I have been thinking lately about what capitalism actually needs us for.
For much of the last century there were some obvious answers. It needed our labor. It needed us to buy things. The state needed our taxes and, at least nominally, our political consent. None of this made working people powerful by itself, but it created dependencies. Organized labor figured out how to turn one of those dependencies into political power.
And Butte is an obvious place to think about that. The Company needed copper, and it could not get copper without miners. Thousands of people had to go underground, drill and blast rock, move ore, operate hoists and smelters, and keep an enormous industrial system running. When miners stopped working, copper stopped moving. That material fact gave working people leverage, and unions gave them a way to consolidate and use it.
This is part of what made organized labor such a remarkable political technology. The workplace concentrated people who individually had little power and placed them somewhere the system depended on them. A strike mattered because something important stopped. Butte became the Gibraltar of Unionism not simply because its miners had radical politics. The mine needed them.
That history makes me think about the robot dog up at Montana Tech. SPOT is being taught to enter abandoned mines, navigate underground, and search for critical minerals. I like the project. If a machine can enter an unstable mine instead of a person, it should. The same is true of the larger movement toward autonomous mining. Fewer people exposed to ground failures, explosives, dust, diesel exhaust, and massive equipment is progress.
But there is a contradiction here that I cannot quite get around. We have spent generations imagining technology freeing people from dangerous, exhausting, or unnecessary work. We should want that. Yet we built a political economy in which access to food, housing, health care, security, and much of our political influence remains tied to being useful to production. Technology can free the miner from the mine while simultaneously freeing the mine from the miner.
The current rush for critical minerals makes this especially visible in Butte, but mining is only one local example of something much larger. AI and automation are reaching into administrative work, software, logistics, manufacturing, transportation, science, writing, design, and many other forms of labor. I do not know how far or how quickly this will go, and the argument does not require believing every job will disappear. The more important question is what happens if increasingly valuable systems simply require fewer people.
There is an older concept that helps here: the resource curse. Countries rich in oil, minerals, or other extractive wealth can become less accountable to their populations because wealth can flow from the ground without depending as heavily on broad taxation or broad participation in the economy. Empire has long operated through a similar relationship, valuing a territory for what can be extracted from it while treating the people living there as secondary, politically marginal, or simply in the way. Butte has always lived close to this paradox. The ground can be enormously valuable while the people living above it are treated as secondary to extraction. The mine matters more than the miner.
A newer analogy, sometimes called the intelligence curse, asks whether AI could reproduce this relationship on a much larger scale. What happens if enormous wealth can be generated with less and less human labor? The population may remain subject to the system without being equally necessary to it. Workers lose some of the leverage that came from being indispensable to production even as their dependence on capital and infrastructure continues.
David Graeber spent a lot of time questioning our strange devotion to work. Technological advances should have reduced the amount of compulsory labor needed to sustain a society, yet many of us remain occupied with work that even the people doing it struggle to justify. I do not want to preserve bad jobs just so workers can retain political power. But Graeber's work on debt points toward another possibility. Wage labor is only one form of dependence. We could become less necessary as workers while remaining very useful as renters, mortgage holders, medical debtors, utility customers, insurance customers, and platform users.
Less necessary as producers, but still valuable as debtors.
Yanis Varoufakis's idea of technofeudalism is useful here even if the label overstates things. It shifts attention from production toward ownership, rent, access, and enclosure. More and more wealth flows to those who control the platforms and infrastructure everyone else must pass through. There is an odd weakness in the feudal analogy, though. A feudal lord still needed peasants to work the land. What happens when the owner of productive infrastructure needs fewer and fewer people at all?
There is something especially strange about seeing all of this through technologies we describe as immaterial. We talk about "the cloud" as though computing has escaped physical reality. It has not. AI requires data centers, electrical generation, transmission lines, water, concrete, semiconductors, fiber, copper, rare earth elements, cooling systems, roads, technicians, and mines.
The cloud touches the ground, and increasingly the mine may not need the miner.
That is a bleak picture, but I do not think the answer is nostalgia for the industrial economy. The labor movement offers a much more useful lesson. Working people found a place where powerful institutions depended on them and organized around that dependence. The question is where that kind of power can be built next.
The answer probably will not be one institution or one traditional workplace. A tenant has little leverage over a large landlord, but hundreds of tenants have a different relationship to that owner. Debtors experience debt individually, but creditors depend on widespread repayment. A data center may automate enormous amounts of work, but it still depends on electricity, water, land, transmission infrastructure, maintenance, and political permission. A mine may need fewer miners, but it still needs roads and power, occupies a watershed, produces waste, and exists somewhere among people and other living things.
None of these dependencies automatically creates power. Neither did working in a mine. People must recognize their shared position and organize around it.
That brings me back to another part of Butte's history. The union was not powerful simply because the Company needed labor. Workers also needed relationships strong enough to act together. A strike required trust. People had to refuse to cross the line. Families had to survive without wages. Information had to circulate. Butte had union halls, neighborhoods, churches, ethnic societies, lodges, clubs, bars, newspapers, families, and countless informal networks. The Company may have controlled much of the formal economy, but it did not control every relationship through which people organized their lives.
Robert Putnam calls some of this social capital. The term sounds a little bloodless compared with the thing itself, but the insight matters. Networks of trust and reciprocity give people capacities they do not possess alone. His distinction between bonding within groups and bridging across groups matters too. Strong communities can sustain solidarity, but they can also become suspicious, exclusionary, and rigid. The relationships we build need to be strong enough to support people without turning solidarity into conformity.
A line from Joyful Militancy has been echoing in my head: "Recovering and sustaining deeper forms of friendship and kinship are indispensable for undoing empire's hold."
Friendship sounds awfully small next to artificial intelligence, autonomous mines, global capital, and planetary supply chains. Maybe that is because we have been trained to treat industrial infrastructure as real and social infrastructure as sentimental. Industrial infrastructure moves copper, electricity, information, and money. Social infrastructure moves trust, care, knowledge, obligation, and solidarity. The labor movement needed both.
Scott Crow makes a related point in Joyful Militancy: political practice cannot become rigid around what it thinks resistance is supposed to look like. We need to cast off some of our ideas about what it "should" be and help make it what it "could" be. There is no reason the forms of collective power built around industrial labor should be the only ones we can imagine after the economy that created them has changed.
That does not mean abandoning unions. Wherever capital still depends on workers, organizing labor remains one of the clearest ways to turn dependence into leverage. But it may mean learning to organize other relationships too: tenants with landlords, debtors with creditors, communities with utilities, watersheds with extractive industries, and people with the digital and physical infrastructure on which concentrated wealth depends.
Two strategies emerge. We can identify where powerful systems still depend on us and organize around those dependencies. At the same time, we can reduce our own dependence on systems capable of disciplining us through scarcity and access. Mutual aid, cooperatives, shared tools, repair skills, public goods, community institutions, friendships, and networks of care will not replace organized labor or defeat global capital by themselves. But they can make people harder to isolate, harder to coerce, and more capable of acting together when leverage appears.
The miners of old Butte knew where they stood in the productive system. They knew what the Company needed from them and what stopped when they stopped. Just as importantly, they inhabited a social world that made acting together possible.
The mine of the future may not need many miners. AI may make many other forms of labor less important as well. The challenge is not to recreate the political economy of a century ago so working people can recover the leverage they once had. It is to understand where the new system depends on us, reduce the ways it makes us dependent on it, and build the relationships that let us act together.
If the old fulcrums of working-class leverage are disappearing, we need to find the new ones.