要材料、器件、算法、精密制造等多方面协同推进；合成生物学的发展也高度依赖计算科学、化学工程、智能制造等交叉支撑。与此同时，一些科研平台和创新资源开放共享不足，跨学科人才培养、联合攻关和成果评价机制不健全，科研评价和资源配置方式还较多沿用传统学科逻辑，对长期交叉协同研究支持不够，难以形成持续稳定的系统创新能力。未来产业竞争，很大程度上比拼的是复杂系统创新能力和跨领域协同能力，如果缺少真正意义上的交叉融合，就难以形成具有原创性和引领性的重大突破.
In the final section, Zhang offers some solutions to all these challenges. He argues that:
High-level research universities must tightly integrate serving the country’s future-industry layout with strengthening basic research, take the initiative to distil major scientific questions from the trends of industrial evolution and to mine directions for basic research from the country’s major strategic needs, and truly connect the scientific frontier with the industrial frontier…Especially in frontier directions such as artificial intelligence, quantum technology, brain-computer interfaces, and biomanufacturing, they must not only participate in technological breakthroughs but also strive to become important incubators of new theories, new methods, and new paradigms. 高水平研究型大学必须把服务国家未来产业布局与强化基础研究紧密结合起来，主动从产业演进趋势中凝练重大科学问题，从国家重大战略需求中挖掘基础研究方向，真正把科技前沿与产业前沿贯通起来...特别是在人工智能、量子科技、脑机接口、生物制造等前沿方向，不仅要参与技术突破，更要努力成为新理论、新方法、新范式的重要孕育者.
High-level research universities must accurately grasp the laws governing the development of future industries, maintain strategic resolve and historical patience, and attach greater importance to long-term stable support and continuous capability building. 高水平研究型大学必须准确把握未来产业发展规律，保持战略定力和历史耐心，更加重视长期稳定支持和持续能力建设.
Finally, high-level research universities should play an even greater leading role in promoting interdisciplinary integration. They must break down traditional disciplinary boundaries and organisational barriers, drive interdisciplinary crossover to shift from the splicing of projects toward deep integration and from short-term cooperation toward deepened long-term collaboration, build cross-disciplinary collaborative research systems around major scientific questions, and work hard to refine a talent-cultivation system suited to the needs of future-industry development. 高水平研究型大学作为学科门类最齐全、创新资源最集聚的创新主体，更应在推动交叉融合方面发挥引领作用。要打破传统学科边界和组织壁垒，推动学科交叉从项目拼接向深度融合转变，从短期合作向长期协同深化，围绕重大科学问题构建跨学科协同攻关体系，着力完善适应未来产业发展需要的人才培养体系.
Third, another article that’s interesting is this one by Sun Maosong, who is the Executive Vice Dean, Institute for AI at Tsinghua University. Sun writes about the impact of generative AI on knowledge production.
Sun’s primary argument is that LLM’s have created, and will create, tremendous opportunities and are opening new vistas in terms of knowledge production and dissemination. But there are serious challenges:
First, there is an issue with regard to hallucination, and humans basically not knowing when and what is hallucination, particularly if one is not a subject-matter specialist.
Second, there’s a design challenge, i.e., the fact that they function on probabilistic consensus in terms of outcomes rather than generating content that reflects individual character and distinctiveness. This creates a tendency toward mediocrity and homogenisation in knowledge production. Sun is referring to more than AI slop when he says this. I think what he is talking about is the lack of depth, tacit knowledge, practical understanding of day-to-day living, or how bureaucracies actually function in the case of public policy.
Third, Sun posits the argument that in a sense with LLM’s and generative AI, we might be again turning towards the “one-to-many” form of digital engagement. I think this too reflects a deeper anxiety, i.e., a certain loss of cognitive autonomy and shifting of knowledge authoritativeness in favour of LLMs as opposed to human expertise.
To Sun’s arguments I want to add a further concern. As a growing number of people come to rely on LLMs to carry out their writing and thinking, they accelerate the pace of knowledge production — but in doing so they invariably bypass the labour through which knowledge is absorbed. This matters greatly if humans are to remain the subject, as Sun — and frankly I — desire. Without having distilled, absorbed, and ruminated upon knowledge, and then added their own distinctive tacit and human understandings, people will struggle to cultivate the very capabilities that would allow them to leverage LLMs to their fullest potential. To use these tools well, one must know which connections to make, which questions to ask, which inferences are possible, and which solutions are feasible. This depends on the maintenance of cognitive autonomy, which in turn relies on the labour of knowledge absorption.
Anyway, here’s a breakdown:
Knowledge production is inseparable from two key links, namely the creation of knowledge and the dissemination of knowledge. These two are mutually reinforcing: good knowledge is the driving force behind good dissemination, while good dissemination, through its civilising effect, can generate more good knowledge as people apply what they have learned to the practical problems of production and daily life. Throughout history, this process has operated as a continuous cycle, without a clear beginning or end, steadily enriching humanity’s accumulated stock of knowledge. 知识生产与两个关键环节密不可分：知识的产生和知识的传播。两者相辅相成：好的知识是好的传播的动力源，而好的传播又能通过知识的“教化”作用，在人们利用所学解决各类问题的生产和生活实践中催生更多好的知识。古往今来，上述过程“环复转运，终始无端”，不断丰富着人类的知识宝库.
Over the long arc of history, the production and expansion of knowledge bore the distinctive imprint of a small number of ‘wise people’. As the ages progressed, this group gradually grew, but remained a minority relative to the population as a whole. Entering the modern era and especially the age of the Internet, this situation changed dramatically, and ordinary people could contribute to knowledge production through Internet information technology. The evolution of knowledge dissemination has been even more astonishing—from oral transmission before the invention of writing, using the spoken word as the carrier of knowledge; to the hand-copying of texts on clay tablets, papyrus, parchment, oracle bones, bamboo strips, and silk several thousand years ago; to the printing of books on paper over the past thousand years and especially the past several centuries; to the production of analogue and digital content via radio, film, and television over the past century or more; and finally to the cross-modal production of digital content via the Internet and social networks over the past thirty years. Knowledge production has driven the advance of human civilisation through the enormous energy released by its successive paradigm shifts. Looking back across history, whether paper-making and printing, or analogue and digital communications and the Internet, each represented a major scientific and technological invention of its era. Each great leap forward in science and technology has brought about a new paradigm innovation in knowledge production. 在漫长的历史演进中，知识的生产与拓展，长期带有少数“智者”留下的鲜明印记。随着时代的发展，“智者群”的数量会渐次扩大，但在整个人群里相对来说仍是少数。进入现代社会特别是互联网时代，这种状况发生了急剧变化，普罗大众可以依托互联网信息技术对知识的生产作出贡献。知识传播模式的演进更是令人叹为观止，从文字产生前以语音为知识载体的口口相传，数千年前以泥板、莎草纸、羊皮卷、甲骨、竹简、绢帛等为载体的文本手抄，到1000多年前尤其是数百年来以纸张为载体的图书印刷，百余年来以广播、电影、电视为载体的模拟及数字内容生产，再到近30年来以互联网和社交网络为载体的跨模态数字内容生产，知识生产以其范式持续更替所引发的巨大能量推动着人类文明进步。纵观历史，无论是造纸术、印刷术，还是模拟和数字通信、互联网等，无一不是当时的重大科技发明。科技的一次次重大进步，促成了知识生产的一次次范式创新.
Generative AI based on large language models that has flourished in recent years will undoubtedly bring about yet another paradigm shift in knowledge production. Unlike all previous knowledge production paradigms, in which human beings were always the sole subject, this shift introduces for the first time an entirely new ‘subject’, i.e., the machine. The subjects involved in dissemination can now be either human or machine, and the two are interwoven with each other, forming an extraordinarily complex web of multiple relationships. It is well known that large language models possess powerful language generation capabilities, as well as strong open-ended language comprehension. That a machine can achieve this is a situation without precedent across the centuries. A series of dazzling technological innovations—including vector representations, self-attention mechanisms, reinforcement learning, and chain-of-thought reasoning—combined with the powerful support of large parameters, large-scale data, and massive computing power, as well as the ‘mysterious’ effects of emergence in complex systems, have converged to create this miracle. While basically learning human language, large language models have also integrated into themselves, in parameterised form, the vast stores of worldly knowledge and logical reasoning capabilities embedded in enormous corpora of text. Yuval Harari, author of Sapiens, has said: ‘Human culture is based on language. And because AI has cracked language, it can now begin to create culture’. Events seem to be bearing out this judgment. At the start of this year, an open-source agentic framework called ‘OpenClaw’—nicknamed the ‘Lobster’—took the world by storm, driven at its core by the excellent comprehension and generation capabilities of large language models. The large language model is in a sense the ‘brain’ of OpenClaw, forming in a certain respect an ‘operating system’ for AI, capable of building and running various ‘skills’ to complete all manner of tasks, including knowledge production. This offers a rare developmental opportunity for the broader application of AI across economic, cultural, and social domains. 近年来勃兴的以语言大模型为基础的生成式人工智能，无疑会促成知识生产的又一次范式变革。与此前所有知识生产范式中人类始终是唯一主体不同，这次破天荒地增加了一个全新的“主体”——机器。传播模式中所关涉的主体，现在既可以是人，也可以是机器，并且两者相互杂糅，构成了异常复杂的多重关系。众所周知，语言大模型具有强大的语言生成能力，同时具有较好的开放式语言理解能力。机器能做到这一点，是千百年来未曾有过的局面。向量表征、自注意力机制、强化学习、思维链等一系列令人眼花缭乱的技术创新，加之大参数、大数据、大算力的强力加持，以及复杂系统涌现机理的“玄妙”作用，诸多要素的风云际会造就了这一奇迹。语言大模型在基本学会人类语言的同时，还将海量语料中蕴含的万千世界知识及逻辑推理能力，以参数化的形式整合到模型中。《人类简史》的作者尤瓦尔·赫拉利曾表示：“人类文化基于语言。而因为人工智能已经破解了语言，它现在可以开始创造文化。”事态的发展似乎正在印证这个判断。今年年初，被称为“龙虾”的OpenClaw开源智能体框架火爆全球，其核心驱动正是语言大模型优良的理解和生成能力。语言大模型可以说是OpenClaw的“大脑”，一定意义上形成了人工智能的“操作系统”，可用来构建并运行各类“技能”，以完成包括知识生产在内的各种任务。这为人工智能在经济文化社会等领域的推广应用，提供了难得的发展机遇.
At the same time, however, it must also be recognized that large language models are far from perfect. In essence, they are probabilistic models. In language understanding, they cannot guarantee the same level of precision and reliability as human beings; in language generation, they exhibit an ‘innate’ phenomenon of ‘hallucination’, namely the generation of false information. In other words, OpenClaw’s ‘brain’ becomes confused from time to time. More problematic still is that it is difficult to predict when it will become confused and what kind of confusion it will produce. As a result, when we grant OpenClaw the authority to take ‘actions’, it cannot guarantee that it will not make mistakes. It is rather like ‘catching mice in a porcelain shop’, i.e., even with extreme caution, breaking one or two pieces of porcelain is hard to avoid. That is to say, OpenClaw inherently possesses security and safety risks. In addition, the Internet corpora used to train large language models often contain various human prejudices and misconceptions, causing the models themselves to possess prejudices and misconceptions as well. A random-sample survey showed that, as of March 2026, the number of articles on the Internet generated by artificial intelligence was roughly equal to the number written by humans. Although this survey was relatively preliminary and suffered from shortcomings such as an excessively small sample size, it nevertheless reflects a phenomenon from another angle: nowadays, a considerable number of articles on the Internet are written by machines, yet we may be completely unaware of this and be influenced by them unconsciously. This process is still only at its very beginning. If necessary governance measures are not implemented, it may become increasingly intense in the future and could even involve issues such as ideological security, responsibility and ethics, academic misconduct, and intellectual property rights. When generative AI engages in knowledge production, it presents us with new challenges from many perspectives, including authenticity, reliability, and security. 但同时也要看到，语言大模型远非完美。本质上它是概率模型，在语言理解上，不能保证如人类一般精准可靠；在语言生成上，会出现“与生俱来”的“幻觉”现象，即生成虚假信息。换言之，OpenClaw的“大脑”时不时会犯糊涂，更麻烦的是，何时犯糊涂、犯怎样的糊涂，很难预料。其结果是当我们将“动作”的权力赋予OpenClaw时，它不能保证不出错，就好像在“瓷器店里捉老鼠”，即使非常小心，打碎一两件瓷器也在所难免，也就是说，OpenClaw天然存在安全隐患。此外，训练语言大模型使用的互联网语料中，往往掺杂人类的各种偏见与谬见，致使模型也存在偏见与谬见。一项随机抽样调研显示，截至2026年3月，互联网上由人工智能生成的文章数量与人类撰写的文章数量基本相当。尽管这个调研比较初步，存在采样规模过小等不足，但还是能从侧面反映出一种现象：现如今互联网上的文章不少是机器写的，我们却可能浑然不觉，无意识地受其影响。这个进程尚处于刚刚开始的阶段，如果不予以必要治理，今后会愈演愈烈，甚至会涉及意识形态安全、责任伦理、学术不端和知识产权等问题。生成式人工智能在进行知识生产时，从真实性、可靠性、安全性等诸多角度对我们提出了新的挑战.
Deeper problems also arise from all this. For example, large language models are trained on enormous volumes of text and can be said to have aggregated and synthesised the knowledge of all people. The highest-performing models are likely to win the widest possible user base, and so the model comes to resemble a single ‘wise person’. In the closed-loop iteration of ‘everyone trains the model and the model serves everyone’, the ‘many-to-many’ dissemination mode of the Internet era may paradoxically regress to the ‘one-to-many’ mode of the print and television eras; except that the ‘one’ is now a machine rather than a human being. Furthermore, large language models tend to generate content reflecting greater probabilistic ‘consensus’ while easily overlooking content of individual character and distinctiveness, leading to a certain tendency toward mediocrity and homogenisation in knowledge production. Research has shown that generative AI can enhance individual creativity while reducing the collective diversity of ‘novel’ content. This apparently paradoxical conclusion is in fact not difficult to understand. 这里面还会生发一些更为深刻的问题。例如，语言大模型是通过海量文本训练出来的，可谓汇总、融通了所有人的知识，性能顶尖的大模型可望得到最广泛人群的青睐和使用，于是乎它仿佛成了唯一的“智者”，在“人人为模型，模型为人人”的闭环迭代中，互联网时代“多对多”的传播模式反而有可能退回到过去印刷时代与电视时代“1对多”的传播模式，只不过此时的“1”已变成了机器，而不再是人了。此外，语言大模型会倾向于生成概率意义上“共识”较大的内容，而易忽视具个性禀赋的内容，导致知识生产出现某种平庸化、同质化倾向。有研究表明，生成式人工智能可以提升个体创造力，却降低了“新颖”内容的集体多样性。这个似乎有些矛盾的结论，其实是不难理解的.
The pervasive issues encountered in the innovation of knowledge production compel us to engage in deep reflection: First, how should we navigate the relationship between humans and machines during the processes of knowledge production and dissemination? There is no doubt that human beings should always occupy the dominant position in this process. Generative AI should be a good assistant and partner to humanity, but must never ‘let the guest usurp the host’s place’. In other words, the human being must naturally remain the subject and core. Second, given the basic condition of human-machine coexistence, how can we more actively bring into play the human capacity for initiative in knowledge production? Third, generative AI can accelerate the discovery and integration of knowledge, as well as the dissemination of knowledge that is present at all times and reaches everywhere regardless of distance. On this basis, how can a new paradigm of knowledge production be effectively reconstructed, enabling it to advance with the times and advance together with the world? Fourth, generative AI has lowered the barriers to knowledge production; how do we seize this opportunity to better mobilise and tap the knowledge-producing capacities of the general public, forming a complementary relationship with expert knowledge production? 知识生产创新中面临的此类带有普遍性的问题，促使我们必须深入思考：第一，在知识生产与传播进程中，如何处理好人和机器的关系？毋庸置疑，在这个进程中人始终应占据主导地位。生成式人工智能应成为人类的好助手、好伙伴，但绝不能“反客为主”，人是当然的主体和核心。第二，在人机共存的基本条件下，如何更积极地发挥人在知识生产中的主观能动性？第三，生成式人工智能可加速知识发现与整合，以及无时不在、无远弗届的知识传播，如何在此基础上有效重构知识生产新范式，使其与时、与世俱进？第四，生成式人工智能降低了知识生产门槛，如何以此为契机更好调动和挖掘大众知识生产能力，与专家知识生产形成互补?
On the whole, the rapid development of large language models, and even generative multimodal large models, has brought unprecedented opportunities and challenges to innovation in knowledge production. Facing the profound transformations brought about by technological iteration, we should always hold high human subjectivity, explore the path of innovation in knowledge production through collaboration between humans and machines, and enable generative AI truly to become an accelerator of the progress of human civilisation. 总的来看，语言大模型乃至生成式多模态大模型的快速发展，给知识生产创新带来前所未有的机遇和挑战。面对技术迭代引发的深刻变革，应始终高扬人的主体性，在人与机器协同中探索知识生产的创新之路，让生成式人工智能真正成为人类文明进步的加速器.
Although I am not covering it in detail, the foreign affairs article in the journal this time around is about the threat of Japan’s “neo-militarism”. The author identifies the following as the reasons for the “growing emergence of Japan’s neo-militarism”, while terming it as a major threat to the regional order and global order.
First, an incomplete historical reckoning within Japan.
Second, the intergenerational transmission of conservative politics, i.e., “From Nobusuke Kishi to Yasuhiro Nakasone, from Shinzo Abe to Sanae Takaichi, Japan’s postwar conservative politics has been threaded throughout by the political fixation on ‘breaking free of the post-war system’, and its family lineages and political mentorships can mostly be traced back to the pre-war ruling establishment.” 从岸信介到中曾根康弘，从安倍晋三到高市早苗，日本战后保守政治始终贯穿着“摆脱战后体制”的政治执念，其家族谱系与政治师承大多可上溯至战前统治集团。
Third, the accumulation of prolonged economic stagnation and deep-seated social difficulties is the social soil in which ‘neo-militarism’ breeds and spreads.经济长期低迷与社会深层困境的累积，是“新型军国主义”滋生蔓延的社会土壤. “Faced with the difficulties of economic and social development, Japan’s right-wing forces have chosen not to deepen regional cooperation and jointly pursue peaceful development, but instead have insisted on externalising internal contradictions and seeking a sense of presence through military expansion. Following the logic that ‘a strong country is bound to seek hegemony’, they regard China’s development as a ‘threat’, play up the ‘China threat theory’, incite narrow nationalism, and drum up support for military buildup and war preparation. On this basis, the Japanese government has further made the development of the military industry and the export of weapons into a new growth point, with the defence industry expanding rapidly, which in turn drives continued increases in military budgets and expansionist policies, adopting exactly the same path of industrial militarisation as the pre-war militarist approach of ‘using war to feed war’. Economic and social difficulties do not in themselves necessarily lead to militarism; what truly causes it to grow is Japanese right-wing politicians’ handling of the problems of economic and social development with a militarist mindset, and this is also the deep soil in which ‘neo-militarism’ is able to grow.” 面对经济社会发展困境，日本右翼势力没有选择深化区域合作、共谋和平发展，反而执意将内部矛盾外部化，以军事扩张谋求存在感。他们因循“国强必霸”逻辑，视中国的发展为“威胁”，渲染“中国威胁论”、煽动狭隘民族主义，为扩军备战造势。在此基础上，日本政府进一步将发展军工、出口武器作为新的增长点，防卫产业快速扩张，反过来推动军事预算与扩张政策持续加码，这与战前军国主义“以战养战”的产业军事化路径如出一辙。经济社会困境本身并不必然通向军国主义，真正使其滋长的，是日本右翼政客以军国主义思维处理经济社会发展难题，这也是“新型军国主义”得以生长的深层土壤.
Finally, the United States’ long-standing cultivation and indulgence of Japan is the external condition for the accelerated rise of ‘neo-militarism.’ Since the end of the Cold War, in order to sustain its strategic posture in the Asia-Pacific, the US has long pushed Japan to take on a greater military role, gradually acquiescing in its lifting of the ban on collective self-defence and its development of counterstrike capabilities, promoting the building of a joint combat system between US forces stationed in Japan and the Self-Defense Forces, and supporting the revision of Japan’s ‘three security documents’, thereby accelerating Japan’s militarisation. In this process, Japan has assiduously exploited the US’ strategic needs, actively free-riding, and has turned ‘every crevice’ of external security cooperation into an opportunity for its own military breakthroughs, from densely deploying long-range offensive weapons in the southwest direction to discussing the revision of the ‘three non-nuclear principles’ and even seeking ‘nuclear sharing’, thereby pulling the US-Japan alliance from its defensive arrangement of the early post-war period onto the track of its own military expansion, with the ‘neo-militarism’ agenda being substantiated step by step in the process. 美国长期对日扶植纵容，是“新型军国主义”加速抬头的外部条件。冷战结束以来，美国为维系亚太战略布局，长期推动日本承担更多军事角色，逐步默许其解禁集体自卫权、发展反击能力，推动驻日美军同自卫队联合作战体系建设并支持日本“安保三文件”修订，使日本军事化进程加速推进。日本在这一过程中处心积虑地利用美国的战略需要，主动搭便车，把外部安全合作的“每一处缝隙”都转化为自身军事突破的契机，从在西南方向密集部署远程进攻性武器，到探讨修改“无核三原则”乃至谋求“核共享”，一路把美日同盟从战后初期的防御性安排拉向自身军事扩张的轨道，“新型军国主义”议程在这一过程中得以一步步坐实.