AI Superpowers

by Kai-Fu Lee

  By 2015, these people were beating down our door—in one case, literally breaking Sinovation’s front door—for the chance to work with us. That group included scrappy high school dropouts, brilliant graduates of top universities, former Facebook engineers, and more than a few people in questionable mental states. While I was out of town, the Sinovation headquarters received a visit from one would-be entrepreneur who refused to leave until I met with him. When the staff told him that I wouldn’t be returning any time soon, the man lay on the ground and stripped naked, pledging to lie right there until Kai-Fu Lee listened to his idea.

  That particular entrepreneur received a police escort rather than a seed investment, but the episode captures the innovation mania that was gripping China. A country that had spent a decade dancing around the edges of internet entrepreneurship was now plunging in headfirst. The same went for Guo Hong. While creating the Avenue of the Entrepreneurs, Guo caught the entrepreneurial bug himself, and in 2017 he left the world of Chinese officialdom to become the founder and chairman of Zhongguancun Bank, a financial “startup” modeled on Silicon Valley Bank and dedicated to serving local entrepreneurs and innovators.

  All the pieces were now in place for the flourishing of China’s alternate internet universe. It had the leapfrog technology, the funding, the facilities, the talent, and the environment. The table was set to create internet companies that were new, valuable, and uniquely Chinese.

  HERE, THERE, AND O2O EVERYWHERE

  To do all of this, the Chinese internet had to get its hands dirty. For two decades, Chinese internet companies had played a role similar to that of their American peers: information nodes on a digital network. Now they were ready to dive into the nitty-gritty details of daily life.

  Analysts dubbed the explosion of real-world internet services that blossomed across Chinese cities the “O2O Revolution,” short for “online-to-offline.” The terminology can be confusing but the concept is simple: turn online actions into offline services. E-commerce websites like Alibaba and Amazon had long done this for the purchase of durable physical goods. The O2O revolution was about bringing that same e-commerce convenience to the purchase of real-world services, things that can’t be put in a cardboard box and shipped across country, like hot food, a ride to the bar, or a new haircut.

  Silicon Valley gave birth to one of the first transformational O2O models: ride-sharing. Uber used cell phones and personal cars to change how people got around cities in the United States and then around the world. Chinese companies like Didi Chuxing quickly copied the business model and adapted it to local conditions, with Didi eventually driving Uber out of China and now battling it in global markets. Uber may have given an early glimpse of O2O, but it was Chinese companies that would take the core strengths of that model and apply it to transforming dozens of other industries.

  Chinese cities were the perfect laboratory for experimentation. Urban China can be a joy, but it can also be a jungle: crowded, polluted, loud, and less than clean. After a day spent commuting on crammed subways and navigating eight-lane intersections, many middle-class Chinese just want to be spared another trip outdoors to get a meal or run an errand. Lucky for them, these cities are also home to large pools of migrant laborers who would gladly bring that service to their door for a small fee. It’s an environment built for O2O.

  The first O2O service other than ride-hailing to truly take off was food delivery. China’s internet juggernauts and a flood of startups like Wang Xing’s Meituan Dianping all made O2O food delivery plays, pouring subsidies and engineering resources into the market. Crowds at Chinese restaurants thinned out, and streets filled up with swarms of electric scooters trailing steam from the hot meals they carried on board. Payments could be made seamlessly through WeChat Wallet and Alipay. By the end of 2014, Chinese spending on O2O food delivery had grown by over 50 percent and topped 15 billion RMB. By 2016, China’s 20 million daily online food orders equaled ten times the total across the United States.

  From there the O2O models became even more creative. Some hair stylists and manicurists gave up their storefronts entirely, exclusively booking through apps and making house calls. People who were feeling ill could hire others to wait in the famously long lines outside hospitals. Lazy pet owners could use an app to hail someone who would come right over and clean out a cat’s litter box or wash their dog. Chinese parents could hire van drivers to pick up their children from school, confirming their ID and arrival home through apps. Those who didn’t want to have children could use another app for around-the-clock condom delivery.

  For Chinese people, the transition took the edge off urban life. For small businesses, it meant a boom in customers, as the reductions in friction led Chinese urbanites to spend more. And for China’s new wave of startups, it meant skyrocketing valuations and a ceaseless drive to push into ever more sectors of urban life.

  After a couple of years of explosive growth and gladiatorial competition, the manic production of new O2O models cooled off. Many overnight O2O unicorns died once the subsidy-fueled growth ended. But the innovators and gladiators who survived—like Wang Xing’s Meituan Dianping—multiplied their already billion-dollar valuations by fundamentally reshaping urban China’s service sector. By late 2017, Meituan Dianping was valued at $30 billion, and Didi Chuxing hit a valuation of $57.6 billion, surpassing that of Uber itself.

  It was a social and commercial transformation that was powered by—and which further empowered—WeChat. Installed on more than half of all smartphones in China and now linked to many users’ bank accounts, WeChat had the power to nudge hundreds of millions of Chinese into O2O purchases and to pick winners among the competing startups. WeChat Wallet linked up with top O2O startups so that WeChat users could hail a taxi, order a meal, book a hotel, manage a phone bill, and buy a flight to the United States, all without ever leaving the app. (Not coincidentally, most of the startups WeChat picked to feature in its Wallet were also the recipients of Tencent investments.)

  With the rise of O2O, WeChat had grown into the title bestowed on it by Connie Chan of leading VC fund Andreesen Horowitz: a remote control for our lives. It had become a super-app, a hub for diverse functions that are spread across dozens of different apps in other ecosystems. In effect, WeChat has taken on the functionality of Facebook, iMessage, Uber, Expedia, eVite, Instagram, Skype, PayPal, Grubhub, Amazon, LimeBike, WebMD, and many more. It isn’t a perfect substitute for any one of those apps, but it can perform most of the core functions of each, with frictionless mobile payments already built in.

  This all marks a stark contrast to the “app constellation” model in Silicon Valley in which each app sticks to a strictly prescribed set of functions. Facebook even went so far as to split its social network and messaging functions into two different apps, Facebook and Messenger. Tencent’s choice to go for the super-app model appeared risky at the start: could you possibly bundle so many things together without overwhelming the user? But the super-app model proved wildly successful for WeChat and has played a crucial role in shaping this alternate universe of internet services.

  THE LIGHT TOUCH VERSUS HEAVYWEIGHTS

  But the O2O revolution showcased an even deeper—and in the age of AI implementation, more impactful—divide between Silicon Valley and China—what I call “going light” versus “going heavy.” The terms refer to how involved an internet company becomes in providing goods or services. They represent the extent of vertical integration as a company links up the on- and offline worlds.

  When looking to disrupt a new industry, American internet companies tend to take a “light” approach. They generally believe the internet’s fundamental power is sharing information, closing knowledge gaps, and connecting people digitally. As internet-driven companies, they try to stick to this core strength. Silicon Valley startups will build the information platform but then let brick-and-mortar businesses handle the on-the-ground logistics. They want to win by outsmarting opponents, by coming up with novel and elega
nt code-based solutions to information problems.

  In China, companies tend to go “heavy.” They don’t want to just build the platform—they want to recruit each seller, handle the goods, run the delivery team, supply the scooters, repair those scooters, and control the payment. And if need be, they’ll subsidize that entire process to speed user adoption and undercut rivals. To Chinese startups, the deeper they get into the nitty-gritty—and often very expensive—details, the harder it will be for a copycat competitor to mimic the business model and undercut them on price. Going heavy means building walls around your business, insulating yourself from the economic bloodshed of China’s gladiator wars. These companies win both by outsmarting their opponents and by outworking, outhustling, and outspending them on the street.

  It’s a distinction captured well by comparing well-known restaurant platforms in two countries, Yelp and Dianping. Both were founded around 2004 as desktop platforms for posting restaurant reviews. They both eventually became smartphone apps, but while Yelp largely stuck to reviews, Dianping dove headfirst into the group-buying frenzy: building out payments, developing vendor relationships, and spending massively on subsidies.

  When the two companies went into online ordering and delivery, they took different approaches. Yelp moved late and went light. After eleven years as a purely digital platform that lived off advertising, in 2015 Yelp finally took a baby step into deliveries by acquiring Eat24, an ordering and food-delivery platform. But it still asked restaurants to handle the majority of deliveries, just using Eat24 to fill in gaps for restaurants that didn’t have delivery teams. The lightweight process offered restaurants few real incentives to participate, and as a result, the business never fully took off. Within two and a half years, Yelp had given up, selling Eat24 to Grubhub and retreating to its lightweight approach. “[The sale to Grubhub] allowed us to do what we do best,” explained Yelp CEO Jeremy Stoppelman, “which was to build the Yelp app.”

  In contrast, Dianping went into commerce early and went very heavily into food delivery. After four years in the trenches of the group-buying wars, Dianping began piloting food delivery in late 2013. It spent millions of dollars hiring and managing fleets of scooter-riding teams that delivered orders from restaurant to doorstep. Dianping’s delivery teams did the legwork, so every mom-and-pop shop suddenly had the option of expanding its customer base without having to hire a delivery team.

  By throwing tons of money and people at the problem, Dianping could attain economies of scale in China’s dense urban centers. It was an expensive and logistically taxing endeavor, but one that ultimately improved efficiency and reduced costs for the end customer. Eighteen months after debuting its delivery service, Dianping doubled down on those economies of scale by merging with archrival Meituan. By 2017, Meituan Dianping’s valuation of $30 billion was more than triple that of Yelp and Grubhub combined.

  Other examples of O2O companies in China going heavy abound. After driving Uber out of the Chinese ride-hailing market, Didi has begun buying up gas stations and auto repair shops to service its fleet, making great margins because of its understanding of its drivers and their trust in the Didi brand. While Airbnb largely remains a lightweight platform for listing your home, the company’s Chinese rival, Tujia, manages a large chunk of rental properties itself. For Chinese hosts, Tujia offers to take care of much of the grunt work: cleaning the apartment after each visit, stocking it with supplies, and installing smart locks.

  That willingness to go heavy—to spend the money, manage the workforce, do the legwork, and build economies of scale—has reshaped the relationship between the digital and real-world economies. China’s internet is penetrating far deeper into the economic lives of ordinary people, and it is affecting both consumption trends and labor markets. In a 2016 study by McKinsey and Company, 65 percent of Chinese O2O users said that the apps led them to spend more money on dining. In the categories of travel and transportation, 77 percent and 42 percent of users, respectively, reported increasing their spending.

  In the short run, this cash-flow stimulated the Chinese economy and pumped up valuations. But the long-term legacy of this movement is the data environment it created. By enrolling the vendors, processing the orders, delivering the food, and taking in the payments, China’s O2O champions began amassing a wealth of real-world data on the consumption patterns and personal habits of their users. Going heavy gave these companies a data edge over their Silicon Valley peers, but it was mobile payments that would extend their reach even further into the real world and turn that data edge into a commanding lead.

  SCAN OR GET SCANNED

  As O2O spending exploded, Alipay and Tencent decided to make a direct bid for disrupting the country’s all-cash economy. (In 2011, Alibaba spun off its financial services, including Alipay, into a company that would become Ant Financial.) China had never fully embraced credit and debit cards, instead sticking to cash for the vast majority of all transactions. Large supermarkets or shopping malls let customers swipe a card, but the mom-and-pop shops and family restaurants that dominate the cityscape rarely had point-of-sale (POS) devices for processing plastic cards.

  The owners of those shops did, however, have smartphones. So China’s internet juggernauts turned those phones into mobile portals for payments. The idea was simple, but the speed of execution, impact on consumer behavior, and resulting data have been astonishing.

  During 2015 and 2016, Tencent and Alipay gradually introduced the ability to pay at shops by simply scanning a QR code—basically a square bar code for phones—within the app. It’s a scan-or-get-scanned world. Larger businesses bought simple POS devices that can scan the QR code displayed on customers’ phones and charge them for the purchase. Owners of small shops could just print out a picture of a QR code that was linked to their WeChat Wallet. Customers then use the Alipay or WeChat apps to scan the code and enter the payment total, using a thumbprint for confirmation. Funds are instantly transferred from one bank account to the other—no fees and no need to fumble with wallets. It marked a stark departure from the credit-card model in the developed world. When they were first introduced, credit cards were cutting edge, the most convenient and cost-effective solution to the payment problem. But that advantage has now turned into a liability, with fees of 2.5 to 3 percent on most charges turning into a drag on adoption and utilization.

  China’s mobile payment infrastructure extended its usage far beyond traditional debit cards. Alipay and WeChat even allow peer-to-peer transfers, meaning you can send money to family, friends, small-time merchants, or strangers. Frictionless and hooked into mobile, the apps soon turned into tools for “tipping” the creators of online articles and videos. Micro-payments of as little as fifteen cents flourished. The companies also decided not to charge commissions on the vast majority of transfers, meaning people accepted mobile payments for all transactions—none of the mandatory minimum purchases or fifty-cent fees charged by U.S. retailers on small purchases with credit cards.

  Adoption of mobile payments happened at lightning speed. The two companies began experimenting with payment-by-scan in 2014 and deployed at scale in 2015. By the end of 2016, it was hard to find a shop in a major city that did not accept mobile payments. Chinese people were paying for groceries, massages, movie tickets, beer, and bike repairs within just these two apps. By the end of 2017, 65 percent of China’s over 753 million smartphone users had enabled mobile payments.

  Given the extremely low barriers to entry, those payment systems soon trickled down into China’s vast informal economy. Migrant workers selling street food simply let customers scan and send over payments while the owner fried the noodles. It got to the point where beggars on the streets of Chinese cities began hanging pieces of paper around their necks with printouts of two QR codes, one for Alipay and one for WeChat.

  Cash has disappeared so quickly from Chinese cities that it even “disrupted” crime. In March 2017, a pair of Chinese cousins made headlines with a hapless string of
robberies. The pair had traveled to Hangzhou, a wealthy city and home to Alibaba, with the goal of making a couple of lucrative scores and then skipping town. Armed with two knives, the cousins robbed three consecutive convenience stores only to find that the owners had almost no cash to hand over—virtually all their customers were now paying directly with their phones. Their crime spree netted them around $125 each—not even enough to cover their travel to and from Hangzhou—when police picked them up. Local media reported rumors that upon arrest one of the brothers cried out, “How is there no cash left in Hangzhou?”

  It made for a sharp contrast with the stunted growth of mobile payments in the United States. Google and Apple have taken a stab at mobile payments with Google Wallet and Apple Pay, but neither has really attained widespread adoption. Apple and Google don’t release user figures for their platforms, but everyday observation and more rigorous analysis both point to massive gaps in adoption. The market research firm iResearch estimated in 2017 that Chinese mobile payment spending outnumbered that in the United States by a ratio of fifty to one. For 2017, total transactions on China’s mobile payment platforms reportedly surpassed $17 trillion—greater than China’s GDP—an astounding number made possible by the fact that these payments allow for peer-to-peer transfers and multiple mobile transactions for items and services throughout the chain of production.

  LEAPING FROGS AND TAXI DRIVERS

  That massive gap is partly explained by the strength of the incumbent. Americans already benefit from (and pay for) the convenience of credit and debit cards—the cutting-edge financial technology of the 1960s. Mobile payments are an improvement on cards but not as dramatic an improvement as the jump straight from cash. As with China’s rapid transition to the mobile internet, the country’s weakness in incumbent technology (desktop computers, landline phones, and credit cards) turned into the strength that let it leapfrog into a new paradigm.