More stories

  • in

    When will the semiconductor cycle peak?

    AMID A CHIP shortage that has hobbled producers of everything from toys to wind turbines, chipmakers are on a spending spree. On January 13th Taiwan Semiconductor Manufacturing Company (TSMC), the world’s biggest contract manufacturer, said it would spend up to $44bn on new capacity in 2022. That is up from $30bn last year, triple the number in 2019 and ahead of earlier plans to spend over $100bn in total over the next three years. Intel, an American rival, plans to burn through $28bn this year. On January 21st it said it would build two big new factories in Ohio by 2025 at a total cost of $20bn. An option to build six more later would take the overall price tag to $100bn. Samsung of South Korea, TSMC’s closest technological rival, has hinted that its capital spending for 2022 will surpass last year’s $33bn. Smaller firms, such as Infineon in Europe, are also splurging.IC Insights, a research group, reckons that, across the industry, capital spending rose by 34% in 2021, the most since 2017. That torrent of money is welcome news for the industry’s customers, who have been struggling with shortages for over a year. For the industry itself, it is the latest iteration of a familiar pattern. Bumper revenues, like those reported by Intel on January 26th and Samsung the next day, compel companies to expand capacity. But because demand can change much more quickly than the two or more years needed to build a chip factory, such booms often end in busts. The chip business has swung between over- and undercapacity since it emerged in the 1950s, observes Malcolm Penn of Future Horizons, a firm of analysts (see chart). If history is a guide, then, a glut is in on the way. The only question is when.Soon, many analysts think. Demand for smartphones may be cooling, especially in China, the world’s biggest market. Sales of PCs, which boomed during covid-19 lockdowns, also seem poised to weaken, says Alan Priestley of Gartner, a research firm. A survey by Morgan Stanley, a bank, found that, partly thanks to the shortages, 55% of chip buyers were double-ordering, which artificially inflates demand. High inflation and looming interest-rate rises could hit economic growth—and chip demand with it. Mr Penn expects the cycle to turn in the second half of 2022 or in early 2023.This time the glut, when it comes, may not affect all chipmakers equally. TSMC’s boss, C.C. Wei, said this month that a correction could be “less volatile” for his firm thanks to its position at the technological cutting-edge. Much of its new capacity is already booked up in long-term agreements with customers such as Apple, which needs a regular supply of the most sophisticated chips for its newest iPhones.The current cycle may differ from previous ones for another reason. The shortages, and America’s tech-flavoured trade war with China, have reminded politicians how vital chips are to the modern economy—and how over-reliant their supply is on a few giant firms. Worries about the sector’s excessive concentration have led trustbusters to challenge the $40bn acquisition by Nvidia, an American chip designer, of Arm, a British one—successfully, if news reports this week that the deal is being scrapped are to be believed.But governments’ favoured way to deal with the over-reliance is to lure more chipmaking home, mostly from East Asia, with subsidies. On January 25th America’s Commerce Department issued a report to that effect, urging Congress to pass a bill, already approved by the Senate, that includes $52bn in handouts for chipmakers. Mark Liu, TSMC’s chairman, was frank in 2020 when he said such subsidies were vital to persuade his firm to build a new plant in Arizona, one of only a few outside Taiwan. Intel chose Ohio for its factories partly because of incentives offered by the state. Pat Gelsinger, its boss, has been touring rich places that have made similar offers.The EU is keen to match the Americans, potentially putting itself on the hook for tens of billions of dollars of its own. It aspires to double Europe’s share of chipmaking, currently around 10%. In May South Korea’s government talked of a national mission to provide $450bn of capital spending over ten years to protect and expand its national industry. In November Japan unveiled a scheme of its own, with TSMC thought to be getting some $3.5bn. China has long nurtured ambitions—invigorated by American sanctions but so far unsuccessful—to build a fully fledged chip industry.Adding taxpayer cash to chipmakers’ already rich spending plans, says Mr Penn, could lead them to build even more excess capacity than usual. That should give politicians and chip CEOs pause. The bigger the boom, the deeper the subsequent bust. ■For more expert analysis of the biggest stories in economics, business and markets, sign up to Money Talks, our weekly newsletter.This article appeared in the Business section of the print edition under the headline “Party on” More

  • in

    Purpose and the employee

    WHAT IS THE meaning of mayonnaise? For Unilever, a consumer-goods giant whose products are all meant to stand for something, the purpose of its Hellmann’s brand is to reduce food waste by making leftovers tasty. For Terry Smith, a fund manager fed up with Unilever’s dipping share price, this is crazy. “The Hellmann’s brand has existed since 1913,” he wrote earlier this month. “So we would guess that by now consumers have figured out its purpose (spoiler alert—salads and sandwiches).”Mr Smith’s concern is the financial performance of Unilever (in the face of investor disquiet, the firm is now planning management cuts and an overhaul of its operating model). But his underlying point, that doing the obvious job well can be purpose enough, is one that has much wider application. For it is true of colleagues as well as condiments.The very idea of a purposeful employee conjures up a specific type of person. They crave a meaningful job that changes society for the better. When asked about their personal passion projects, they don’t say “huh?” or “playing Wordle”. They are concerned about their legacy and almost certainly have a weird diet.Yet this is not the only way to think about purpose-driven employees. New research from Bain, a consultancy, into the attitudes of 20,000 workers across ten countries confirms that people are motivated by different things.Bain identifies six different archetypes, far too few to reflect the complexity of individuals but a lot better than a single lump of employees. “Pioneers” are the people on a mission to change the world; “artisans” are interested in mastering a specific skill; “operators” derive a sense of meaning from life outside work; “strivers” are more focused on pay and status; “givers” want to do work that directly improves the lives of others; and “explorers” seek out new experiences.These archetypes are unevenly distributed across different industries and roles. Pioneers in particular are more likely to cluster in management roles. The Bain survey finds that 25% of American executives match this archetype, but only 9% of the overall US sample does so. Another survey of American workers carried out by McKinsey, a consulting firm, in 2020 found that executives were far likelier than other respondents to say that their purpose was fulfilled by their job.This skew matters if managers blindly project their own ideas of purpose onto others. Having a purpose does not necessarily mean a desire to found a startup, head up the career ladder or log into virtual Davos. Some people are fired up by the prospect of learning new skills or of deepening their expertise.Others derive purpose from specific kinds of responsibility. Research by a couple of academics at NEOMA Business School and Boston University looked at the experience of employees of the Parisian metro system who had been newly promoted into managerial roles. People who had been working as station agents before their elevation were generally satisfied by their new roles. But supervisors who had previously worked as train drivers were noticeably less content: they felt their roles had less meaning when they no longer had direct responsibility for the well-being of passengers.Firms need to think more creatively about career progression than promoting people into management jobs. IBM, for example, has a fellowship programme designed to give a handful of its most gifted technical employees their own form of recognition each year.Another mistake is to conflate an employee’s commitment with good performance. A recent paper from Yuna Cho of the University of Hong Kong and Winnie Jiang of INSEAD, a business school, describes an experiment in which groups of people with managerial experience listened to two actors playing the part of colleagues. One group heard an “employee” saying that he was looking forward to retirement; another group heard the employee saying that he did not want to retire at all. In all other respects the conversations were the same. The observers assigned a bigger bonus and a higher raise to the employee who appeared to have more passion.There is some logic here. Employees with a calling could well be more dedicated. But that doesn’t necessarily make them better at the job. And teams are likelier to perform well if they blend types of employees: visionaries to inspire, specialists to deliver and all those people who want to do a job well but not think about it at weekends. Like mayonnaise, the secret is in the mixture.For more expert analysis of the biggest stories in economics, business and markets, sign up to Money Talks, our weekly newsletter.This article appeared in the Business section of the print edition under the headline “Purpose and the employee” More

  • in

    Lakshmi Mittal transformed steelmaking. Can his son do it again?

    LAKSHMI MITTAL has two passions: the steel industry and his family. His embrace of the first turned a poor boy from Rajasthan into the “Carnegie from Calcutta”, a man who built the world’s second-biggest steel empire from scratch, culminating in a takeover in 2006 of Arcelor, a European champion. The second sometimes sounds like tabloid fodder: lavish weddings in Paris; family homes—one known as the Taj Mittal—on London’s “Billionaire’s Row”. Yet Mr Mittal’s family knows the steel business inside out. Last year Aditya, his 46-year-old son, became CEO of ArcelorMittal. It now falls to him to transform the industry again.That is because about half of ArcelorMittal’s revenue comes from Europe, where pressure to decarbonise steel production, source of up to a tenth of global carbon-dioxide emissions, is becoming irresistible. The region is laden with coal-burning blast furnaces, the carbon-heaviest of steelmaking technologies. Many are on their last legs. Rather than refurbishing them, some firms are opting to replace these with new direct-reduced-iron (DRI) and electric-arc-furnace (EAF) plants. Blast-furnace steelmaking is doubly carbon-intensive: it uses coking coal to soak up oxygen from iron ore, as well as dirty energy to heat the furnaces. DRI-EAF technology, hitherto dependent on natural gas, can use hydrogen and renewable energy instead. Once scaled up, it could mark a revolution in steelmaking. By jettisoning their once-cherished blast furnaces, European steelmakers hope to start slashing emissions this decade in order to become net-zero by mid-century.Aditya Mittal still has his 71-year-old father, ArcelorMittal’s executive chairman, by his side. But the challenge ahead is uniquely tough. Whereas the older Mr Mittal made his own luck, Aditya is not master of his own destiny. He needs a vast infrastructure of hydrogen and carbon capture to emerge from nowhere to achieve his ambitions, not to mention a market for expensive “green steel”. Unlike his father, who made his fortune by taking privatised steelworks off government hands, he will not succeed unless ArcelorMittal receives taxpayer support. He is not alone in seeking that. The whole industry believes that rapid decarbonisation will be impossible unless governments foot part of the bill. History, however, suggests the state and steel are unpromising bedfellows.ArcelorMittal starts with some advantages. For decades the elder Mr Mittal bought mini-mills in different parts of the world that used DRI pellets and EAFs rather than blast furnaces and basic oxygen furnaces. The technology is still only a bit-player in Europe. Fuelled by hydrogen and renewable electricity, it could become the dominant one within a decade. ArcelorMittal is not the most advanced among European steel companies in developing zero-carbon mills. It has three low-carbon DRI-EAF projects under way, in Spain, Belgium and Canada. SSAB of Sweden is ahead of it. Yet it has reduced debt to shore up its balance-sheet, giving it the flexibility to increase spending. Moreover, its presence in poorer countries such as India, where steel use per person is a fraction of its level in the West, gives it plenty of growth opportunities.The transition will be costly, though. McKinsey, a consultancy, estimates that decarbonising steel requires investment of $145bn a year on average for the next 30 years, and could push the cost of making the stuff up by 30%. ArcelorMittal says its three low-carbon plants will cost $10bn in total by 2030, which is doable for a company with annual capital expenditure of about $3bn. However, its strengthened balance-sheet is raising investors’ hopes of higher payouts, and it needs to weigh their demands against big investments in green steel. Even with modest government support for capital and operating expenditures, says Jefferies, a bank, returns would be too low to justify a normal steel project.That is why the industry believes hefty state backing is essential. ArcelorMittal expects governments to fund about half of its $10bn decarbonisation commitments over the next ten years. Investors argue that subsidies for operational expenses such as electricity bills should be thrown in, too. The same, they say, goes for aid to ramp up production of clean hydrogen, whose price must fall by 60% for clean steel to become cost-competitive with the alternatives, according to McKinsey. On top of that, government money is needed to speed up the roll-out of more renewable energy required to power the clean furnaces. Jefferies estimates that total electricity demand by EU steelmakers will more than double by 2030. The developing world’s blast furnaces, which are younger than Europe’s, will probably be fitted with carbon capture and storage rather than replaced. That nascent technology, too, needs a leg-up from the government.It goes beyond that. By the mid-2020s, Europe’s steelmakers will begin losing the free allocations of carbon permits they receive under the EU Emissions Trading System. To compensate, they await the introduction of a carbon-border-adjustment mechanism, starting in 2026, which will protect them further from importers selling cheaper dirty steel. They also need governments to help kick-start demand for green steel. Some sectors, such as carmakers, are keen to buy it, believing that they can pass the costs on to carbon-conscious consumers. But the construction industry, the steel firms’ biggest market, is not nearly as enthusiastic. Hence steelmakers say they need lots of public works built with low-carbon steel to justify their investments.Kicking the coke habitSome state action is warranted. In the long run subsidies for electric vehicles may curb emissions by less than curing the steel industry’s coal addiction. But the cure must be judicious. It is all too easy for a closer relationship with governments to degenerate into job-safeguarding schemes, protectionism and a revival of the old revolving door between bureaucrats and business. That is what happened the last time the state and steel were intertwined. Until, that is, the elder Mr Mittal made his fortune prising them apart. ■For more expert analysis of the biggest stories in economics, business and markets, sign up to Money Talks, our weekly newsletter.This article appeared in the Business section of the print edition under the headline “The greening of steel” More

  • in

    Will web3 reinvent the internet business?

    LIKE SEEMINGLY everyone these days, Moxie Marlinspike has created a non-fungible token (NFT). These digital chits use clever cryptography to prove, without the need for a central authenticator, that a buyer owns a unique piece of digital property. Alongside cryptocurrencies such as bitcoin, NFTs are the most visible instantiation of “web3”—an idea whose advocates and their venture-capital (VC) backers hail as a better, more decentralised version of the internet, built atop distributed ledgers known as blockchains. Digital artists, celebrities and even the occasional newspaper have issued and sold them to collectors, often for hefty sums (the immaterial version of The Economist’s cover image fetched over $400,000).Although it looked cryptographically sound like any other NFT, however, Mr Marlinspike’s token could shift shape depending on who accessed it. If you bought it and viewed it on a computer, it transformed into a poop emoji. After a few days the NFT was taken down by OpenSea, a marketplace for digital artefacts. This played into Mr Marlinspike’s hands. For his aim was not to raise cash but to raise awareness. His token showed that NFTs are not as non-fungible as advertised. And OpenSea’s reaction illustrated that the supposedly decentralised web3 has its own gatekeepers.The Marlinspike caper was the latest turn in perhaps the biggest controversy to erupt in tech world for several years. On one side sit techno-Utopians, firms offering assorted web3 services and their VC backers. They claim that web3 is the next big thing in cyberspace, that it is truly decentralised—and that it promises juicy returns to boot. Globally, the value of VC deals in the crypto-sphere reached $25bn last year, up from less than $5bn in 2020 (see chart). Last week Andreessen Horowitz (a16z for short), one of Silicon Valley’s most illustrious VC firms and its biggest web3 champion, was reported to be raising a $4.5bn web3-related fund, to add to three existing ones worth a total of $3bn. A senior partner left a16z this month to set up her own firm focused on web3.Pitted against them are the sceptics. They range from Mr Marlinspike, highly respected even among the techno-Utopians for creating the secure-messaging app Signal, to Jack Dorsey, who founded two platforms of the sort that web3 promises to supersede (Twitter in social media and Square in payments). They argue that a truly decentralised internet is a pipe dream—“You don’t own ‘web3’. VCs and their [limited partners] do,” Mr Dorsey warned last month. And a dangerous one at that for the unwary investor: since November some $1trn of the value of cryptocurrencies, the most mature province of web3, has gone up in flames.The feud may seem abstruse. But the stakes are big. It could change the trajectory of the internet—and the multitrillion-dollar business models that it has enabled.The centre cannot holdThe history of modern computing is a constant struggle between decentralisers and recentralisers. In the 1980s the shift from mainframes to personal computers gave more power to individual users. Then Microsoft clawed back some of that power around its proprietary operating system. More recently, open-source software, which users can download for nothing and adapt to their needs, took over from proprietary programs in parts of the industry—only to be reappropriated by giant technology firms to run their mobile operating systems (as Google does with Android) or cloud-computing data centres (including those operated by Amazon, Microsoft and Google).The web3 movement is a reaction to perhaps the greatest centralisation of all: that of the internet. As Chris Dixon, who oversees web3 investments at a16z, explains it, the original, decentralised web lasted from 1990 to about 2005. This web1, call it, was populated by flat web pages and governed by open technical rules put together by standards bodies. The next iteration, web2, brought the rise of tech giants such as Alphabet and Meta, which managed to amass huge centralised databases of user information. Web3, in Mr Dixon’s telling, “combines the decentralised, community-governed ethos of web1 with the advanced, modern functionality of web2”.This is possible thanks to blockchains, which turn the centralised databases to which big tech owes it power into a common good that can be used by anybody without permission. Blockchains are a special type of ledger that is not maintained centrally by a single entity (as a bank controls all its customers accounts) but collectively by its power users. Blockchains have outgrown cryptocurrencies, their earliest application, and spread into NFTs and other sorts of “decentralised finance” (DeFi). Now they are increasingly underpinning non-financial services.The portfolio of a16z offers a glimpse of this wild new world. It already includes more than 60 startups, at least a dozen of which are valued at more than $1bn. Many are developing the infrastructure for web3. Alchemy offers tools for other firms to build blockchain applications, much as cloud-computing provides a platform for developers of web-based services. Nym has built something called “mixnet”, a decentralised network to mix up messages in a way that means literally no one else can tell who is sending what to whom.Other a16z investments are serving end users. Dapper Labs creates NFT applications such as NBA Top Shot, a website where sports fans can buy and sell digital collectables such as key moments in basketball games. Syndicate helps investment clubs to organise themselves into “decentralised autonomous organisations” (DAOs) governed by “smart contracts”, which are rules encoded in software and baked into a blockchain. And Sound.xyz allows musicians to mint NFTs to make money.What all these companies have in common, explains Mr Dixon, is that it is hard for them to lock in customers. Unlike Google and Meta they do not control their users’ data. OpenSea, in which a16z also has a stake, and Alchemy are just pipes to the blockchain. If their customers are unhappy, they can move to a competing service. Even if he wanted, he could not keep them from leaving, says Nikil Viswanathan, Alchemy’s boss. “As a business, I would love to have proprietary choke points. But there aren’t any. We tried to find them.”The idea is that this makes web3 companies try harder to satisfy customers and keep innovating. Whether they can do this while also making pots of money is another matter. It is not clear how much demand exists for truly decentralised projects. That was the problem of early web3 offerings (then called “peer-to-peer” or “the decentralised web”). Services such as Diaspora and Mastodon, two social networks, never really took off. Their successors could face the same problem. A service like OpenSea would be much faster, cheaper and easier to use “with all the web3 parts gone,” says Mr Marlinspike.Or can it?A more fundamental problem is that even if web3 worked as smoothly as its immediate predecessor, it may nevertheless lend itself to centralisation. Lock-in, reckons Mr Marlinspike, tends to emerge almost automatically. The history of the internet has shown that collectively developed technical protocols evolve more slowly than technology developed by a single firm. “If something is truly decentralised, it becomes very difficult to change, and often remains stuck in time,” he writes. That creates opportunities: “A sure recipe for success has been to take a 1990’s protocol that was stuck in time, centralise it, and iterate quickly.”Centralisation and lock-in have been incredibly lucrative. In fact, a16z has made billions from Meta, in which it was an early investor; one of a16z’s founders, Marc Andreessen, sits on Meta’s board to this day. Web3’s VC boosters may be counting on something like this happening again. And to a degree, it already is. Despite being a relatively recent phenomenon, web3 already exhibits signs of centralisation. Because of the complexity of the technology, most people cannot interact directly with blockchains—or find it too tedious. Rather they rely on intermediaries such as OpenSea for consumers and Alchemy for developers.Albert Wenger of Union Square Ventures, a VC firm that started investing in web3 firms a few years ago, points to other potential “points of recentralisation”. One is that the ownership of the computing power that keeps many blockchains up to date is often very concentrated, which gives these “miners”, as they are called, undue influence. It could even allow them to take over a blockchain. In other systems the ownership of tokens is heavily skewed: at recently launched web3 projects, between 30% and 40% is owned by the people who launched them.These dynamics, combined with the latest crypto crash that may cool enthusiasm for the sector among investors, suggest that web3 is unlikely to displace web2 altogether. Instead, the future will probably belong to a mix of the two approaches, with web3 occupying certain niches. Whether or not people keep splurging on NFTs, for example, such tokens make a lot of sense in the metaverse, where they could be used to track ownership of digital objects and to move them from one virtual world to another. Web3 may also play an important role in the creator economy, another buzzy concept. Li Jin of Atelier, a VC firm, points out that NFTs make it much easier for creators of online content to make money from their wares. In this limited way, at least, even the masters of web2 see the writing on the wall: on January 20th both Meta and Twitter integrated NFTs into their platforms. For more expert analysis of the biggest stories in economics, business and markets, sign up to Money Talks, our weekly newsletter. More

  • in

    Where next for air travel?

    WORK AND shopping have, for better or worse, been permanently altered by the pandemic. The airline industry hopes that its own covid-19 disruption proves temporary. Luckily for those deprived of holidays, visits to family and friends, or even the odd business trip, flying in 2022 will look a bit more like the pre-pandemic jet age—with differences between domestic and international routes, short-haul and long-haul ones, and east and west.The numbers taking to the skies have risen steadily since March 2020, when the pandemic first grounded flights. Most forecasters expect that by 2024 as many passengers will fly as did in 2019. IATA, a trade body, reckons that 3.4bn people will buckle up in 2022. That is nearly double the number in 2020, though still some way shy of 2019, when 4.5bn took to the air.Uncertainties remain, however, not least the pandemic. Consider the Omicron variant. Ed Bastian, boss of America’s Delta Air Lines, has described navigating the past few weeks as “hellacious”, after some 8,000 of his staff, about 10% of the total, contracted the virus. Crew shortages, tighter travel restrictions and bad weather conspired to force the cancellation of 60,000 flights worldwide between December 24th and January 3rd, calculates Cirium, an aviation-data firm. That corresponds to roughly one in every 40 flights. The fact that the worst Christmas period for a decade still made December the busiest month of 2021 illustrates just how far the industry has to go.Covid-19’s unpredictable course shows that even bright spots can cloud over. Large domestic markets, unaffected by international travel bans and other unco-ordinated border restrictions over vaccinations and testing, have led the recovery. Within America, the world’s biggest internal market, demand for seats has nudged above 80% of pre-covid levels. In China it has exceeded pre-covid times on occasions over the past year, thanks in part to the country’s strict “zero-covid” strategy. Although lockdowns to snuff out recent outbreaks in the run-up to the Winter Olympics in Beijing next month have slapped the chock blocks back on, China’s aviation regulator still expects domestic traffic at around 85% of pre-pandemic levels in 2022.The plans for restoring capacity among the world’s airlines give a sense of the likely shape of improvement on international routes, which IATA predicts will reach only 44% of pre-crisis demand this year. Some low-cost airlines serving short-haul connections in America and Europe, where travel restrictions may soon be relaxed, could surpass pre-covid capacity, reckons IBA, another aviation-research firm. America’s big three network carriers will also benefit from the reopening of the lucrative transatlantic market, which this year is expected to bounce back to where it was in 2019. Delta will approach pre-covid capacity in 2022, and United may exceed it. Some of Europe’s legacy airlines may benefit, too. IAG, owner of British Airways, is expected to restore all of its flights across the Atlantic by summer 2022.Airlines in the Asia-Pacific region are likeliest to remain stuck. Many governments, relying on isolation to control the virus, have toughened already strict travel rules to contain Omicron. Capacity is still around 60% below previous highs. Singapore Airlines will run at half of its pre-covid capacity for at least the first couple of months of 2022; Australia’s Qantas may operate at just 45% this year.Even if Omicron were the last of covid, airlines have other things weighing them down. As Andrew Charlton of Aviation Advocacy, a consultancy, notes, governments have doused beleaguered airlines with cash to keep them aloft. Much of that—around $110bn, says IATA—needs to be paid back. And that is on top of new debts owed to private-sector creditors. Moreover, so long as demand remains weak airlines will find it hard to pass the rising cost of fuel on to passengers. The industry’s net losses will narrow from the staggering $138bn in 2020 and $52bn in 2021. Collectively, airlines are expected to lose another $12bn this year. Better—but hardly stellar. ■For more expert analysis of the biggest stories in economics, business and markets, sign up to Money Talks, our weekly newsletter.This article appeared in the Business section of the print edition under the headline “Flight tracker” More

  • in

    Can China create a world-beating AI industry?

    “SOUTH OF THE Huai river few geese can be seen through the rain and snow.” In classical Chinese this verse is a breakthrough—not in literature but in computing power. The line, composed by an artificial intelligence (AI) language model called Wu Dao 2.0, is indistinguishable in metre and tone from ancient poetry. The lab that built the software, the Beijing Academy of Artificial Intelligence (BAAI), challenges visitors to its website to distinguish between Wu Dao and flesh-and-blood 8th-century masters. Anecdotal evidence suggests that it fools most testers.The system, whose name means “enlightenment” and which can emulate lowlier types of speech, derives its power from a neural network with 1.75trn variables and other inputs. GPT-3, a similar model built a year earlier by a team of researchers in San Francisco and deemed impressive at the time, considered just 175bn parameters. As such Wu Dao represents a leap in this type of machine learning, which tries to emulate the workings of the human brain. That delights fans of classical literature—but not as much as it does the Communist authorities in Beijing, which have put AI at the heart of China’s technological and economic master plan first set out in 2017. It spooks Western governments, which worry about AI’s less benign applications in areas like surveillance and warfighting. And it intrigues investors, who spy a huge business opportunity.On the face of it, the plan is off to a good start. The logistics arm of JD.com, an e-commerce group, operates one of the world’s most advanced automated warehouses near Shanghai. In May Baidu, China’s search giant, launched driverless taxis in Beijing. SenseTime’s “smart city” AI models—urban surveillance cameras that track everything from traffic accidents to illegally parked cars—have been deployed in more than 100 cities in China and overseas. China has been deploying more AI-assisted industrial robots than any other country. And in 2020 it surpassed America in terms of journal citations in the field.The five most prominent listed Chinese AI specialists are collectively worth nearly $120bn (see chart 1). The biggest of them, Hikvision, has a market value of $60bn. SenseTime, which went public in Hong Kong on December 30th, is worth $28bn. Two more are expected to list soon. In 2020 investments in unlisted AI startups reached $10bn, according to the AI Index compiled by researchers at Stanford University. In its prospectus SenseTime forecasts that revenues from AI-assisted image-recognition and computer-vision software, the most mature part of the market, could hit 100bn yuan ($16bn) by 2025, up from 24bn yuan in 2021 (see chart 2).Look beyond the headlines or Wu Dao’s elegant verses, however, and things look more complicated. Yes, China has made progress on AI, and even the occasional big splash like Wu Dao. But it almost certainly still lags behind America in terms of both investment and cutting-edge innovation. In 2020, three years into the master plan, privately held Chinese AI firms received less than half as much investment as their American counterparts. And a lot of the public and private money pouring into the sector may end up being wasted.China’s five-year-old AI master plan set out a number of goals. For example, by 2025 the country is to create an industry with global revenues of 400bn yuan, achieve “major breakthroughs” in technology and lead the world in some applications. Five years later it is to dominate the industry (by then worth $1trn in sales), having written its ethical code and set its technical standards, just as Europe and America defined the contours of the Industrial Revolution.Elements of the Communist Party’s approach are characteristically prescriptive. The Ministry of Science and Technology has instructed China’s tech giants with existing ventures in certain subdisciplines of AI—Tencent in medical image recognition, Baidu in autonomous driving—to double down on these. That said, the plan is less hands-on than some of the country’s other development projects, observes Jay Huang of Bernstein, an investment firm. In the words of Huw Roberts of Oxford University and five co-authors, the blueprint acts chiefly as a “seal of approval” which “derisks” assorted AI initiatives championed by central-government entities, local authorities and the private sector.In practice, the derisking involves doling out lots of public money. Some of this takes the form of tax breaks and subsidies, as in the “little giants” programme to nurture 10,000 promising startups across various sectors, including AI. Local governments, even in poor rustbelt provinces such as Liaoning in the far north-east, have also dangled similar incentives in front of AI-curious companies.Another type of support comes from government procurement. Firms do not disclose how much revenue they derive from public-sector contracts. But the share is likely to be significant. Central and local authorities use SenseTime’s surveillance technology. Megvii, which also specialises in image recognition, has extensive dealings with state-owned enterprises.The state is also investing in AI companies directly. The central government runs several tech-investment vehicles. Local governments are increasingly creating their own, often armed with billions of dollars. Tianjin, a coastal metropolis, announced a $16bn AI fund in 2018.Government capital is increasingly helping plug a gap left by foreign investors scared away by American sanctions against some of China’s AI darlings, which are seen as being too close to the Communist Party. A fund run by the Cyberspace Administration of China, a regulator, has acquired an undisclosed stake in SenseTime, which last month was hit by another round of American sanctions over its alleged involvement in government repression of the Uyghur ethnic minority. (SenseTime says that the sanctions are based on a “misperception” of its business.) A separate vehicle, the Mixed-Ownership Reform Fund, accounted for $200m of the $765m that the firm raised in its initial public offering (IPO). Local governments chipped in another $220m.Lost in translationState dosh, combined with access to plentiful public data, has helped turn Chinese AI firms into powerhouses in certain niches. According to Bain, a consultancy, by last June the cloud division of Alibaba, China’s e-commerce behemoth, was offering 62 AI-enabled services, from voice recognition to video analytics, compared with 47 from its closest Western rival, Microsoft. SenseTime and Megvii mass-produce computer-vision software and hardware that can be adapted to and installed in individual factories. Despite being locked out of most Western markets by the American sanctions, SenseTime raked in 762m yuan in overseas revenues in 2020, compared with 319m yuan two years earlier, mostly from South-East Asia.For all these successes, though, China’s AI industry trails the West in important ways. Despite leading America in the overall number of AI-related publications, China produces fewer peer-reviewed papers that have academic and corporate co-authors or are presented at conferences, both of which are typically held to a higher standard. It ranks below India, and well below America, in the number of skilled AI coders relative to its population. These shortcomings are likely to persist, for three reasons.First, capital may not be being allocated efficiently. It is unclear, for example, how much of Tianjin’s $16bn kitty has actually been deployed. More damaging, Beijing has created a system for rewarding local officials that favours debt-fuelled spending and seldom punishes wastefulness.Many state AI investments have been “reckless and redundant”, says Jeffrey Ding of Stanford University. Zeng Jinghan of Lancaster University has documented the rise of firms that falsely claim to be developing AI in order to suck up subsidies. One analysis by Deloitte, a consultancy, estimated that 99% of self-styled AI startups in 2018 were fake. Such boondoggles not only burn through public cash, Mr Ding notes, but also consume scarce human capital that could more usefully have been deployed elsewhere.China’s second problem is its inability to recruit the world’s best AI minds, especially those working on high-level research. A study in 2020 by MacroPolo, a Chicago-based think-tank, showed that more than half of top-tier researchers in the field were working outside their home countries. America and Europe look more appealing to such footloose brainboxes, including many Chinese ones. Though about a third of the world’s top AI talent is from China, only a tenth actually works there. A shortage of non-Chinese researchers further handicaps China’s capabilities, notes Matt Sheehan of the Carnegie Endowment for International Peace, a think-tank in Washington.Even more problematic for the party, its master plan ignored the cutting-edge semiconductors that power AI. Since its publication Chinese companies have found it ever more difficult to get their hands on advanced computer chips. That is because virtually all such microprocessors are either American or made with American equipment. As such, they are subject to restrictions on exports to China put in place by Donald Trump and extended by his successor as president, Joe Biden. It will take years for Chinese companies to catch up with the global cutting-edge, if they can do it at all.These challenges will continue to bedevil all of China’s high-tech industries for years to come. It could leave its AI businesses stuck in a rut—successfully rolling out relatively unsophisticated products while trailing Europe and America in paradigm-shifting developments of greater financial and strategic value. Consider Wu Dao 2.0. Although it was a huge improvement on GPT-3, it did just that—improve an existing technology rather than break new ground. No amount of Chinese taxpayers’ money is likely to change that. ■This article appeared in the Business section of the print edition under the headline “In search of mastery” More

  • in

    Making sense of the East-West divide in tech

    THANKS TO A venture-capital (VC) boom, it is no longer unusual to find tech unicorns, as unlisted startups valued above $1bn are known, springing up in middle-income countries. However, two coming from Turkey are particularly strange creatures. First, they are big. Trendyol, an e-commerce company, is valued at $16.5bn, giving it the status of a “decacorn” worth $10bn or more. Getir, a pioneer of “superfast” grocery delivery, is reportedly close to joining that select group. Second, they are battle-hardened. Both come from a country wracked by inflation, currency instability and barmy economic policies, any of which can be kryptonite for investors. Most striking, their founders bear no resemblance to archetypal tech bros. Trendyol’s Demet Mutlu is a 39-year-old woman. Getir’s Nazim Salur is a 60-year-old man.And yet look closely at their two companies, now worth more than almost any listed firm in Turkey, and the differences outweigh the similarities. Fittingly for a country that sees itself as a gateway between the Orient and the West, their view from the Bosporus is Janus-like. One takes its inspiration from China, the other looks to Europe and America. One shuns the spotlight. The other craves it. One wants to turn women into go-getters. The other has the male-sounding mantra of “democratising the right to laziness”. They encapsulate several different dimensions of the tech divide. That makes them intriguing to compare and contrast.Start with the division between East and West. In simple terms, this represents a choice between Asian-style super-apps and Silicon Valley-style blitzscaling. Trendyol’s biggest backer is Alibaba, and the Chinese e-emporium’s influence runs deep. The Turkish firm shares Alibaba’s marketplace model: it accounts for more than a third of e-commerce in Turkey and provides a platform for trading about $10bn a year of merchandise. Unlike Amazon, the American giant, it sells only a few of its own goods. Like Alibaba, it calls itself a super-app, aiming to offer a variety of services, including payments, on its platform, and it puts the importance of its small-business sellers, who are everywhere in Turkey, on a par with buyers. International expansion, when it comes, will probably be to emerging markets, such as those in eastern Europe and the Middle East. It believes, as Alibaba does, that the super-app potential is greatest in such young, mobile-mad places.By contrast, Getir’s first international backer was Michael Moritz of Sequoia Capital, an American VC firm. Aptly, its strategy borrows from the Silicon Valley playbook: blitzscale first, make money later. Founded in 2015, Getir claims to have invented the business of delivering groceries in under ten minutes (unsurprisingly in Istanbul, where few people live more than ten minutes from a shop, many of Mr Salur’s friends wondered at first why they would need it). Discounts help get customers hooked, Mr Salur says. Then, he hopes, the temptation to treat Getir like a personal butler will take over. With competition from America’s Gopuff and Germany’s Gorillas growing, speed is of the essence. Since launching its first international operation in Britain a year ago, the firm has moved through the developed world almost as fast as its purple-and-yellow-clad moped riders dash through the streets of London. It is now in 40 cities in Europe and America, from Barcelona, via Bristol, to Boston.Mr Salur has long set his sights on penetrating America—and eventually listing the firm there. “If you’re a startup guy, you want to succeed where the startups are,” he says. In true American style, he revels in media attention. Getir welcomed your columnist to a brightly lit depot (“dark store” is a misnomer) under railway arches in South London to see baskets of biscuits and avocados whizzing out the door. Only when discussing the financials of a cash-guzzling business is Mr Salur guarded. He declines to comment on its latest valuation, which Bloomberg reports to be as high as $12bn. “When money is in the bank, you will hear about it.”Ms Mutlu could not be more different. She has put a China-like media firewall around Trendyol and mostly shuns interview requests. One of the few nuggets commonly repeated about her is that she dropped out of Harvard Business School to set up Trendyol in Turkey. And yet she is more remarkable than that. Besides founding Trendyol, she co-founded another Turkish unicorn, a gaming company sold to San Francisco-based Zynga for $1.8bn in 2020. To put that into perspective: PitchBook, a data gatherer, calculates that of 1,335 unicorns globally, only 185, or just under 14%, have at least one female founder.Furthermore, Ms Mutlu is described by an investor as “maniacal” about tech. Having started out selling fashion items on Trendyol, she is a champion of Turkey’s textile industry. She is also an advocate (albeit a media-shy one) for women in the digital economy. Women make up about half of Trendyol’s employees, including some software engineers, and many of her buyers and sellers. Those who know her say she struggled to be taken seriously as she built her business. Adding to the frustration, she did not know whether it was because she was a woman, or Turkish, or both.Ottoman empire-buildersThese are heady times for startups everywhere. Both companies are aware that they have thrived at a time when VC funding across the globe is frenzied—and sometimes indiscriminate. Neither is likely to do an initial public offering soon, at least until the valuation shortfall of public versus private markets narrows.Yet they have also benefited from growing up in Turkey’s school of hard knocks. Living amid galloping price increases prepares them for a world that is reawakening to the menace of inflation. In a country where VC funding was negligible until 2021, they learned to operate leanly. And they stand proudly behind names that are hard-to-pronounce in English. As Mr Salur quips: “Remember Arnold Schwarzenegger? He didn’t change his name.” It may be time to get used to them. ■For more expert analysis of the biggest stories in economics, business and markets, sign up to Money Talks, our weekly newsletter.Read more from Schumpeter, our columnist on global business:TikTok isn’t silly. It’s serious (Jan 15th 2022)Glencore’s message to the planet (Jan 1st 2022)The billionaire battle for the metaverse (Dec 18th 2021)This article appeared in the Business section of the print edition under the headline “East v West, Venus v Mars” More

  • in

    Unilever’s £50bn health cheque

    WHEN UNILEVER bought Bestfoods for $20.3bn at the turn of the millennium, it was one of the largest cash acquisitions ever. After two failed bids, the British consumer-goods giant dug up an extra $2bn to sweeten the deal. It divested 700 of its brands in the year that followed but replenished its larder with Bestfoods’ Knorr soup and Hellman’s mayonnaise. Now, in pursuit of another mega merger that could be four times as big, Unilever has been prepared to dispose of the larder entirely.Unilever’s new target has been the consumer-health unit of GlaxoSmithKline (GSK), a British drugmaker. On January 15th it emerged that the soup-to-soap group was offering to pay £50bn ($68bn) for the business. GSK, which has been keen to ditch the division in order to focus on more lucrative prescription medicines, refused to bite. The markets choked: Unilever’s share price fell by 7% the next trading day. Analysts are almost uniform in their view that the deal is a bad idea, arguing that it presents more risk than Unilever, with a market capitalisation of £94bn, can stomach. Selling lagging categories like food may not be enough to fund the transaction, of which nearly £42bn would be in cash. Fitch, a ratings agency, warned that Unilever could lose its A credit rating if it took on too much debt.Alan Jope, who took over as chief executive three years ago, sees the future of consumer goods in health and hygiene products rather than food. Hand sanitiser and paracetamol have certainly sold well during the pandemic. Moreover, Unilever has a big presence in developing countries, which could create new markets for GSK’s brands such as Sensodyne toothpaste and Advil painkillers. Still, on January 19th the company, possibly having read all the warning labels about the deal, said it would not raise its offer above £50bn, which GSK’s bosses said undervalued their division. This may end the pursuit.It won’t end Mr Jope’s troubles. He is under immense pressure to improve the group’s performance. The affable Scotsman has so far been unable to reignite growth in his three years in charge. Unilever’s share price has declined in the pandemic even as those of rivals such as Nestlé, a Swiss giant, or Procter & Gamble (P&G), an American one, have gone up by more than 20% (see chart). A career-defining deal might have set him apart from his predecessor, Paul Polman, who was known for eschewing financial engineering. If the £50bn transaction came to pass, it would be one of Britain’s biggest-ever.There is also a growing sentiment that Unilever’s zeal for purpose-driven brands, first instilled by Mr Polman, has run out of steam. From ethically sourced tea and fighting deforestation with sustainably-sourced palm oil to marketing Dove soap as a women’s-self-esteem project, the firm has sought to connect with shoppers on their values and draw investors interested in environmental, social and governance (ESG) factors as well as profits. Although ESG remains popular, hints of a backlash against it are appearing. This month Terry Smith, an asset manager who is among Uni lever’s top ten shareholders, groused that the firm has “lost the plot” by pursuing sustainability medals at the expense of financial performance. A hard-headed pivot to a more profitable health business could, if successful, allay such worries.The deal would have been problematic, and not just because it looked like a heavy lift for Unilever. Megamergers seldom work out as advertised, and Mr Jope’s firm is not renowned for stellar execution. Moreover, the consumer-health market is expanding but incumbents’ share of it is not. Established brands have a place—people need to brush their teeth—but growth in the sector increasingly comes from a new pharmacopoeia of clever products and services, many of them with digital features. Even in good years GSK’s consumer-health division has grown at best in single digits. The long-term growth prospects for its brands look pale. Antacids and nicotine patches have only limited potential, even in emerging markets.Unilever’s rivals have been more discerning with their acquisitions. In 2020 Nestlé acquired Aimmune, a novel peanut-allergy medication, and a year later it bought Nuun, a challenger in the sports-beverage market. Both deals gave the Swiss firm a foothold in profitable, underdeveloped niches. P&G is dabbling in premium skincare, one of the industry’s fastest-growing categories, with its latest acquisitions Tula Skincare and Farmacy Beauty. If Unilever does end up disposing of its food business, it may also miss out on the boom in alternative proteins, notes Bruno Monteyne of Bernstein, a broker. Meat substitutes appear certain to become more popular with time and companies like Unilever stand to benefit, given their mix of solid research-and-development base and brands beloved by consumers.Unilever says it has another, undisclosed initiative up its sleeve to improve performance. It had better. The pandemic boost notwithstanding, the entire consumer-goods industry has experienced slower growth over the past decade. With the exception of Nestlé, European companies have done poorly. Unilever needs some refreshing, but more toothpaste won’t do the trick. ■For more expert analysis of the biggest stories in economics, business and markets, sign up to Money Talks, our weekly newsletter.This article appeared in the Business section of the print edition under the headline “Health cheque” More