Reverse centaur: how AI turns humans into a biological appendage of the machine
Imagine a courier in a van, where the algorithm doesn't just plot a route but totally controls every movement, down to banning singing while driving. This isn't the plot of a dystopia, but the everyday reality of the modern economy. In his new work, "The Reverse Centaur's Guide to Life After AI," writer Cory Doctorow exposes a troubling trend: we are moving from a symbiosis of human and machine toward a form of outright parasitism, where technology subjugates the operator.
Inversion of the Myth: From Symbiosis to Subjugation
The classic "centaur" is a human enhanced by technology. A bicycle, a hearing aid, or autocomplete in an editor are examples where the machine remains the "body" executing the will of the "head." This model, popularized by Garry Kasparov after his famous match with Deep Blue, has proven its effectiveness: the pairing of a grandmaster with AI consistently beats both humans and machines individually.
However, Doctorow describes a frightening inversion of this model. The "reverse centaur" is a mechanical head mounted on a human body. Here, it's not the human using the tool, but technology using the human as a biological appendage, working at an inhuman, machine-like pace. The question shifts from "can AI help?" to "whom does this AI obey, and in whose interests does it act?"
Where We Encounter Reverse Centaurs
This isn't abstract philosophy, but very concrete sectors of the economy. The most striking example is logistics. Amazon drivers are forced to check in with an algorithm that sets unrealistic quotas, causing them to skip meal and bathroom breaks. In warehouses, it's the same picture: AI distributes tasks and penalizes slowness, turning the human into an appendage of the conveyor belt.
An equally troubling situation is developing in intellectual professions. Developers using AI assistants like Copilot risk turning from code creators into code verifiers. They no longer control the process but merely check and correct the machine-generated result, while the pace is set by the algorithm's productivity, not human capabilities.
Particularly cynical is the "human-in-the-loop" system, common in law and finance. A specialist is forced to process vast amounts of content at "superhuman speed," while bearing full legal and financial responsibility for AI errors. In essence, they "absorb the friction" and take on all the risk, stripped of the creative component and under constant stress.
The Economics of Exploitation and the AI Bubble
Doctorow links the spread of this model to simple "fun arithmetic" for the employer: replace ten employees with three armed with AI. The savings are achieved not through productivity growth, but by redistributing the workload and shifting risks onto the remaining workers. The executives making decisions about technology implementation never become reverse centaurs themselves—for them, AI remains a management tool.
Significantly, the writer estimates the current AI boom at $1.4 trillion, calling it a bubble that will inevitably burst. The seven largest companies make up more than a third of the stock market, but their models are held up by mutual expectations, not real profits. After the crash, according to Doctorow, only a "useful residue" will remain—cheap local models for routine tasks.
My analysis: Doctorow's observations accurately describe what I see in the labor market. The problem isn't AI itself, but the asymmetry of power in its implementation. As long as decisions are made by a narrow circle of people who bear no responsibility for the outcome, we will witness not the liberation of labor, but its intensification. The key paradox the author exposes is that the cost of verifying AI results is often comparable to the cost of doing the task independently. The economic effect arises only when the employer shifts these costs onto the employee's shoulders, forcing them to work more for the same money. It is at that moment that the reverse centaur becomes not just a metaphor, but a source of superprofits, and this is the main challenge for regulators and unions in the coming years.