Wu Qi/Wen

In the past few years, discussions on the integration of artificial intelligence (AI) and manufacturing have rapidly heated up. Intelligent quality inspection, predictive maintenance, AI scheduling, digital twins, industrial intelligent agents, etc. have become high-frequency terms in policy documents, industry forums, and enterprise planning in various regions. The discussions surrounding these applications mostly point to the same goal: improving quality, reducing costs, and increasing efficiency.

This is certainly not wrong, but if the understanding of China's manufacturing industry stays at this level, it is easy to mistake a more profound change for another familiar round of digital transformation. AI is not only rewriting several aspects in factories, but also the value distribution logic of manufacturing competition.

In the past, the advantages of China's manufacturing industry were mainly manifested in cost, efficiency, supporting facilities, engineering, and scale; In the future, these advantages will still be important, but they will no longer automatically correspond to higher value-added positions.

The combination of big models, industrial software, industry models, and data infrastructure is driving manufacturing competition from "who is better at production" to "who is better at organizing knowledge, data, and scenarios into capabilities". This means that a part of future competition will no longer only occur between the upstream and downstream of the industrial chain, but will also shift towards the ability level.

This is not an abstract technical judgment, but a very real industry problem. China still has a complete industrial system, dense industrial clusters, rich manufacturing scenarios, and strong engineering capabilities. These conditions give China a natural advantage in the era of industrial AI.

But the problem is that if these advantages are still only understood as the advantage of "making things", they may be re priced in the new competitive structure. What policies really need to answer is not just how to make factories smarter, but how to enable Chinese manufacturing to occupy a higher position in the new value distribution structure.

Reassess the logic of competition

In the past few years, companies have been repeatedly asked the same question: What is your AI strategy? Most answers focus on efficiency: faster document generation, more labor-saving customer service, more automated analysis, and more efficient R&D assistance. These changes are all real, but they are not the most important part yet.

What is truly worth noting is that AI is changing the underlying logic of enterprise competition: which ones are starting to appreciate and which ones are starting to depreciate; What still constitutes a moat, and what is losing its moat attribute.

In the past few decades, the basic logic of corporate competition has been quite stable: whoever can control the scarce key resources may gain higher profits. Capital, channels, information, experts, experience, scale, and organizational capabilities are valuable not because they are naturally noble, but because they are scarce and slow to replicate.

Especially in the manufacturing industry, many truly valuable things are not a single point of technology, but implicit knowledge accumulated over a long period of time: how to judge the process window, how to identify early defects, how to balance delivery, cost, and quality, and how to see anomalies before they occur.

The real change brought about by artificial intelligence is that some cognitive activities that used to heavily rely on expert experience and organizational hierarchy are rapidly reducing costs. In the past, activities such as analysis, comparison, screening, induction, pattern recognition, anomaly detection, scheme generation, and auxiliary judgment were expensive because they relied on long-term training of the human brain, complex collaboration, and repeated trial and error. Today, a considerable portion of them are being significantly accelerated by models, algorithms, and data systems, and are beginning to be supplied on a large scale.

We can summarize this change as' cognitive abundance '. Cognitive abundance in this article refers to AI significantly reducing the cost of acquiring, processing, and generating cognitive resources for enterprises, leading to the popularization of some previously scarce cognitive activities. It's not that judgment is no longer important, nor is it that experience has become ineffective, but rather that some highly scarce, expensive, slow, and difficult to replicate cognitive activities in the past are losing their original exclusivity. What is shaken is not the value of cognition itself, but the acquisition cost and barrier structure of a part of cognitive activities.

This is particularly crucial for the manufacturing industry. The manufacturing industry may seem like a world of 'things', but behind it lies a large number of high-value cognitive processes: product definition, process development, parameter optimization, defect diagnosis, quality judgment, supply chain collaboration, and operation and maintenance return. What is being rewritten today is not the physical foundation of manufacturing, but these high cost cognitive activities that have long been encapsulated in experience, processes, and organizational stability.

Therefore, the real question that China's manufacturing industry needs to answer today is not just whether it can continue to grow and strengthen, but how to continue to appreciate its advantages when the rules of the industrial ecosystem are rewritten and some cognitive activities are no longer as scarce as in the past.

The advantages of China's manufacturing industry are re stratification

Understanding the future of China's manufacturing industry cannot be separated from its practical starting point. The biggest advantage of China today is not a single point of technology, but a complete set of established real capability networks.

Firstly, there is a complete industrial system. In the past, a complete industrial system meant complete supporting facilities, low cost, and fast response; In the era of AI, it also means richer training scenarios, denser process feedback, faster model validation, and stronger capability diffusion.

Industrial AI is not "imagined" in the laboratory, but rather "ground" by real equipment, real working conditions, real defects, and real customers. Whoever has more real industrial scenarios has a better chance of turning models into capabilities. In this sense, the integrity of China's industrial system is not only a manufacturing advantage, but also becoming a training and validation advantage.

The reason why new energy vehicles, lithium batteries, photovoltaics and other fields are more prone to rapid iteration largely comes from this high-density cycle of "scenario data feedback diffusion".

Next is industrial clusters. The close range, multi-level, and fast response network formed by the Yangtze River Delta, Pearl River Delta, and several key industrial belts is still scarce globally. This physical density cannot be replaced by models, but will further appreciate due to intelligent scheduling, data backflow, and scheme sharing.

The discovery of a process problem, the testing of a solution, and the adjustment of a supply chain node can spread and form a closed loop at an extremely fast speed in many manufacturing clusters in China. This kind of practical organizational ability is precisely difficult for many countries to replicate. For industrial AI, this means that once a capability is validated in a factory, it can often quickly spread among neighboring companies, without the need for every company to start from scratch.

The third is the super large scale market. In the past, it mainly meant diluting costs and scaling up, but today it increasingly means verifying advantages. The maturity of industrial AI cannot rely solely on the R&D department, but must constantly trial and iterate in real products, real customers, and real operating environments. The significance of a super large market is not just about buying more, but also about enabling models to experience more boundary situations faster, expose problems earlier, and complete corrections faster.

However, at the same time, some traditional advantages of China's manufacturing industry are also undergoing changes, the most typical of which is the labor efficiency advantage. China's manufacturing industry is no longer primarily dependent on the lowest labor costs, but more on the large-scale supply of high-quality industrial workers and engineering technicians.

But the problem is that AI and automation will not only improve the efficiency of Chinese companies, but also reduce the manpower density required for other economies to catch up. In other words, China's efficiency advantage will not naturally disappear, but its sustainability will increasingly depend on whether experience can be digitized, processed, and modeled, no longer just relying on manpower scale and experience thickness.

There is another type of advantage that, if conversion is not completed, will exist in name but at a substantial discount, the most typical of which is industrial data.

Chinese companies do not lack data, what they lack is infrastructure that makes data available, accessible, trainable, and verifiable. Without this layer of transformation, data resources are only accumulated and will not form capabilities. If data cannot form a compatible, reusable, and traceable structure across devices, production lines, and factories, it will be difficult to support true industry models and industrial intelligence.

There is also a differentiation worth noting here: the impact of AI on consumer goods manufacturing and industrial goods manufacturing is not in the same direction.

For consumer goods manufacturers, AI enhances the power of "deconstruction": demand perception is faster, flexible production is stronger, and platform based enterprises and quick responders are more likely to erode existing brand barriers.

For industrial product manufacturers, AI is more likely to enhance customer lock-in: the accumulated working condition data, optimization models, and operation and maintenance knowledge from the continuous operation of equipment on customer sites will turn one-time deliveries into long-term relationships.

For consumer goods companies, AI is more like forcing them to enhance product definition, brand autonomy, and user direct connectivity capabilities; For industrial enterprises, AI is more like driving them from selling equipment to selling continuous services and solutions.

Therefore, a more accurate judgment is not whether China's advantages still exist or not, but that the advantages of China's manufacturing industry have begun to re stratify: some advantages are being amplified, some are being weakened, and some will gradually depreciate if the transformation of form is not completed.

The fundamental logic of manufacturing being rewritten

The most common misjudgment in the manufacturing industry is not the inability to understand technology, but the underestimation of the power of technology to rewrite the original competitive logic.

The first one that is loosening is the boundary between efficiency and flexibility. The manufacturing industry has long been constrained by a classic dilemma: large-scale standardization to lower costs, and small batch flexibility to meet diverse needs, both of which are often difficult to achieve simultaneously. Every switch means time loss, parameter restructuring, and yield fluctuations. But when the scheduling system can dynamically optimize the switching sequence, the detection system can adapt to new standards faster, and the parameter system can continuously adjust based on real-time feedback, the boundary of switching costs will be continuously lowered. So, the two tracks that were naturally divided into "high-end small batch" and "low-end large batch" in the past began to blur. A manufacturer with strong AI capabilities may compete on two tracks simultaneously.

The second barrier that is loosening is the experience barrier. The so-called "experience" in the manufacturing industry is essentially a large part of information barriers, but it has long been encapsulated in the tacit understanding of mentors, engineers, and organizations. One of the important things AI does is to make these barriers explicit: since it is essentially a high cost information processing and pattern recognition activity, it has the potential to be digitized, modeled, and algorithmic. Experience will not fail, but if it cannot be structured, it will continue to depreciate.

The third point that is loosening is the definition of quality advantage. Machine vision, state monitoring, multivariable process control, and digital twins are pushing quality management from "maintaining stability through experience" to "continuously approaching the optimum through the system". Once more and more companies can achieve a high level of consistency, quality consistency itself will turn from a competitive advantage into an entry ticket. True differentiation will migrate upwards, shifting from 'making equally good' to 'creating things that others cannot create'.

The fourth point that is loosening is the scale of optimization. The manufacturing management of the past two hundred years has essentially been focused on finding better local optima. The difference of AI lies in its increasing ability to simultaneously handle the relationships between multiple variables such as raw materials, equipment, scheduling, energy consumption, delivery time, and quality, thereby pushing optimization problems from local to global. Competition is no longer just about 'how to do better', but about 'optimizing what is right'.

AI will not eliminate manufacturing barriers, but it will rewrite them. In the past, knowing how to do it was more difficult than knowing what to do; In the future, as "how to do it" becomes easier, "what to do" and "why to do it" will become the real competitive high ground.

What will determine the position of China's manufacturing industry in the future is not only the height of the industrial chain, but also the position of the capability level

In the past 30 years, the main line of upgrading China's manufacturing industry has been to climb vertically along the industrial chain: from low-end to high-end, from assembly to independent research and development, and from following to leading. This route is generally problem free and has achieved significant results.

But the change brought by AI is that future competition may not necessarily be reflected in "who is at the higher end of the industry chain", but more likely in "who has stronger control and organizational power in key capability layers".

If the manufacturing industry is placed in a new competitive framework, it can be roughly divided into four layers. The first layer is the physical execution layer, which includes manufacturing, delivery, engineering, and large-scale response capabilities. The second layer is the industry translation layer, which refers to the ability to translate general AI, general algorithms, and general digital capabilities into usable solutions for a specific industry. The third layer is the intelligent infrastructure layer, which includes industrial software, digital twins, data infrastructure, industry models, and engineering knowledge base. The fourth layer is the rules and standards layer, which includes data formats, interface protocols, security authentication, model evaluation, and the ability to shape ecological rules.

It should be pointed out that this framework is not a universal value ladder, and not all industries are "more valuable at the top". In fields such as special materials, precision machining, and hot end components of aircraft engines, physical execution capability itself may be the highest value. The purpose of this framework is not to require all industries to 'move up', but to help businesses determine where the key layers of future competition lie in their industry.

But for the overall Chinese manufacturing industry, one judgment still holds: today's global competitive advantage is mainly built on the physical execution layer and the ultra large scale application layer. These advantages are still strong at present, but they are no longer automatically equal to future advantages. The future is not only determined by the position of the industrial chain, but also by the position of the capability layer.

Policy priorities for industrial AI in China

It is precisely on this issue that the development path of China's industrial AI should not simply copy the United States or Europe. It's not because we deliberately 'walk our own path', but because the starting points, advantages, and weaknesses of the three parties are different, and the optimal path will naturally not be the same.

The advantages of the United States lie in its basic models, leading platforms, computing power ecosystem, developer ecosystem, and standard spillover capabilities, making it easier to take the path of "occupying the intelligent core first, and then penetrating into industry".

Europe's advantages lie in industrial software, industrial equipment, industrial automation, engineering standards, and institutional governance, making it easier to follow the path of "first industrial systems and regulatory frameworks, and then promoting widespread adoption".

China's starting point is completely different. China's biggest advantage is not a single platform or software system, but a complete industrial system, ultra large scale manufacturing scenarios, industrial cluster density, and rapid engineering capabilities.

From the perspective of upgrading Chinese manufacturing, the most advantageous model is not to copy the platform route of the United States or the rule route of Europe, but to take a combination path that is more in line with its own comparative advantages. From the perspective of practical urgency, this path can be roughly divided into two categories: one is the foundational ability that must be filled as soon as possible, and the other is the strategic ability that determines the future high-level competitive position.

The first category is led by application and scenario traction. The biggest practical advantage of China is that it has the most complete industrial scenarios, the longest industrial chain, and the richest real working conditions. The top priority should not be to abstractly pursue the 'strongest universal model', but to transform massive real-world industrial scenarios into high-quality data, industry models, and replicable solutions. This will enable China to quickly form an advantage in the "industry AI translation layer".

Secondly, it is to place industrial data infrastructure in a higher strategic position than general digitization. China is not without data, but rather with a large amount of data that is stored but not used, difficult to circulate, difficult to connect, and difficult to form a high-quality training loop. Industrial data infrastructure is not auxiliary engineering, but a new type of infrastructure for manufacturing upgrading in the AI era.

Thirdly, it is based on industrial clusters rather than individual enterprises as the fundamental unit for promoting industrial AI. This is a unique condition of China and an advantage that is difficult for both the United States and Europe to replicate. When an AI solution is successfully validated in a factory, it can quickly spread to surrounding similar enterprises because they use similar equipment, face similar process problems, and even share supplier and engineer networks. Advancing on a cluster basis is more likely to generate economies of scale and diffusion effects than letting each enterprise explore independently.

The latter category is more directly related to whether Chinese manufacturing can move from execution advantage to high-level competition.

Firstly, industry solutions and intelligent product layers are the main battlefield for value migration. If China only uses AI to improve factory efficiency, it may remain in a position of "strong execution, weak platform" for a long time.

A more favorable path should be to promote leading enterprises to move up towards industry models, intelligent products, continuous services, and platform nodes. Not just creating products, but turning them into a system of continuous perception, continuous optimization, and continuous service. Siemens, Caterpillar, and other companies are worth paying attention to not only because they use AI, but also because they are embedding AI into their product systems, operations systems, and customer relationships, thereby transforming one-time deliveries into sustained value.

Secondly, we cannot give up on the weaknesses at the bottom level, but our strategy should focus more on the "bottleneck layer" rather than fully expanding it. China cannot bypass the underlying issues of advanced chips, core industrial software, and critical design toolchains. But the optimal strategy is not to fully replicate all the underlying layers at the same time, but to identify the most critical and influential low-level links that affect industrial capabilities and make high-intensity breakthroughs, while using algorithm efficiency, scenario advantages, and industry data to amplify the boundaries of available capabilities.

Finally, it is important to enter the standards and rules layer as early as possible, rather than waiting for the technology to mature before participating. If China only values applications and neglects rules, even if its industrial scale continues to lead in the future, it may still suffer losses in value distribution. The data format, interface protocol, security authentication, model evaluation, and trustworthy system of industrial AI are still in the early stages. China cannot only be the largest application market, but also strives to become an important rule participant.

It should be pointed out that the above paths have different meanings for different industries.

For industries that already have global competitiveness, the key is not to continue amplifying execution advantages, but to transform hardware, scale, and scenario advantages into intelligence and service layer advantages.

For industries that are still in the catching up stage, the key is to use AI to accelerate cognitive accumulation, rather than mistakenly believing that it can replace basic research and process precipitation.

For fields where there is still an intergenerational gap, it is even more important to identify the nodes where AI can truly rewrite the path, rather than using it as a shortcut to replace long-term basic research.

For industries with a unique global position, the key is to transform existing positions into higher value solution capabilities and regulatory influence.

At the end of the day, the policy focus should not only be on promoting more enterprises to "invest in AI projects", but should also shift towards data infrastructure, industry translation layers, the upward movement of intelligent products, the filling of gaps in key bottom layers, and the early adoption of rules, which truly determine their positions.

The biggest risk is that the old model is still effective

The biggest tension currently facing China's manufacturing industry is not the lack of direction, but the fact that the old model is still creating real returns, so the new direction is naturally prone to being delayed.

The successful model in the past can be summarized as: based on large-scale physical execution capability, with cost, efficiency, and response speed as the main competitive means, and continuous investment and scale expansion as the main growth methods. Today, this model has not failed. China's manufacturing scale still ranks first in the world, and its industrial chain advantages are still obvious. The international competitiveness of many industries has not been weakened. Because the old model is still effective, resource allocation, organizational attention, and strategic discussions naturally tend to continue amplifying existing advantages.

But what really needs to be vigilant is precisely this. The main reason for missing the paradigm shift is often not because they have not seen the new direction, but because the old model is still successful, and the investment in the new model always seems "not urgent enough".

The reason why today's window period will not last forever is because it has a specific closing mechanism. If the international standards for industrial AI are dominated by others, it will form a regulatory lock up; If key industries take the lead in forming strong platform network effects, they will form market lock-in; If industrial data infrastructure does not make substantial progress in critical stages, it will form infrastructure lock-in. Once the three factors are combined, China's most unique manufacturing endowment today may not be able to smoothly transform into high-level capabilities in the next stage.

For China's manufacturing industry, the key in the next three to five years is no longer whether it can continue to strengthen its position as the world's factory, but whether it can gradually become a key capability node and rule participant in the future industrial intelligence ecosystem while maintaining its position as the world's factory.

This is not a posture issue, but a resource allocation issue: the watershed of China's manufacturing industry in the future may not necessarily depend on how much more can be produced, but more likely on which layer can occupy in the new value distribution structure.

Of course, the premise for this judgment to hold is that AI will continue to penetrate into the core links of the manufacturing industry, and platformization and competition at the capability level can indeed partially form in the physical world. At present, these two premises themselves are still evolving.

The window still exists. But what is truly meaningful is not knowing that the window is still open, but knowing where to move resources from and to.

(The author is the former President of Roland Berger China and Vice Chairman of Accenture Greater China)