Qiao Hong, an academician of the Chinese Academy of Sciences: Human emotions cannot be replaced by machines.

In the past year, what surprised Qiao Hong the most was the development level of domestic humanoid robots. "The gap compared with the global leader is very small." The short time it took for domestic young science and technology innovation enterprises to catch up made her realize that "the seemingly cool technology demonstrations abroad are not that mysterious."

She also said bluntly that the biggest bottleneck in the industry lies in how to achieve the leap from following to leading. Otherwise, there is still a risk of being "strangled". In the fields of large models and embodied intelligence, companies like DeepSeek and Unitree have grown into star companies and even achieved their own original technological breakthroughs. "However, many advantages in original technologies such as chips still lie abroad."

Qiao Hong is a member of the National Committee of the Chinese People's Political Consultative Conference, an academician of the Chinese Academy of Sciences, and a leading expert in the field of robot theory and application. She has long been deeply involved in research on robot decision-making, perception, control, and structural design. Currently, she serves as a researcher and doctoral supervisor at the Institute of Automation of the Chinese Academy of Sciences, the director of the National Key Laboratory of Multimodal Artificial Intelligence Systems, and also concurrently serves as the president of the World Robot Cooperation Organization.

During this year's Two Sessions, Qiao Hong continued to focus on embodied intelligence and emphasized the need to strengthen basic research. She believes that domestic related enterprises are developing rapidly, but they have been established for a short time and lack core barriers. In the future, they need to rely on the cross - integration of multiple disciplines to build independent and practical technological barriers.

The transformation and upgrading of all industries and the in - depth integration into daily life are all testing grounds for humanoid robots. When people are used to seeing humanoid robots dance, attract customers, and even accompany them for a long time, topics such as human - robot relationships and robot ethics have also become the focus of attention.

Source: China Pictorial Library

On March 5th, around topics such as the development of the humanoid robot industry, AI hardware, and human - robot relationships, Qiao Hong expressed her views to China Entrepreneur. As a female benchmark in the global science and technology innovation field, she also talked about the topic of female growth.

The following is the transcript of the interview between China Entrepreneur and Qiao Hong (with deletions):

On "Embodied Intelligence": Strengthen Basic Research

China Entrepreneur: What topics did you focus on during this year's Two Sessions?

Qiao Hong: During this year's Two Sessions, I focused on the basic theoretical research of embodied intelligence. Many people think that it takes decades for basic theories to be applied in practice, but the basic theory of embodied intelligence is different. It is different from basic disciplines such as astronomy, mathematics, and physics. It can be "used quickly or immediately". At present, although domestic enterprises in this field are developing rapidly, most of them have only been established for two or three years, and their core technological barriers are not strong enough.

China Entrepreneur: Under the goal of high - quality development, what are your most urgent scientific and technological innovation tasks and solutions for breaking the deadlock?

Qiao Hong: The core is to build independent and practical technological barriers. The key lies in the cross - integration of multiple disciplines and the connection with data models and hardware systems. This can solve the contradiction between generalization and performance, and achieve the evolvability, high efficiency, and low cost of the robot system.

China Entrepreneur: Over the years, what explorations have you made to promote the collaboration between industry, academia, and research and talent cultivation?

Qiao Hong: With the support of the Beijing Municipal Government, we established the Institute of Embodied Intelligence to promote the implementation of scientific research results and incubate multiple enterprises. In terms of talent cultivation, we have gathered many excellent university students and industry backbones. We let them start from actual application scenarios, develop useful technologies and products, and then use practice to promote basic research. This not only makes the research "substantial" but also can cultivate leading talents in the industry.

On "Humanoid Robots": Surprises and Bottlenecks Coexist

China Entrepreneur: What was the most surprising change in the field of humanoid robots in the past year? What is the biggest bottleneck it faces?

Qiao Hong: The most surprising thing is that the development level of China's humanoid robots has entered the international first - tier, and the gap with the top is very small. For example, the "Beijing Humanoid Robot Half - Marathon" event well demonstrated our technological strength. Behind this is the joint efforts of the government, society, and scientific research institutions, which also greatly boosted the confidence of the entire industry.

Source: China Pictorial Library

Another deep feeling is that domestic young science and technology innovation enterprises and entrepreneurs can catch up quickly in a short time. This made me realize that the seemingly cool technology demonstrations abroad are not that mysterious.

The biggest bottleneck, I think, is how to achieve the leap from following to leading. Following means that when others come up with or create something, you can also do it or make some improvements; leading means that you are the first to come up with the concept or idea, or you are the first to achieve it, and others can't do it in the short term. This is very important. Otherwise, we still face the risk of being "strangled".

China Entrepreneur: Can you give an example to illustrate?

Qiao Hong: In the fields of large models and embodied intelligence, companies like DeepSeek and Unitree have grown into star companies and even achieved their own original technological breakthroughs. However, original technologies such as chips are still in the hands of foreign countries. In fact, efforts in the chip field have always been ongoing, but it still takes some time to achieve relevant core breakthroughs. This is also one of the key bottlenecks in the development of AI hardware.

DeepSeek's innovation in reducing the use of real - time chips is worthy of recognition. Now, many domestic scientists are also conducting research on miniaturizing large models to better deploy them on robots.

In addition, we cannot rely solely on data - driven methods. Instead, we should combine data - driven methods with mechanism research. Relying on the large - scale scientific research talent reserve in China, we should transform our advantages into real technological barriers. At present, this trend is getting better and better.

On "AI Development and Evolution": How to View the Valuation of Hardware Enterprises

China Entrepreneur: Currently, the capital market is more optimistic about AI Native enterprises, while being relatively cautious about the valuation and investment of AI + hardware enterprises. What do you think of this situation?

Qiao Hong: The investors' judgment is reasonable. Logically, original technologies do have a longer - lasting vitality. However, this does not mean that the AI + model has no value. The AI + model focuses more on the actual implementation of technology. In the process, it will face many difficulties such as operation, management, and financing, but it can also create real value.

China's entrepreneurial ecosystem combines the dreams of scientists, the practical work of engineers, the support of the government, and the promotion of major engineering projects. Everyone works together to form a perfect system of "centralized and decentralized, academic and practical", which is very excellent and great.

China Entrepreneur: Some people say that Chinese large - model companies pay more attention to implementation and quick monetization and have insufficient patience for basic research. Do you agree?

Qiao Hong: It can't be completely said so. Now, we are in the upsurge of artificial intelligence. Some enterprises are really telling stories to the capital. However, many enterprises and scientific research institutions have always adhered to their own rhythm in basic research. Under the upsurge, they also need to take implementation into account. If they don't contact the capital at all, their core talents are easily poached. So now, many enterprises are promoting technology implementation on one hand and persuading the capital to pay attention to basic research on the other hand.

On "Human - Robot Relationship": Robots Are Assistants, Not Subjects

China Entrepreneur: What kind of relationship mode do you hope to establish with humanoid robots in the future? What role will humanoid robots play?

Qiao Hong: Currently, our humanoid robots mainly focus on the to - B fields such as industry, agriculture, and convenience stores. Their core role is to solve practical problems and provide convenience. Serving the elderly is also an important direction, which can reflect social care and help people free up more energy, which is very valuable.

Photography: Kong Yuexin

Personally, I prefer to maintain a natural and green - like state and don't need robots to intervene in my emotional life.

China Entrepreneur: Are you worried that people will develop feelings for humanoid robots?

Qiao Hong: I'm not worried. Everyone should learn to have a dialogue with themselves, know what is important, and distinguish what is real emotion. The relationship between people is based on mutual appreciation and respect for humanity. If deceived, the relationship will fade due to the loss of respect. However, robots are not human beings in essence. Their "kindness" to people is calculated by algorithms and is not a real emotional expression.

Human beings' need for respect can only be obtained from their peers, not from machines. Therefore, personal psychological growth and positive social education are very important, which can help people clarify the boundary between human - robot emotions.

China Entrepreneur: What do you think of the AI companion small robots that became popular last year?

Qiao Hong: These small robots have a memory function. For example, they can simulate the personality and tone of relatives to communicate with people, and people can also confide their past to them. They can provide logical analysis and reference opinions to help people sort out their thoughts. However, no matter what functions they have, they are always just machine assistants, not subjects.

China Entrepreneur: What norms and boundaries do we need to establish to improve robot ethics?

Qiao Hong: The core of robot ethics can refer to the Three Laws of Robotics: First, robots cannot actively harm humans; Second, robots cannot violate human will, unless this will conflicts with the first law; Third, robots can protect their own safety on the premise of not violating the first two laws. Therefore, the top priority of robots is to "not harm humans". On this basis, we can then talk about obeying human will and self - protection.

Based on this, we need to clarify specific implementation norms. For example, the compliant use of data, the ethical boundaries of algorithms, and in what situations the "emergency brake" mechanism needs to be activated. In addition, we also need to pay attention to the problem of emotional drift that may be brought about by the excessive intervention of robots in human society. Through social guidance, we should make it clear to everyone that human will and emotions cannot be replaced by machines.

On Female Power: Not to Be Defined

China Entrepreneur: How should women become better versions of themselves?

Qiao Hong: Whether male or female, everyone should pursue self - growth. The difference between women and men lies more in the role positioning given by society and family. I think the core is "not to be defined". You should believe in your own ability. You can do many things and don't be influenced by others' evaluations. Many men are also very family - oriented. I hope that women don't set so many ceilings for themselves in their careers. Be self - reliant and self - strong, but don't put too much pressure on yourself.

China Entrepreneur: Some people say that there are relatively few female workers, especially female entrepreneurs, in the AI field. What do you think?

Qiao Hong: One of the most difficult things for a person to do is to achieve a balance.

Many women, at a specific age, will give priority to getting married and having children and then give up their careers. In fact, the two are not completely opposite. Many people can achieve both family happiness and career success. The key is to find a solution that suits them. If you can't achieve a balance for the time being, you can make a choice. You can focus on one direction in one stage, but don't give up the other direction and then adjust gradually.

Overcoming this contradiction is itself a kind of growth.

This article is from the WeChat public account “China Entrepreneur Magazine” (ID: iceo - com - cn), author: Li Yanyan, published by 36Kr with authorization.