The Steam Loom and the Data Center

On the afternoon of March 11, 1811, in Nottingham England, British troops broke up a group of protesting mill workers demanding more work and better wages. That night, angry workers smashed up textile machinery in a mill in a nearby village. The attack sparked a movement.

Soon, similar attacks began to occur throughout the textile centers of the English Midlands and north into Yorkshire, first happening nightly and then in sporadic waves. Efforts to suppress them, both by mill owners and the crown, grew increasingly violent as the attacks spread. One of the bloodiest incidents occurred in April 1812, when a large group of protestors stormed a mill near Manchester. The owner of the mill ordered his men to fire on the protestors, killing three and wounding 18. British soldiers called in the quell the unrest killed five more the next day.

The attacks nonetheless kept up, on and off, for another five years, smashing machinery and even setting some mills ablaze. Mill owners as well as the authorities responded with increasingly forceful tactics to try to suppress the movement. Dozens of protesters ultimately were hanged and others sent to penal colonies in Australia.

Yet for all the panic the attacks engendered, the “movement” was never truly organized and had no clear leader. Much of its mystique—at the time and since—stemmed from the shadowy figure of Ned Ludd, sometimes called Captain Ludd, General Ludd, and even King Ludd. He made is first appearance at an attack in Nottingham in November 1811, and popped up at protests all across the Midlands.

But there was no such person, at least not at that time. He was an apocryphal figure cooked up by the protesters and loosely based on an 18th Century figure named Ned Lud, or perhaps Ludham, who himself may or may not have existed. He was purported to be a young apprentice in a textile mill in Leicester in the 1780s. According to legend, he was working one day at a stocking frame—a kind of mechanical knitting machine—when he was admonished by his boss for knitting too loosely and ordered to “square his needles.” Rather than comply, Ned supposedly grabbed a hammer and smashed the stocking frame to bits.

The 19th Century textile workers invoked his legend and took the name “Luddites.” They sent bellicose letters and posted billboards pseudonymously signed Ned Ludd. So effective was their propaganda that British soldiers sent to put down the protests more than once swore they had seen him, sometimes as a spectral figure.

Today, the term “Luddite” is used mainly as a pejorative, to denote someone opposed to technology and resistant to technological innovation. But the Luddites themselves, by and large, were not anti-technology or anti-innovation. Many, in fact, were skilled machine operators in the mills they attacked. Although they’re often remembered today for smashing the newly developed steam-powered looms then being introduced into mills, giving rise to their later reputation for opposing technological innovation, among the most frequent targets of their attacks were in fact stocking frames, a technology developed during the reign of Queen Elizabeth I more than two centuries earlier.

Instead, their real grievance was with production-obsessed mill owners using machines in what the Luddites called “a fraudulent and deceitful manner” to get around standard labor practices, suppress wages, and churn out cheaper, inferior quality goods to feed new and expanding export-driven markets in the aftermath of the Napoleonic Wars. Their principal demands were for machines that made better quality goods, run by workers who had gone through a traditional apprenticeship and paid a better wage, and an end to child labor.

The steam looms and stocking frames the Luddites destroyed were merely the smashable proxies for a new capitalist, mercantile economic paradigm that characterized the British Industrial Revolution, as both a commercial and political project, and into which they had little input and to which they did not consent. It was an approach to automation that was almost entirely extractive, appropriating value from both inputs and outputs, in which the fruits of the increased productivity it yielded fell disproportionately on one side of the industrial-labor fence.

The intercession of state power in support of that new paradigm only deepened their grievance. Rather than addressing the underlying causes of the unrest, the authorities as much as the mill owner regarded the Luddites as mere vandals and criminals, and responded accordingly.

Today, more than two centuries hence, Luddism is again abroad in the land. The focus of its ire this time is not the steam loom or the stocking frame but the data center. Protests against their construction have broken out nationwide, catching developers and the local officials approving granting their permits and zoning waivers off guard. Like the Luddite attacks of the past, the anti-data center protests have erupted spontaneously in disparate communities, with little if any coordination or clear leadership. According to Data Center Watch, which tracks data center development across the U.S., at least 48 such projects, worth $156 billion, were cancelled or delayed last year due to community opposition.

The Baltimore City Council this month became one or more than 60 local governments to impose a moratorium on data center construction. Last month, the Maine legislature became the first to pass a state-wide moratorium on development of new data centers (although the bill was ultimately vetoed by Gov. Janet Mills). A group of more than 60 organizations, ranging from environmental groups to community organizations to labor unions last week sent a letter to Gov. Mikie Sherrill calling for a state-wide moratorium. At least a dozen other states are considering similar moves. On May 10th, Colleton County South Carolina passed a six month moratorium on data center construction with an option to extend it.

Unlike the Luddite movement, however, anti-data center sentiment is widely shared. According to a recent Gallup poll, seven in 10 Americans are opposed to data center development in the local area, and nearly half (48%) are strongly opposed. The opposition cuts across party lines, regional differences, and generational divides. More Americans support building nuclear power plants in their area than support building data centers.

Much of the opposition stems from the very tangible impacts data centers can have on the communities where they’re located. They use enormous amounts of energy, causing utility prices to spike for consumers and businesses in the area; they have myriad baleful environmental impacts, including on local water quality a producing copious amounts of e-waste; they generate noise at a decibel level and frequency range many find noxious; and they undermine nearby property values.

But the ferocity of the data-center backlash points to grievances that go deeper than higher electric bills. Like the steam looms and stocking frames that took the brunt of the Luddite’s ire, the data centers now being rushed into development around the country, often with little input or consent from the community, have become physical proxies for disruptive new economic and political forces unleased by the AI Revolution.

Around the country this spring commencement speakers have been showered with boos at the mere mention of AI, as graduates face grim employment prospects as AI automates away many entry-level jobs and companies lay off middle level workers by the thousands. A recent New York Times-Siena poll found that more than twice as many voters feel AI is “mostly bad” than feel it is “mostly good.” That same poll found that nearly half (47%) of voters under age 30 have a mostly negative opinion of AI, the highest share of any age group, suggesting the most tech-savvy generation is the most resistant.

Tech companies that have sunk hundreds of billions of dollars into building AI infrastructure have been reduced to forcing AI tools into everything from operating systems to office productivity tools to search bars to create the illusion of adoption in the face of customer resistance. The privacy-focused search engine Duck Duck Go, which promotes its lack of AI features, has seen a 30% surge in installs since Google announced an AI-centric overhaul of its core search platform last week, prompting long-time users to flee.

Just as empire and industrialization were inextricably linked in 19th Century Britain, moreover, the current American regime’s techno-imperialist ambitions have aligned state power with the interests of AI makers. The Trump administration has used tariffs and sanctions to try to coerce European countries into altering their digital rules to make them more accommodating to U.S. technology companies. Domestically, he has sought to use federal power to coerce state and local governments into relaxing or repealing laws regulating AI systems and companies.

No less a figure than Pope Leo XIV made the connection between the two eras explicit in his first encyclical since becoming pontiff. His 42,000-word pastoral treatise, Magnifica Humanitas, denounces the “culture of power” driving the AI race and warns against “opaque algorithms,” controlled by “a few” companies.

In his Encyclical Laudato Si’, Pope Francis denounced the growing dominance of a technocratic paradigm in our globalized world: the tendency to let the logic of efficiency, control and profit alone shape personal, social and economic decisions. This makes it clear that technology is not simply a tool. When it becomes the standard by which everything is judged, it begins to dictate what matters and what can be discarded, reducing creation to an object of exploitation and human beings to mere cogs in a system driven toward ever greater efficiency.

For emphasis, Leo signed his encyclical on May 19th, the 135th anniversary of his namesake Leo XIII’s encyclical Rerum Novarum, which called for fair wages and workers’ rights in response to the Industrial Revolution.

For many in the creative community, of course, the AI revolution has always felt extractive and coercive. AI models extract value from all that they ingest in training, generally with neither compensation or consent of whoever may have created it, and return mediocrity at best, to say nothing of the more pernicious and abusive output of which they’re capable.

The question for creators is whether they can effectively channel the broader cultural anxiety over AI into policies and practices that materially improve their lot.

While the Industrial Revolution did eventually create enormous wealth and lifted living standards broadly, the Luddite project was a failure in its own time, undone by a combination of self-defeating tactics and official repression. It did little immediately to improve the material circumstances of the Luddites themselves or of the average laborer, and the Industrial Revolution rolled on.

Whether the AI Revolution will have the same world-changing impact as its predecessor is yet to be seen. While the productivity gains of the Industrial Revolution were fairly obvious and concrete early on, at least for the industrialists, companies that have adopted AI, by and large, are yet to see meaningful productivity gains from the deployment. According to a recent study by the National Bureau of Economic Research, 89% of companies across four countries surveyed reported no productivity gains from AI over the last three years.

With the initial deployments teetering on the edge of the hype cycle’s Trough of Disillusionment, and the cultural backlash to the AI juggernaut’s coercive and disruptive ethos hardening, the moment may be ripe for meaningful change to the prevailing balance of equities of the AI Revolution, and to avoid the fate of the Luddites.

Share