Any suggestion that Big Tech has had a rough time must contend with this fact: Amazon, Facebook and Alphabet, Google’s parent company, spent the past decade growing much faster than the rest of the economy.

Whatever reckonings they’ve faced so far, and whatever backlash they’ve endured — in the press, on Capitol Hill — have been easily absorbed in financial terms. These ostensibly embattled firms recorded billions of dollars in profits, and can reasonably expect to continue to do so. It has been a great worst year ever, and next year is promising to be pretty good, too, as well as much, much worse.

The tech giants haven’t been considered start-ups for years, and in some cases decades. They’re no mere incumbents, either; they’re some of the biggest companies in the world. But it’s not enough to simply take their measurements. They’re diversified conglomerates whose power is greater than even their staggering user numbers suggest. They were expansionary powers, chasing and luring new customers by the hundreds of millions, laying claim to territory and souls with the zeal of missionary explorers. What they’ve become are superpowers, whose imperative for growth has been replaced with a need to fortify, ally and extract. Reaching a billion users is a successful conquest; keeping them, and turning their continued allegiance into lasting power, is empire.

Google is much more than a search engine; Amazon is much more than a simple e-commerce site. This alone complicates the idea of competition. To comprehensively take on Facebook’s expanding coalition of megaservices is implausible; the best a competitor can do is create some sort of service that might steal away time or advertising dollars. Testifying before two Senate committees in 2018, Mark Zuckerberg was asked about his company’s biggest competitor. He struggled to name one, instead gesturing vaguely at “the other tech platforms,” including Google, Apple, Amazon and Microsoft. “We overlap with them in different ways,” he said, answering as if the question were flawed. Competition is something you do in a market you share with others; it is, in the Facebook investor and board member Peter Thiel’s words, “for losers.” Overlap is what might occur when sovereign powers happen to occupy the same space, or lay claims to the same populations in the normal course of conquest.

The tech giants, in becoming tech superpowers, have been growing in every direction beneath our feet, becoming tangled in ways that we cannot easily see and, together, improvising a new world order that is increasingly hard to route around, or to escape. To use the internet, in 2019, is to engage to some degree with the handful of private entities that control it. To start an internet company is to submit to one or many of them from the start. We, and the rest of the internet between us and them, are but subjects on the surface of a planet they’ve fully colonized and terraformed. Unfortunately for us, theirs are empires we’re stuck with for the foreseeable future. Unfortunately for them, they’re also stuck with one another.

At the dawn of our new century, the web was still relatively federated — a messy, sprawling network of sites and services, some serving millions of users, some just a few. Google was still a search engine, Amazon was best known for selling books and Mark Zuckerberg was in high school. This web never fully lived up to its boosters’ ideals of free, mutually beneficial cooperation, of open standards and citizen empowerment, but it was closer than today’s internet. It was built by, and for, people used to accessing networks through computers; for all its professions of freedom and open access, this web was exclusive by default, rooted firmly in an era in which digital participation was economically and socially limited. But whatever else this web was, it was spacious.

The internet is now mobile — available almost anywhere — which has significantly expanded the user base and use cases for all sorts of services: You can buy from Amazon while standing in a Walmart, or check Instagram while you’re watching TV. In the past decade, there was also a wholesale migration of online services to a small group of conglomerate hosting companies, from which start-ups and venerable tech firms alike can rent space and computing power instead of investing in costly, and ultimately inferior, physical infrastructure themselves. On their own, each of these trends is, at most, an airport book. Welcome to the mobile revolution! Allow me to introduce you to network effects. As I’m sure you know, The Cloud is the future, and you’d better get to elevation, fast! The ways in which these trends have combined, however, have produced the peculiar results that have actually reordered things.

The death of the relatively open web, the rise of mobile phones and centralized app stores, the consolidation of media consumption around megascale intermediaries — all of these trends helped to establish what is now the internet’s reigning business model. What emerged is a tech industry that exaggerates the ugliest features of the global economy in which it operates: crushing consolidation of power by an ever smaller number of dominant firms; stratified marketplaces supported by precarious laborers; sector domination as table stakes. The victors in this process captured enormous amounts of attention and capital along the way, which the savviest among them immediately used for — what else? — aggressive, multilateral expansion.

A result is that the largest companies in the Western tech sector don’t operate on the internet so much as they are synonymous with it. The rise of TikTok, the Chinese-owned social app, is an object lesson in the consequences of this arrangement. To a casual observer, it seemed to explode in the manner of a viral sensation, but in reality, it found many of its Western users through an advertising campaign that cost hundreds of millions of dollars in a single year, much of which went straight to American social-media giants — nominally TikTok’s competitors, but also gatekeepers to an unimaginably massive audience. Or ask Netflix, the leading video-streaming company in the United States, the kind of firm that might be confused for a superpower itself, if not for the fact that it outsources “nearly all its computing and storage needs,” in Amazon’s words, to Amazon, which has a video-streaming service of its own, with tens of millions of viewers and its own shelves of entertainment-industry awards.

These, it should be said, are success stories. TikTok gained access to its audience, which it can now try to leverage into money and new forms of influence (though it has recently drawn scrutiny not just from the United States government but from Facebook, which has hosted ads for the service). Netflix doesn’t have to maintain an unwieldy and expensive infrastructure of its own, and is the streaming service to beat (although it recently reported losing users in the U.S. for the first time in eight years).

This is how the internet business works now: If you need to access new populations, you have to deal with Facebook, and buy your way in. If you need to build a new tech company from scratch, you’re going to need to make a deal with one of the empires that has secured access to certain valuable resources. Have a product to sell? Good luck hitting your goals without listing your wares on Amazon, or without setting up a storefront in Apple’s and Google’s app marketplaces. The paths to all but the largest websites run through Google’s territory, which is expanding by the day. The app stores are monuments to the smartphone boom, a gold rush that defined an era — sprawling, peerless capitals through which astonishing volumes of people, products and time pass under the watchful eyes of Apple’s and Google’s bureaucracies.

If it’s still possible to exist online outside the territorial boundaries of these tech empires, it’s nearly impossible not to engage them at all. What little space that remains untouched by the tech empires is still menaced by them. To be online, as a business or as an individual, is to accept their premises or demands. What you end up with is situations like that of Pinterest: a well-known brand that reports to having 300 million monthly users, an online destination, as well as a tool that maps over the rest of the web — another success story.

This same Pinterest, however, exists at the mercy of companies that may not think of it as a primary competitor but would not hesitate to release a competing product. In its investor prospectus, this Pinterest listed its competitors as “larger, more established companies such as Amazon, Facebook (including Instagram), Google, Snap and Twitter.” Elsewhere it listed many of those same companies as risk factors: Google can, and it claims has already, limited visitors to Pinterest through search; Facebook and Google can be used to sign into Pinterest, and changes to Facebook’s login system have already “negatively impacted” user growth and engagement; Pinterest depends on Amazon to host “the vast majority” of its operation, and is committed to spending at least $750 million with the firm over the next four years.

Or maybe you use Spotify, the music-streaming app that can credibly claim to have changed how millions of people listen to music as well as what it means to be a professional musician. It’s an enormously successful operation, practically a generic trademark for music streaming. More than 248 million people around the world use it, and 113 million pay for it, according to the company. But there’s another way of looking at Spotify, and it makes the service look somewhat thinner.

It’s not just that the company is an intermediary between listeners and content that it doesn’t own, though that is true: Spotify, like all but the very largest tech companies, relies on another company — in this case, Google — to host most of its infrastructure. When you hit play on a Spotify song, something happens on servers owned by Google, a privilege and action for which Spotify pays a fee. And for a time after its U.S. launch, in 2011, the only way to sign up for it was by using a Facebook account. Some 70 percent of users access it through an app installed on Apple and Android devices, through app stores run by Apple and Google, each of which has music-streaming services of its own, with comparable pricing and features to the ones Spotify made standard years before. Spotify got its users with Facebook’s help, owns neither the material that it sells or advertises against and in fact negotiates for rights to it alongside much richer competitors that count Spotify as a client.

It might be a stretch to attribute Spotify’s success to Facebook’s largess, or to describe app stores as charitable, but these much more powerful companies have created the conditions in which a company like Spotify has been able to thrive, relatively speaking. In a complaint Spotify filed against Apple with the European Commission, which contends that the App Store policies and fee structures place competitors at a disadvantage to Apple’s own music-streaming services, the company stops short of contending that Apple is deliberately sabotaging its business, or stealing its customers; it simply notes that the conditions created by a world in which Apple has so much power will produce an indistinguishable outcome.

Apple responded publicly by suggesting its operations amount to more than just a company; it’s an entire economy whose success corresponds with the success of its numerous clients, or partners, or participants, or whatever everyone else is. “Spotify seeks to keep all the benefits of the App Store ecosystem,” it said, “without making any contributions to that marketplace.” All Apple wants to do, it went on, is “grow the pie.”

Such lopsided relationships as Apple and Spotify’s provide a preview of the way much bigger conflicts might emerge in the future. The genuine tech empires have always been in competition, but until recently they’ve always had plenty of room to expand in different directions. There were plenty of users to go around, and plenty of ways to make money from each of them. Amazon could dominate retail and hosting. Facebook could dominate social advertising. Google could keep its control over search and mapping and online video, through which it manages a massive advertising marketplace. Their collective success enlarged the tech industry as a whole and created new opportunities within it, as well as nearby.

When imperial ambition came into direct contact — such as when Google tried to launch its Google Plus social network, or when Facebook tried to create a smartphone platform — failures were softened by continued expansion elsewhere. It was a prosperous era for those that were a part of it; less so for any company, or industry, or system, that happened to be in the way. (See: retail; media; democracy.) Everything was growing, and future growth was limitless. Among the American tech superpowers, creeping interdependence has likewise been accepted as a necessary, and even desirable, component of a new digital order. But the end of explosive growth, combined with external pressures stemming in part from the worldwide turn against globalization, will illuminate the ways in which the tech giants have power over one another, and how tenuous are the situations of their smaller proxy states.

The manners in which the tech empires are entangled are occasionally absurd. Apple is currently being sued for supposedly misrepresenting its iCloud product to customers, who may have assumed they were buying storage for their personal photos solely on Apple’s own servers. Apple, the suit contends, had also built iCloud on top of other companies’ cloud services: namely Amazon’s AWS, Google Cloud and Microsoft Azure. In other cases, such interdependence has spilled into conflict. On multiple occasions, Apple has suspended apps created or owned by Facebook for breaking its policies — once even revoking Facebook’s ability to test apps internally, which, among other things, rendered thousands of employees’ phones, all full of nonpublic company apps, temporarily useless. They didn’t touch Facebook’s consumer-facing apps, which provide Facebook with nearly exclusive access to the hundreds of millions of people who use the service, and see its ads, in iOS. But the implication was clear: It could if it wanted to; Apple and Facebook are only partners, or allies, as long as both parties agree.

Facebook could escalate, too, if it wanted, by removing all of its apps from iOS, and preventing Apple users from easily accessing its services. This would be disastrous for Apple, but probably more disastrous for Facebook — and most disastrous for the user who expected to be able to use WhatsApp to talk with her family back home, who then could not. The recent attempts of Apple’s chief executive, Tim Cook, to position Apple as a privacy-forward company, at Facebook’s expense, might sound like grandstanding from the outside, but within Facebook they’re heard loud and clear: Apple believes it is in a position to make demands.

The American tech superpowers are running out of new users to sign up in their mature markets. But more worrying, for them, is the accelerating collapse of their prospects for expansion into some major international markets — particularly China, and countries where Chinese tech companies are snatching up territory of their own. This is a significant revision to their outlooks from just 10 years ago, when they could blithely envision the acquisition of the “next billion” users as nothing more than a matter of engineering resources and time. Now that they can no longer forecast forever growth, the risk of open conflict among these companies in the future seems much higher. Apple and Amazon, for example, might have liked to dominate music streaming — who wouldn’t? — but they never needed to destroy Spotify. Apple was making plenty of money every year selling phones, and Amazon’s main line of business has been growing monstrously, and there were always more markets to expand to anyway. But what happens after a few years of slow growth, when secondary businesses like music streaming suddenly need to make more money?

Even bigger conflicts are possible. It might be unwise, but it would not be impossible, for Amazon to weaponize its hosting services — declaring that it wouldn’t work with firms it considered competitors, or which provide services that “overlap” with its own. (Jeff Bezos’s ownership may have already been weaponized against him, if indeed the Pentagon’s unlikely selection of Microsoft for a $10 billion cloud-computing contract resulted from President Trump’s issues with the C.E.O.) Like any advertising provider, Google and Facebook can choose whom they do business with. What if they declined to accept advertising money from the next social app based in China? (Or what if they were just told to do so by the American government?)

As users — as subjects — we’re not even party to these disputes. What vanishingly little power we did have, to somehow vote with our wallets on services that we generally don’t pay for, is narrowed each time another winner takes all, leaving us with nowhere else to go. We’ve only ever known these empires during periods of expansion. What happens when they run out of land to conquer, and people to claim as their own? How do they rule, and when do they go to war?

John Herrman is a media and technology reporter for The Times. He last wrote a Screenland column about Joe Biden’s gaffes. Maurizio Cattelan is an Italian artist whose work has been the subject of numerous solo exhibitions, including at the Guggenheim Museum in New York and the Centre Georges Pompidou in Paris. Pierpaolo Ferrari is an Italian photographer and, along with Cattelan, is a founder of the magazine Toiletpaper, known for its surreal and humorous imagery. Martha Harbison is an infographics and data-visualization journalist based in Brooklyn.

Additional design and development by Jacky Myint.