GLG 업데이트

GLG China Presents: The Future of Artificial Intelligence


August 26, 2016

Empowering Humans: Breakthrough & Development of Artificial Intelligence

The development of Artificial Intelligence (AI) has catalyzed more varied user demands and created more jobs; human beings have benefited greatly from the development of AI.

On August 26, GLG hosted the event “The Future of Artificial Intelligence” in Beijing, in partnership with China’s top AI experts and industry leaders shared their experiences and insights into the AI field and explored how to effectively apply AI technology in businesses and consumer markets.

Industry leaders who spoke at this event included:

  • Xiang Zhonghong, founder and CEO of RobotZero
  • Yu Zhichen, founder and CEO of Turing Robot
  • Shen Xiuping, Chairman and CEO of Ulucu
  • Liu Wei, Director of the Man-machine Interaction and Cognition Engineering Laboratory at the Beijing University of Posts and Telecommunications
  • Yang Jing, founder of Xinzhiyuan
  • Lu Tao, Asia-Pacific Regional Director of Aldebaran Robotics/Softbank Robotics
  • Ge Hantao, founder of the China Chief Data Officers Alliance and Deputy Secretary-General of Zhongguancun Smart Hardware Industry Alliance
  • Hu Jie, General Manager of Qingdao Krund Robotics Co., Ltd.,

Here are some of the insights shared with us:

The state and future of AI

With the rapid growth of advanced technology, it is hard for anyone to predict the future. However, one thing is certain: AI will definitely influence the future of humankind. As indicated by Mr. Xiang Zhonghong, founder and CEO of RobotZero, in today’s world where use of mobile devices is surging and the tentacles of the Internet reach every corner, we already live in an era where AI, virtual reality and robots coexist. From this perspective, AI is indeed an important aspect of human beings’ future.

Mr. Yu Zhichen, founder and CEO of Turing Robot, pointed out that for the past 60 years, AI’s development has focused on the technology and neglected product innovation and development. According to Mr. Yu, for AI to truly change the world and people’s lives, the industry must make technological breakthroughs, innovate and develop new products.

At the very beginning of AI technology, so-called AI was no more than simple game systems, e.g., games similar to chess. “The past five years and the next period will usher in the third wave of development for AI,” said Mr. Yu. He pointed out that voice/semantic processing will be the main direction of research at this stage. Voice and image can already be processed well, but natural language processing still lacks relevant theory and technical solution. “If the bottleneck of ‘natural language comprehension’ can be removed, I believe the industry will truly enter the era of next-generation AI,” Mr. Yu said. Mr Yu also shared his view on the smart home, a hot topic at the moment. He thought the concept of smart home in the past put too much emphasis on networking and connection and lacked a truly ‘smart’ brain. The emergence of AI technology will turn home appliances from being passively smart to being actively smart, better meeting the needs of users. According to Mr. Yu, consumer electronics was still not very sophisticated over a decade ago. At that time, consumers needed to buy several digital devices such as cameras and MP3 players, in addition to their mobile phones so as to meet their entertainment needs. Now, as the technology advances, a smart phone will be able to meet their needs. “This is a typical route of product evolution.” As Mr. Yu pointed out, domestic robots will develop toward this direction and become an integrated control center for security monitoring, building intercom and lighting.

Should AI be made open source?

Many industry leaders believe that if it doesn’t open-source its code to third-party developers, it may end up being marginalized. Ms. Yang Jing, founder of Xinzhiyuan, noted that after Google open-sourced its TensorFlow software last November, Facebook and Baidu also announced their open-source plans on its AI technology. It is fair to say that open source is the trend of the AI field.

“In the past, domestic businesses have done very little in terms of open source, so the origin of many open-source technologies come from outside of China. But the trend remains that many domestic companies will open-source their technologies in the future,” said Mr. Yu, founder and CEO of Turing Robot. While open-sourced algorithms lowered the technological barrier, it is not enough to rely only on the algorithm when developing a smart product. As highlighted by Mr. Yu, open source is the best way to obtain data, and the value of data obtained through open-sourcing technology far exceeds the value of the technology itself. Only through owning data can AI take off.

Mr. Liu Wei, Director of the Man-machine Interaction and Cognition Engineering Laboratory at the Beijing University of Posts and Telecommunications, expressed a different view about open source. According to him, although Google, Facebook and Amazon open-sourced their AI software so any engineer can obtain the codes and develop new apps, some essential codes and data are still kept secret. He believes we can’t have complete confidence in open source with such secrecy.

How to realize the commercial value of AI?

Should AI business models and solutions target businesses or consumers? Mr. Shen Xiuping, Chairman and CEO of Ulucu, believes that any business model, especially a disruptive one, must focus on its environment. For AI robots to truly enter households, there must be mature technology and a smart environment. For the business-to-business market, innovation and application should continue to be the focus of AI companies.

Mr. Xiang Zhonghong, founder and CEO of RobotZero, believes that, so far both the smart home and service robot sectors are faced with a quandary: they need interior design companies and real estate developers to act as intermediaries and have no direct access to end users. To supplement this is a direction where companies in these sectors can make a big difference in the future.

Mr. Hu Jie, General Manager of Qingdao Krund Robotics Co., Ltd., shared his insight from the perspective of business operation and customer needs. “It is understandable for a company to focus on technology, but how to commercialize its technology is also an important matter that demands close attention,” said Mr. Hu. As per his views, technology by itself cannot create commercial value; commercial value comes from the operation of business model. Take for instance the robot showcased at the Spring Festival Gala, it doesn’t have much technological content, but through its appearance on the Gala stage and appropriate advertising, the lite version of the dancing robots has produced considerable profits. This shows that what really matters for an AI company is not how advanced its technology is, but what kind of pain it can relieve for its customers. “This instead is a field of theory that AI companies should carefully study,” Mr. Hu said.

In the smart home sector, in addition to domestic robots, there will be large numbers of other kinds of smart devices. “As part of the home environment, these devices exist not as individual units, but must be interconnected,” said Mr. Ge Hantao, Deputy Secretary-General of Zhongguancun Smart Hardware Industry Alliance. From the standpoint of the industry itself, more user-friendly and more unified man-machine interaction in the future will bring consumers real smart home experiences.

Mr. Lu Tao, Asia-Pacific Regional Director of Aldebaran Robotics/Softbank Robotics, mentioned that to produce robots, we first need to consider consumers’ trigger points. Generally speaking, consumers’ trigger points have two stages. At the first stage, consumers show interest when laying their eyes on the robot, but such interest is merely the first reaction to the robot’s appearance or cool features. At the second stage, their excitement cools down and they start to think whether and how such product can meet their needs. A robot producer, therefore, should design robots from the perspective of the customers’ trigger points. Meantime, it should pay more attention to consumer application. Many smart home companies are used to thinking from the perspective of business application. A business application is a more “uniform” matter, which is relatively simple, whereas consumer application is much more complex. To meet these individuals’ demands is not a small challenge for many companies. Only the most difficult task is worthy of the greatest success. Smart home companies need to change their perspective and strike a balance between business application and consumer application.