Roundtable Dialogue: In the AI Era, How Does New Power Catalyze Industrial Chain Reactions | 36Kr Future Industry Conference 2025

On September 10, the 2025 36Kr Industrial Future Conference, hosted by 36Kr, grandly kicked off in Xiamen, China. This conference has joined hands with the "China International Fair for Investment and Trade" hosted by the Ministry of Commerce. With the core theme of "In the Era of Intensive Cultivation, the Tides Surge in the Land of Abundance", it strives to create a high - standard, high - value, and high - influence industrial event that combines national vision, industrial depth, and market popularity.

The conference closely aligns with national strategic orientations and the forefront of industrial development. It focuses on five core sectors: artificial intelligence, low - altitude economy, advanced manufacturing, new energy, and large - scale consumption. It brings together top industry players to discuss development paths and envision the future of the industry. During the two - day agenda, with the "industrial collaboration chain" as the main logical line, the conference focuses on the collaborative mechanism among the government, capital, and industry. It delves into how to break down barriers, integrate resources, and precisely address the pain points, bottlenecks, and constraints in industrial development.

36Kr organized a round - table discussion titled "In the AI Era, How Does New Electricity Trigger an Industrial Chain Reaction" at the Industrial Future Conference. Liu Yang, the co - founder of ELU (and its subsidiary Yuanli Wuxian), Jiang Dongyun, a partner at Songhe Venture Capital, Fang Xiaodong, a partner at Huafang Capital, Chen Zhenhao, the investment and financing director at Deep Robotics, Dr. Wang Chao, a partner and the dean of the Forward - looking Institute at Mainline Technology, and Wang Zhiwei, the managing director at Xinding Capital participated in the discussion.

The following is the content of the round - table discussion, edited by 36Kr:

Liu Yang: Good morning, everyone. I'm very glad to be in Xiamen. It's a beautiful day today, and I'm also very happy to discuss the future of AI related to the industry with all the guests here. Hello, I'm Liu Yang from Yuanli Wuxian. Today's topic is "In the AI Era, How Does New Electricity Trigger an Industrial Chain Reaction", which I think is a very interesting topic. As we all know, AI has been very popular recently, but the organizer has directly defined it as "new electricity", which is the driving force for the future. Under this theme, we have invited experts here. First, please introduce yourselves briefly. Let's start with Mr. Jiang.

Jiang Dongyun: Hello, everyone. I'm Jiang Dongyun, a partner at Shenzhen Songhe Venture Capital. Songhe Venture Capital has been investing in hard technology, new materials, and healthcare for more than twenty years. We're veterans in the venture capital industry. In the future, we'll still stick to these three directions. We hope that you'll share new opportunities and ideas with us. Thank you.

Fang Xiaodong: Hello, everyone. I'm Fang Xiaodong from Hangzhou Huafang Capital. We're a CVC. Since Hangzhou Huali Group has been among the top three suppliers of smart meters for the southern and national power grids in China and also exports to places like Kazakhstan, the Middle East, and Central Asia, we've indeed made many arrangements in the power field. Our capital also invests in cutting - edge technologies such as controllable nuclear fusion and solid - state batteries. So, I'm very glad to be at this power - related forum. Thank you all.

Chen Zhenhao: Hello, everyone. I'm Chen Zhenhao, the financing director at Deep Robotics. We're an embodied intelligent robot company. In the past few years, we've been focusing on the R & D, production, manufacturing, and sales of robotic dogs and humanoid robots. This year, due to the popularity of embodied intelligence, we've received attention from the capital market and society. I'm very honored to be invited to communicate with you today. Thank you.

Dr. Wang Chao: Hello, everyone. I'm Wang Chao from Mainline Technology. Mainline Technology is a globally leading provider of L4 - level autonomous driving trucks and intelligent transportation solutions. We mainly target large - scale low - speed closed - loop logistics hubs represented by ports, port areas, and high - speed trunk logistics transportation. Our services cover logistics scenarios such as express delivery, dedicated lines, and bulk cargo transportation. Currently, we've cumulatively delivered and operated nearly a thousand sets of intelligent trucks and intelligent transportation solutions, with a cumulative intelligent transportation mileage of nearly 100 million kilometers. I'm very honored to be invited to this event and share my views with you. Thank you.

Wang Zhiwei: Hello, everyone. I'm Wang Zhiwei from Xinding Capital. We're a leading domestic private equity investment fund. We mainly invest in three core sectors: semiconductors, aerospace, and new energy. We've invested a relatively large amount in the semiconductor field, with nearly 50% to 60% of our funds going into the semiconductor industry. Projects like Cambricon and Hygon Information, which rank among the top three on the Science and Technology Innovation Board, are our investment projects. I'm very glad to be here and hope to interact and communicate with you more.

Liu Yang: First, I'd like everyone to share your views. When we talk about AI empowering the industry, currently, we're mainly talking about using AI to improve efficiency. From your perspectives, apart from the basic goal, what deeper incremental value does AI have in today's manufacturing industry? When we judge this, we first think from the perspective of efficiency improvement. When we mention AI, the first impression is that it's a tool. In fact, in the long - run, if we really reach the AGI era, we don't think AI is just a simple tool; it has deeper meaning. From today's perspective, please share your views. Let's start with Mr. Jiang.

Jiang Dongyun: I think that the cost - reduction and efficiency - improvement effect of AI is just a superficial phenomenon. In fact, AI is reconstructing many production relations and productive forces in the following three aspects:

First, AI is making more ideas possible. We can see that many flexible production lines have emerged. For example, if you want to make a consumer - grade drone, in the past, it required the support of several large enterprises, and the whole process was very long. Now, AI - enabled flexible production line companies can quickly provide various personalized components, enabling R & D ideas in the manufacturing industry to be realized more quickly and flexibly.

Second, AI has led the manufacturing industry from "experience - based manufacturing" to "data - based manufacturing". For example, there's a shortage of experienced workers in many factories. Now, the combination of AI and manufacturing has enabled many production lines to achieve automatic production. For instance, AI - powered quality inspection robots have improved the yield rate of many products.

Third, in terms of supply - chain relationships, AI data can not only trace the current production situation upstream but also see the downstream products and even the usage of end - users. For example, Xingxing Charging at today's venue can not only see the layout of charging piles but also the usage of users. This makes the production relations in the entire production chain closer and greatly improves production efficiency.

Overall, the effect of AI in manufacturing is not just about financial data. It's transforming the entire production elements and components from "manufacturing" to "intelligent manufacturing". Thank you.

Liu Yang: Thank you, Mr. Jiang. You've directly challenged our question and shared your unique views.

Fang Xiaodong: In my opinion, the most important thing in the manufacturing industry is cost - reduction and efficiency - improvement. The question has already asked what else AI can do besides that. As China has developed to this stage, apart from cost - reduction and efficiency - improvement, the first thing the manufacturing industry needs to consider is innovation. Currently, AI has not been widely popularized in the innovation field, especially in the R & D end of the manufacturing industry. In some simple fields, such as 3D printing and some component manufacturing fields, it has indeed been helpful. However, in more industrial environments, there's still a gap in areas such as material simulation, aerodynamics, fluid dynamics, optics, and magnetics simulation. So, we believe that in the future, AI will have great potential in the innovation aspect of the industrial front - end.

Second, in our view, if AI is only used for the quality inspection process, it's not enough to improve the product yield rate. Besides cost - reduction and efficiency - improvement in the industrial environment, we also need to create new products and improve product quality. Improving product quality is not just about screening out defective products; it's about increasing the yield rate from 70% to 80% or 90%. Only with a higher yield rate can enterprises be more competitive. So, I think AI may have greater breakthrough scenarios in the two directions of innovation and improving product yield rate.

Liu Yang: Thank you, Mr. Fang. You've summarized two key points: innovation and quality. Mr. Chen, from your perspective, in the past, when we talked about robots, it was more about programming them to move from point A to point B. But now, with the arrival of the so - called AI era, how do you view this?

Chen Zhenhao: To be honest, currently, the public has slightly over - expected the capabilities of embodied intelligence. In fact, what embodied intelligent robots can do in the industrial field is still in its infancy. In many fields, we're still in the process of learning and innovating. At present, neither our robotic dogs nor the small number of humanoid robots can perform tasks as complex as screwing, let alone replace a large amount of manual labor. Currently, they can only perform some simple and repetitive operations. We can only say that what we're doing now is more complex than what traditional robots used to do and can adapt to certain scenarios. But we're still in the early stage.

I think ultimately, whether it's robots or AI, the goal is to achieve cost - reduction and efficiency - improvement. In the past, we only thought about direct cost - reduction and efficiency - improvement. However, whether it's improving the yield rate or achieving other goals, it ultimately boils down to cost - reduction and efficiency - improvement from a business perspective. It may be achieved in a more long - term and complex indirect way. I think the biggest feature is that it will reshape the entire production relations to a large extent. For example, at present, robots often replace humans in many repetitive and boring jobs, liberating most humans from such work and allowing them to focus more on creative and organization - coordinating work.

I think in the future, people will increasingly find that what they manage is no longer humans but robots. The important thing is no longer how to perform simple and repetitive work but how to unleash human creativity. This may be the significant change that AI will bring to the entire production relations in the future.

Liu Yang: Thank you, Mr. Chen. I really agree with your view. The entire embodied intelligence industry is still in its very early stage. People's understanding of the concept of embodied intelligence and the subsequent development of the industrial chain are indeed still in the early days. All guests can pay attention to companies like Deep Robotics where Mr. Chen works and our Yuanli Wuxian, which are all working in the field of integrated intelligence. Maybe in the next 20 years, we'll see BAT - like and Huawei - like companies emerging in the embodied intelligence industry. As Mr. Chen said, efficiency is just a result. What really matters is the change in the production relations and production chain during the process, which is what we should focus on more.

Dr. Wang Chao: I agree with the previous speakers. Cost - reduction and efficiency - improvement are the common goals pursued by the manufacturing industry and even the entire industry. It's also to create greater economic benefits. In the past, cost - reduction and efficiency - improvement were achieved by simplifying processes, optimizing product quality, and making adjustments to personnel.

Now, the use of AI technology is more about filling the labor gap and freeing up some labor to do more efficient work. As the industry has developed to a certain stage, it's necessary to change and optimize work processes.

For example, as mentioned earlier about the supply chain, we know that artificial intelligence mimics human capabilities and can, to a certain extent, exceed human capabilities. Especially in some complex and large - scale systems with high - dynamic uncertainties, in the past, the management of the entire supply chain might have relied on experienced personnel who were very familiar with the upstream and downstream and had accumulated knowledge over a long time. But now, it's data - driven. As Mr. Jiang mentioned, in areas such as demand forecasting, inventory optimization, in - transit logistics management, and production scheduling, a holistic approach is needed. This is a fundamental innovation, and the same applies to many manufacturing processes.

Some industries with a high degree of standardization have already achieved automated and even unmanned production and manufacturing. The rapid development of AI technology has given small and medium - sized manufacturing enterprises the opportunity to transform into intelligent manufacturing and can even empower the development of the entire manufacturing industry. As Mr. Fang said, AI has provided an opportunity for the manufacturing industry to overtake on the curve, similar to the development of intelligent vehicles.

Thank you.

Liu Yang: Thank you, Dr. Wang. Mainline Technology where Dr. Wang works is a typical company that applies AI very well. We know that new energy is a national strategy. When combined with AI, we not only have an opportunity to overtake on one curve but also on another. Now, let's invite Mr. Wang.

Wang Zhiwei: Hello, everyone. I think that in China, a leading AI company needs to emerge in the field of AI digitization. How can this happen? I think we need to build an AI - enabled digital and industrial platform. This industrial platform is not just about a company's software development and application or digital transformation. It's more about integrating the data resources of the industrial sector, providing open interfaces, and creating a good foundation. Just as in the Internet era, SaaS (software as a service) was popular, in the industrial Internet era and the industrial digitization era, it may be about digitization and application. China needs a leading AI company to emerge, like Google and Apple, which can open up a platform for everyone to participate. It will also integrate experts from various industries, standardize the processes of experienced experts, design them into the platform to form standards, and include open interfaces to integrate the experience of many industries into the platform. Ultimately, this can truly solve the problem of industrial intelligence in various sub - industries, including improving cost - reduction and efficiency - improvement and increasing the yield rate.

Liu Yang: Thank you, Mr. Wang. Now, we're moving on to the second stage: solutions to key problems. First, I'd like to ask Mr. Jiang. We know that Songhe Venture Capital has invested in Youai Zhihe, which is widely used in high - end manufacturing fields such as semiconductors. What do you think is the key to convincing customers to introduce AI - enabled industrial mobile robots into factory production lines?

Jiang Dongyun: For mobile robots like Youai Zhihe that we've invested in, there's no real issue of "convincing" customers to introduce them into factory production lines. There are two key points for all these robots to enter factories: "I can meet your needs" and "the numbers add up".

Regarding meeting needs, let me give you some examples.

When moving a product from completion to the warehouse, for small consumer products, regular robots can handle it. However, when it comes to industrial - grade or automotive - grade components weighing up to 600 kilograms, it's beyond the capacity of ordinary robots. Robots equipped with Youai Zhihe's technology can solve this industrial - grade pain point with their load - bearing capacity, flexibility, and safety.

There are also some extreme application scenarios. For example, in a port environment with a poor Wi - Fi signal, can a mobile robot solve the transportation problem? In a mining area with bumpy and dusty roads, can it transport coal or parts? In extremely high - or low - temperature environments, such as in a cold - storage warehouse at minus ten degrees Celsius, can a robot perform transportation tasks? These are all pain points for the industry. As long as you can meet these needs, there's no need to even talk about the cost - benefit analysis; it's enough for them to give it a try. Currently, we can see many such robots emerging.

Second, regarding the cost - benefit analysis, there are two types of calculations: calculating "current costs" or "future costs". Calculating current costs involves simple labor substitution. However, we can see that many large enterprises and industry leaders are calculating future costs. These enterprises say that they have industrial - grade pain points, technological bottlenecks, and a desire to increase the value of the industrial chain in the future. They're willing to invest a large amount of money and give you enough time. They believe that in the future, they'll get a return on their investment.

Third, although the combination of AI and manufacturing is still in its early stage, I'm relatively optimistic. We can be patient and wait for it to gradually improve.

Liu Yang: Thank you. Mr. Jiang, you've mentioned a key word, "future", which matches our today's theme. Regarding the future, I have a question for you. How do you balance the weight of AI algorithm capabilities and years of manufacturing experience in the early - stage evaluation?

Jiang Dongyun: In the AI - manufacturing industry, we've seen two types of representative teams: one is composed of people from large companies like Tencent and Alibaba with expertise in AI algorithms; the other is from the manufacturing industry, who have identified many pain points in manufacturing and want to solve them. We've encountered both types of teams in our investment process. It's difficult for