Text: "China Science News" reporter Liu Runan
He is already an associate professor at Zhejiang University, but he chose to go to Germany to re-study for a doctorateHe has a great interest in scientific research, but he chose to work in the company for 11 years, and then returned to universityAfter becoming a member of the European Academy of Sciences and the Alexander von Humboldt Chair Professor of Artificial Intelligence in Germany, he received a lot of support, but at this time he chose to return to China full-time to ......No matter how you look at it, Jin Yaochu is a little "out of common sense".
Recently, the scientist, who has been "tossing" back and forth between China, Germany, the United States and the United Kingdom for 27 years, chose to return to his homeland and join Westlake University full-time to create the "Trusted and General Artificial Intelligence Laboratory". Jin Yaochu shared his life experience and scientific research story with China Science News.
Throughout the interview, Jin Yaochu's "golden sentences" came out frequently. Referring to the first degree, he said that setting thresholds is an obstacle for young people;When it comes to **, he says that ** text is not omnipotent, but it is not possible to do without **;Referring to himself, he said that he did not settle down until he was almost 40 years old, before the family was wandering, and the furniture was picked up from the road;Referring to parental opinions, he said that young people can listen to their parents' advice, but they cannot fully follow their ...... either
Jin Yaochu Photo provided by the interviewee (the same below).
The following is an exclusive interview with Jin Yaochu by China Science News.
Don't pay much attention to short-term material conditions, just have food to eat
China Science News:You've been living and working abroad for 27 years, why do you choose to return to China now?
Jin Yaochu:
In recent years, the domestic scientific research environment and scientific research investment have been greatly improved. Both academia and industry attach great importance to artificial intelligence, and the project is highly efficient. That's a big attraction.
By chance, someone introduced me to President Shi Yigong, and he invited me to visit Westlake University. Due to the epidemic, it was not possible until March this year. During the three and a half days, I visited the campus and met with many teachers and administrators, and I had a first-hand impression that there was an academic atmosphere that encouraged long-term exploration and was free and relaxed. In addition, many teachers have come back from overseas studies and work, so we had a good conversation.
In addition, before going abroad, I studied and worked at Zhejiang University for 12 years, so I am very familiar with Hangzhou, and it will be very convenient to live.
China Science News:In 1996, you received your Ph.D. from Zhejiang University, and then you were given the title of associate professor shortly after staying there. Why give up all this and go to Germany to do another PhD?
Jin Yaochu:
I am a part-time Ph.D. student who stayed on after graduating from graduate school. At that time, I had the opportunity to visit Germany for a year, and the associate professor was awarded during the visit.
During the course of my visit, I gradually discovered that due to the scientific research conditions and environment at that time, there was a big disconnect between our theoretical research and practical application. The research theory is limited to computers**, and there is no condition to test it on real robots. Germany has a good mix of theory and practical application, so I hope to learn more about Xi in the field of artificial intelligence.
In addition, the doctoral program is relatively long, so you can have enough time to learn how Xi do research independently and adapt to this field. Postdoctoral research, on the other hand, is mostly project-driven, and the research that needs to be done is relatively clear, without much freedom. So I decided to do another PhD.
He received his Ph.D. in Germany in 2001.
China Science News:When you were studying for a PhD in Germany, you were married and had children, and your financial situation was not particularly good, but you didn't seem to be under much pressureEven during his doctoral studies, he took a year off to do a doctorate in the Department of Industrial Engineering of The State University of New Jersey in the United States
Jin Yaochu:
Material conditions, age, ......I didn't really care about that at the time. It's probably a trait of me personally, I don't care about short-term things.
When we first visited Germany, our wages were very low in China. During the same period, almost all of the people who visited Germany went to work part-time, because even working in a supermarket could earn a lot more money than in China, but I didn't work a day. I spent all my time on scientific research and tried to get the approval of my supervisor, so I did it for a year before he agreed to let me stay for a PhD.
I didn't go to work because I was in good condition. I came from a peasant family and had a very difficult life, but I didn't pay much attention to short-term material conditions, as long as I had food to eat and could get by. I don't want to give up what I really want because of something in front of me.
If I could, I would even like to do a third Ph.D., in biology or medicine (laughs).
China Science News:What do you think of these "age thresholds" at the moment?
Jin Yaochu:
The current "age threshold" has created a great obstacle for researchers who start later. I don't think it's reasonable to simply set an "age threshold". One could learn from a youth programme in the UK, which stipulates how many years after graduating from a PhD to be eligible to apply, not before what age.
In China, it is often emphasized that young people should be given opportunities, but these opportunities are stuck by such and such thresholds, including the threshold of the first degree, etc., which are unnecessary obstacles set up for young people.
I was 35 years old when I finished my PhD in Germany. The eldest in the family is 5 years old, the second child was just born, there is no right of abode, let alone a house, our furniture is picked up from the road, and I don't know if I will live in ** in the future, but I was not anxious at that time. I don't know what would happen if I were in this environment.
Big ecology, small ecology
China Science News:You worked in research at Honda Research Institute for 11 years and became the chief scientist, why did you choose to return to university?What is the difference between a business and a university job?
Jin Yaochu:
In 2003, Honda established an R&D center to lead the R&D and manufacturing divisions with a focus on application-oriented basic research. It can be said that it is closer to application than doing scientific research in universities.
At the same time, the company is flexible in the management of the project. We have a new idea, and after a brief report to the institute, we evaluate it internally, and if it is appropriate, we can get support. But in universities, whether in the United States, the United Kingdom, or other countries, it is very important to apply for programs. Many people may have to write a dozen or so applications a year, some of which may not be what they really want to do, and may end up with limited research output. I wanted to leave my time for research that I really wanted to do, so I chose to work at Honda Research Institute.
I worked until I was in my 40s, and I don't know if I encountered what everyone called a "midlife crisis", but I felt like I had reached the "ceiling". The limitations of the institute are gradually highlighted, such as the company may not want you to have frequent academic exchanges, all cooperation projects are very directive, and so on. It's perfectly in the company's interest, but for me personally, it's a limitation.
During my time at Honda, I made some achievements and published a lot, and with my experience in industry, I was able to go back to university and teach at the University of Surrey in the UK.
China Science News:At graduate level, you moved from the field of automatic control to the field of artificial intelligence. Soon after, artificial intelligence experienced a cold winter, and at that time, did you think about changing careers?How did you persist in your research in this field for decades?
Jin Yaochu:
I didn't think about changing careers, that's what interests me. During my master's degree, I liked to go to the library to read books, and I happened to see a book by an Israeli scholar on robot trajectory planning and genetic algorithms, and I felt very interested, so I consciously Xi it. Later, I joined the Computational Intelligence Society under the IEEE (Institute of Electrical and Electronics Engineers).
The computational intelligence research I do has always been a relatively niche field. The Computational Intelligence Society, to which I belong, has always had six or seven thousand members, and there has been basically no change. At that time, when I went to the company to talk about how good evolutionary algorithms and artificial neural networks are, they didn't believe in it and thought that traditional mathematical methods were more reliable. In recent years, artificial intelligence has become hot, and everyone's attention to artificial neural networks has increased.
When it has not yet been paid attention to by the public, it is important to have such an ecosystem as the Computational Intelligence Society. Regardless of the popularity, this organization has always been there, and you know that there are so many people who have been doing relevant research all the time, and that will give you strength.
Compared to the Gakudo, the Honda Research Institute is a smaller ecology. There, I was in charge of the "Evolution and Xi" group. There are also "computational neuroscience" and "humanoid robots" research groups in the institute, and a small group of us study each other. A big ecology, a small ecology, so I have never felt the cold of "winter", and I have been working in this field.
China Science News:Your social ** avatar is the cover of "Computer Evolution of Nervous System and Morphological Development: Towards Evolutionary Development of Artificial Intelligence", what is the story behind this book?
Jin Yaochu:
When I was at Honda Research Institute, the company wanted us to do some basic research, so I started teaching myself computational biology and collaborated with many universities to conduct basic research on the co-evolution and development of the nervous system and morphology. After returning to university to do research, I also won related research projects, but this is not my mainstream research direction. For more than ten years, it has been done piecemeal.
Later, after signing a contract with the publishing house, I never had time to write. Until the beginning of the epidemic in 2020, the time-consuming international travel was gone, and there was a whole block of time to think and sort out. In the end, I finished 3 books. In 2021, he published a book on machine Xi and evolutionary algorithms for data-driven optimization, and in 2022 he published a book on federated Xi and privacy, and in 2023 he published this one. Therefore, I always have a feeling of "if you lose your horse, you don't know if you are blessed".
2022 Humboldt Chair Award Presentation Ceremony (second from left, Kim Yaochu).
The life principle of "convergence - divergence".
China Science News:It sounds like you have a very good mentality and can face difficult things calmly, so are there any times in scientific research and life that make you feel particularly difficult and frustrated?
Jin Yaochu:
Can't tell. I don't seem to have had a painful experience for a long time, I can't really put it into words.
In the control system, we often talk about convergence and divergence, and I often use them to describe my research and life.
On the one hand, we must insist on doing scientific research and make progress in the professional field, which is the process of convergenceOn the other hand, look around and absorb new ideas. When convergence reaches a dead end, it will not be able to move if it converges again, and it must diverge. This is true for scientific research, and it is also true for life.
I think it's important to set a directional goal, not a specific goal. For example, if you have to become a full professor in 10 years, this kind of goal is too specific and may make people anxious. After 10 years, it is more feasible to be able to do the research you like and make a breakthrough in your own research field.
China Science News:How do you typically train students?
Jin Yaochu:
I hope that every doctoral student can do 3 weighty tasks. It could be in different directions related to a problem, or it could be different sides of the same direction. When doing research, be problem-oriented, not goal-oriented. But you still need to have 2 or 3 good ** during the doctoral period, otherwise you will apply for a project or job after graduation, and introduce yourself very well, but there is no **. It's definitely not going to work.
*Not a panacea, but without** is not possible either.
China Science News:What advice do you have for young people?
Jin Yaochu:
I have found that my students are very constrained by their parents when they choose their future career after graduating from the PhD.
I understand very well that there is an only child factor in this, or that filial piety to parents is our traditional virtue. But I would like to say that when it comes to studying, researching, and working, you can listen to your parents' advice, but don't listen to them completely.
I want their parents to give their children enough freedom. Speaking of which, I am very grateful to my parents, they let me do whatever I am willing to do, and never say "it's time for you to start a family, it's time to go back to China, what should you do".
On Christmas Day 2013, Jin Yaochu (fifth from right) with students.
The Humboldt Chair Award is Germany's most prestigious international research award, with no more than six scientists awarded each year in the field of artificial intelligence. Each chair receives 350,500,000 euros for his or her first five years of research at a German university.
Thanks to our reporter Wen Caifei for her support for this article).