“AI” isn't a runaway god or a doomsday machine. It's dead labor that has been resurrected by capital. The real question is, who does it serve? I found some striking insights from Marx—or thought I did—but I’m not an expert in that area. Fortunately, I know a leading expert: my friend and collaborator Professor Richard Wolff. My questions and thoughts led to this rewarding conversation.
You’ll find selected quotes and a transcript below. Before then, however, one more thing: In the video I quoted a speech given by Ernesto “Che” Guevara in February 1963 in Camagüey, Cuba, but I obtained the quote from a secondary source and used their translation. I wanted to be sure that it conveyed its meaning correctly, and I wanted to understand the context, so I tracked down the original speech and did my own translation (however imperfectly). Here’s my version. Think of it as a first pass; translation is a fine art which demands time and concentration.
We must remember that machines do not in any way compete with workers or enslave them the way they do under capitalism. All workers should see machines as liberators of their force. Machines will serve man when we end the exploitation of man by man. That’s what we seek. We see machines as instruments of liberation for the peasant, giving them time to develop in every way. Our machines should free us to accomplish everything we’re capable of accomplishing. They should help us reach the full extent of our human potential. That’s what we are all fighting for.
The human beings of the future will need hearts that are as pure and simple as the hearts of human beings today. But they will also need the ability to perform great mental abstractions so they can discover whatever new treasures nature may reveal—to all humanity, for the benefit of all humanity.
We have a gigantic task. We must struggle against imperialism, be an example for all the Americas, and carry on a fight to the death with no surrender. At the same time we must develop our own technology with our own engineers, so that we can continue to progress without outside help; so that we can stand on our own two feet ,with our own inventions, our own strength, and our own way of life as a strong and rich country.
This work is only possible thanks to reader support. If you can, please consider a paid subscription through Substack or Patreon. We really do need the support. Many thanks.
SELECTED QUOTES
Richard Eskow:
“All online activity is labor.”
“A runaway freight train is not ‘deciding’ to annihilate a town. It is not the super-powered human being of the future. It’s just badly managed technology.”
“We are all laboring in a massive global machine: the data matrix, the data economy.”
“The mythos of AI and of Silicon Valley seems to be an unconscious attempt to refute the fundamental reality of labor-capital relations.”
“I hope we can convene some serious conversations about what it would look like to really view our technology as liberators of human force and spirit rather than as enslavers of them.”
Richard Wolff:
“You’re not going to get out of the high-tech machine anything other than what you have programmed to go in there... there is no magic here.”
“People without machines can live. We know that because for thousands of years they did. Machines without people sit there and gather rust and dust.”
“In a capitalist system, technology is installed if and when and to the extent that it is privately profitable. And if it isn’t, it won’t be.”
“Don’t ever think that the way technology is installed is in any sense neutral or fair or equitable. It isn’t.”
“You have to see through the ownership logic to get to the basic labor logic. Labor produces.”
TRANSCRIPT (lightly edited using—yes, I confess—AI)
Richard Eskow: We’re going to do something a little different today and see how it works out. Everybody nowadays is hearing alarmist talk about AI, and those who follow this show, the Zero Hour Report, or my writing for various magazines may know that I’ve written a lot over the years about technology, about AI, about medical technology, about the intersection of human experience and technology.
It’s a very timely issue right now. I want to write about it some more, and of course I will. But I want to be informed by some inspired, even prophetic thinking from the left, primarily the Marxist tradition. I’m not an expert in that. I read, but Professor Wolff is. So I’ve found a few pieces of commentary that I think are very germane to what’s happening now. He’s kindly agreed to help me with this.
Before we do, I need to set the stage. I’ve been getting asked a lot about the panic talk, the hysteria talk. And this is one of the reasons I found myself struck by some of the Marxist writings, not only of Marx and Engels but of those who followed, Marcuse and others. Because every technological change is a kind of dialogue between human culture and technical machinery.
There’s a lot of thought that’s been put into that over the years. What we’re seeing now, I think, is a dominant billionaire, Silicon Valley culture that has a religion. And that religion is a millennialist belief that there’s going to be a new future where human beings are obsolete. You may remember that Peter Thiel had to hesitate when he was asked in the New York Times whether he wants humanity to succeed, to survive. There was actual hesitation, because they believe that what they’re building with software, with technology, is going to be better than humanity — that it’s going to replace humanity. Many of them believe that despite there being no evidence for it. A child loves a doll and may think the doll is alive. It’s not alive. It’s a doll. There’s a lot of projection going on. That’s part of what we’re seeing in the panic.
The other part of it is, as with so many faiths—let’s say the faith of the Church of England as practiced by the aristocrats of the 16th centur— it just so happens that their deeply felt faith intersects with their own economic interests. That’s what we’re seeing here, in my view, with Silicon Valley billionaires. And a lot of people are getting roped into it, including Steve Bannon on the right and my old boss Bernie Sanders on the left. I think he’s bought into it a little too much too.
I think the reality is that this is dangerous technology, just as a runaway freight train is dangerous. But a runaway freight train isn’t “deciding” to annihilate a town. It isn’t the human being of the future. It’s technology out of control. That’s what we’ve seen in these recent examples. But if you want to believe it’s a superbeing, you tend to believe it’s a superbeing. And if it’s in your financial interest to make everybody else believe that too — isn’t that convenient? Because now, all of a sudden, we’re not talking about the AI bubble. We’re not talking about the fact that they seem to be building excess capacity. We’re not talking about the flaws, the slop, the incorrect data coming out.
These companies were headed off a cliff, I strongly believe, in terms of economic reality — stock valuations, everything else. Now, thanks to their theology, they get to say: hold on a second, we’re building something so amazingly powerful you have to slow us down. Well, isn’t that convenient? It keeps the value of their stock inflated, because it’s going to be the future, just not yet.
So that’s my short version of what’s happening now. But as always, the working people of this planet, this country, this society, are getting shafted. That’s the other piece of it. My final thought, which I’ve said before, is that all online activity is labor. Everything you do online is labor. You send an email: it’s labor, because it’s feeding the large language models they call artificial intelligence. You scroll Facebook, stop here, stop there, comment, watch a video: labor. We are laboring in this massive global machine that is the data matrix and the data economy.
And that’s what I wanted to talk to Professor Wolff about. Rick Wolff, if I made you sit there too long while I rambled on, I’m sorry, but I needed to set the stage.
Richard Wolff: No, this is helpful. Every time we talk about this topic, you help me understand an area I know I have to learn more about than I’ve so far been able to. Very interesting. Let me react briefly to a couple of the things you said.
Richard Eskow: Please.
Richard Wolff: I was struck by the similarity to one point you made: that slowing down is a clever way for them to consolidate. These are companies that, in a very short historical period, have become monstrous, billionaire-creating global colossus enterprises. And now, in my language, I hear them wanting to consolidate their monopoly.
Richard Eskow: Correct.
Richard Wolff: And you have to have the government working with you, because you don’t want the government to break you up into smaller companies, which is how you could handle a monopoly situation. You don’t want that. You don’t want them funding competitors. You don’t want that. So, under the heading of making everybody safer, you can now consolidate so three or four companies that overlap anyway control the field.
I’ll give you a historical model of how that happened. At the turn of the century — the end of the 19th, early 20th century — when agriculture was going through a big consolidation in this country, millions of small farms were disappearing, replaced by what we’ve since learned to call agribusiness, or mega-corporations in agriculture.
We had, understandably, a flood of adulterated food. Desperate farmers, large and small, threatened with extinction, cut corners. They didn’t spend as much time making sure the animals were clean, that the people who plowed the land didn’t defecate where they were putting the seeds, and all the rest of it. We had terrible problems, which, by the way, are recurring now that we’ve cut away all the inspection, like Mr. Elon Musk with his chainsaws. So we’re all getting sick. This morning I read that we can’t buy ice cream anymore because it’s making everybody sick. Walmart, Target, the other big supermarkets, can’t sell certain ice creams because they’re contaminated.
What did the big companies do? They rushed in and endorsed government inspections and government mandates for the machinery needed to make sure the milk is clean, the broccoli is clean, the animals are clean. Why? Because the mandated machinery was much more expensive than any medium or small-sized farmer could ever afford. Only the big agricultural companies could afford it.
So they used the safety argument, and to some extent the genuine safety improvements they made, as a very useful way to consolidate a monopoly. And this smells to me like all of that.
Richard Eskow: Absolutely. They’ve been pursuing antitrust relief. A perfect example.
Richard Wolff: Yeah. And the other thing that came to mind. Before I did economics, my parents pressured me toward the natural sciences, math and physics, and I did a lot of that even into college before I said I don’t want to do this anymore. There was a rule in physics: you can’t get out what you don’t put in. It may appear to you that that’s what’s happening, but that’s because you don’t understand everything that’s going into it because there’s a lot going in that you didn’t put in.
Richard Eskow: Right.
Richard Wolff: It doesn’t mean it isn’t put in. You’re not going to get out of the high-tech machine anything other than what’s been programmed into it, whether you’re conscious of what you put in there or not. That’s another matter. But there is no magic here. You’re not going to get something out of nothing.
That’s crucial to Marx. Marx, early in the 19th century, was battling theories of economics before him that had magical outputs. People looked at agriculture and were overwhelmed. By the time Marx is writing, agriculture is the number one industry everywhere in the world. People spent most of their time cultivating land for vegetables, for meat. And they kept being mystified: you only have so many seeds, but you plant them, and the plants that grow generate many more seeds than the number you started with.
Richard Eskow: Oh.
Richard Wolff: So a surplus is what nature gives you. No, you only get out what you put in. You have to understand that things are being put in. You didn’t put them there, but they’re required, like the sun, like the water —
Richard Eskow: Right.
Richard Wolff: You take those away, guess what? It’s not there. So it depends on what your labor can do. You become not the origin but the handmaiden for nature to do what’s built into it. You’re not violating the laws of thermodynamics, where something happens out of nothing.
Richard Eskow: Right.
Richard Wolff: The AI machines cannot give you anything other than what you put into them.
Richard Eskow: Yes—and what you compelled or persuaded others to put into them. Which gets me to the first quote I wanted to raise, bearing in mind my point that I believe all online activity is labor. More and more, our social as well as our personal and work activities are shaped by AI, directly or indirectly. We get instructions on how to do things. I built something for my cat; it’s a long story, but the instructions were clearly AI-produced, and that shaped the hours I put into it. People are obviously being forced to align themselves with AI at work. The way we interact with social media is increasingly influenced by AI. And I should add: I don’t think it’s intelligence. But since you mentioned the 19th century, Karl Marx strikes me as what Norbert Wiener or some of the 20th-century thinkers would describe as a systems thinker.
Richard Wolff: Mm-hmm.
Richard Eskow: He was able to see the workers in the factory and the machines of the factory as a whole, as an organic entity. And what struck me as a good way to start talking about AI is here. I believe this is from Part 3 of Capital. If not, you’ll probably correct me. Let me try to quote it as best I can.
“Past labor in the automaton” — meaning what I just described — “and the machinery moved by it, steps forward as acting apparently in independence of living labor.”
Think about AI now, acting apparently in independence of living labor.
“It subordinates labor instead of being subordinate to it.”
And then this powerful phrase:
“It is the iron man” — we’d say digital person, talking about AI — “confronting the man of flesh and blood.”
Later in that paragraph he writes:
“Dead labor” — meaning the labor of those who’ve gone before — “has been endowed with movement.”
It’s been reanimated, like Frankenstein.
“And living labor only continues to be present as one of dead labor’s conscious organs” — one of its limbs.
It strikes me that this is increasingly what AI and these platforms are doing with their reward-punishment systems, their little dopamine hits, the other ways they direct our activity. Dead labor — or dead so-called recreation that’s really still labor — has been endowed with movement, and we’re the limbs of this resurrected thing. Does that make sense to you? Does it fit with what Marx had in mind?
Richard Wolff: Absolutely. Marx wanted the working class to understand — and I’ll put it a little more starkly than you — that the capitalist says: I’m going to give you a job. I’m going to put you to work with this raw material, this tool, this equipment, this machine. It’s a system. And the way the machines work, you have to stand over there, move your arms this way, look at that screen while you’re doing it. You’ve become an appendage of the machine. And since it’s the capitalist who presents you with the machine — you didn’t have to bring it to the factory, you didn’t have to buy it, the capitalist took care of it — the capitalist then makes the key move: I’m entitled to a piece of the output because I contributed the machinery, the tools, the equipment, and you’re just a small part of this story. You press a button to animate what I’ve brought.
And Marx says: but where do you think the raw material comes from? Who made the tool? The capitalist? Absolutely not — he doesn’t know the first thing about it. His great achievement was to buy what other workers, dead now, embodied in these machines, produced. Or, as I put it to my students: people without machines can live. We know that, because for thousands of years they did. Machines without people just sit there and gather rust and dust. If they’re not oiled, supervised, repaired, fueled, they don’t do a thing.
Richard Eskow: Right — if no one flips the switch. That’s enlightening, thank you. And the “dead labor endowed with movement” also struck me this way — take a lumber mill. Before there were mills cutting wood into planks and shipping them around the world, there were people doing that by hand. And it’s from watching people do that, long dead now, that investors, capitalists, got the idea to build a machine to do it. So I think of those dead woodworkers as also being part of the dead labor. Is that right?
Richard Wolff: That’s right, absolutely. It’s not only that — it’s how the workers did it that shaped the technology. That human behavior is what was in the minds of whoever developed the first machine, built the way it was built, because of what they saw and heard from the people who did it with their hands. That’s why the technology changes over time — as people learn other skills, they go back and realize, oh, I could make this machine a little differently, we could process more planks, or a different kind of wood. Over time this builds up. But there’s never an excuse for it to justify giving the person who bought the inputs a piece of the output. The output comes from the input — it doesn’t come from whoever owns the input.
Richard Eskow: Right.
Richard Wolff: This whole ownership logic — you have to see through it to get to the basic labor logic. Labor produces. That’s why, for Marx, analyzing the value of things has to go back to the one ultimately scarce item, which is labor. Every community only has so much labor. It has to allocate it. You make the food, the clothing — every society has to distribute the labor. That’s the key question. Not the equipment it has — the labor. Because labor can make the equipment if you need it.
Richard Eskow: And what struck me as you were speaking, Rick, is that — in a way I don’t think they themselves understand — the mythos of AI and Silicon Valley seems to be an unconscious attempt to refute the fundamental reality of labor-capital relations. What I mean is: one of the foundational beliefs is that either now or at some point soon, these so-called artificial intelligences will no longer be designed by us — they’ll be designing themselves. And if they design themselves, bootstrapping themselves out of nonexistence into existence, then they can claim there’s no dead labor involved — it created itself. But of course this is a fantasy. Even when a system can design another program — which they do — that’s not because they’re creating something new. It’s because previous programmers, most of them hired, salaried workers, came up with the methodology for writing programs. And when the AI systems write programs, they don’t execute that methodology very well — that’s a separate, technical question. But the idea that they’re creating something out of nothing is, nice try — there’s labor going into it. Does that make sense to you too?
Richard Wolff: Absolutely. It’s another way of explaining that, in the end, it’s the labor that keeps the whole thing going — which people intuitively understand. Someone has to turn the machine on and off.
Richard Eskow: Right.
Richard Wolff: Oil the machine. Turn it off so you can oil it. You have to do this, you have to do that. It is human beings who have produced this totality. The notion that something can happen independent of these human beings — that’s the Frankenstein model. The notion that something dead can return to life.
Richard Eskow: Right.
Richard Wolff: We kind of all know that can’t be done, at least not yet. I can’t predict the future any more than anyone else. Could there come a time when you could bring someone back who had, in some sense, died? Maybe. But for now, we don’t have that capacity. They certainly didn’t have it when Mary Shelley wrote Frankenstein. But she wants to talk about what Marx wants to talk about: how capitalism, in a peculiar way, endows past labor, dead labor embodied, with power over us.
Richard Eskow: Right.
Richard Wolff: We get together and produce something, and instead of we who made it happen deciding what to do with it, we can’t turn to the dead people who did the inputs, because they’re gone. That happened before. We have to replace it. So we take a part of the output and use it to pay for the inputs, to give the people who make the inputs the means to replace what we used up. Marx is very clear on that.
Richard Eskow: Right.
Richard Wolff: We should take the output, use the part needed to replace the used-up inputs, and give the rest to the people who made it happen. We don’t do that. We take another portion, called surplus value, and give it to people whose only claim is ownership. What does that mean? It means they have the right to withhold from us what the community of workers produced. And that, as Marx would insist, is an outrage against the human community—that a subgroup of it can take into its own hands the fruits of other people’s labor, then threaten to withhold it from production unless they get a cut of the output, which then has to come from the living laborers who would otherwise have received it as wages. Therein lies the exploitation he rails against.
Richard Eskow: And it’s a good segue to the next paragraph, which I was struck by. Marx was writing at a time of transition from the worker as artisan, as independent craftsperson, to the worker as factory unit. Next paragraph:
“The living connection of the whole workshop no longer lies here in cooperation” — meaning among autonomous people.
“Instead, the system of machinery forms a unity” —
and here I’m thinking of AI as an engine of social media that we’re all connected into, a factory that virtually everyone is working in right now.
“The system of machinery forms a unity set in motion by the prime motor and comprising the whole workshop, to which the living workshop is subordinated” —
that means us, insofar as we’re workers. And then:
“Their unity has thus taken on a form which is tangibly autonomous and independent of them.”
In my notes I’ve underlined “autonomous and independent” — “a” and “I,” “AI” — because this so-called AI can act autonomously and independently, but only because it’s a cognitive digital workshop comprising all of us and our collective labor. It’s taken this subordination to a global scale. What do you think?
Richard Wolff: Yep. I love that the concepts translate so well from the original to now. You’re applying Marx’s categories and adjusting them for the technology that’s emerged since Marx died. It takes a different form now. But I agree with you. My instinct is that this is just the latest stage of automation: not the last, not the first, somewhere in the middle. It will continue. And each time a discontinuous jump is made, these same fantasies—for lack of a better term, these Frankenstein fantasies—begin to emerge: we’ve set in motion something independent of ourselves.
Richard Eskow: Right.
Richard Wolff: And instead of us controlling it, it’s going to come back and get us.
Richard Eskow: Right, right. One more quote, if you have time?
Richard Wolff: Sure.
Richard Eskow: This one isn’t from Marx — it’s from Che Guevara, from a speech he gave in 1963. Remember, in the 1960s there was a big scare about automation too.
Richard Wolff: Yeah.
Richard Eskow: And the sociologist David Riesman was warning a few years later about a coming “crisis of leisure” — that people would have too much free time and it would create terrible problems. It hasn’t worked out that way. Che Guevara was speaking to a group of sugarcane workers in Camagüey, Cuba, about automation under a socialist system. This is what he said — a vision of what might be:
“We must remember that machines do not, as in the capitalist system, compete with workers or enslave workers. Workers must view machines as liberators of their force” — liberators, freeing their capabilities—“Machines are placed at the service of workers as soon as the exploitation of man by man” — we’d say person by person now — “stops. And that is what we are striving for. We are trying to turn machines into liberating instruments for peasants, so they will have more time for leisure, more time to study, to develop in every sense, to achieve the most important thing we must fulfill: individuals developed to the full. This is the aim we are all struggling for.”
At some point, I hope we can convene some serious conversations about what it would look like to carry this vision forward — to view technology as a liberator of human force and spirit rather than an enslaver of it. That’s my closing thought. What’s yours?
Richard Wolff: Before I answer — where does that Che quote appear? How do I find it?
Richard Eskow: I’ll find it for you. Havana, Cuba, 1963 — I believe February.
Richard Wolff: He—
Richard Eskow: I’ll get it for you. It’s not a common one — it was in a book, but I double-checked it, so I know it’s real.
Richard Wolff: Yeah, if you could send me whatever citation you have, I’ll track it down.
Richard Eskow: Isn’t it great?
Richard Wolff: My closing comment is a short parable I use in my teaching, and have for years. I’ll use ChatGPT as the example. ChatGPT enables a person to complete several reports, or several emails, in a fraction of the time it used to take — twice as productive. Here’s a simple version: a factory has 200 workers, and the employer pays wages to all 200. He hears about the latest AI, buys it, because it’s advertised as making each of the 200 workers twice as productive as before. If he’s a capitalist, and nothing else changes, here’s what he does: he says to 100 of the 200 workers, you’re fired — here’s your check, don’t come back Monday. I don’t need you, because the remaining 100, with this new tool, can do what it took 200 of you to do before. Same output, same price, same revenue — but now, instead of whatever profit I had after paying 200 workers, I have much more profit, because I no longer pay the wages of the 100 I fired. That’s extra profit for me. And doesn’t that make sense for me to do? I’m grateful to AI — it’s increased my profits.
Then I ask my students: what’s wrong with this story? They have no idea. So I tell them. The employer had a choice. He could have bought the machine and said to the 200 workers: I bought a machine that makes all of you twice as productive. So starting Monday, the work week is 20 hours for each of you. In those 20 hours, you’ll produce what used to take you 40 hours, because you have the new AI I purchased for you. I’ll produce the same amount, sell at the same price, earn the same revenue — but I’ll still pay you your full wages. My profit doesn’t go down, but it doesn’t go up either. Installing AI gave you leisure.
Richard Eskow: Right.
Richard Wolff: It gave you half of every day to write poetry, to spend time with your spouse, your children, your hobbies, your passions. It’s a transformation of your life. And since you’re the majority relative to me, the employer, I’m giving an immense benefit to the majority rather than to myself in enhanced profit. But — and I glare at my students here — don’t you ever think that the way technology is installed is in any sense neutral, fair, or equitable. It isn’t. In a capitalist system, technology is installed if, when, and to the extent that it’s privately profitable. If it isn’t, it won’t be. All the opportunities to benefit the way Che Guevara described are excluded from the conversation.
Richard Eskow: Absolutely. And of course, the design of the technology itself is shaped by these same economic imperatives. With that, I really enjoyed this discussion. Thank you for helping me clarify my thoughts on this, Richard Wolff — and as always, thanks for coming on the program.
Richard Wolff: Thank you for pushing this as a topic. It’s one I know I have to learn from, and you’re a good teacher. This has been a good experience for me. Thank you.
