China’s Authoritarian State Has an Edge in Artificial Intelligence Development
A wealthy repressive state can press ahead with tech innovation unimpeded by privacy concerns
China made enormous economic strides in the past quarter-century by manufacturing everything from toys to tires inexpensively and exporting them overseas. To become a truly wealthy nation, it must move beyond its role as a low-cost manufacturer and become an economic innovator itself.
Can a repressive state, led by a central government specializing in five-year plans and surveillance of its own people, make such a leap? The odds against success look steep. Economic history includes few examples of authoritarian states becoming innovative business leaders. But China aims to make that jump in artificial intelligence—or high-level machine learning—with an unusual approach that can’t be dismissed.
Beijing is bankrolling a big effort in AI, in part, to keep better track of homegrown individuals it considers criminals and dissidents, and to intimidate would-be opponents. That work involves fundamental research in image recognition, data collection and sorting that could have commercial spinoffs in the software used to run complex systems. A city with millions of self-driving cars, for instance, would need data analysis and the ability to recognize, say, that a ball bouncing across the street might be followed by a child chasing it.
“What gives China an edge is there is more of a sense of urgency,” says Paul Triolo, a technology research manager at Eurasia Group.
Last summer, China’s central government published a comprehensive plan for artificial intelligence development that aims to make the nation “the world’s primary AI innovation center” by 2030.
In China such plans aren’t simply blueprints, but indications of central government priorities, which work down to localities, state-owned firms and entrepreneurs. So far, local governments, looking to turn themselves into software centers, have pledged about $7 billion in funding for AI development, estimates a congressional panel that focuses on China.
Mix that with genuine enthusiasm for the technology among Chinese scientists and companies—and world-class talent—and you have a formidable force, say U.S. researchers who have studied China’s AI program.
“There’s an enthusiasm for AI and culture of dynamism that we don’t see as much in the U.S., at least outside of Silicon Valley,” says Massachusetts Institute of Technology researcher Erik Brynjolfsson.
RELATED
Among the most innovative—and threatening—of the Chinese research involves surveillance. The Wall Street Journal has detailed how Beijing has turned its western Xinjiang regioninto a warren of facial scanners to track millions of Uighur minorities. Police there use hand-held devices to search smartphones for encrypted chat apps. A new twist: mobile facial-recognition units mounted on eyeglasses that police use to search crowds for fugitives, among other uses.
Artificial intelligence works by the collection of vast amounts of data used to “train” computer software to recognize patterns and reach conclusions, precisely the kind of skills needed to track a small number of dissidents among a sea of Chinese faces.
“A repressive state can be an engine of innovation,” says Robert Atkinson, president of the Information Technology and Innovation Foundation, a tech policy center.
In the past, Chinese efforts to innovate have stumbled over the state’s desire to centralize control, denying researchers and companies the freedom to follow their own ideas. China, for instance, has tried to jump-start its semiconductor industry with projects dating as far back as the Cultural Revolution, but still hasn’t produced a company that can compete with global leaders like Intel or Samsung. Beijing is trying once again in computer chips, putting together a massive government fund, but the U.S. government so far has blocked China’s efforts to acquired advanced Western technology.
AI may be different, say U.S. technology researchers. China’s AI effort is being joined by globally competitive internet powerhouses including Tencent Holdings Ltd. , which is focused on medical imaging, and Alibaba Group Holding Ltd. , which wants to help create smart cities that use sensors, cameras and computers to manage traffic. So far, the government is following their lead in technology development, AI researchers say, whatever the goals laid out in the government plan.
Chinese researchers may not have political freedom, they say, but they have the economic freedom to chart their own course. And the Chinese government doesn’t throw up roadblocks to the technology on the grounds of privacy, as occurs in the West.
“Freedom is very important,” says Mr. Brynjolfsson. “But there is more freedom in China to do startups than most people realize.”
Becoming an innovator in high technology and making a business success of those innovations in international markets is something that only a few countries have done, says Loren Graham, an emeritus MIT science historian.
“All of them—so far—are open, democratic societies in which government companies do not dominate,” he said.
He says China’s push in AI represents a great economic experiment.
“Can a country like China with lots of money combine repression, creativity and economic success based on that creativity? If the answer is yes, then we will have to rethink everything.”
Write to Bob Davis at bob.davis@wsj.com
Appeared in the February 26, 2018, print edition as ‘Authoritarianism Is China’s Edge in AI.’