Shangtang Wang Xiaogang: Iron Triangle and independent R&D are very important for AI landing applications.


Sina Technology Myrica rubra
During the World Artificial Intelligence Congress in 2019, Wang Xiaogang, co-founder of Shangtang Science and Technology, president of Mobile Intelligence Business Group and Dean of research institute, was interviewed by Sina Science and Technology to share the impact and application of the Industry-University-Research model on Shangtang Science and Technology.
“Talent recycling ecology is very important for Shang Tang”
Shang Tang has chosen and adhered to the mode of integration of industry, education and research since its inception. Wang Xiaogang once said, “The integration of industry, education and research is the source of innovation of Shang Tang.”
So, how does the Industry-University-Research model work in Shang Tang? Wang Xiaogang said, first of all, there is a big difference between academia and industry.
On the one hand, the academia has a deep and long accumulation in a certain direction, which may be more than ten years; while the industry landed in a certain direction, often with a short product cycle, it may be necessary to achieve landing in a year or two.
On the other hand, the academic atmosphere and research model are relatively free and open, and they have the opportunity to try different directions.
In addition, the academic group is geared to the cultivation of doctoral students. When facing academic problems, the doctoral students have a systematic methodology of thinking, defining and solving problems, and the academic community is also systematic in the cultivation of people.
This is the advantage of academia over industry.
Specifically, Shang Tang Society cooperates with academia from the perspectives of talent, technology and so on.
Wang Xiaogang disclosed that from the talent aspect, the academic circles will convey talents to Shang Tang, at the same time, reverse promote Shang Tang’s training of its own R&D personnel. For example, in the process of cooperation with universities, R&D personnel have also been trained and promoted.
In addition, according to the mechanism of Shang Tang, the introduced teachers can return to the university after a period of time, and the company will send its own researchers to the school to serve, and import talents to the school. “For example, there are some young employees in Shangtang who are willing to go to school for further study and Ph.D. Now that we have graduated with such a Ph.D., we are back in Shangtang. This is a talent cycle, and I think this ecology is very important for Shang Tang.
From a technical point of view, Wang Xiaogang believes that if you don’t want to let the academia make products or land at once, what the academia does is more prototype, raw materials, to product this step requires a lot of research and development work input, because the product does not pass this industrial red line is not feasible.
“The school achieves 60%, the innovation is enough, this is its demand. The remaining 40% is actually very difficult, to invest a lot of resources, which is not what academia is good at doing. So if the industry forces academia to make this product-level, it is unrealistic.
What kind of research results can be landed? Iron triangle and independent research and development
“Our academic research results are not immediately aimed at landing, or have a broad direction of judgment, and then in this direction we hope that they can do more exploration.”
Wang Xiaogang said that there are two important points. The first is to have an iron triangle. The second is to have independent research and development. A team that can master the rhythm of research and development by itself can go forward lightly.
“Over the past few years, we have come to the conclusion that the most important thing is to have an iron triangle.” That is, business, products and research, these three elements can closely cooperate with each other, and they also interact and restrict each other.
First of all, from the first step of landing, enterprises should understand the needs and discuss with customers. “In response to the needs of these customers, research has a role to play in telling you where the boundaries of technology lie.” Wang Xiaogang said.
“For example, when we make products on mobile phones, mobile phone manufacturers will tell us how to define them, and then we will follow the needs of our customers. In fact, some AI products and technologies on mobile phones we are pushing today are not made by Shang Tang alone, but by working with customers, because this kind of communication and interaction is very important.
On the other hand, the research results can only be landed if the product finds the right entry point; moreover, the product will drive the research.
Wang Xiaogang pointed out that customers will give the team a lot of research and development projects, as a result, customers will immediately ask what the next innovation is, customers have always wanted new things.
“Early on, the team’s energy was more on project delivery, but when it was finished, it was found that there was no innovation reserve, which was definitely not good.” Wang Xiaogang revealed that there is still a wave of people in the team moving forward to do more advanced research and development independent of the project.
When a team is doing R&D, it may not be able to achieve 100% technology, only 80% technology, then the remaining 20% will need to be iterated in depth with the product to complete, which is very resource-consuming.
Wang Xiaogang revealed that his team’s independent innovation R&D requirements are 80%, and then do the next, new technology reserves to show customers, customers feel which is good, and then polish with customers. “If every item is 100% of the product level, the team will be exhausted, energetic, completely dragged down, and it needs to be lightly loaded into battle.”
It is reported that Shangtang’s initial founding team is more research team, and later joined more and more people who understand business, business and products.