May, 2018

US, China vie for edge in artificial intelligence

Original article: Business Weekly

Competition is intensifying between the U.S. and China in the emerging field of artificial intelligence, which consultancy McKinsey reckons can create annual value of $3.5 to $5.8 trillion annually across major sectors of the global economy. The U.S. remains ahead of China is most areas of artificial intelligence for now, but Beijing's massive state-backed push to dominate AI by 2030 has raised eyebrows in Washington. The White House and Congress are moving to tighten scrutiny of tie-ups between American and Chinese firms in the AI field. 


For years, America's technology giants were sacrosanct. Silicon Valley, everyone agreed, represented America as a shining city on a hill for the modern age: the world's premier technology hub and birthplace of some of its largest companies by market capitalization. 

The idealized view of the U.S. tech sector changed after the 2016 election, amidst allegations that Russia used Facebook to manipulate American voters. To his own chagrin, U.S. President Donald Trump's "America First" slogan took on new meaning. Now it was about national security, not just a gutted manufacturing sector. 

Amidst the tech backlash and Sino-U.S. trade spat, Washington may step up scrutiny of cooperation between U.S. and Chinese tech firms in artificial intelligence, according to an April Reuters report. Some of the paramount U.S. firms working together with Chinese partners in AI are chipmakers like Nvidia, Qualcomm and Intel. They have long operated under the assumption that technology supply chains are inherently global. These firms invest heavily in China, establishing R&D labs in the country, because they believe the market offers them attractive long-term growth prospects. 

Chinese entities have been accused of widespread IP theft, but sometimes American companies voluntarily give them valuable technology. Last year, Nvidia gave away samples of a new graphics-processing unit that powers data centers, video games and digital currency mining to 30 AI scientists, three who work for the Chinese government. An April report in The South China Morning post states that such practices are standard for a firm like Nvidia, which depends on China for 20% of its business. "Offering early access helps Nvidia tailor products [for the China market] so it can sell more" there, the report says. 

U.S. firms in the AI field are eager to tap a potentially enormous market opportunity: Beijing says it plans to create a domestic AI industry worth US$150 billion by 2030. In an April interview with Wired, ex-Google China chief and tech startup investor Kai-Fu Lee said: “The State Council paper laid out China’s desire to be a hub of AI innovation by 2030, and these papers have teeth in terms of very strong local execution." Waxing lyrical about China's treasure trove of AI data, Lee said that "the data advantage is a huge one," which "can be cranked through the AI engine" for more optimal predictions, efficiency, profits, labor productivity and cost structure.

Unfortunately for U.S. tech firms, Beijing's techno-nationalism doesn't excite Washington in the same way. In fact, President Trump is weighing the invocation of the International Emergency Economic Powers Act, a federal law from 1977 that allows the president to regulate commerce after declaring a national emergency in response to "an unusual or extraordinary threat." 

Invoking the law would allow Trump to review or even terminate informal partnerships between U.S. and Chinese companies, any Chinese investment in a U.S. tech firm, or the Chinese purchase of property near sensitive U.S. military sides, sources told Reuters. 

It is unclear whether Trump will invoke the law, given its potential to upset the applecart. Nevertheless, U.S. firms in the AI field need to think carefully about their partnerships and investments in China, especially Nvidia. The California-based chipmaker should keep in mind that China's Ministry of Science and Technology last year explicitly challenged Chinese researchers to create a chip 20 times faster than Nvidia's.