Bengaluru: Silicon Valley just can’t crack China. In July, Facebook won approval to open a subsidiary in China; a day later, its licence was revoked. Earlier this month, news surfaced that Google was working on censored versions of their products that comply with China’s Great Firewall, the system used to control internet access. Backlash from Google’s own employees forced the company’s leadership to clarify that the project was exploratory.

The desperation of Google and Facebook to enter China isn’t driven by a hankering for growth alone. At stake is the opportunity to control the global economy at a scale not previously possible. The US has always been at the centre of the technology universe. While there are examples of successful tech companies founded in other countries, American companies defined the technology direction. The rise of China’s tech has challenged that position.

In 2018, among the leading internet-based companies globally, 9 were Chinese (Figure 1). The remaining 11 were American companies, highlighting the dominance of the two nations in technology leadership. China’s tech giants even have their own acronym, BAT (Baidu, Alibaba and Tencent) to rival the American FAANG (Facebook, Apple, Amazon, Netflix and Google).

Within start-ups, China’s share of venture funding has grown from 6% of total funds in 2011 to ~22% today. The country is among the top three in the world for fundraising in high-growth sectors like virtual reality, AI, drones and autonomous vehicles. Last year, investors ploughed almost half of all AI-related funding into China. The country is also fast overtaking the US in AI and machine-learning (ML) R&D, as measured by patent filings.

This shift in China’s technology sector—from low-cost manufacturer to innovation-leader—is fuelled in part by the changing nature of the Chinese economy. Domestic consumption now accounts for almost 62% of China’s GDP growth. In e-commerce, China outstrips the US, both in absolute numbers and in growth rates (Figure 2).

The Chinese government, never a passive spectator, has also limited competition by restricting entry to foreign firms. In the absence of any real competition from Google, Amazon or Facebook, China spawned domestic alternatives. Building on this progress, the government has set very aggressive targets to become an innovation driven economy by 2020. The 13th five year plan seeks to double the number of patents issued, increase the share of high-tech in GDP and establish the country in the “top 15" in innovative nations, among other targets. A large local market, high internet penetration, limited foreign competition and supportive government policies have helped China’s tech companies flourish.

For almost a decade, US and Chinese technology companies avoided any direct conflict. Chinese companies were focused on winning the local market. US tech companies, barred from China, expanded into other markets. But with China’s technology firms expanding globally, US tech faces its fiercest challenge yet.

Big fish in a big pond

Leadership in technology matters. Tech companies now make-up more than 25% of S&P index, the largest of any sector. Recent research has also shown a strong link between technology intensity and economic growth. But the technology sector is unlike any other with a chameleon-like ability to pervade every aspect of our lives. From self-driving cars and digital health to robo-investors and smart farming, digital disruption bends the cost-curve downwards, improves productivity and displaces incumbent companies.

Seen in this light, China’s rapid march towards technology leadership has been impressive. But displacing the US isn’t easy.

For one, compare the top American and Chinese companies in key technology verticals (Figure 3). It’s striking how mutually exclusive these two lists are. US tech, which has no presence in China, is leader in most segments globally. Facebook for social media, Google for search, Amazon for e-commerce, Netflix for streaming video; US tech dominates. Chinese companies need to grow outside the mainland to maintain their growth rates. In the past few months fears of a slowing economy have weighed on superstar tech stocks. Tencent, once Asia’s most valuable company, has seen its market cap fall by more than $140 billion since January.

But instead of exporting their products to international markets, China’s tech giants have taken a different route by investing billions in fast-growing start-ups. Tencent has built significant stakes in Snap, Tesla and Go-Jek, while Alibaba has invested in Lazada, Magic Leap and Paytm, among others. A bulk of the BAT-trio’s investments have focused on the Southeast Asian market, where conditions are similar to those in China a decade ago.

Part of this is strategic. Chinese companies don’t have the know-how to compete in non-Chinese markets. By working with companies that are successful overseas, Chinese firms hope to learn about newer markets. Over time, there may be opportunities to offer services or combine products as a way to enter new markets.

Building data stockpiles

The lack of global reach also translates to another limitation that Chinese companies need to overcome—access to data. The next wave of software will be powered by intelligent algorithms using machine learning and AI techniques. Access to data is fundamental to building these algorithms.

On the surface, it would seem that this is one area that Chinese companies have a leg-up over their Western rivals. In other countries privacy concerns have tempered the application of technology for mass government programs. But in China, the government has embraced technology to build out its surveillance state. The willingness of the government to use technology at scale means Chinese companies are generating data, which their US counterparts will never have access to.

The country’s social credit system, for instance, links data across applications, from payment apps to social media platforms and rates citizens. Yitu Technologies, whose facial recognition software is used to screen visitors at ports and airports, claims to have scanned almost 320 million faces. SenseTime has scanned and authenticated the faces of 400 million consumers. The sheer volume and use in real-world scenarios are invaluable in training their algorithms.

But there’s a limitation. By virtue of accessing only data on Chinese citizens, these algorithms are trained on a homogenous data set. Research into the accuracy of AI systems has highlighted the limitation of not training the machine on a diverse sample. In a recent independent test, Amazon’s facial recognition software wrongly identified 28 members of the US Congress as police suspects. A majority of the matches were people of colour. Amazon dismissed the results saying the researchers used an 80% match confidence instead of the recommended 99% match confidence for law enforcement. But the fallibility of the machine is real—a machine is only as intelligent as the data that it is trained on.

American companies may not have the same level of data access with government authorities, but their users willingly give them their data. Consider the countless photos, videos and posts uploaded on Facebook or WhatsApp across the world. Google has intimate knowledge of its users via their search activity, emailing patterns, photo shared and much more. Even better, this user-generated data is global—the widespread use of these services means that Facebook, Google, Amazon and others are collecting data on a global scale. The diversity of this data makes it invaluable in training more accurate, and ultimately more reliable, algorithms.

Predictions of technology dominance rest on a country’s ability to attract the best talent

To understand the value of data, consider the recent deal that CloudWalk Technology, a Chinese start-up, signed with the Zimbabwean government. In return for providing a mass facial recognition program, CloudWalk gets access to population data whose racial mix is far different from China’s.

Finally, predictions of technology dominance rest on a country’s ability to attract the best talent. Despite the recent moves by the Trump administration, the US remains the preferred destination for talented developers. Getting developers to move to China is a lot harder due to cultural and language barriers. In recent times Alibaba, Baidu and Tencent have opened offices in the US mostly for recruitment purposes.

$100 billion dollar baby

Any analysis about technology winners and losers is incomplete without considering SoftBank. Since launching early last year, SoftBank’s $100 billion Vision Fund has shaken up start-up investing. The fund takes large positions in fast-growing start-ups, often investing in competing companies. With a stated goal of removing the constraint of capital for innovative companies, the fund has become a market-maker in its own right.

Ostensibly a Japanese operation, SoftBank’s fund is hard to place. Of the $100 billion committed in the Vision Fund, the largest investor is the Saudi Arabian government with $45 billion, followed by SoftBank and Dubai’s sovereign wealth fund at $28 billion and $15 billion respectively. SoftBank’s approach shows another approach to technology dominance—instead of building companies, place bets on a portfolio of companies.

In the battle for technology dominance, the US retains an edge. Chinese technology companies have grown under sheltered conditions, with a government that’s kept foreign competitors at bay and prioritized data sharing over privacy concerns. Other governments won’t be as indulgent. RBI’s recent directive to Paytm to stop signing new customers was based on concerns around Alibaba’s ownership of Paytm. Europe’s strict privacy laws require explicit consent from consumers for any data collection. China’s technology companies have shown that they are quick learners. They’ll need to show the same agility if they are serious about toppling US tech.

Shailesh Chitnis is head of product at Compile Inc., a data intelligence company.

Also read | US readying to take on China on the tech front

Also read | Trade, technology and Xi Jinping’s question

Also read | Hold that handshake, trade wars have just begun