You're not ready for how AI will refactor society

Some worry about a "permanent underclass" or "useless class" but the truth is actually darker. Much darker.

I have bad news, and I have worse news.

The bad news is that we’re heading for techno-feudalism.

The worse news is that this outcome is all but guaranteed, in a pattern that you could call “overdetermined” or “inevitable.”

Yuval Noah Harari has popularized the term “useless class” but if you’re not familiar, I’ll catch you up—the idea is that AI is obviously about to outstrip the intelligence of most humans, and then all humans. Which means if your livelihood, contribution, or differentiating factor is your intelligence, then you’re about to economically useless.

Forever. The end.

In online circles and social media, the term is a bit darker and more realistic. They say “permanent underclass” usually in the context of ‘prepare to be a member of the permanent underclass’ or ‘you have a year or two to escape the permanent underclass.’

But both of these terms are too generous.

Why?

Because they include the term “class” which implies civic standing that may not actually exist.

History is rife with examples of disempowered, marginalized, and disenfranchised people. Calling them a “class” is charitable at best, and outright dangerous at worse.

Chinese peasants under corvée systems were pressed into service and died in droves to build ancient China’s megaprojects, including canals and the Great Wall. They were not “slaves” insofar as their service was not permanent. But it was forced, and often lethal.

Peasants were a “permanent underclass” and they were still necessary to work the fields. You will not be necessary for anything.

Why? Because China could afford to waste lives. The fertile river valleys produced an annual “crop” of humans, a renewable resource that warlords and emperors could treat as disposable.

Keep in mind that they still needed humans under this regime, and mortality rates in these projects was often north of 50%.

Under a fully automated economy, where robotic soldiers fight the wars and AI agents run the businesses, your body is a net negative to the economic system. The Nazis used an even darker term, which is more apt to what you will be in this system; useless eater.

Technically, the Nazis used this term to described disabled persons—lovely, I know.

But it’s worth naming the anxiety outright. I am merely giving words to the fears, and articulating them clearly. Personally, I was resistant to this worldview. That is, until I started researching it for my next book.

You see, we all live under Enlightenment-era delusions that there are so-called “natural rights of man” but in reality, those are legal fictions. Historically speaking, such grandiose language of the Enlightenment and other romantic thinkers came after the blood had been spilled. We remember events like the American Revolution and Glorious Revolution proudly, but make no mistake, they were bloody affairs.

Wealthy thinkers and poets came after, and wrote pretty language to justify the bloodshed and the new way of being post-hoc.

This is literally a 1:1 substitution for a human. The robot cannot say no. It will serve whatever master, without complaint, and without political agency.

That’s why my next book centers on what I call a realist theory of rights. All human rights, all human dignity, has been secured through coercive force. Sometimes that means strikes and revolts. In other cases it requires mass withdrawal. Sure, there are plenty of arguments to made that we are all endowed with some divine rights by the Creator, but those rhetorical flourishes mostly serve as Schelling points (ideals to rally around), but there is no substitute for coercive extraction of concessions.

Rather than borrow the Nazi terminology, I have my own term for how most humans will be treated under a fully automated regime: redundant biomass.

The darkest possibility that some outline is a self-sealed system where the owners of capital, the elites, simply… ignore us. They secure every step of production, from mines and farms to luxury goods, with AI and robotic security, monitor the rest of us with automated surveillance, and simply wall us off.

I would have once scoffed at such ideas—such outcomes were basically impossible under a democracy. Right?

Having looked at the historical record, I’m no longer willing to bank on the goodwill of power. When the British Empire let millions of Bengalis starve simply because they wanted the food for other purposes, and wrote off the loss of life as a temporary labor shortage, that was an act by a so-called “civilized” people.

I gave a talk at Duke University last year, and during that talk I warned that human nature is still dangerous. One student disagreed, saying that we’ve “evolved” beyond that. Evolution is not that fast. And also, human nature will reassert itself when push comes to shove. Yes, it’s true that we Westerners generally live in very stable, safe environments. Many have done for many generations, so it’s easy to forget how humans can be.

When I said that the outcome was all but assured, I used the word “overdetermined” and here’s what that means: there are many forces at play, powerful trends that all pull us in the same direction.

The first and most powerful is geostrategic competition between the US and China. This is the Cold War 2.0, and it guarantees that AI, robotics, and automation will be treated as first-class political and capital priorities for decades to come. In that time frame, automation will surpass all human capabilities.

China and US have both declared they will globally dominate AI and robotics. That is a race condition. There can be only one winner.

The second most powerful force pushing us in this direction is capitalism and neoliberalism directly. The data center and AI buildout is the largest private megaproject in human history, full stop. By 2030, we will have spent trillions of dollars on AI infrastructure, which will eclipse every megaproject in history, public or private. We’ve already surpass the Apollo program and Manhattan Project. The market has already spoken.

The third major force is rational economic decisions combined with creative destruction. When push comes to shove, every business, household, and government makes the economically rational decision to go with the cheaper option, even when the cheaper option is worse. As AI and robotics ramp up, automated goods and services will be significantly cheaper. They will also end up being better and faster. Now multiply that preference by 340 million Americans, millions of businesses, and 200 nations across the planet.

This creates what’s called an attractor state. The inevitable outcome is a maximally automated economy. That includes all production and military. Why? Because humans are slow, dumb, and expensive. Humans are always the weakest link where technology is concerned.

This is a hyper-scaler data center under construction. If you cannot afford to build or buy one, congratulations! You’re already a member of the permanent underclass!

Now, it’s not all gloom and doom.

I am a techno-optimist and techno-progressive.

But I am also a structural realist, which means I take incentive structures and attractor states and macro trends very seriously.

We’re arguably already living under techno-feudalism. I had my YouTube channel demonetized because my feudal lord Google decided I violated some arbitrary policy. Well, there goes more than a quarter of my income! Whoops! And my appeal process is an automated process that is “checked” by some non-English speaking worker across the globe just checking a few boxes.

Tech bros in Silicon Valley want you to believe “the market will work it out.”

We are heading for full automation, which has the potential be beautiful and glorious. Imagine fully automated cancer research, space exploration, and environmental rehabilitation. Solar farms that build themselves, schools that solve Bloom’s two-sigma problem, and more. That is the promise of automation.

But, all technology is a double-edged sword.

The self-same automation that can liberate us from drudgery and create abundance also, by definition, makes us economically, structurally, and politically irrelevant.

That is obviously not good.

And this is the focus of the next leg of my work.

For now, my book Labor/Zero, which looks exclusively at the economic and household income question of this Post-Labor world we’re building, comes out later this year. We’re aiming for July or August launch.

I’ve already begun on my next book, which takes a hard look at the civics, power, and governance of a post-labor world.

The outlook is grim. But there is still a path to a brighter outcome. After all, the owners of capital are still dependent upon human labor, for now. That means we have a collective veto. We can halt all production.

Make no mistake, we are losing that veto power, inch by inch, to the automation.