The era of liberal democracy has passed without much fanfare. With feudalism’s fall, historians could point to Luther beginning the process and the Fall of the Bastille, more or less, ending it. More recently, with the end of Soviet communism, we witnessed the heartfelt joy as the Berlin Wall crumbled. But today, as we live through the end of another historical era, fate has not graced us with any memorable milestones, much less brave markers of resistance for our new dark age.
However, in our age of narcissism and self-consumed distraction, that feels appropriate. More than that, though, perhaps our lack of recognition stems from the fact that we were merely living through a transition period, a purgatory-like space between liberal capitalism and whatever we’re going to call the digital dystopia piecing itself together in the palm of our hands.
And if that world slowly becomes clearer every time we click “accept the terms” — now, with the help of AI, the pace of corporate tyranny has only accelerated. Any freedoms we used to enjoy, from surveillance and bureaucrats dictating every term of our existence, are rapidly disappearing. The process has become so rapid and all-encompassing that it’s clouding our understanding not only of what’s possible, but what’s human.
For if AI is a new form of intelligence, it also represents a novel form of life itself. To be sure, it cannot actually think or reason, whatever the tech-bro positivists might claim. Yet in its clear-eyed lack of empathy, it nonetheless symbolises an utter break with the civilisation that came before. How else to explain the fact that AI chat systems “accidentally” end up telling depressed people to commit suicide, kill their rivals at work, or start a cocaine habit? To them, such patterns seem logical. The customer is always right after all. And what does the AI know? Killing yourself certainly will end your misery. Cocaine will indeed make you feel good. It’s hard to argue with the validity of pattern recognition.
The logic of an organism using binary code, no matter how sophisticated, can only be self-interested and sociopathic in pursuit of its creator’s wishes. Computer systems will never replace friendships, much less family or life-fulfilling work. Whether we’re talking about Facebook or some future virtual reality system resembling the Star Trek holodeck, digital programmes will never be a cure for liberal individualism’s connection crisis. Nor will their implementation improve economic production to the point where “fully automated luxury communism” becomes inevitable.
It’s no mystery what our AI future will bring. Occam’s razor applies. Whether near or distant, AI will do precisely the same thing they are doing in the present — extract every last element of economic and ecological value available and feed it to wealth-hoarding individuals and corporations. Whether made of skin and bone, or composed entirely of metal and microchips, these entities will then reinvest enough of their extracted resources to create even “better” (more addictive) tools of distraction.
These distraction tools will then be used to extract even more value — in the form of behavioural data and prospective consumer habits — from the citizenry. As for those poor saps, they’ll eventually function as little more than defecating and urinating batteries for an economic system that neither improves existence for most people, nor reflects humanity’s spiritual or moral values. God is indeed dead; Nietzsche’s great error was being too early.
Welcome, then, to the Age of Maximum Extraction. Yanis Varoufakis calls our new order “techno-feudalism”, though if we’re going to nitpick, “hyper-feudalism” feels a bit more on the nose. With omnipresent iPhones, smartwatches, and now internet-connected refrigerators and coffeemakers, our new order is hyperactive in every sense of the word — hyper-consumptive, hyper-distractive, and, certainly not least, hyper-destructive. Like a lecherous porn director, AI both participates and enables our own self-destructive behaviour in a manner demeaning to any honest sense of what’s consensual, much less healthy or environmentally sound.
At the current rates, the average UK student today will spend 25 years of their life on their mobile phones, which, very soon, will be mostly interacting with AIs of one sort or another. This compulsive behaviour — an average of more than six hours a day for today’s university cohort — is almost entirely useless except for the purposes of value extraction on the part of big-tech firms. The amount of time now thrown into the cosmic trash bin by those addicted to their phones is staggering to behold. For instance, if an English speaker were to practice Spanish or Italian for six hours a day, most learners could become close to fluent within four months or less. That same amount of time is now spent, month after month, year after year, simply staring — transfixed — at Orwellian-named “smart” devices.
But self-consumed, time-wasting is only one of many horrifying elements to arise from the mass adoption of handheld computers. When an influential data scientist earnestly prognosticates that the Anglo-American world is likely headed toward a mostly illiterate future, one where reading will soon become a class marker, practiced regularly only by the well-educated and the wealthy, the concept of the “attention economy” threatens to soon take on a new, more grotesque meaning. Orwell himself is rightly celebrated for his insight regarding the strategic use of language to control thought. In Nineteen Eighty-Four, the Ministry of Truth constantly revises written rhetoric not only to distort the past — but the present too. But what if the written word itself effectively becomes a historical artefact, something only those in power, or else fictional dissidents like Orwell’s Emmanuel Goldstein (a kind of regime-sanctioned scapegoat), still bother to engage with? The mere act of reading could soon become a mark of deviancy.
“The mere act of reading could soon become a mark of deviancy.”
If the human consequences here aren’t bad enough, meanwhile, the environmental costs of Maximum Extraction are equally staggering. As the move to insert AI into everything takes shape, few appear aware that a typical ChatGPT or Anthropic query uses roughly 10 times more energy than an old-school Google search. Scaled to roughly 700 million queries per day, AI chats already use the same amount of electricity in one 24-hour period as 35,000 American homes do annually.
During a time of raging wildfires and terrifying drought — where the need for large storehouses of water is an absolute necessity to stave off disaster — cooling our existing AI servers already requires hundreds of thousands of gallons of water per day. As AI chats replace most browser searches, that water demand is going to get infinitely worse. Even at the present levels, which are probably underestimates anyway, AI data centres are set to withdraw 4.2-6.6 billion m³ of water by 2027: more than the total water withdrawal of the UK for an entire year.
The obscene ecological waste surrounding the production of AI is merely a drop in the bucket of the hyper-destruction these systems will wreak. Let’s start with the misery they’ve already created for average workers and consumers. During the pandemic, even as millions were dying, and scores of people were thrust out of work — their businesses literally banned from operating effectively — residential rental prices still skyrocketed in cities right across the US. How did this make any sense? AI use, of course.
One rare courtroom win for regular people came when the real estate software outfit RealPage was recently taken to federal court for helping spark the rent-gouging crisis through AI tools offered to landlords. The suit was brought by the Biden administration’s Department of Justice, alongside eight state attorneys general. RealPage’s software, the suit alleges, allowed landlords to collude their pricing via an AI algorithm. No need for backroom deals: just plug in your units, make a few mouse clicks, and let the AI system do the collusion for you — setting rent just high enough to crush tenants with nowhere else to go, because everyone else in the area is doing the same.
Even landlords not using the software profited from the artificial price hikes. Think of it as a labour union, but without any risks involved and with all the benefits flowing to management or, in this case, the rentier class. Even if the government’s lawsuit against Real Page prevails — and, under a presidential administration headed by a real estate tycoon, the motivation to succeed is likely waning — the AI tools are becoming so sophisticated, and the Constitution remains so woefully out of date, that the prospect of prosecutors consistently holding trillion-dollar tech giants in check seems dims. Yet in the Age of Maximum Extraction, rent fixing is only one of many elements of everyday life made dramatically worse by AI.
Companies like Amazon, Uber, DoorDash, and other so-called “gig” employers currently use algorithms to set their pay rates “dynamically”. In other words, an employee’s pay changes minute-to-minute or task-to-task, and is calculated in real time using data that responds to supply, demand, location, and even worker behaviour or time of day. An AI algorithm — not an actual human manager or clear set of agreements — decides how much a worker earns for a given job or task, often without disclosing its formula in any intelligible way.
The AI wage formulas used by these firms are entirely opaque. Almost no one understands exactly how they work, and their deployment makes it impossible for workers to anticipate how much they’ll earn on any given day — let alone week to week or month to month. The effect is to turn the worker into a kind of human turbine, spinning endlessly yet going nowhere, all while sinking deeper into debt. And all just to stay afloat in an attention-starved existence where nearly every aspect of daily life serves the needs of value-extracting tech overlords.
It’s tempting, here, to contrast our coming reality to the old-style feudalism stamped out between Luther and the French Revolution. In truth, though, even that might be too generous a comparison for those at the bottom of the digital pile. Medieval serfs and even 19th century slaves had far more stable and predictable lives. Most US slave masters, as Marxist historian Eugene Genovese noted, were motivated to keep their workers reasonably nourished and safely housed. Purchasing new slaves was extremely expensive, not to mention time-consuming. In today’s Age of Maximum Extraction, conversely, the vagaries of globalisation ensure developed economies have a near-unlimited supply of cheap (and often illegal) labour.
Not that the situation is entirely hopeless. President Trump may be faulted for a great many things, but his treatment of the country’s immigration crisis ought not make the list. Draconian they may have been, his policies have seen border detentions fall by as much as 90%. Even if that government figure proves an exaggeration, Trump will still have fulfilled his promise to voters. A message to the global south has been sent: the US is no longer a place of open immigration.
Unfortunately, border enforcement and immigration restriction is the only tangible thing Trump has done to counteract the forces of maximum extraction. Not that Trump’s opponents are much better. The Democrats, too, are fully captured by wealth-extracting exploiters, as evidenced not only by their support for the previous regime’s reckless open immigration policies, which greatly undermined the negotiating power of working people, but also the fact that they too receive a great deal of their financial support from the big tech giants. The voraciously anti-labour Elon Musk famously “donated” over $250 million to Donald Trump’s reelection bid; yet Musk and Peter Thiel’s GOP partisanship is more the exception than the rule. PAC donations from Amazon, Google, Microsoft, alongside nearly all the big tech giants, are generally split evenly between both major parties. The same double-dealing donation pattern can be seen with Uber, Doordash, and the rest of the “gig” economy profiteers.
To be fair, there are some resisters to our new digital dystopia. But their efforts are often disorganised, and typically offered by woke politicians with repulsive social politics. In this vein, Rashida Tlaib and Greg Casar lately introduced legislation intended to halt the use of AI tools and algorithms in coordinating prices and suppressing wages. Everything laid out in their proposed bill is reasonable and indeed highly necessary. But with the Democrats out of power on every national front — and recent polls confirming the party remains at its lowest ebb in three decades — the legislation has no chance of passing.
Even worse, the bill arguably provides tech lobbyists with advance warning for how their opponents seek to attack the AI mothership. A moment of what Michael Oakeshott might call rationalist politics at its most unmoored: abstract ideals offered in place of real strategy, leaving power unthreatened but forewarned.
Given Tlaib’s recent publicity stunt of banging pots and pans on the Capitol steps, meanwhile, it’s difficult not to wish that someone — anyone — other than her had fronted the legislation. Unfortunately, Tlaib and her colleagues in the progressive caucus are reflective of how “resistance” in the Age of Maximum Extraction is offered: purposeless cross-country speaking tours and TikTok sketches, coordinated by 20-something social media “experts” obsessed with their own maximum algorithmic reach. An Emmanuel Goldstein meme served up for mockery by the very tools it seeks to undermine.
