China Rising in Age of Applied AI

Category: blog | Write Comment
Posted by Heather Simmons | February 20, 2019 | Tags: ,

When it comes to deploying artificial intelligence, China is building a formidable early lead. That’s what the startling data from PriceWaterhouseCoopers’ (“PwC”) 2019 survey of 1,378 global CEOs points to.

Let’s start with what the survey says about the countries’ expectations for AI. Expectations matter, because expectations about impact drive investment. Chinese and American CEOs diverge significantly in their expectations for the impact of AI. 84% of Chinese CEOs felt that AI would have a larger impact on the world than the internet. Just 38% of American CEOs felt the same.

Source: The Washington Post, “In Davos, US executives warn that China is winning the AI race,” January 23, 2019

The survey also showed that, while North America boasts some of the world’s most elite AI research talent (Geoffrey Hinton, Yoshua Benigio, and Yann LeCun), China’s companies are ahead when it comes to AI implementation. 25% of Chinese business leaders said they had implemented AI broadly across their organizations, whereas just 5% of American CEOs said the same. PwC estimates that AI rollouts will add $15.7 trillion to global GDP by 2030. China will capture $7 trillion of that total, to North America’s $3.7 trillion.

“The West may have sparked the fire of deep learning, but China will be the biggest beneficiary of the heat the AI fire is generating. That global shift is the product of two transitions: from the age of discovery to the age of implementation, and from the age of expertise to the age of data.” Kai-Fu Lee, AI Super-Powers: China, Silicon Valley, and the New World Order

When did China become so hell-bent on leading the AI revolution? China’s Sputnik moment in AI came in May 2017, when Google’s AI machine, AlphaGo, defeated Chinese teenager Ke Jie. Ke Jie was considered the best player in the world at Go, an ancient Chinese game representing one of the four art forms Chinese scholars were expected to master. The Chinese government’s response was swift, as detailed in Dr. Kai-Fu Lee’s excellent book, AI Super-Powers, China, Silicon Valley, and the New World Order. Two months later, the central Chinese government launched an ambitious plan for building artificial intelligence capabilities, including greater funding, policy support, and centralized coordination for AI development. By 2017, Chinese venture-capital investors made up 48 percent of venture capital investment globally, surpassing the US for the first time. In contrast, between Q1 2012 and Q2 2016, US AI firms had received $17.9 billion in funding, compared to $2.6 billion in funding for Chinese AI startups.

Chinese AI startups raised $4.9 billion in 2017, compared to $4.5 billion for US AI startups
Source: ABI Research, https://www.techrepublic.com/article/chinese-ai-startups-raised-5b-in-vc-funding-last-year-outpacing-the-us/

In his book, Lee identifies four keys to China’s AI dominance: scrappy entrepreneurs, abundant data, AI scientists, and an AI-friendly policy environment. We’ll cover the first two in this blog, and dive into the latter two next week.

Let’s take that first one – scrappy entrepreneurs. China’s entrepreneurs grew up in the most cutthroat competitive environment, one where money is revered over the Silicon Valley obsession with “mission,” copying is commonplace, and the only way to survive is to constantly innovate on yesterday’s version of your product.

“…I’ve spent decades deeply embedded in both Silicon Valley and China’s tech scene, working at Apple, Microsoft, and Google before incubating and investing in dozens of Chinese startups. I can tell you that Silicon Valley looks downright sluggish compared to its competitor across the Pacific…Every day spent in China’s startup scene is a trial by fire, like a day spent as a gladiator in the Coliseum.  The battles are life or death, and your opponents have no scruples.”  Kai-Fu Lee, AI Super-Powers: China, Silicon Valley, and the New World Order

As an example, Lee describes Wang Xing, a Chinese entrepreneur who shamelessly copied Groupon’s business model, turning it into Chinese group-buying site Meituan. When the group-buying craze faded, Groupon did as well, to a $2.6 billion market cap in late 2017. Meituan, meanwhile, had developed an automated payments system that got cash into the hands of sellers faster, and expanded into dozens of new lines of business, from noodles to movie tickets to hotel bookings. Meituan Dianping was valued at $30 billion in 2018, making it the fourth most valuable start-up in the world. Chinese entrepreneurs are not content to rest on “one big idea,” secure in the knowledge that an army of patent lawyers can protect that idea, at least for a time. They are continually tweaking the product, or the model, to better meet customer needs.

Let’s move now to the second factor, an abundance of data. Once compute power and engineering prowess reach a certain level, quantity and quality of available data becomes the determining factor in the capability and accuracy of an algorithm. China has 5 times the US’ mobile internet subscribers and 5X the US’ active social media users. By some estimates, Chinese mobile payment spending outnumbers that of the United States by a ratio of 50 to 1.

Source: Hootsuite and We Are Social 
Source: Hootsuite and We Are Social

In addition to the sheer volume of data, China also benefits from the ubiquity of WeChat, which, unlike Twitter or Facebook or SnapChat, has very broad applications to modern life. WeChat users is the “universal social app” in China, and is used for sending messages, paying for groceries, booking appointments, unlocking shared bikes for transport, and even paying taxes. It’s not an app, it’s an “operating system” for daily life in China. All of those transactions create data, the “new oil” which fuels AI algorithms and models, used to predict consumer behaviour among other things.

We are moving from an era of AI experimentation to AI implementation, and from the power of algorithms to the power of data. China is out to a commanding early lead. To catch up, North America needs to build mammoth, usable data sets that cross organizations while protecting user privacy, relentlessly improve the user experience, take more risks than corporations are used to, and establish government policies that protect user privacy AND encourage corporate investment in AI.

Like this post? See Mensa Canada’s blog for more posts about AI by author Heather Simmons.

Leave a Reply

Your email address will not be published. Required fields are marked *