​iAsk x China Daily | How To Be A Unicorn Company?

本文章刊登在CHINA DAILY-艾问专栏


At present, mankind is at a time of transformation from the big data era to the big computing era.

——余凯 地平线创始人


In 2012, he took the lead among the top Chinese scientists to return China and was the first to start deep learning research in China. In 2013, he led the Baidu team to pioneer the research in automatic driving in China. In 2015, he started a business and was the first to propose chip researches in the software field.


Yu Kai, founder and CEO of Horizon Robotics, former executive director of Baidu Research Institute, national distinguished expert of under the “Thousand People Plan” initiated by the Organization Department of the Central Committee of the CPC and member of the Next Generation AI Strategy Advisory Committee of the Ministry of Science and Technology …


Over these years, more and more labels have been attached to Yu Kai, and he has also tried to look deep into himself.


“Many people say that scientists are not suitable for starting a business, but I think great entrepreneurship is actually related to one’s characters Scientists tend to think carefully and get everything ready before they start doing things. I realize that no matter how smart you think you are, how good propositions you think you’ve made and how well you know the competitive landscape in advance, once you get into the battlefield, the situation always changes much faster than you have already experienced. The only thing you have to do is be brave.”


When Horizon Robotics was founded in July 2015, few people understood deep learning of artificial intelligence, and few people thought a processor company could become a unicorn.


But as the first company in China to produce artificial intelligence processors, Yu Kai knew from the beginning that what he was going to do was extremely difficult, but he was not afraid and felt that he must doit.

三年,地平线已经成为了让中国骄傲的AI芯片企业。在达沃斯最新公布的中国AI 50榜单中,地平线成为登陆该榜单的Top14独角兽企业。

It takes three years for Horizon Robotics to become an AI chip company and the pride of China. In the latest list of AI 50 Chinese companies released in Davos, Horizon Robotics became a top14 unicorn company.


Yu Kai and Horizon Robotics have inspired a lot of curiosity. This issue of IASK LEADERS will bring you a story of artificial intelligence.



Gloria Ai: Some people say making chips requires heavy investment, isn’t it? It will take a long time and you have to face the challenge of the industrial giants. It seems that many negative speculations have been around you and the horizon.

余凯:对的。但是我认为就是说世界上面的话,如果你说创新什么叫创新?创新一定是做一件事情是带来这个世界new dimension(新次元),如果你没有一个new dimension 的话,也就是说你这个公司存在或不存在,这个世界是一样的,说明这个是公司是没有价值的。

Yu Kai: Yes. But I think that is to say what’s already known to the world. What is innovation? Innovation means bringing about a new dimension in the world. If you don’t create a new dimension, that is to say, the world remains the same whether your company exists or not, then this company is worthless.


Therefore, an innovative company must be one of your unique ideas. Most people don’t understand it or agree with it, and then you do it, so that you can enjoy a period of loneliness and then create real value during that period. By the time everyone realizes its value, your lonely battle is actually over.


Gloria Ai: What irreplaceable value does Horizon Robotics create?


Yu Kai: When Horizon Robotics was founded in July 2015, it was the first company in China to produce artificial intelligence processors. At that time, many people said that I had just understood what is deep learning and how algorithm works in 2015, and then suddenly decided to do processors; it sounded very risky. However, what I saw is a huge opportunity. Why? Because I studied and practiceson artificial intelligence algorithms, especially deep learning. I am one of the first scholars in the world to do researches in deep learning.


In 2015, I deeply realized that it’s hard to use software algorithms to change the world. In different scenarios, such as automatic driving and camera, the real room for improvement lies in the breakthrough in the processor. It must have less power consumption and smaller size.


We must take up this very daunting challenge which many people think is hard, because it has value. We don’t care if we are short of resources or we have a long way to go; we just must do it. That’s what I think.


Gloria Ai: But the problem is when people suddenly realize this era is going to be dominated by artificial intelligence which needs chips, when everyone sees this opportunity, aren’t you nervous? You are no longer the first-mover.


Yu Kai: For me, I saw this opportunity two years ago and then delved into it. Today, everyone begins to agree with it. For me, I feel very good about myself. This is the first point I want to make.


Secondly, in fact, I’m not thinking much about competition. I think everyone should mind their own pace. If you are running a marathon, you can’t finish it by looking at people around you. You must enjoy the process, listen to your footsteps and adjust your pace.


Gloria Ai: I know that the most common question about Horizon Robotics is what it is doing. When you can’t explain specifically, you just say you are making chips in the robot’s brain. What does that mean?


Yu Kai: Sometimes I do feel a little challenging to explain what we are going to do and our business model. Since it has never existed before, you can only see it if you believe it, right? That is to say, it is difficult to describe, so sometimes you enjoy this kind of intellectual loneliness, but I think a real entrepreneur and innovator must be different.


Gloria Ai: In terms of the applying AI chips, what scenarios do you plan to prioritize?


Yu Kai: Horizon mainly considers applying it to automatic driving in urban settings. It’s a technology that can differentiate us from others hence showcasing our value.


Gloria Ai: There are many kinds of robots in the world, why choose car?


Yu Kai: Actually, I’m thinking that there will be a lot of robot applicationsin the future, but I’m going to make breakthroughs in one area first. We choose self-driving car, because it may be the most difficult of all robot applications.


The other reason is that it actually has the greatest impact on the quality of human life. In terms of the time I spend every day, driving actually accounts for a large part of my life. It affects our efficiency and quality of life. I believe that in the era of artificial intelligence, AI may have the greatest impact on human life in two dimensions, one is travel and the other is medical care.


Gloria Ai: What kind of surprise will the future robots and Horizon bring tous?

余凯:我们在2018年的话,我们的核心就是show me the product, 就是产品落地,2019年我们展现给大家的一定不只是芯片,一定是something(产品),但是芯片inside(植入在里面)。每年都在进步,每年都还蛮有意思。

Yu Kai: In 2018, our key task is to show me the product, that is, the product manufacturing. In 2019, we will show you not only the chip, but also something (a product) with a chip inside. Every year we are making progress, and every year is quite interesting.