Table of Contents
There is something almost comedic about imagining Karl Marx hunched over a laptop, typing prompts into ChatGPT. The man who spent decades in the British Museum scribbling notes about the exploitation of factory workers would now be asking a machine to summarize his own theories. But the question is not really about Marx the person. It is about whether the ideas he left behind, dusty and debated as they are, have anything useful to say about the moment we are living through right now.
Because something strange is happening. For the first time in history, machines are not just replacing muscle. They are replacing thought. And that changes everything.
The Ghost in the Machine
Marx was obsessed with machines. Not because he was a technophile, but because he understood that every major shift in technology rearranges who has power and who does not. The spinning jenny did not just make cloth faster. It made skilled weavers irrelevant. The steam engine did not just move trains. It moved entire populations from farms to factories. Technology, for Marx, was never neutral. It was always a weapon in an economic war most people did not even know they were fighting.
His core insight was straightforward. When you introduce a machine that can do what a worker does, the worker loses bargaining power. The owner of the machine gains it. This was true in 1850. It was true in 1950. And it is very much true in 2026, except now the machine writes poetry, drafts legal contracts, and diagnoses skin conditions from a photograph.
What makes this moment different from every previous wave of automation is scope. The industrial revolution displaced physical labor. The digital revolution displaced routine cognitive labor. The AI revolution is displacing creative and analytical labor. We have run out of categories of human work that are obviously safe.
The Productivity Paradox Nobody Talks About
Here is where things get interesting, and where Marx becomes unexpectedly relevant to a conversation most economists are having without him.
Productivity has been the holy grail of economic thinking for centuries. Make more stuff with fewer inputs. Grow the pie. Everyone benefits, or so the theory goes. AI promises the greatest productivity leap in human history. A single person with the right AI tools can now do work that would have required a team of ten just five years ago. Graphic design, market research, software development, financial analysis. The list grows weekly.
But Marx would ask a question that rarely gets asked in Silicon Valley boardrooms: productivity for whom?
This is not a rhetorical flourish. It is the central economic question of our time. When a company uses AI to replace half its customer service team, productivity goes up. Revenue per employee goes up. Stock price goes up. But the fired workers do not experience a productivity miracle. They experience unemployment.
The standard rebuttal is that new technologies always create new jobs. The automobile killed the horse carriage industry but created mechanics, gas station attendants, highway engineers, and suburban real estate agents. This is historically true. But it contains a hidden assumption that is starting to crack: the assumption that the new jobs will arrive fast enough, pay well enough, and be accessible to the people who lost the old ones.
Marx had a term for what happens in the gap. He called it immiseration. Not a fun word, but the concept is brutally simple. Technological progress under capitalism tends to concentrate wealth among those who own the technology while pushing down wages and security for those who compete with it. The pie gets bigger, but your slice gets smaller.
Look at the data from the last forty years and tell me he was entirely wrong. Productivity in the United States has roughly doubled since 1980. Median wages have barely moved. The gains went somewhere. They just did not go to the people doing the work.
The Alienation Update
One of Marx’s most famous ideas is alienation, the sense that work under capitalism becomes meaningless because workers are disconnected from the products they make, the process of making them, and ultimately from themselves. A factory worker bolting the same part onto the same widget eight hours a day does not experience work as fulfilling. They experience it as something that happens to them in exchange for money.
Now consider the modern knowledge worker using AI. A copywriter who used to craft language for a living now prompts an AI and edits its output. A junior developer who used to write code now reviews code that a machine generated. A graphic designer who used to create visual concepts now curates outputs from image generators.
The work still gets done. Arguably it gets done faster and sometimes even better. But something has shifted in the relationship between the worker and the work. You are no longer the creator. You are the supervisor of a machine that creates. And there is a quiet existential crisis hiding inside that shift that we have not yet fully reckoned with.
This is not just philosophical navel gazing. It has real economic consequences. When your job becomes supervising AI output rather than producing something yourself, your value proposition changes. You become easier to replace, because the skill of prompting and reviewing is more generic than the skill of creating from scratch. The barrier to entry drops. And when barriers to entry drop, so do wages. Marx would nod knowingly.
There is a strange parallel here with the world of professional chess. When Deep Blue beat Garry Kasparov in 1997, many predicted the death of competitive chess. Instead, something unexpected happened. Chess became more popular than ever, partly because AI analysis tools made the game more accessible and the analysis richer. But the economics of being a professional chess player changed dramatically. When anyone with a laptop can access superhuman chess analysis, the premium on human expertise shrinks. The game thrives. The professionals struggle. This is a pattern we should pay attention to.
The Ownership Question
If Marx were alive and scrolling through tech news, the thing that would make him pound the table is not AI itself. It is who owns it.
The means of production, in Marxist terms, used to be factories, land, and raw materials. Today the means of production are algorithms, training data, and compute power. And they are owned by a remarkably small number of companies. OpenAI, Google, Meta, Anthropic, and a handful of others control the foundational AI models that an increasing portion of the global economy depends on.
This is a concentration of productive power that would make the nineteenth century robber barons blush. At least when Carnegie owned the steel mills, you could theoretically build a competing steel mill. Building a competing large language model requires billions of dollars, access to massive datasets, and specialized hardware that is itself controlled by a near monopoly. The barriers to entry are not just high. They are atmospheric.
Marx argued that the fundamental conflict in capitalism is between those who own the means of production and those who must sell their labor to survive. AI is turbocharging this dynamic. The companies that own the models capture an enormous share of the value they generate. The workers who use those models to do their jobs are in an increasingly precarious position, because the same tool that makes them productive today could make them redundant tomorrow.
And here is the part that would really fascinate Marx. Many of the people being displaced are not blue collar factory workers. They are the educated middle class, the lawyers, accountants, writers, analysts, and programmers who were told that education was the ticket to economic security. The professional class is discovering what the working class learned two centuries ago: your skills are only valuable until a cheaper alternative comes along.
What Would Marx Actually Do With ChatGPT?
If we are being honest, Marx would probably use ChatGPT extensively. He was a relentless researcher who spent years compiling data about factory conditions, trade statistics, and economic theory. An AI that could summarize parliamentary reports and cross reference economic data would have saved him decades of work in the British Museum.
But he would also see it for what it is: a tool whose social impact depends entirely on who controls it and who benefits from it. A hammer can build a house or break a skull. AI can democratize knowledge or concentrate power. The technology does not decide. The economic system it operates within decides.
And that is the part of the conversation we keep avoiding. We talk endlessly about what AI can do. We talk very little about what it should do, and even less about who should decide. The technical capabilities advance at breakneck speed. The social, legal, and economic frameworks that are supposed to govern those capabilities move at the speed of committee meetings and legislative sessions.
Marx would see this gap between technological capacity and social organization as the defining crisis of the era. He would argue that we are building tools of incredible power within a system that distributes the benefits of that power according to the logic of capital accumulation rather than human need. And he would say, with considerable historical evidence on his side, that this mismatch tends to produce instability.
The Question We Are Actually Asking
The real question behind “Would Marx use ChatGPT?” is not about one dead philosopher and one chatbot. It is about whether we can manage the most powerful technology ever created without repeating the patterns of every previous technological revolution, where the gains were captured by the few and the costs were borne by the many.
History says we probably will repeat those patterns. The optimists say this time is different. Marx would say the optimists always say that.
But here is the genuinely interesting thing. The fact that ordinary people are debating the economic implications of AI, asking who benefits and who loses, questioning the distribution of technological gains. That is itself a form of the consciousness Marx thought was necessary for change.
Whether that consciousness leads anywhere depends on choices that have not been made yet. Policy choices, institutional choices, and individual choices about what kind of society we want to build with these extraordinary tools.
Marx would use ChatGPT. But he would want to know who owns the server it runs on. And he would want to know why you do not.
