Direction Of Tide: Jingdong Technical Spectrum

The shell shot out of the cast-iron barrel. Two people beside the cannon, one covered his ears and the other waved his arms. Spring tide surging, waves surging, the center of the vortex, looming high-rise metropolis

These are two works from Bob Dylan's personal art exhibition this year. Their inspiration comes from the two lyrics of the Nobel Prize winner for literature, "the answer is in the wind" and "the storm is coming".

From an artistic point of view, such a creation is a precious inspiration and manuscript of the author. But from a technological point of view, it can become a very interesting thing. What is the creative ability of technology? Is there a boundary between science and art? Where is the imagination space of the combination of the two?

At present, there is no standard answer in the industry, but Jingdong has begun its own exploration. On November 25, at the 2020 Jingdong global science and technology Explorer conference, Zhou Bowen, chairman of the technical committee of Jingdong group, President of Jingdong Zhilian cloud and IEEE fellow, announced the establishment of Jingdong exploration and Research Institute for new technologies in the next ten years. Science and art is one of the six major research directions.

Of course, Jingdong's exploration of future technology is not limited to this. "Artificial intelligence", "quantum computing", "data science, engineering and management", "decentralized computing", "technical ethics" are also important research fields.

Roughly speaking, several directions set by Jingdong exploration and Research Institute are frontier and grand, even some metaphysical. But it doesn't mean uselessness. "What I can't create, I don & apos; t understand." In an exclusive interview with the 21st century economic reporter, Zhou Bowen quoted a famous saying of physicist Richard Feynman, "in order to truly subversive innovation, the scale or priority of research will change."

For JD, such an opportunity is reflected in the construction of a digital intelligent social supply chain and the formation of a new generation of infrastructure in the future. Photo by Gan Jun

The direction of the tide

Today is not the former, the future is as fast as the wind. The world is changing rapidly.

But behind the stream, there is always a direction, and the tide of Commerce and technology is no exception. "Standing at the current time point, looking forward to 2030, I think there will be two certainty, one is that technology will continue to iterate, and the other is that enterprises must expect their own continuous development and growth." Zhou judged that by 2030, all enterprises will become technology enterprises.

In fact, the forecast data released by IDC recently also supports this judgment. According to its prediction, by 2022, 65% of global GDP will be driven by digitalization; by 2023, 75% of organizations will have a comprehensive digital transformation (DX) implementation roadmap, much higher than the current 27%; by 2025, driven by the global turbulent environment, 75% of business leaders will use digital platforms and ecosystem capabilities to adjust their value chains.

"The digital transformation of industry represented by artificial intelligence, big data, cloud computing, Internet of things and other technologies is our biggest certainty at present and the biggest opportunity for us in the next 10 years." Zhou Bowen pointed out. For JD, such an opportunity is reflected in the construction of a digital intelligent social supply chain and the formation of a new generation of infrastructure in the future.

It needs a new generation of technologies, including artificial intelligence, 5g, IOT, etc., to provide "nourishment", which is also the driving force for the establishment of Jingdong exploration and Research Institute.

It should be noted that some research directions of Jingdong exploration and Research Institute are not popular words like "big data" and "blockchain", but "data science, engineering and management" and "distributed computing" instead. Even, "technical ethics" and "science and art" are not popular in the current technological context.

"We don't simply pursue the current hot spots." In this regard, Zhou Bowen threw a word to the ground.

This is the long-term thinking behind Jingdong's technology. For example, in terms of engineering management, Zhou Bowen mentioned that in order to truly transform AI into productivity, the engineering of AI is very important. This includes both secure storage, AI call center, and various business scenarios such as voice outbound call. It can also help decision-making and working levels improve management efficiency.

"At present, our cost in the field of artificial intelligence is relatively high. If we do better in engineering, it may not be this number." Zhou Bowen revealed.

Another example is that after seeing that artificial intelligence has more and more creative potential, Jingdong is also considering how it can understand the subjective emotions behind poetry, music, calligraphy, etc. "Artificial intelligence itself has no emotion, but from a deeper level, it should also have emotions. We hope to further deepen the understanding of artificial intelligence, which is to continue to expand the service of technology for human beings, and it is also a very meaningful research direction."

Technical function

How to treat the usefulness and uselessness of evaluation technology is an interesting proposition.

The technology direction for the next ten years proposed by Jingdong Exploration Research Institute focuses on disruptive innovation at the basic theoretical level. In the 2020 emerging technology maturity curve released by Gartner, responsible AI, a subdivision of "technical ethics", is still in its infancy. Similar to it is the "bi-directional brain computer interface", which as an important breakthrough of autonomous system and even future cyberberg, is still in the early stage of technology germination.

Neither quantum computing nor distributed computing has even appeared in Gartner's curve. This means that there is a long-term cycle behind every term of the Institute.

Although they seem to be investing heavily and it is difficult to have a direct impact on the business in the short term, Zhou believes that such investment is worthwhile.

From the perspective of the big environment, when the boundaries of the traditional Shannon theory, von Neumann classical computing architecture and Moore's law are approaching, the exploration of these cutting-edge technologies becomes more difficult or even urgent. At the same time, the investment in technology in the next decade not only reflects Jingdong's determination and confidence in technology, but also collides with and complements the "functional" technology research of JD.

"There is no doubt that there is great commercial value in generating images and videos from text. At present, we are able to help models change their clothes online without violating their feelings, which is based on the enhancement of our technology to understand pictures Zhou Bowen said. Such technical practice also includes the first large-scale emotional artificial intelligence customer service launched by Jingdong, which can better serve and respond to users.

"But first of all, we need technology to better understand and express emotions and emotions. From a deeper level, we need to solve the problem of how artificial intelligence and human society live in harmony." Zhou Bowen pointed out.

In the world's top technology enterprises, such innovative ideas are not uncommon. Jim pinkelman, senior director of Business Expansion Department of Microsoft Research Institute, once told the 21st century economic reporter that at the beginning of its establishment, Microsoft Research Institute did not focus on product landing, instead, it replaced it with the most cutting-edge computer technology, sometimes even more than ten years ahead.

A specific example is that Microsoft research tried to publish a paper on how to recognize sound in motion in 1999. However, the paper was rejected because there was no space for this technology. However, the Xbox technology has made it more effective for players to interact with other games in 2010.

However, from the perspective of the actual research, it does not mean that the research of Jingdong will be advanced. For example, in the field of retail, it is necessary to conduct in-depth research on people's impulse shopping, so as to guide people's rational consumption with technical means, which is also the space for the application of technological achievements in the scene.

However, the research methods adopted by Jingdong Institute are closely related to the research methods of human beings. We can see the impact and collision of different ideas brought about by this diversity, and this is a very interesting thing

?