A year ago, Jeffrey Ding, a DPhil researcher from the University of Oxford started a newsletter “
The book outlines a Chinese national strategic initiative on AI. Ding thought it would be a pity if the content remains in Chinese and a wider audience is unable to appreciate its insights. Ding’s effort was well-received, not only by his peers but also other readers. As interests mount, Ding began to turn other Chinese publications of similar nature into English.
He compiles them into a (nearly) weekly newsletter, which has now more than 2800 subscribers. Ding celebrated the first anniversary of his newsletter at the beginning of the month. He summarized some of the differences he had noticed, in terms of AI cultures between the west and the east, over the past 12 months and published them on MIT Technology Review.
From distorted understanding to the unprecedented hype
One of the first things Ding highlighted was the “language asymmetry” existed between the English and Chinese-speaking AI communities. He said the latest AI news and big developments are often translated from English into Chinese within a day or two. Most researchers in China are also capable of reading English. As such, the Chinese have a good grasp of the global AI landscape.
However, the opposite seldom takes place. AI breakthrough and progress within China experience a lower tendency to be translated into English. Thus, western observers may have an unrealistic view of what is going on with, for example, a Chinese tech tycoon like Tencent at a technical level. Coupled with the spread of false information on the internet and media sensationalism, the western audience may have developed a rather exaggerated
What often regard as the “AI arms race” between the East and the West may not be entirely true especially in the eyes of the Chinese. Previously, some Chinese writers had detailed the less remarkable sides of Chinese technology. These include the lack of engineers to improve on present algorithms and only a handful of the technology companies are investing on AI etc. To them, a misunderstood or distorted perception of Chinese technology will put the western World in a less advantageous position in the long run.
AI at a government or ethics level
The other aspect which the western audience fail to realize is the Chinese government sees AI as a tool for social governance. Based on a report drafted by a local consulting firm, Yiou Intelligence, security and AI firms occupy most of its top 100 AI company list.
Some of the enterprises, such as successful facial recognition startups like Sensetime and Megvii (Face++), are part of the Chinese government surveillance over Xinjiang. An autonomous territory in the Northwest mostly dominated by ethnic minority groups. Others may be part of the “Silk Road Economic Belt” conglomerate which exports surveillance technology to countries in Central Asia.
Besides, the Chinese do care about AI ethics and privacy as much as their western counterparts. Two technology giants – Tencent and Huawei, were involved in a dispute of user privacy earlier. Last summer, the Shandong Province brought charges against 57 individuals and eleven data companies for violating personal information. The incidents had created an extensive debate on an improved standard of personal information protection.
The China-US AI collaborations
Ultimately, Ding believes AI research will benefit greatly from a China-US collaboration. Over time, ideas and talents exchange between the west and the east. This made rooms for one to ponder what makes a company or technology “American” or “Chinese”. A fine example will be Microsoft Research Asia (MSRA) in Beijing, one of the largest research centers the company has outside Redmond, Washington.
For the past 20 years, MSRA had played a pivotal role in many of Microsoft research as it thrives in the Chinese AI ecosystem. MSRA not only trained many of the local AI engineers and technology experts. Projects like Microsoft XiaoIce, a popular social chatbot in China, tapped onto the country massive user base to cultivate on its AI competencies.
A science writer with data background and an interest in current affair, culture and arts; a no-med from an (almost) all-med family. Follow on Twitter.