Technofeudalism — Martin Stellinga

People tell me AI has become inevitable. Maybe so, but do people realize what that means? Let’s speculate on the future Big Tech is pushing, because that’s not about tech and not inevitable.

Disclaimer

I can’t predict the future. I am not a big tech insider, and I don’t have an inside line to secret military plans. I’m a blogger and a software engineer. I am well-versed in science and science fiction, I know how AI works, how political systems work, and I have a good brain to think things through.

Anyway, you have been warned. This is a thought piece; an opinion. I hope it helps you think. And if you see flaws in my reasoning, please point them out. Preferably backed by data.

Not all AI is equal

So, second disclaimer. There’s AI, and then there’s AI. It’s become an umbrella-term for all kinds of useful tech, a lot of bullshit tech, and some very nasty applications. Generative AI, e.g. ChatGPT and Gemini, is the type of AI I have issues with. Generative AI steals, is power hungy, and it what companies like OpenAI are mostly peddling. But an algorithm to predict the weather is also AI, and that is very useful and doesn’t steal.

So, let’s dig into what AI could mean for the future. First, the environment.

1. Environmental collapse

AI is power hungry. Really power-hungry. From the internet:

Data centers in the US used somewhere around 200 terawatt-hours of electricity in 2024, roughly what it takes to power Thailand for a year.

- Lawrence Berkeley National Laboratory

That’s a lot. And it will get worse going forward. The International Energy Agency estimates that by 2030, world-wide AI energy use could go up to 1.000 TWh (terawatt-hours). That’s 3% of the 30.000 TWh we use world-wide each year. Note that this is new power use; a few years back that demand was almost 0. And look at the chart below.

Renewable energy is at nearly 10 TWh and rising. So, yeah, growing, but we don’t know what the world-wide slide to the right will do to these numbers (Trump’s “drill, baby, drill” is not helping). More importantly, we’re at the brink of irreparable damage. It’s like you’re smoking 30 cigarettes a day and the doctor says you will die if you don’t quit. So, on the one hand you cut down, but on the other you think ‘well, why not smoke some extra from this new brand’. But, it’s not the total disaster I thought. Not quite.

On top of that, though, AI data centers need water for cooling. Numbers vary, and AI companies like to juggle numbers and tell outright lies. By 2028 AI might be using a trillion liters of water. That’s the water usage of a country like Jordan. In stark contrast to this: we have far too little water in the world.

Then there’s the heat island problem. AI is generating a lot of heat in areas around their data centers, and that will affect hundreds of millions of people worldwide who live within a few miles of those data centers. And when the AI chips break after a few years, they result in e-waste. We currently cannot recycle all the parts coming out of the data centers. Meaning we produce more garbage, and need to have more polluting mines — if there is even enough to mine.

In other words, we were already heading toward a world-wide climate disaster, and AI is making it worse at a time we should be doing everything we can to stop it.

The positive

On the flip side, AI may help with the energy transition. Some say that those gains will outweigh the net draw of energy. AI groupies and companies will pull out this research when you say bad things about AI and the climate, so I shouldn’t ignore it. There are two interesting things to note about the research I linked.

The first is that this is only partially about things that AI can do, and partially a prediction about what AI might do. And that ‘might do’ part is speculation. So are the energy drain predictions above, of course, but an extrapolated energy-use line that you can back by contracts for future data centers is a far more reliable prediction than future advancements based on speeches by Sam f*cking Altman.

The second, and probably more important point, is that the MIT research I linked is about AI models controlling energy grids and predicting weather patterns. Like I said above, not all AI is equal. These predictions do not involve the great power guzzler that is generative AI. We could kill off chatGPT and Claude, prevent much of the power draw, and still have those advancements.

The net result

In any case, my personal outlook on saving the planet and the role generative AI has in that has not been swayed to positive. The future is bleak.

What this means is people will die. Millions of them, if not billions. Part of that was already inevitable, but it will be more because of AI. Countries will flood. There will be more extreme weather; forest fires, floods, droughts. And famines or water shortages. Depending on where you live there’s a good chance you will experience one or more of those in your lifetime. And don’t expect to be able to afford chocolate or coffee in the future.

2. Capitalism as a religion

But AI is the future! This is how capitalism works! People vote with their wallet and they apparently want AI! Yes, I’ve heard this.

But voting with your wallet is a mixed bag at best. Yes, it sort of does something, but not what you might expect. Heck, try voting with your wallet against AI. There isn’t a choice of ‘no AI’ to vote for. Believe me, I’m trying. If there was a safe friendly AI vendor, I’d be all for it. But there isn’t.

The problem is capitalism itself. Many people have this idea in their head that capitalism is good and leads to a just, free, democratic society. It does not. Let’s look at the Wikipedia definition:

Capitalism is an economic system based on the private ownership of the means of production and its use for the purpose of obtaining profit.

- From Wikipedia

The two main components are ‘private ownership’ and ‘profit’. Note that neither of these things have anything to do with justice or democracy. Private ownership has a link with freedom, but as I’ve argued often: equality is more important than freedom. Furthermore, as I wrote in my explanation of libertarianism, completely free market forces lead to gatekeeping. What Cory Doctorow and Rebecca Giblin call chokepoint capitalism.

So capitalism can be used in a free equal democratic society, but only if you have governments that ensure the playing field is level, everybody has an equal chance to succeed, and there are basic social safety nets. Because people think capitalism itself does this, they often believe removing all regulations will create a better world. And the rich and powerful are very eager to keep that myth going, because this idea gives them more power and riches.

All of the above is not an AI problem. However, generative AI is being used to push this ideology on steroids. Big Tech owns the data centers, the compute power needed to train generative AI, and the know-how on how to apply it. They also have the money to train it — AI’s dirty secret is that it required a lot of manual (exploitative) labor. Big Tech owns it all, and they will milk us for profit. And worse.

You see, Big Tech did not own the writing they used to train their AIs, but here you see how unbridled capitalism is about power: they just took it and decided that copyright law only applies to lesser beings.

Some scenarios

There are two possible views on how this will play out.

Either AI can replace humans, and human productivity will go through the roof. In this scenario AI will be everywhere… and big tech will charge us through our noses for it. Gatekeep AI to all business and all consumers. Us poor saps at the bottom will have to vie for the shit jobs and work for scraps, or take a second job to make ends meet. And the only jobs are shit jobs, because all the fun ones are taken by AI. Great future.

Or, AI can not really replace humans. Still, companies can pretend it can. They can force employees to do more work for less money ‘because AI’. In that scenario, big tech will still gatekeep the shit out of everything too. At the heart of these two scenarios is power. It’s about funneling money to big tech and the billionaires behind it. Profit! Because capitalism good.

There is a third and fourth scenario, of course: if AI increases productivity, we use that to improve everybody’s lives. If AI doesn’t, we kill it and do not hand big tech more money. But that requires regulations; it requires curbing power of the billionaires. And we’re not doing that at the moment. Even if the AI bubble pops, there is a good chance Big Tech will try to push the damage back onto everybody except the rich elite. And that push could lead to fascism.

Religious end game

The problem is the religious adherence to unregulated capitalism. Property gives power. Power makes money. And money grants more property. A feedback loop that leads to a modern form of technofeudalism: a small elite own the means of production, and the masses swear fealty and perform backbreaking labor for the scraps the elite deign to hand them of what they produce.

An ideology — almost a religion — has sprung up around unregulated capitalism and tries to convince us that ‘this is the way’. Where medieval kings were supposedly ordained by god, current-day billionaires seem to believe they have been ordained by money. The wallet is all and government regulations are preventing the rise of the Great AGI god who will deliver paradise.

3. End of paid art

As I wrote above, Big Tech has used art to train their mighty generative AI. They took books and pictures, crawled the internet, and took all that data to train their mathematical model. Is this legal under copyright law? Big Tech claims it is. Yeah…

Even if we ignore that they didn’t even pay for the works they used, that argument is pretty shaky. Feeding a lot of copryighted works into a mathematical equation that can reproduce those works almost verbatim makes it highly dubious.

But the spirit of copyright is that the creator of a work has the right to determine how it is used so they can make a living. Buying a book gives you the right to read the work, and buying a painting gives you the right to hang it up and look at it. But feeding it to an equation that is then marketed to put you out of work. Well, the spirit of the law requires Big Tech to pay for that, even if you could pretend the wording implied something else.

The creative industry has waged a war on creatives for decades, and generative AI might be the end game. Yes, you heard that right. Writers barely make an income from writing. The same goes for painters, musicians, actors, and screenwriters. While the creative industries have been thriving, those doing the actual work are not.

Creatives in poverty

Again, unregulated capitalism is to blame. The problem is: creatives don’t have a lot of bargaining power. The industry likes to speak about “content” and any creative who can produce “content” is as good as any other. Amazon deliberately inserted themselves as gatekeeper for the book market. They can keep prices low and ship more ‘units’ and take a bigger cut of those units, then charge creators again for better placement in their store (so, ads). And if a creative balks at that, they can be thrown by the wayside and replaced with others, quality optional.

This is of course, horrible. We all partake of creative output (books, tv shows, and games), so why do we keep those we create that output in poverty? We apparently value billionaires like Zuckerberg and Musk millions of times more than creatives, even though they help instigate genocides and endorse world hunger.

Creative pursuits are becoming a hobby of the rich. And that is unfair, and bad for our society. Creatives are a powerful counterweight to dictators and abusers. Stories can help fight inequality and abuse, by shedding a light on the bad things.

AI content

Gen AI is a new front in this gatekeeping wars. It helps creative industries create the ‘content’ without needing a creative at all. This is their holy grail. All the benefits, but none of the costs. It might even work, but we will lose a lot. TV shows have a social component, and you break that when you switch to AI.

We’d also put Big Tech in charge of our content, and push anybody else out. Already it’s almost impossible to get a job in game journalism, and other journalism and writing and art jobs might soon follow. It really says something about us as a species. As Chuck Wendig put it

“When the robots make the music and the humans are cleaning up the dog shit — WTF?”

- Chuck Wendig

And a fun fact about capitalism: if all the consumers are poor, who is going to generate the profit?

One other interesting catch-22 is that the world changes, and our language does too. That means AI needs retraining. However, if there are no creatives any more, how are you going to retrain your AI? You can’t feed it its own slop (which also says something about what it produces). So… steal more from the last die-hards refusing to give in? Hire writers on the cheap to write article upon unpublished article for a pittance to feed the AI beast? It seems… bleak.

4. End of democracy

Like I wrote above, AI is taking over the arts. That undermines democracy, but it goes a lot further than this.

ChatGPT is quickly becoming the primary way people partake of the news, just like social media. That makes Gen AI and social media a very important — if not critical — part of the fourth estate. An independent press is a requirement for a functioning democracy. And we are seeing that fourth estate in the hands of Big Tech. As I wrote before, the medium is the message, and the message of AI isn’t a good one.

Even if a model is ‘neutral’, it still pushes the status quo. Everything is political, and mathematically generating the most ‘default’ thing is conservative. So AI-generated images doctors will be male, and stories will have white male heroes. Capitalism is the way of the world.Systemic inequality is built in.

Worse. Already Elon Musk has been promoting lies that benefit him and his fascist friends through Grok. There’s a good chance the billionaire elites try to push the message that ‘they are a better breed of man’. Like the church would tell the medieval peasants the king was chosen by god in their sermons. It’s already happening, and most are more subtle than Musk.

A world of lies

There’s even more, unfortunately. Generative AI blurs the line between lies and truth. It will lie as convincingly as it tells the truth, and moves from truth to lie without blinking. When our primary media source is a peddler of lies, we shouldn’t be surprised that lies become the norm. Lies benefit those who peddle lies, and those who are best at peddling lies are the toxic narcissists and those without a conscience: the dictators and psychopaths of this world.

A world where truth no longer matters is a world where democracy cannot exist. And no, that isn’t just AI’s doing. Social media has already chipped at the concept truth for two decades. Why do you think there is growing trend towards extreme right views and a growing fondness for totalitarians?

5. Every home a warfront

The last few years we’ve seen the face of war change. It used to be easier, and you could recognize war by troops physically crossing borders. But wars have turned hybrid, involving hackers and sabotage on a number of invisible fronts. AI is taking this to a new level. A level predicted by scifi writers for decades, and now it’s here.

In Ukraine we’ve seen fleets of autonomous drones driven by AI, fought by jamming and other countermeasures. Missiles rain down from the sky, even as Russia and Ukraine both fight across the physical and digital terrain. A country like Iran can defend itself against traditional warfare by fleets of cheap drones draining expensive missile systems and disrupting international trade.

Meanwhile, the Orange King of Western Lalaland (I’d call it the US, but that doesn’t do justice to the chaos) has started committing war crimes in international waters and on foreign soil, alongside we-pretend-babies-are-terrorists-to-kill-them Israel. Note how Iran’s leaders were recently taken out by well-timed missile strikes, after tracking them through hacked street cams and AI powered target selection. A school of children in Iran was bombed by Claude selecting the wrong bombing target based on incorrect data. A war crime in itself, but also a signal flare illuminating the future. Maduro from Venezuala was kidnapped from his home and is now in some US prison — another violation of international law.

And yes, Iran leadership was horrible. So was Maduro’s reign. But the Pandora’s box is now open, and it won’t stop at bad people. The Orange King could decide to take the President of Spain next, or bomb the Netherlands for housing the International Court. Or kidnap EU functionaries for daring to regulate Big Tech.

It means there is no safety. If anybody with enough resources decides to target you, you are screwed. We’ve seen Israel and Russia take out dissidents all over the world. Imagine that, but times ten. And it’s not just countries.

And no, this isn’t just about AI, but AI is a big part of it.

The powers that be

And this isn’t about countries alone. Remember the crusades? Those were not countries waging a war, but the Catholic church going on a rampage. So why couldn’t Elon Musk, emperor of crap-with-X-in-the-name, start a crusade? Or Jeff Bezos. Or Robot Zuckerberg? They own the US already, and have more power than most countries. There is now an Orange King, but above him are an oligarchy of Big Tech Emperors.

In this age of hybrid warfare, the fronts are not neat. The fighting is not visible. But Facebook pushing a genocide in Myanmar should be a sign. No soldier was sent, but people died. Musk had as much effect on the Ukraine war as international leaders. The US is trying to push its vision of uncontrollable megacorps and untouchable billionaires on everybody; their religion of feudal capitalism. Trump as King, and Musk as the pope of the church of X, fighting crusades against democracy and freedom, the rest of us serfs in their game of thrones.

Nowhere is safe

But the thing that scares me, that truly scares me, is that it touches everything. Back in the day, an army had to physically go somewhere to cause trouble. The Viking raiders at the gates had to be truly at the gates. These days, my MacBook is a battlefield. My home is on the front; all our homes are.

You could be walking through the street, passing some hacked doorbells, and a drone could come swooping in and shoot you, powered by a facial recognition algorithm that mistook you for a criminal. Or rather, who certain billionaires or kings dubbed a criminal. And that’s just the flashy version of this.

It can be more subtle. My blog could trip the wrong AI agent, and have my mortgage application yanked. Big Tech could nudge the careers of everybody who doesn’t believe in their version of feudalism. Heck, if I continue to fight AI, may job prospects could be in trouble. That can be done at scale, in all places in society. No human intervention needed. I wrote about this before, but things seem to only have gotten worse. AI is put into everything. Who is to say I might not end lower on the list for medical attention because of my political views? Payments for black people could start to get delayed or vanish.

The end game is a society where no place is safe, and we are nudged every second, through every interaction with any computer. The most horrible oppression in human history, where every man, woman, and child is actively targeted and spied on. And we don’t even know we are being attacked and victimized. A thousand tiny AI nudges are hard to separate from bad luck. AI could become the tool to lock us into fully automated totalitarian states. A religious feudal empire where the peasants are kept in line by computers. The critical nudged into obscurity, and the sycophants promoted to power. AI-powered religious fervor for the church of tech feudalism.

And if that doesn’t give you nightmares, imagine forced brain implants for our children from Elon Musk to make the control complete.

And that is what keeps me awake at night.

Summary

Am I painting a bleak future? Environmental collapse, a dystopian feudal society forever under the boot of an elite of evil white men. I guess I am.

But that is unfortunately the future you choose when use something like ChatGPT. You don’t need ChatGPT. You don’t need Claude. Gemini is a choice. You can live without Copilot. When you do use them, you are choosing this future I described. I know, you don’t want that future, and you’re not voting for that. You just want a fun cat picture, or a nice report for your school assignment.

And maybe what I describe can be stopped. But that’s only if we choose to change it. That’s why I’m writing this. Not because I hate the tech, or want to scare people. It’s not about turning back the clock. It’s about regulating away the nightmare world I described.

Just, please don’t think ‘it’ll never happen’ or ‘somebody will fix it’. Because there is no such guarantee. It is up to all of us to make the world better. To care. To change the laws so they are fair. Fight inequality and racism, and billionaires taking control. Accept that you might be supporting these things accidentally with your actions. It’s okay, it doesn’t make you a bad person, but you are helping bad people, and now you know why.

So, REGULATE AI. Say no to being shackled to Big Tech. Better yet, say “f*ck you!” Then go out and make real human art. Read a real human book you buy directly from the author. Go to a real concert. Live, and make real art and real human connections.

Originally published at https://martinstellinga.com on April 17, 2026.