The Inflection Point Has Emerged, All Companies in the Future Could Become Robot Companies

Investing in stocks? Look to Golden Kylin Analysts’ Reports—authoritative, professional, timely, comprehensive—helping you uncover potential thematic opportunities!

(Source: Huxiu APP)

Cover Photo | Shot by Huxiu

The convergence of AI and robotics is driving a transformation in the physical industry. From specialized mechanical arms to general humanoid robots, the industry is evolving from industrial to commercial and household applications, sparking widespread discussion.

In this revolution, a core question remains unresolved: what is the methodology behind robot development?

Is it sticking to specialized paths, or evolving toward general-purpose robots? How will automotive companies entering the field impact traditional robot companies?

With rapid technological iteration, how should companies determine which investments will translate into real value?

In this episode of “Huxiu·AI Paradox,” we invite Pan Jing, Chairman and General Manager of Titanium米 Robotics (CEEMBA 2017), and Chang Li, Co-founder and Vice President of Jieka Robotics (CEEMBA 2025), both from the medical and collaborative robot fields. Today, we will explore the following topics through the real-world applications of Titanium米 and Jieka:

● Who will first bring specialized and general-purpose robots into daily life?

● Will automotive companies building robots disrupt traditional robot enterprises?

● Overseas expansion and challenges in the robotics industry?

● When will robots enter households and become part of daily life?

● What will the robot industry look like in ten years?

The host for this episode is Hao Junsheng, Chairman of Rehuo Group and investor (CEEMBA 2017). Below is the transcript of the discussion (edited):

Hao Junsheng: Recently, Tesla’s Optimus news has been hotly discussed. Do you think robots must be humanoid?

Chang Li: I don’t think they necessarily have to be humanoid. Cars are one of humanity’s greatest inventions, and robots share many similarities with vehicles, but they don’t have to be humanoid. Functionally, what’s most important is how to operate them most stably, quickly, and effectively.

Chang Li, Co-founder and Vice President of Jieka Robotics

Pan Jing: Yes, I started working on humanoid robots around 2005. In the long run, our phones, watches, even glasses might disappear from our lives someday because their functions are replaced by other technologies. But the form of robots will definitely see different experiments. For products like ours, humans may develop some special emotional attachments. Some theoretical studies suggest this could evoke fear, but also closeness. Currently, humanoid robots are still a platform to showcase technological development capabilities.

Pan Jing, Chairman and General Manager of Titanium米 Robotics

Hao Junsheng: Which do you think will enter our daily lives faster—specialized robots or general-purpose robots?

Pan Jing: I believe technology will drive robots from specialized to general-purpose. As general technology matures and costs decrease, it will empower various specialized fields, leading to continuous evolution. For example, collaborative robotic arms emerging from specialized production lines to interact with humans.

Chang Li: I share a similar view. Currently, robots are applied mainly in three scenarios: industrial, commercial, and household. As some general technologies, like AI, become more capable and cross-disciplinary integration increases, we will see leapfrogging. But almost all of this will be built on a foundation of specialization. I believe that whether specialized or general, the key depends on autonomous decision-making ability in future scenarios—that is, the level of intelligence.

Pan Jing: Let me give a metaphor: I think we’ve collected “four Dragon Balls,” but only three and a half are complete.

The first Dragon Ball is mature batteries, motors, and control systems. The second is miniaturization, portability, low power consumption, and low cost of sensors. The third is the development of chips and semiconductors, greatly reducing power consumption and increasing computing power. The last half is human-machine interaction after AI emerged—this is the real methodology for achieving robot generalization. In the past, engineers wrote code to solve kinematic and dynamic equations for trajectory smoothing and control. Now, with large language models, text is tokenized, and through training, different token arrangements generate value sequences. Replacing text tokens with pixels creates images.

The challenge with the other half is that, although there is abundant data for training large language models, the fundamental constraints of robot motion—preconditions and specific data—are lacking. We need data that corresponds to concrete applications, bridging silicon-based and carbon-based systems.

As for timing, I believe the robot industry cycle is around 15–30 years, but increased capital attention and focus could accelerate this process.

Hao Junsheng: Elon Musk says 2026 will be a significant milestone, and the next few years will be painful. Do you feel pain in developing robots?

Chang Li: I feel both pain and joy. Over the past two years, the environment and market for robots have improved. On the other hand, robotics is an interdisciplinary field requiring time and accumulation—product development depends on enterprise effort, skilled personnel, reasonable customer expectations, and openness and innovation. This process requires industry-wide and even national effort.

Pan Jing: Human pain comes from two points: one is unfulfilled desires; the other is uncertainty. This “pain” is especially evident in robotics, particularly at the industry level. For example, after so many years, are there truly mature technologies with strong generalization capabilities? Not really. We see many superficially “solved” problems, but in reality, they are far from being fully addressed.

Also, if a robot capable of doing household chores existed today and cost 2 million RMB, would anyone in Shanghai buy it? Certainly. But does it exist now? No.

So, I think if everyone is in pain because the technology isn’t mature enough, the focus should be on identifying where the gaps are and how to form collaborative efforts for innovation—insight is key. You need to understand the industry’s essence to set realistic expectations and reduce anxiety.

Hao Junsheng: Tesla is making robots, and many domestic EV manufacturers are eager to try. Will their entry impact traditional robot companies?

Chang Li: First, the fact that automakers are making robots indicates a positive industry trend. But Musk’s goals keep changing—from 100,000 to a million units, and now he firmly states Tesla will become a robot company. Li Auto also emphasizes robots and AI. Their entry can accelerate current technological progress, but the application scenarios for intelligent robots are still unclear.

Pan Jing: I agree. The robot industry is not yet fully industrialized, so cross-industry issues are limited. The robotics industry naturally aligns with the automotive industry. Car manufacturers use industrial robots to ensure standardization and consistency in production, especially in complex manufacturing processes. When automakers develop robots, the logic is to leverage their manufacturing advantages to produce cheaper, more reliable robots. I believe that whether automakers or traditional manufacturers enter the field, it’s beneficial because collective effort accelerates progress.

Hao Junsheng: I understand that capital bubbles push technology forward. Compared to the US, China’s computing power, GPUs, and top-tier technology are still largely US-based. But China has a strong manufacturing base capable of producing tens of thousands of robots in a month or even a day—something the US cannot do. Do you think China can catch up or even surpass through a curve overtaking?

Pan Jing: I believe we have two core advantages. First, from 0 to 1, generalizable robots will emerge in China because software ultimately translates into hardware. This requires a strong manufacturing foundation, which China has. Second, China has vast private data sources. For embodied intelligence, large-scale data is crucial for training products. Foreign companies might need ten years to accumulate such data, but China’s huge market makes it easier.

In China’s industrial context, much of the private data is controlled by government as a social coordinator. This centralized advantage can be transformed into industrial strength.

Chang Li: In collaborative robotics, many new players entered in recent years—around 200 globally—but only a few have succeeded. Success depends on R&D capability, small-batch production, scale, and commercialization. In the early stages, competition between China and the US is more obvious, but in later stages, China has advantages.

For example, Jieka Robotics’ major client strategy initially targets automotive scenarios, which provide rich, clear use cases. These clients have specialized processes, automation, and R&D teams, making integration easier. I believe in the future, all companies could become robot companies; data will be the most valuable resource.

Hao Junsheng: For AI companies in Shanghai, Beijing, Hangzhou, and Shenzhen, what should they consider when investing in robotics and transitioning to manufacturing?

Pan Jing: In Shanghai, there are over 66,000 companies with robotics as part of their scope; in Beijing, over 70,000; in Shenzhen, nearly 90,000. But their revenue scales are mostly below 20 million RMB. The current industry lacks clustering; it’s fragmented. The scale and impact are far from the trillion-level industry many imagine. Also, Jieka is focusing on automotive supply chains, which means the industry is still in a relatively weak position.

Some internal data: the popular robot dog market, for example, has global sales under 100,000 units. At a price of about 10,000 RMB per unit, that’s less than $2 million in total market size. Despite capital, technology, and talent investments, this is a long-term cycle.

Hao Junsheng: What kind of companies could become industry leaders in robotics?

Chang Li: Besides traditional investment models, the founder’s background and management team are crucial—they determine how problems are solved. The core support for long-term growth comes from the people, their culture, and values.

Hao Junsheng: Broadly speaking, the first thing we look at is the industry track. Robotics is a huge field, and at a specific point, I want to work with a team that I believe in, with a market potential of trillions or even tens of trillions.

Pan Jing: If there’s a trillion-dollar market, there will be trillion-dollar companies. Leveraging market forces can accelerate industry development.

Hao Junsheng: What strategies are your teams currently pursuing internationally?

Chang Li: In Jieka, the underlying culture and values are very pure, professional, and resilient—these qualities will help us sustain long-term. For example, whether we go public by 2025 or continue focusing on robotics, the team quickly reached consensus within a day. We prioritize long-term growth.

Pan Jing: Going public is just a round of fundraising; it’s not necessarily good or bad. Not going public isn’t necessarily worse. The team should keep doing what they do best.

Hao Junsheng: How are your companies planning to expand overseas?

Chang Li: Since 2019, China’s share of the global robotics industry has grown from about 30–35% to around 55–60%. Currently, about 30% of Jieka’s business is overseas, 70% domestic. Our principles are: first, secure local clients before expanding abroad; second, approach from near to far—starting in Southeast Asia and Japan, then Europe, then North America. We’ve established four overseas hubs: Japan (with major automotive clients), Malaysia (focused on electronics and semiconductors), Nuremberg (close to Siemens and Schneider, covering Europe), and the US.

Hao Junsheng: How do you handle competition when expanding in Japan?

Chang Li: We follow the first principle: secure clients, serve them well, earn their trust and reputation. Business practices in Japan differ from other regions. Once clients recognize us, they can recommend us within the industry. Competition comes from local companies and Chinese exporters. We stick to quality and service.

Pan Jing: Our story is different. For example, selling medical robots isn’t just about individual units—they’re part of a larger ecosystem. I compare Titanium米’s AI models for hospitals to a building; the robots are just the bricks. Currently, our overseas revenue is zero. If we were selling buildings, we’re not there yet. In China, hospitals like Zhongshan and Ruijin have hundreds of our robots in use—surgical rooms, ICUs, patient care, etc. We provide comprehensive intelligent solutions, acting more as consultants, helping hospitals optimize workflows and meet regulatory standards. Overseas, we lack the capacity to truly engage with decision-makers. There’s also concern about Chinese companies’ core innovation capabilities, especially in healthcare, which affects trust. So, we’re cautious and prioritize sustainable profitability and focused R&D. International expansion requires a conservative approach, assessing our strengths and strategies carefully.

Hao Junsheng: Is the robotics industry high barrier?

Pan Jing: There’s a term called “hand-crafted robots,” which relies on industry chain maturity—sensors, AI algorithms, hardware design—often developed through trial and error. Also, bubbles create opportunities, and small teams can innovate. But we must be cautious about how sustainable these small teams are, especially when profitability is elusive.

Chang Li: Yes, “hand-crafted” is fine, but who will buy these products? Who will sell them? While it injects innovation, scaling up and certified production lines are long-term challenges—moving from small-scale to commercial mass production is a lengthy process.

Hao Junsheng: When will I have a truly problem-solving household robot? A robot costing 2 million RMB is unaffordable for most. If someday, a robot costing 2,000 RMB or even 200 RMB could solve small household issues, that would be a huge market.

Chang Li: Household problem-solving robots already exist—like vacuum cleaners and floor washers. But the “all-purpose humanoid robot” is still a distance away. Household environments, ethics, safety are complex. Future robots in homes will need certification and regulation.

Pan Jing: From the robot’s perspective, household intelligence should be a two-way effort. Why would I need a robot with complex fingers to press a rice cooker button? It’s unnecessary. A simple IoT module on the appliance can enable voice control. Innovation should not be one-dimensional.

I believe home scenarios require a dual approach—solving real problems, like the over-65 population in Shanghai exceeding 30%. We focus on elderly care at home, helping seniors live dignified lives. Integrating different industry players and designing smart solutions from the robot’s perspective is key.

The essence of a certified operator’s license is to define responsibility boundaries in uncertain business environments. This is a new market that needs a closed-loop system.

Hao Junsheng: What will the robot industry look like in ten years?

Chang Li: I think it will become bigger and more diverse. I hope Jieka can maintain its entrepreneurial spirit.

Pan Jing: The next decade will see accelerated industry development. Changes won’t be limited to a single company but will reflect broader environmental shifts.

In ten years, I hope technology allows everyone to avoid crowded top hospitals and access high-quality healthcare at local community clinics. Even at home, people could benefit from robots and AI in health management. As aging intensifies, more people will consider whether to hire a caregiver or rent a robot. When these decisions become common, the industry will grow.

We hope that in ten years, everyone will have at least one more option like this.

This content is authorized by the author; opinions are solely those of the author and do not represent Huxiu’s stance. If you have objections or complaints about this article, please contact tougao@huxiu.com.

Original article from Huxiu, link:

Massive information, precise analysis, all on Sina Finance APP