In 2011, Princeton graduate and Google-trained product manager Wang Yi gave up on the “American dream”. Bidding farewell to Silicon Valley, he returned to China minus a solid plan but with a strong desire to set up a business that “can better meet consumer demands in one of the world’s largest mobile internet markets”.
Now, riding the wave of China’s national plan to achieve global leadership in artificial intelligence (AI), the 38-year-old is perhaps a living example of the “Chinese dream”, part of President Xi Jinping’s national drive to achieve dominance in a wave of cutting-edge technologies to drive future economic growth.
Wang’s Shanghai-based LAIX, which operates AI-powered English-learning app Liulishuo, has grown from a three-man shop into a US-listed company with more than 2,000 employees in less than six years. More than 7 million Chinese now practice English on Liulishuo with their personalised AI teachers online on a monthly basis.
“Government policies work like batons in China. It [the government] can rapidly guide capital and talent towards the direction it wants,” said Wang, talking about his experience of setting up an AI business in a country that has prioritised the field.
But it is this flowering of AI start-ups, covering applications from education to security and finance, pushed forward by the “baton effect” of government policy, which has fanned Western accusations of unfair state intervention in China's economy.
China’s national plan for AI development, a three-step road-map towards worldwide dominance in AI by 2030, is just one of the more than 100 top-down state-led policies designed to propel industrial and technological development. This has not only unnerved the West, it has become central to the ongoing trade war between the US and China.
“Empirically, we have no examples of large authoritarian economies that succeeded in relatively modern times in exploiting new technologies and innovations in a broader and enduring way to become rich and commercially successful,” said George Magnus, an associate at Oxford University's China Centre and a former chief economist at UBS.
“China could be different – and certainly the first to break ranks. But to do so it needs to show that it can use its scientific and engineering capabilities top down to generate multiple new innovations and change bottom-up across a wide range of industrial and service producing sectors,” said Magnus, who recently published a book Red Flags: Why Xi Jinping’s China is in Jeopardy.
Over the past two decades, China has achieved rapid technological progress, thanks in no small part to its massive investment in research and development, which amounted to 2.2 per cent of GDP last year.
Xi has also, on numerous occasions, called for the construction of an ‘innovation-driven economy’ and has been forthright about his ambition to make the country a global leader of innovation by 2035.
But this top-heavy approach contrasts sharply with the US’ belief in the “invisible hand of the market” – the naturally emerging force that, according to influential economist Adam Smith, regulates a free market. Can innovation be instilled from the top as opposed to emerging from the dynamism of freewheeling capitalism that promises rich rewards for those who succeed?
Although China is not standing shoulder-to-shoulder yet with the West on core disruptive technologies, it has emerged as an up-and-coming challenger, thanks in large part to a wide range of supportive industrial policies – particularly Beijing’s sensitive “Made in China 2025” blueprint.
This 2015 document – which has become a lightning rod in the escalating trade war between the US and China – called for an upgrade of China’s manufacturing model to better take on the US in key strategic industries such as robotics, aerospace and new energy vehicles.
The US has seized on the plan as an example of what it sees as unfair state intervention in China's economy.
“Unfair trade, security and industrial policies, tolerable in a smaller developing economy, are now combined with China’s immense, government-directed investment and regulatory policies to put foreign firms at a disadvantage,” said James Lewis, director of technology policy programs at the Washington-based Center for Strategic and International Studies.
“China now has the wealth, commercial sophistication and technical expertise to make its pursuit of technological leadership work,” he was quoted as saying in a March report by the US Trade Representative Office.
“The fundamental issue for the US and other western nations, and the IT sector, is how to respond to a managed economy with a well-financed strategy to create a domestic industry intended to displace foreign suppliers.”
Substantial funding – including research grants, loans, tax discounts and subsidies – is arguably one of the strongest weapons a top-down approach to innovation has and is perhaps the weakest area for a bottom-up free market economy, say some experts.
“China’s strengths here relate to the ways in which a broad technology strategy can be formulated, and responsibility for implementation can be distributed among lower tiers such as state enterprises, local governments, and designated private firms,” said Magnus in response to emailed questions.
“Plus, of course, companies can be given special loans and [favourable] fiscal and commercial treatment. There may not be capital constraints or funding pressures. In theory it should be possible to pool science, education, laboratory and data resources to deliver desired outcomes.”
There is no official data regarding the total amount of money the Chinese government has showered on its favoured industries. But in the electric vehicles (EV) industry alone, China’s central and local governments allocated US$7.7 billion in subsidies in 2017, both to makers of the cars and the consumers that buy them.
With average subsidies of about US$10,000 per electric vehicle in China, the country has cemented its position as the world’s largest EV market with 770,000 EVs manufactured and sold in China in 2017.
“The Chinese system makes it better positioned in the top-down approach due to the government’s efficiency in execution,” said Jeffrey Towson, professor of investment at the Peking University Guanghua School of Management, adding that the US has no mechanism to support a top-down model.
“The approach is especially effective in building infrastructure, such as high-speed rail or next generation mobile networks. But is it a better way for innovation? I am not so sure,” said Towson, adding the US has a long track record of producing innovation and disruptive technologies via its bottom-up approach to entrepreneurship.
According to Towson, markets tend to be much smarter than top-down decision makers, and only allow the best business ideas to succeed and float to the surface.
Meanwhile, China is also directing its educational system to train talent in preparation for an AI-powered future.
“The government is able to take a whole-of-society approach to implement educational reforms, from constructing an AI academic discipline, to establishing AI majors, and creating AI institutes,” said Jeffrey Ding, an AI researcher at Oxford University’s Future of Humanity Institute.
For some experts, a top-down model can be used to perform some “catch up” action, such as rapidly scaling up adoption of certain technologies. But using it to produce groundbreaking innovation is an entirely different undertaking.
“The problem is if you look at government plans to bolster strategic industries for years, they just haven’t worked,” said Samm Sacks, a cybersecurity policy and China digital economy fellow with New America.
“The semiconductor industry is a prime example of throwing billions of dollars of state money at a sector – plus massive policy support – and still when the US threatens to cut off access to components, such as under the ZTE ban last spring, you have a national champion nearly brought to its knees,” she said.
Christopher Thomas, a partner at McKinsey & Co who co-leads the company’s semiconductor practice in Asia, says it takes three to four technology generations of chip design to get to world-class starting from scratch, and therefore it is hard to predict whether the Chinese approach will work or not with limited evidence.
“There are obvious benefits of a top-down approach, but it is so difficult for government to get it right,” he said, adding that in terms of building a national champion, the hardest and trickiest thing is to choose and finance a company that can win.
Thomas does not think it’s a binary decision between top-down and bottom-up approaches to innovation – it is more a question of degree.
“There are many top-down innovation programs in China, but there is also a lot of bottom-up [innovation] going on,” he said.
For Liulishuo’s Wang, it is the attention brought by the national AI strategy rather than financial support that has helped the company the most.
“Anyone aged between 7 and 70 in China knows the term AI,” says Wang. “We don’t need to explain who we are and what we do when we hire, people know that we represent a very promising industry in China.”