余凯:中国市场永远存在着不可思议的泡沫

2019-02-19 13:00
北京

本文章刊登在CHINA DAILY-艾问专栏

“逻辑的尽头,不是纯粹理性下的秩序王国,而是用生命去奉献的热爱。” 

"The end of logic is not the kingdom of order ruled by pure reasons, but love devoted with life."

2019年,美国拉斯维加斯,艾诚与余凯关于国际消费类电子产品展览会(International Consumer Electronics Show,简称CES)的一年之约。这一次,余凯,带着蒸蒸日上的地平线,带着边缘AI计算的最新研究成果,前来赴约。

2019, in Las Vegas, United States. Gloria Ai and Kai Yu met again at International Consumer Electronics Show ("CES") to keep their promise made one year ago. This time, Mr. Yu brings along his prospering company, Horizon Robotics, and latest research findings on AI over Edge Computing.

AI领域跌宕浮沉,这阵风吹到大热又冷却了很多次,太多的投机者频繁进出这一风口,最终剩下来的人数量极少,他们不是AI精英,就是狂热爱好者,而余凯,这位曾经的科学家,二者皆是。

There have been ups and downs in AI field. This storm went overheated and cooled down several times. Too many speculators piled in and out frequently, left only few survivors, who are either AI elite or fanatic AI fans. While Kai Yu, used to be a scientist, falls into both groups.

在AI风口的边缘观望了20年,他参与了“AI进化史”的每一重大事件,离开了美国NEC研究院,离开了微软,离开了斯坦福计算机系讲台,离开了西门子和百度。2015年,余凯终于创立了自己的理想国——“地平线机器人”,这一次,他要自己到风口处飞一把了。

Watching by the edge of AI for past twenty years, he involved in every milestone of "AI evolution". He worked in NEC Research Institute (US), Microsoft, Computer Science Department of Stanford University (as Adjunct Professor), Siemens and Baidu. In 2015, Kai Yu finally built his own Utopia, "Horizon Robotics". This time he wants to fly with the storm himself. 

“没有软件和生态的AI芯片只能是一块石头”,余凯称,我们中国人的“房子”建在别人的“地基”上面,因为Foundation不是自己的,所以房子摇摇晃晃的,真正的产业界并没有深刻理解AI芯片,也没有跟上深刻的“软硬结合”这个思路。这是目前我们国家AI界最大的问题,也是余凯成立地平线的最大原因。

"AI chip without software and ecology system is only a single piece of stone." said Kai Yu, the "house" of our Chinese people is built on a "foundation" laid by someone else. Since we did not do foundation work, so the house is rickety. The true industry has not thoroughly understood what exactly AI chip is, nor followed the profound thinking about "combining software and hardware". This is the most important problem Chinese AI community is facing now, and also the main motive for Kai Yu to establish Horizon Robotics.

“自己去做中国的AI芯片”。不鸣则已,一鸣惊人,低调了20年的余凯正式高调起飞,地平线机器人,致力于研制一颗"机器大脑",打造万物智能时代的"AI Inside",给人们日常生活的无数设备和产品装上“大脑”。

"I will develop China's own AI Chips." It never rains but it pours. Being low-profile for two decades, Kai Yu officially took off with a high profile. Horizon Robotics is committed to develop a "machine brain", to create "AI Inside" for this era of everything smart and empower numerous devices and products used in our daily life with "brains".

2018年的CES,艾诚也曾对话余凯,与那时相比,虽然仍然是那个孤独创业的科学家,但他明显多了一丝沉稳,眼神也更坚定了。

In 2018 CES, Gloria Ai had also conversed with Kai Yu. Compared to that time, though Mr. Yu is still a scientist entrepreneuring alone, much obviously he is calmer now and with firmer eyes.

因为他所做的,是他的毕生热爱,他最真切的理想。

Because what he is doing, is what he has loved for all his life and is his most real ideality.

“中国的市场永远存在着不可思议的泡沫”

"Chinese market is always blowing incredible bubbles"

艾诚:距离上一次来CES访问你,已经一年之久,我最明显的感觉,是你比上一次更自信,而且更愉悦。但是我清楚地记得,在一年前你还说过,创业就是一个很孤独的旅程,在很多人都不理解甚至不支持的时候,你要享受孤独,等很多人理解你的时候,可能战争已经结束了。所以,当你今年再来到拉斯维加斯时,你是什么样的感受,作为地平线创始人,还孤独吗?

Gloria Ai: It has been one year since our last interview in CES 2018. Obviously to me you are more confident than last time, and more cheerful. But I still remember it clearly one year ago you told me, entrepreneurship is rather a lonely journey. You have to enjoy loneliness when many others cannot understand you nor support you; when many understand you at last, perhaps the battler has been over. So when you come to Las Vegas this year, what do you feel? As the founder of Horizon Robotics, are you still lonely?

余凯:我觉得分两个层面。一个层面,从我们的业务推进来讲的话,我们越来越受到产业界的认可,地平线倡导的AI芯片、边缘计算,也越来越成为一个产业界的潮流,所以这一点,感觉好像不那么孤独了。

Kai Yu: I think we could look at it from two sides. On one side from perspective of our business development, we are increasingly recognized among the industry. And AI chips and Edge Computing initiated by Horizon Robotics are turning out to be the industry trend. So looking from this side, I seem feel not that lonely.

余凯:另一方面,作为企业家本人,永远是希望在整个的噪音背景下面去找到那一丝信号,这个信号是通向未来的,但这一定不是每个人都看到的,因为每个人都看到的东西,它往往是没有价值的,今天不是讲反共识吗?所以企业家本身天然应该去享受孤独,利用孤独,去找到通向未来的那个非常孤独的路径。

Kai Yu: But on the other side an entrepreneur is always hoping to find out the tenuous signal under a noisy background. That signal is leading to future but certainly not everyone can see it. Usually something everyone can see is worth nothing. So entrepreneur himself should naturally enjoy and take advantage of loneliness to find the super lonely path which is leading to future.

艾诚:2018年,AI的芯片市场越来越热,但热的同时也意味着泡沫,你自己在市场上观察到了什么是泡沫,什么是真相吗?什么可为?什么不可为?

Gloria Ai: Market of AI chip was getting hotter in 2018. But hot also means bubbles. What have you observed in the market as bubbles, or as the truth? What could be done and what could not?

余凯:中国的市场永远存在着这种不可思议的泡沫。我个人认为,中国整个的思维,从众的思维比较占主流,但特立独行的这种思考力不是那么占主流。地平线的话,还是不要去太多地被这些所影响,要从噪音中去看到信号,怎么去思考自己的战略,怎么去打造强大的组织,更好地去连接我们跟客户,这些是核心去思考的问题。至于资本市场怎么看我们,或者媒体怎么看,不应该那么关注。

Kai Yu: Chinese market is always blowing incredible bubbles. Personally, I think in the overall mindset of Chinese people, following the crowd prevails. While the ability of doing critical thinking is not that popular yet. It's better for Horizon Robotics not to be affected too much in this way. We still have to find the signal from noises, to think about our strategy, how to build a strong company, how to better connect us with our clients. These are core questions for us to think about. As for how the capital market will see us, or how the media will see us, we should not pay so much attention to it.

一年前,余凯曾对艾问表示,“很多人说科学家不适合创业,跟性格特质有关。越是科学家,越喜欢深思熟虑、万事俱备再去做事,但我认识到一件事,无论你认为自己多么聪明、做多么充分的设想、事先对战场的状况多么了解,一旦冲入战场,形势的变化会远超你已有的经验。你唯一要做的事是勇敢。”

One year ago, Kai Yu told iAsk that "many said scientists are unsuited for entrepreneurship due to their personality traits. The more you are as a scientist, the more you prefer deliberating and trying things until all necessary conditions are ready. But I have learned one lessen, no matter how smart you think you are, how comprehensive your proposal is, how well you know the situation, once getting into the battle, the ever changing circumstances will overwhelm your existing experience. The only thing you should do is to be brave."

从成立到产品落地,地平线仅用了三年,且技术成果被广泛运用在智能驾驶、智慧城市管理、智能安防、智慧零售等领域。

It took Horizon Robotics only three years from inception to their product debut. Their technologies have also widely applied in various areas such as smart driving, smart city management, smart security and smart retail. 

2019年的CES上,中国创业公司地平线Matrix的自动驾驶计算平台获得创新奖。事实证明,勇敢的科学家余凯找到了他想找的“信号”,也暂时得到了他想要的“认可”。

In 2019 CES, the Matrix Autonomous Driving Computing Platform by Chinese startup company Horizon Robotics won the Innovation Award. The reality proved that brave scientist Kai Yu has found the "signal" he is searching for and received the "recognition" he wants.

追求理想的道路注定是孤独的,但这种孤独注定不平凡。为了打造中国AI自己的Foundation,余凯在这个领域观望了20年,拒绝一切大平台的“橄榄枝”,创造自己的“理想国”,无论是不理解不支持,还是多变的战场以及形势,带给他的孤独感一定是巨大的,但他还仍然沉浸在这里,勇敢地享受孤独,利用孤独,去突破。

Journey of pursuing ideality is destined to be lonely, but such loneliness is destined to be extraordinary. In order to build China’s own foundation of AI, Kai Yu kept observing this area for twenty years and turned down the "olive branches" offered by all big platforms to construct his own "Utopia" instead. Both the misunderstanding and nonsupport from others, or variable challenges and circumstances, all brought him surely the feeling of immense loneliness. But he still submerges himself in it, even enjoys it with courage and take advantage of it to make breakthroughs.

作家刘同曾说过一句话:你的孤独,虽“败”犹荣。这句话,只适合描述余凯这样内心强大的人。

 Writer Tong Liu once said: You are lonely; it is a glorious "defeat". This quote is only suitable to describe people like Kai Yu who are strong inside.

“他提到AI的时候,眼睛里面闪着光”

"When he is talking about AI, his eyes are shining"

2018年被称为人工智能技术规模应用的拐点,作为核心的人工智能芯片备受关注,谷歌、苹果、微软、Facebook、英特尔、高通、英伟达、AMD、阿里巴巴等巨头纷纷开始自主研发人工智能芯片。面对挑战,余凯说,地平线这种边缘计算平台要想真正发挥价值,一定要有合作伙伴,一定要有生态,所以需要更积极一点,互帮互助,通过差异化的能力寻求共存,通过重合来把地平线跟最终的应用场景拉的更近。

Year of 2018 is called the inflection point of large-scale application of AI technology. AI chips at core place attracted great attention. Industry giants such as Google, Apple, Microsoft, Facebook, Intel, Qualcomm, NVIDIA, AMD and Alibaba started one after another to develop their own chips. In the face of challenges, Kai Yu told us, if edge computing platforms like Horizon Robotics want to add true value, they must secure partners and related ecology system is also inseparable. Therefore, they need to be more active and to help each other. They should strive for coexistence by differentiated capabilities and narrow the gap between Horizon Robotics and final application scenarios through overlapping.

他可是潜伏了20年才出来“搞事情”的余凯,他的目光,不可能短浅。虽然,余凯现在总是让自己忘掉科学家的身份,时刻用一个公司CEO的标准来要求自己,“衡量利弊”、“务实”是他应当具备的主要属性,但余凯的身上,至少目前为止,还存留着科学家的“理想主义”,那份知识分子所与生俱来的“入世”的心。

He is no one else but Kai Yu who has been latent for 20 years before "turning up to do something". His vision surely will reach far enough. Although for the time being, he keeps reminding himself to ignore the fact that he is a scientist and asking himself with the standards of a company's CEO, who should have major characteristics such as "balancing" and "pragmatic". But at least so far, he still preserves the "idealism" of a scientist in him and a born "world-engaged" heart of intellectual.

“地平线本身的话,从创立第一天就希望成为一个平台级企业,我们做的是说AI的Edge Computing的平台,所以Horizon Robotics 实际上就是Horizontal platform for Robotics。要达到这样一个目标,地平线可能要花至少15年的时间,我们有足够的耐心去干这个事情,去集聚势能转化成动能。短期来讲的话,我们没有任何的考虑说被收购,我们希望自己成为巨头。”

"From the very first day of Horizon Robotics, we hope to develop it into a platform. What we are doing is the platform of AI over Edge Computing. So in fact Horizon Robotics is the Horizontal Platform for Robotics. To achieve this goal, from my point of view it might take us at least 15 years. We have enough patience to do this, to collect and transform potential energy into kinetic energy. In the short term we don’t have any consideration of being acquired. We hope we can become a giant ourselves."

余凯想做的,是将地平线打造成为世界一级的芯片公司,资本只是他的一个工具,一个他想要实现自我去架构出一个生态的媒介。他不是说只想要发财,他更想真的做出突破,打造出这个产业中国自己的“地基”,成就自己的一番事业。所以,当提到“收购”这个话题时,他语气中的抵触是真的,眉间稍纵即逝的颦蹙也是真的。

What Kai Yu wants to do is to develop Horizon Robotics into a world leading chip company. Investment is merely one of his tools, a vehicle which he could make use of to build an ecology system through devoting himself. It might not be accurate to say he does not want to be rich. What he desires much more is to make some breakthrough, creating China’s own "foundation" in this industry and achieving his career goal. So, when it comes to talking about "acquisition", some reluctance IS revealed in his response, and the ephemeral frowning on his face IS also visible.

艾诚:我明白叫盛名之下,后面接的一般是难副其实的,我不打算让难副其实给你,你打算怎么副其实?

Gloria Ai: I see it as under a prestigious name, it is usually not what it should be. I don't want you to show us such an image. What are you going to do to match your name?

余凯:一定要有清晰的商业路径。企业家,一定是理性跟感性要结合在一起,然后他的理想主义跟现实主义一定要结合在一起,以最现实主义的态度,去做最具有梦想、最具有颠覆性的理性主义的这样的事情。

Kai Yu: You must have a clear business path. I think entrepreneur must accommodate both sense and sensibility. His idealism and realism must be merged together. To do something most fantastic, revolutionary, and rationalistic while with the most realistic attitude.

艾诚:如果让你做两个预见,一个是预见十年之后的自己,一个是预见十年之后的地平线,你会如何预测未来?

Gloria Ai: If you are asked to make two predictions, one about yourself ten years later, and another one about Horizon Robotics ten years later. How would you forecast the future?

余凯:十年以后,希望我本人成为一个不是说是多有钱的企业家,最重要的是成为一个更加有思考力的企业家。

Kai Yu: I hope ten years later I myself would be an entrepreneur regardless how rich I will be, but most importantly I want to be a more thoughtful entrepreneur.

艾诚:那十年之后的地平线呢?还是那个口号吗?要做世界第一的芯片公司?

Gloria Ai: Then how about Horizon Robotics after ten years? Still with the same slogan: To be the number one chip company in the world?

余凯:十年以后地平线当然还是这样的一个目标,我们会希望成为全球在边缘的人工智能最大的平台。

Kai Yu: We hope this will still be our goal ten years later. We hope to become the largest platform of AI over Edge computing in the world. 

无论是作为科学家还是企业家,余凯的理想主义是务实却也感性的。可以说,逐梦的人都是感性的,主观感受远大于物质需求。

No matter as a scientist or as an entrepreneur, idealism in Kai Yu is both practical and sensible. It can be said that people who chasing their dreams are all emotional, with far greater subjective feelings than material needs. 

在访问中,余凯说的最多的两个词是“我”和“人工智能”,可能对于他来说,这是生命中最重要的,如果仔细观察就会发现,每当他提到这些,眼镜下面那对聪明的眼睛,闪着坚定的光。

During our conversation, two words he said most were "I" and "AI". Maybe to him these are most important things in life. Observe carefully and you will notice the pair of smart eyes under his glasses are shining firmly when he is talking about these.

人生的挚爱,足够支撑起所有逻辑,也就是余凯所说的“信号”,当它出现了,世界都会变成红色。余凯有一个很大很大的理想,他不畏惧说出来,就像他不畏惧拥有这个理想,可能,他甚至都不怕失败,他只是想朝着这个方向再走远一点。

Love of life is strong enough to support all logic, and this is the "signal" mentioned by Kai Yu, at the moment which appears, the whole world will turn red. He has a magnificent dream and is not afraid to speak it out just as he is not afraid to keep it. Or probably, he is even not afraid of failure. He just wants to go further toward that direction.

“有地平线的地方一定是个辽阔的地方,在天地的尽头,梦想的终点。向着地平线延伸的方向奔跑的人,一定是胸怀梦想的人。天有多广大,地有多开阔,地平线就有多绵长。”

"A place with horizon mush be a vast one. It is at the end of sky and earth, it is the end of dream. A man running towards where horizon is reaching out to, must be a man with dream. Compatible to vast sky and open earth, the horizon will extend as far as it could be."

END

    特别声明
    本文为澎湃号作者或机构在澎湃新闻上传并发布,仅代表该作者或机构观点,不代表澎湃新闻的观点或立场,澎湃新闻仅提供信息发布平台。申请澎湃号请用电脑访问https://renzheng.thepaper.cn。