Exponential: Order and Chaos in an Age of Accelerating Technology

by Azeem Azhar

A new technology might at first cause a small social change – but one that eventually spirals into major repercussions for the whole of society. When these ripples – or ‘feedback loops’, in the jargon of complexity science – start to spread, they can feel uncomfortable. One need only glance at the pages of a newspaper from the turn of the twentieth century to realise that sudden change is anxiety-inducing. A quick survey of New York Times articles from a century ago reveals that Americans were apprehensive about elevators, the telephone, the television and more.

Complexity scientists refer to moments of radical change within a system as a ‘phase transition’.5 When liquid water turns into steam, it is the same chemical, yet its behaviour is radically different. Societies too can undergo phase changes. Some moments feel abrupt, discontinuous, world-changing. Think of the arrival of Columbus in the Americas, or the fall of the Berlin Wall.

We often assume that tech is somehow independent of humanity – that it is a force that brought itself into being and doesn’t reflect the biases and power structures of the humans who created it. In this rendering, technology is value-free – it is made neutral – and it is the consumers of the technology who determine whether it is used for good or for evil. This view is particularly common in Silicon Valley. In 2013, Google’s executive chairman Eric Schmidt wrote, ‘The central truth of the technology industry – that technology is neutral but people are not – will periodically be lost amid all the noise.’6 Peter Diamandis, an engineer and physician – as well as the founder of a company that offers tech courses, Singularity University – wrote that while the computer ‘is clearly the greatest tool for self-empowerment we’ve yet seen, it’s still only a tool, and, like all tools, is fundamentally neutral’.7 This is a convenient notion for those who create technology. If technology is neutral, its inventors can concentrate on building their gizmos. If the tech starts to have any insidious effects, society – rather than its inventors – is to blame. But if technology isn’t neutral – that is, if it has encoded some form of ideology, or system of power – that might mean its makers need to be more careful. Society might want to manage or regulate the technologists and their creations more carefully. And those regulations might become a hassle. Sadly for these engineers, their view of t...

First, new technologies are being invented and scaled at an ever-faster pace, all while decreasing rapidly in price. If we were to plot the rise of these technologies on a graph, they would follow a curved, exponential line. Second, our institutions – from our political norms, to our systems of economic organisation, to the ways we forge relationships – are changing more slowly. If we plotted the adaptation of these institutions on a graph, they would follow a straight, incremental line. The result is what I call the ‘exponential gap’. The chasm between new forms of technology – along with the fresh approaches to business, work, politics and civil society they bring about – and the corporations, employees, politics and wider social norms that get left behind.

The second is the transformation of warfare. As the world gets re-localised, patterns of global conflict will shift. Nations and other actors will be able to make use of new adversarial tactics, from cyber threats to drones and disinformation. These will dramatically reduce the cost of initiating conflict, making it much more common. A gap will emerge between new, high-tech forms of attack and societies’ ability to defend themselves.

Put simply, exponential growth refers to an increase that compounds consistently over time. Whereas a linear process is what happens to your age, increasing by a predictable one with each revolution of the earth around the sun, an exponential process is like a savings account with interest. It increases by a fixed percentage, say 2 per cent every year. But next year’s 2 per cent applies not just to your original savings but to your savings plus last year’s interest. Such compounding starts slow – it’s even slightly boring. But at some point, the curve turns upwards and starts to take off. From this point onwards, the value leaps higher at a dizzying pace. Any number of natural processes follow an exponential pattern: the number of bacteria in a petri dish, or the spread of a virus through a population, for example. A more recent development, however, is the emergence of exponential technologies. I define an exponential technology as one that can, for a roughly fixed cost, improve at a rate of more than 10 per cent per year for several decades. A mathematical purist, of course, would argue that even a 1 per cent compounding change is an exponential change. Strictly speaking, it is. But a 1 per cent annual change takes a lot of time to get going. A figure compounding at 1 per cent per annum would take 70 years, the better part of a lifetime, to double.

Picture yourself changing your car after ten years of ownership. And imagine if the main features of the car – its top speed or fuel efficiency, say – had improved at 10 per cent per annum. Your new wheels might have double the fuel efficiency or more than double the top speed. Generally, that doesn’t happen. But for many of the technologies discussed in this book, it is exactly what happens. In fact, a number of these technologies effectively improve at rates of 20 to 50 per cent (or more) per year.

For each, he has established how long it took to reach 75 per cent penetration of the American market, meaning three-quarters of adults (or households, if appropriate) have access to it. While each product is different, there are common themes to the ways they take off. The spread of most technologies follows a ‘logistic curve’, or S-curve. At first, the uptake of a technology is slow. Early adopters are experimenting with it; while producers are figuring out exactly what to make and how to price it, and are building up their capacity. At some point the product hits an inflection point, and its rate of spread increases very rapidly. So the first two parts of the curve look like a classic exponential curve: slow and boring at first, then rapid and exciting. Unlike a pure exponential curve, however, the S-curve has a limit. After all, there are only so many cars or washing machines a family can own. As the market gets saturated, the uptake tapers off: there are fewer families who don’t own a digital camera or microwave, or fewer steelmakers who haven’t moved to electric arc furnaces. The steep part of the graph starts to flatten. In other words, the pattern of uptake looks like a lazy ‘S’.

In Kurzweil’s view, however, at any one time there are multiple technologies following an S-curve. As one S-curve reaches its steepest gradient, another curve begins. Once our first curve starts to flatten out, the younger technology is approaching the explosive phase of its acceleration – and takes up the mantle of rapid growth. And, most importantly of all, these different technologies nourish one another – innovations in one sector inspire developments in the next. When one technology reaches the limits of its potential, a new technological paradigm is waiting in the wings to pick up the slack. The result is that, across a society, there is a quickening in the pace of technological progression – even if the development of individual technologies is consistently slowing off.

Exponentiality has become widespread in four key domains of technology, which between them form the bedrock of the global economy. Computing, of course, is one. But so too are the domains of energy, biology and manufacturing. Each is undergoing a spectacular, thrilling transformation. The costs of the key technologies in each area are falling dramatically, by the equivalent of a factor of six or more every decade.

Additive manufacturing, or 3D printing – I’ll use the terms interchangeably – is an exponential technology that delivers the individual detailing of subtractive manufacturing, without the waste. Typically, objects are crafted through computer-aided design. The process involves creating a new object from scratch: by putting together layers upon layers of melted material, using a laser or a device a little like an inkjet printer. The material can range from glass to plastic to chocolate. It marks a fundamental break with many millennia of subtractive manufacturing, and thousands of years of casting and moulding.

Part of the reason GPTs are so transformative is the way they have effects beyond any one sector. Consider one of the key GPTs from the start of the twentieth century: the car. To reach their potential, cars needed suitable roads – physical infrastructure that spanned nations. But cars also needed fuel and spare parts, and drivers needed sustenance – creating the demand for fuel stations and roadside cafés. Cars forced changes to the urban environment and so cities started to change, with precedence going to motorised vehicles. Over time, suburbs developed, and with them came the gradual reshaping of consumer practices: reasonably priced hotels for vacationers, and big-box retail stores. New rules slowly emerged, including a slew of safety regulations for drivers. In short, the GPT changed everything. And this hints at why the exponential revolution in our four key sectors is so important. We are witnessing the emergence of a transformative new wave of GPTs. Not a single GPT, as in the time of the printing press. Nor even three GPTs, like the early twentieth century’s offering of the telephone, the car and electricity. In the Exponential Age, we’re experiencing multiple breakthrough technologies in the four broad domains of computing, energy, biology and manufacturing.

This means Wright’s Law has an edge over Moore’s Law. Both describe the way the cost of technology diminishes exponentially. But Moore’s Law simply describes performance improvements over time. There are scenarios it can’t account for – those striking microchip-factory workers, for example. Wright’s Law, meanwhile, connects progress to the quantities produced. Say that for every doubling of production, the unit cost of a gadget drops 20 per cent. If production doubles every two years, costs will drop by 20 per cent every two years. If production doubles every year, costs will drop by 20 per cent every year. Wright’s Law holds true even during that mythical strike: if production stops, the reduction in cost stops.

Silicon chips get faster as you make their components smaller. And since chips live on a square wafer, each time you shrink a component you get the efficiency gain squared. If you have a wafer that is 100 square millimetres and you can fit one component onto every millimetre, then a single wafer can hold 10,000 components (100 × 100). If you shrink the components by 50 per cent, so that you can fit two onto every millimetre, you end up with 40,000 components (200 × 200) on the same die. This process holds true for many exponential technologies. Even mighty wind turbines are not immune to such effects. The generating capacity of a wind turbine is proportional to the area that the blades sweep through. That area grows by the square of the length of the blades, so if you are able to manufacture a blade twice as long, you get quadruple the impact.

But the greater cause of the newfound power of Wright’s Law lies in economics. Previously, the S-curve of demand tapered off when a market reached saturation. Today, that point of market saturation is much more distant – because global markets are so much larger. And this means that the process explained by Wright’s Law can continue for much longer, and the exponential gains can continue to mount up. As we’ll see time and again in this book, the rise of a global market for products is one of the great changes of the last 50 years. The volume of world trade increased 60-fold, from $318 billion to $19,468 billion by 2020. And this pushed us ever further down the path of Wright’s Law. Bigger markets mean more demand; more demand means more efficient production; more efficient production means cheaper goods; cheaper goods mean bigger markets. The cycle has an inherently exponential logic.

But today, it would be trivial – and cheap, measured in tens of dollars – to give Tom what he wanted: accurate, up-to-date photographs of almost any part of the planet. Such images are used by hedge funds and commodities traders in precisely the way my boss envisaged. These companies count cars in shopping-mall parking lots to estimate customer demand, and ultimately assess how well the retail sector is going. Or they analyse the shadows cast by oil tankers to estimate their load – and hence the global demand for oil. Lemonade, an American insurance company, uses satellite imaging to estimate the risk of forest fires before it offers home insurance policies.

Here, too, standardisation helped: by 1965, the International Standards Organisation had agreed a specification for shipping containers (8 feet wide, 8½ feet high, and 10, 20 or 40 feet long). Containers could be handled by ports the world over and rolled straight onto flat-bed trucks. And in March 1966, the first ships holding such containers departed the US.43 That first ship held 226 containers. A year later, a container ship holding 609 containers started to ply the seaway between Oakland on the west coast of the US and Cam Ranh Bay, an American military base in Vietnam. And these ships only got bigger. By 2020, the world’s largest container ship, the HMM Algeciras, fully laden, could hold 23,964 of the boxes.

Exponential technologies are being driven by three mutually reinforcing factors – the power of learning by doing, the increasing interaction and combination of new technologies, and the emergence of new networks of information and trade. But for this picture to be complete, we also need to understand the economic and political context. As you may have noticed, none of these driving forces can be explained without reference to a wider set of shifts in politics and economics. In particular, the process of globalisation. Our continual ability to learn by doing is based on the ever-growing international demand for goods. The standardisation of parts, software and technology protocols is partly down to the emergence of global standards bodies. The rise of trade networks has as much to do with the development of new markets as with the advent of new technology. Technology, politics and economics are intertwined.

But in the 1970s, everything changed. It’s a well-known story. Rich economies were assailed by a toxic combination of low growth and high inflation – known as ‘stagflation’. Waves of strikes and fuel crises eroded trust in Western governments. And, as faith in government reached a new low, voters and policymakers cast around for a new way of doing things. The scene was set for the emergence of a new school of economics, whose standard-bearer was University of Chicago professor Milton Friedman. Friedman’s belief was that markets would work better if the government got out of the way. Since the end of World War Two, Western governments had adopted a fairly interventionist approach to their economies. Until 1979, the top rate of income tax, payable on investment earnings, was 98 per cent in the UK and 70 per cent in the US. In the UK, the basic rate of tax was 33 per cent by the latter half of the 1970s. And with high taxes came hands-on government: nationalised industries, heavy regulation and interventionist industrial policy. But Friedman’s acolytes advocated a different approach.

At the heart of Amazon’s success is an annual research and development budget that reached a staggering $36 billion in 2019, and which is used to develop everything from robots to smart home assistants. This sum leaves other companies – and many governments – behind. It is not far off the UK government’s annual budget for research and development.1 The entire US government’s federal R&D budget for 2018 was only $134 billion.2 Amazon spent more on R&D in 2018 than the US National Institutes of Health. Roche, the global pharmaceutical company renowned for its investment in research, spent a mere $12 billion in R&D in 2018.3 Meanwhile Tesco, the largest retailer in Britain – with annual sales in excess of £50 billion (approximately $70 billion) – had a research lab whose budget was in the ‘six figures’ in 2016.4 Perhaps more remarkable is the rate at which Amazon grew this budget. Ten years earlier, Amazon’s research budget was $1.2 billion. Over the course of the next decade, the firm increased its annual R&D budget by about 44 per cent every year.

The most basic cause of the exponential gap is simple: we are bad at maths.

Why do humans consistently underestimate the power of exponential change? The shortest answer is perhaps the best. Most processes we go through – ageing from one to two to three years old; moving incrementally along a metro line – follow a linear scale. And there are good evolutionary benefits behind this linear mentality. Our minds developed for a world that, in general, hadn’t yet discovered the power of rapid change. The rhythms of the hunter-gatherer routine were slow: we adapted for a life that was largely seasonal, based on repeated patterns throughout the year. And this slow pace defined our existence until remarkably recently.

These norms are also a type of institution. They bring stability that frames our collective behaviour. All these institutions have something in common. They are largely not cut out to develop at an exponential pace or in the face of rapid societal change. In the most extreme cases, they’re not cut out to adapt at all.

For the first three decades of my life, there was a set of clear rules about how business worked. Dozens of very large companies dominated the economy, and each had a specialty. Exxon and BP invested in oil rigs and drew oil from the ground, selling it at a market price that was above the price of extraction. General Motors and Ford assembled raw materials and components, stitched them together and sold them for a profit. Successful companies grew bigger, by dint of better products or keener pricing. They might benefit from economies of scale – as they got bigger, their costs, relatively speaking, went down. But these companies also encountered forces that held them back. As they became bigger, they might become too complex to manage. The flows of information between bosses, managers and employees might get entangled. Headquarters might lose track of what the front line worker was doing. If you have ever worked at a very large company, you’ll be familiar with how bureaucracy creeps in and slows everything down.

Some of these companies were dominant until relatively late: WordPerfect was the market leader in word processing in 1995, when it had 50 per cent market share. But once Microsoft had established dominance in the operating system market, it became increasingly hard to take on. There was a powerful network effect: once everybody else was using Windows, and exchanging Word documents and Excel spreadsheets, it became much easier for you to use them too. The network effect driving Microsoft’s success allowed it to spread inexorably from one market area to the next – first operating systems, then word processors, then spreadsheets.

The economy of the machine age was dominated by things you could stub your toe on. Cars, washing machines, telephones, railway lines – for centuries, the most important products were built in the physical world. And the companies that supplied these products were the corporate titans of their time. You could work out a company’s value by focusing on the land and buildings it owned, the machines in its factories, and the stocks of products waiting to be sold. This physical stuff is known as a company’s ‘tangible assets’. In the Exponential Age, value is rather more nebulous. Many of today’s firms do not get their value from the physical products they move around. That is not to say that during the Exponential Age companies won’t build physical things. Hugely expensive, customised factories to build semiconductors are clustering in high-tech areas; shallow seas with stable, strong wind patterns are being populated with grids of titanic windmills. But the true source of these companies’ profits is more ethereal. The software that runs Google’s search engine, the data that represents the network of friends in Facebook, the designs and brand identity of Apple, the algorithms that recommend your evening’s viewing on Netflix – this is where their true value lies. Economists call these non-physical assets ‘intangible’.22 The speed of the shift to intangible assets has been remarkable. In 1975, about 83 per cent of the market value of companies on the Standard and Poor 500 stock ma...

Netflix has followed a similar pattern. In 2010, Netflix had sales of $2.1 billion and 2,100 employees. Each one of them could account for slightly less than $1 million in revenue. A decade later, the firm had grown more than 10 times in size, with revenues approaching $25 billion. Its 9,400 employees counted, now, for nearly $2.7 million in revenue each. The dynamics of increasing returns is reflected in Netflix’s content production, as well. In 2011, 14 years into its existence, Netflix had zero original content titles under its wing. In 2012, it launched one. By 2019, it released more original shows and films than the whole TV industry’s yearly output prior to 2016.32 With 37 Oscar nominations under its belt in 2021, the company is close to breaking a 1940 record of 40 nominations for a single studio. In these moments, it becomes apparent how increasing returns to scale aren’t just rewiring business, but also culture and wider society.

But are Apple’s efforts worth 30 per cent of a developer’s efforts? Many developers don’t think so. Epic Games, which makes Fortnite – a video game beloved of teenagers and loathed by their parents – was one.39 After challenging Apple’s fees and attempting to bypass the 30 per cent cut, they were kicked off the platform. (Apple did eventually back down, sort of – a few weeks after the fracas, they halved fees for some developers, albeit only for those with much lower revenues than Epic.) The problem is that we can’t really know what a fair price is in the absence of a competitive market. Given that Apple has an 80 per cent market share among the young people Epic targets, you can’t just turn to the market to work out the ‘true’ value. In some segments of the economy, Apple is the market.

This has long-term consequences for innovation. Research shows that breakthrough inventions are more likely to come from individual inventors or smaller teams – one group of Ph.D. researchers analysed 65 million papers, patents and software products from 1954 to 2014, and found that ‘while large teams do indeed advance and develop science, small teams are critical for disrupting it’.43 And while, as we’ve seen, a few companies take a very ambitious approach to research – Google springs to mind – the data suggests that those corporate giants can also come to narrow the focus of research in fields where they dominate. Researchers who analysed 110,000 papers in the broad field of AI concluded that much of the research conducted by companies tended to be very narrow compared to academic research. Commercial research, which seemed to be laser-focused on approaches which had already been shown to be successful, tended to reduce the number of pathways being explored.

And there’s a final problem with the tendency towards monopoly. Bork doesn’t mention it much, but companies are not just important to consumers – they are also part of the wider functioning of society. Most obviously, they are supposed to pay tax. And for the first giants of the Exponential Age, our tax codes have been generous. The bulk of the assets for Exponential Age firms reside in intangibles, and intangibles are prone to sliding across borders – and easier to slip past the taxman. According to The Economist, in the years to 2020, the largest American tech firms paid a tax rate of around 16 per cent of their profits, far below the then 35 per cent corporate US tax rate.

But the notion of a jobless future – the ‘robopocalypse’ of tabloid headlines – is overstated. It is alluring and it makes the news, yet it is muddle-headed. Historically, our economies have become more automated. And historically, employment levels have tended to increase. How is this possible? Because automation has the potential to create more work than it destroys. Yes, there may be unemployment in the short term as some sectors wane. But automation creates jobs which often require new and distinctly human skills: from the programmers who develop systems to those who operate and maintain them. Over time, automation will end up inventing whole new sectors of the economy – ones that we can now only imagine.

This dynamic means that, in practice, much of the labour economists consider ‘unskilled’ might prove difficult to automate. A workplace manual is usually barely even a rough guide to what a job is – it doesn’t cover half of the things you need to be successful. And when such tacit knowledge exists in a workplace, it makes artificial intelligence that can do the job very hard to build. An AI system needs a clear and unambiguous goal, and modern systems need to be trained on that data. If the know-how about a job is largely hidden, an AI system will be trained on only half the picture. In short: if an environment is designed for a person, it is likely to be too complicated for machines now or any time in the near future.

These people and millions of other gig workers are managed by computers. Their work is scrutinised through a stream of quantitative performance assessments. Rideshare drivers may only have 10–20 seconds to respond to an offered ride, without knowing in advance where they’re expected to go or how much they can expect to make. If they refuse too many rides in a row, they can be kicked off the platform.71 And this is not limited to ride-sharing. The warehouse workers who put together your grocery delivery are commanded by scanners, which tell them how much time they have to collect each item.72 None of Taylor’s lackadaisical soldiering here. Deliveroo drivers, meanwhile, can be punished if they arrive later than the time of arrival calculated by an algorithm – regardless of the local weather conditions or traffic. This system of management works like Taylorism on steroids – with all the dehumanising downsides but without the commensurate high pay or job stability.

The cause of the decline of labour’s share of the economic pie is multifaceted. But it is closely related to the shift to the exponential economy. Four key causes stand out. Globalisation, which drove down wages in the West as companies offshored jobs to cheaper locations across the world. The decline of unions, which meant workers lacked the bargaining power to stop economic rewards going to the owners of capital (we’ll return to this in a moment). The rise of the intangible economy, which reduced the relative value-add of the average worker – more value was being created by know-how, software and data, stewarded by smaller numbers of highly specialist workers, than by the human sweat of the larger parts of the workforce. And superstarification – as markets consolidated around ever-fewer superstar firms, there was less competition for labour and so workers had less leverage.77 These are all hallmarks of the rise of exponential technologies – the emergence of a global, high-tech, intangible economy, dominated by a handful of big firms. According to one study, more than two-thirds of the loss to workers results from the transition to intangibles and the growth of superstar firms.

Perhaps the word ‘union’ just has too much baggage in the Exponential Age. It conjures up images of striking miners and steelworkers more than of striking Deliveroo drivers. Yet collective action remains the best way to guarantee a workplace that is fair for employees. If we are to build an employment settlement that is dignified, flexible, secure and equitable, workers will need to get organised. Only unions can collectively bargain on behalf of workers. And unions are also better than individual workers at developing the expertise required to digest the complex technical issues emerging in a rapidly changing economy. But in the world of trade unions, too, digital technology is both the problem and the solution. The creation of new networks of information gives workers an opportunity to identify their shared experiences, discuss how to respond, and organise. Finding like-minded people to collaborate with has always been a barrier to unionisation – but this problem is less pronounced in the Exponential Age when anyone with a smartphone becomes a potential comrade. Today, the process of unionisation often begins with informal collectivisation on WhatsApp groups and online forums. While workers in each industry have their own demands, the use of technology for collectivisation is uniform.90 In West Virginia, where union membership rates hover at around 10.5 per cent,91 teachers looking to organise set up an invite-only Facebook group. Close to 70 per cent of the state’s 35,0...

Pix Moving was using only the most modern manufacturing methods – ‘dematerialised’ techniques. Rather than exporting cars, Yu explained, they ‘export the technique that is needed to produce the cars’.1 Vehicles are not loaded onto container ships and sent to their destination. Rather, the company sends design blueprints over to colleagues in the US, who use additive manufacturing techniques to print components locally. From those components, the finished product can be assembled. Yu’s approach could skirt around customs inspectors (and tariffs). Additive manufacturing lets him build wherever his customers are, trade conflicts be damned. Pix Moving is a start-up that reflects how manufacturing will change as we move further into the Exponential Age and, in doing so, unpick our assumptions about globalisation.

And so here we can make out a new system of global trade. Gone will be the world of poor countries manufacturing goods for rich countries, and shipping these products across the world. Instead, each rich country will begin to make its own goods at home, for a domestic market. But as ever in the Exponential Age, the rewards of new technology are not evenly distributed. The diminishing dependence of rich countries on poor countries’ commodities may fundamentally destabilise the economies of much of the developing world.

Even a global pandemic seemed unable to stop the rise of urban life. The early stages of the Covid-19 pandemic saw a flight from the cities – The Economist reported that 17 per cent of Parisians left the French capital as the country went into lockdown in March 2020.35 Yet cities bounced back with remarkable rapidity. Researchers at the Wharton School of Business found that most of the major shifts away from cities early in the pandemic were temporary. While some particularly expensive areas such as Manhattan might see a more long-lasting exodus, cities generally held on to their residents.

The story of the de-localisation and re-localisation of the internet is perhaps the neatest summary yet of the trajectory of the Exponential Age. Digital technologies have the potential to transcend national borders; but, as we have seen, they are just as likely to strengthen them. And that is especially true at a time when many global governments are turning back to nationalist politics. All this poses a problem for policymakers. The institutions that guide global policy and security were built during an era in which it was assumed that locality was not only dead, but perhaps didn’t matter. From geopolitics to global economics, institutions like the IMF and the WTO have long been the missionaries of globalisation, establishing a template for what participation in the world economy would like. These assumptions seem out of place in a world of national self-sufficiency in energy and commodities, localised manufacturing, and cities of ever-growing importance.

Cities are often at the forefront of tackling problems caused by exponential technology. It was places like London that first had to contend with the sudden growth of gig-working platforms such as Uber. Barcelona was among the earliest to reckon with the explosion of digital accommodation marketplaces like Airbnb, and all the economic effects that brought. And, as we have seen, cities will be the engine of the Exponential Age economy. The solution may be to develop more federal models of national politics, which give more power to regions and cities to manage their own affairs. They need the ability to attract people and regulate the quality of their citizens’ lives, by increasingly governing their own energy, resources and climate agendas. This might seem likely to exacerbate the rural-urban divide, but in fact it may help close it. Many of the divisions in our society result from cities and nations – which have vastly different economic interests and political outlooks – being tied together under a single, over-centralised government.

More nebulous forms of attack become more feasible too. In the twentieth century, for a state to orchestrate a wave of misinformation they would have had to infiltrate newspaper offices, radio and TV studios and broadcast transmitters. Today, it’s merely a matter of making malicious posts go viral on social media.

But that is changing fast. The North Korean cyber army included anywhere between 5,000 and 7,000 troops as of 2016 – all under the direct command of leader Kim Jong-un.18 The importance of this organisation lies not only in its attempts to disrupt South Korean infrastructure, but in its involvement in illegal state activities. The North Korean cyber force has drawn blood with online scams, fraud and attacks on online banks and cryptocurrency exchanges.19 International organised crime and cyber warfare are thought to have fuelled the growth in North Korea’s GDP since 2015.20 A UN report estimated that North Korean cyber experts have illegally ‘raised’ up to $2 billion for the country’s weapons programs. This is a much cheaper form of extraterritorial activity than the traditional modes of warfare – which involve sending troops to the other side of the world. All the North Koreans need to wreak havoc is a computer and a talented hacker. This dynamic means many relatively poor nations can have a disproportionately impactful foreign policy.

Along the way, these cheaper drones brought wholly new actors into military conflicts – battlefield warfare became affordable to any number of terrorist groups and small states. Military drones represent the inverse of traditional flagship military technologies – the ballistic missiles, nuclear weapons, aircraft carriers and stealth fighters that remain the privilege of a handful of rich nations with mature defence industries. Increasingly, anyone can build a drone army. In a 2018 attack, the Kurdistan Workers’ Party (PKK) targeted several Turkish government locations with drones that anyone could buy on Amazon for under $300, having packed them with explosives.50 Then, in 2019, Houthi rebels from Yemen launched drone attacks against the world’s largest oil processing facility in Saudi Arabia, stopping about half of the kingdom’s output for a day.51 In the language of military theory, the development of drones has driven an increase in ‘asymmetric’ conflict – between powerful, well-established actors like nation states, and upstart combatants with fewer resources.

This is a phenomenon with three key features. The first, best exemplified by Facebook’s arbitrary controls of conversation as discussed above, is the emergence of new, private rule-makers – whose growing power amounts to the privatisation of the ‘public sphere’. This term, first coined by the German philosopher Jürgen Habermas, refers to the arena in which private individuals come together to discuss the needs of society and the laws that govern it. As the name suggests, this is supposed to take place in public. We have long thought that laws should be made by accountable, elected officials; and they should be scrutinised transparently and collectively. Today, however, a handful of private organisations increasingly dominate that public sphere. They, rather than elected parliaments, make the laws that govern our lives. And our public conversation is conducted on a small number of private platforms, rather than in a wide array of newspapers and broadcasters, think tanks and coffee shops. The second is the encroachment of markets into our private lives. The private sector and our private lives were once very different. We bought and sold our houses, our food, our labour. But there were parts of our lives that we wouldn’t sell. The conversations between lovers or within families. Essential, personal information about our health, or our relationships, or our sex lives. Increasingly, private companies monitor – and lay some kind of claim to – all of these. As our conversations ...

What is now Facebook started as FaceMash, the result of a teenage college student’s peccadillo. Mark Zuckerberg and some of his friends built a website to judge the attractiveness of female students. It soon morphed into TheFacebook, an embryonic social network that played up to the dating-and-mating culture of American colleges. Naturally, the early developers decided that it was important to ask for someone’s relationship status during registration – and that one of those relationship statuses would be ‘It’s complicated’. Both of those decisions, small in themselves, made sense in TheFacebook’s early context. But as Facebook became more successful, this dorm-room thinking became the global standard. If you wanted to enjoy what Facebook offered, you had to subscribe to that standard. And Facebook exported these particular sociocultural qualities to societies where such questions were inappropriate, or meaningless. People like my mother, then a septuagenarian of Muslim heritage, were offered the choice of an ‘It’s complicated’ relationship. In her context, and doubtless the contexts of hundreds of millions of other people, it was a nonsensical, perhaps even offensive, question. But the code had become law.

At the time of writing, the democratic world remains blissfully social-credit-system free. However, companies do still use data to profile us – often in ethically dubious ways. Every time you take out a loan, or buy an insurance plan, the terms will be determined by data about similar consumers – and the more sophisticated our data becomes, the more dependent on these methods insurers and banks will be. But this creates a problem: we are all being profiled on the basis not just of our behaviour, but also of our demography. And demographic profiling leads, inexorably, to prejudice. Racial bias in particular tends to seep into data-informed decisions.

Consider the way people make friends online. Social networks endeavour to link us with people – the ‘people you may know’ who pop up on the side of your Facebook feed. These people are those with common traits: they went to the same university, or love the same soft rock band, or share your friends. As a consequence, social networks are more homophilous and clustered than the rest of society. And they are more cliquey, ultimately, for business reasons. People who are connected with like-minded people online seem to use social networks more. And groups with similar interests are more useful to advertisers: similar types of people tend to buy similar products, and that means they are easier to sell to. Thus there is a commercial incentive to group similar people into progressively more precise, and discrete, segments. In practice, however, these methods can end up pushing people into ever more isolated and extreme groups.

According to James Boyle, a law professor at Duke University, ours is the era of the ‘second enclosure’. The first enclosure largely took place between the fifteenth and nineteenth centuries in Europe, as lands once held in common – fields used for grazing cattle, forests used to forage for wood – were taken into the hands of private owners through a series of enclosure acts. Many economists argue the enclosure movement laid the ground for the industrial revolution, allowing farmland to be used more productively. But it was a drawn-out and often violent process – one that destroyed the social fabric that had held together village life. In the words of one leading economic historian, enclosure meant the rich ‘literally robbing the poor of their share of the common’.34 In the age of the second enclosure, we once again witness the commodification of new spheres of our lives. ‘Things that were formerly thought of as either common property or uncommodifiable are being covered with new, or newly extended, property rights,’ writes Boyle.35 The laws that define our lives are made by private companies; our public conversation takes place on private platforms. The most intimate details of our private lives are bought and sold for profit. And the ways we meet people and forge social bonds are dictated by the proclivities of giant corporations.

Imagine that the resources we need to sequence a single human genome reduce by a factor of 100, pushing the cost of sequencing down much further. If, as a result of this lower cost, we end up sequencing a million times more genomes, our overall use of resources will have increased. The risk is that reduced costs lead to gross increases in consumption.