Reverse centaur: how artificial intelligence turns humans into a biological appendage

Imagine a truck driver where the algorithm doesn't just plot the route but controls every breath, fining for a careless gesture or even singing behind the wheel. This isn't a dystopian plot but an everyday reality that author and futurist Cory Doctorow calls the "inverted centaur problem." In his new book, he examines how technologies meant to augment humans turn them into subordinate elements of the machine.
Inversion of Symbiosis: From Kasparov to Algorithmic Servitude
The concept isn't new. The term "centaur" entered common usage after the famous match between Garry Kasparov and Deep Blue, when it became clear that the symbiosis of a grandmaster and AI is always stronger than either of them alone. In this model, the human remains the "head" making strategic decisions, while the machine is the "body" performing routine operations.
However, Doctorow describes a frightening inversion of this model. The "inverted centaur" is a mechanical head on a human body, where the human is reduced to the role of an executor of the algorithm's will. It's a system in which technology doesn't serve the human but uses them as a cheap biological processor for tasks the machine can't yet handle on its own.
Industrial Hell: From Warehouses to Law Firms
The most striking examples of this inversion are in areas where algorithms dictate the pace and rules. Amazon drivers are forced to work at the limit of human capability, skipping breaks to meet AI-set quotas. A similar situation exists in warehouses, where algorithms assign tasks and penalize slowness.
A more alarming trend is "human-in-the-loop." In legal and medical firms, AI generates document drafts, and the specialist merely reviews them at "superhuman speed." Meanwhile, all responsibility for the algorithm's errors falls on the human. They become a verifier, absorbing the system's "friction" and bearing all the risk. This isn't just a loss of autonomy; it's shifting responsibility from the machine to the human with no control whatsoever.
The Bubble Economy and the Price of Verification
Doctorow exposes the economic underpinnings of this phenomenon. Employers benefit from "cheerful arithmetic": replacing ten employees with three armed with AI tools and offloading the workload and risks onto them. Savings are achieved not through productivity growth but through intensifying the labor of those who remain and unpaid overtime.
The author pays special attention to the AI market bubble, valuing it at $1.4 trillion. He predicts this bubble will inevitably burst, leaving behind only cheap local models. However, until that moment, corporations will be interested in maintaining the myth of AI's omnipotence to keep profiting from "inverted centaurs."
My analysis: Doctorow's argument is impeccable from a labor ethics standpoint, but it misses one important aspect. In the software development industry, where I work, tools like Copilot have already become the standard. The key question isn't whether to use AI, but who controls the implementation process and task distribution. If a developer decides for themselves when to use an assistant, that's a "centaur." If management sets impossible KPIs based on AI productivity, that's already an "inverted centaur." The line runs not along a technological but a managerial principle. And it's this line worth fighting for before the bubble bursts.