Reverse centaur: how AI turns humans into a biological appendage of the machine

Imagine a truck driver where algorithms don't just plot the route but control every move, down to banning singing behind the wheel. This isn't a futuristic scenario but the harsh reality of logistics giants. Here, it's no longer the human driving the vehicle—it's the system driving the human, turning them into a "biological supplement" for an as-yet-imperfect artificial intelligence.
In his new work on life after the total implementation of AI, writer and technology critic Cory Doctorow introduces the term "the inverse centaur problem." This concept isn't just a philosophical metaphor but a precise diagnosis of the current economic model, where technology doesn't enhance humans but subordinates them.
The Inverse Centaur: Mechanics of Reversed Symbiosis
Doctorow draws a clear line between two models of interaction. The classic "centaur" is a symbiosis where the human remains the "head" making decisions, while the machine serves as the "body" executing commands. This is the ideal of automation: a bicycle, a hearing aid, or autocomplete in an editor—all are tools that expand human capabilities.
The "inverse centaur" is a complete inversion. It's a "mechanical head on a human body," where the human's role is reduced to that of an executive organ. The system dictates the pace, volume, and methodology, while the human merely provides physical presence for operations the machine can't yet perform on its own. The term is borrowed from Garry Kasparov, who proposed the "advanced chess" format after his defeat by Deep Blue in 1997. However, as it turned out, the symbiosis that beat pure AI often becomes a tool of exploitation in the real economy.
Where We Encounter "Inverse Centaurs"
The reality is that we're surrounded by these hybrids. The most striking examples are in areas where algorithms have already taken over managerial functions:
- Logistics and Warehousing: Amazon drivers work under the watch of AI cameras that fine them for singing or taking extra pauses. The algorithm sets an inhuman rhythm, demanding "superhuman endurance," and the human merely executes its commands, stripped of the right to a break.
- Software Development: The adoption of AI assistants like Copilot has led to programmers being forced to produce impossible volumes of code. The developer turns from a creator into a controller who merely checks and fixes what the machine generated, with the pace set by AI productivity rather than human capability.
- The "Human-in-the-Loop" Sphere: The most insidious format. In law, medicine, and finance, AI prepares draft decisions, and the expert is obliged to verify them at "superhuman speed." Meanwhile, all responsibility for algorithmic errors falls on the human, who becomes a hostage to the system's unpredictability.
- Media: Newsrooms cut staff but not workload. Three journalists with AI are forced to do the work of ten. They no longer write but filter and edit the "stream of plausible verbiage" spewed out by chatbots.
The Economics of Paradox and the AI Bubble
The key question is why this happens. The answer lies in the employer's "funny arithmetic." Replacing ten employees with three using AI tools means redistributing the workload and shifting risks onto those who remain. Savings are achieved not through increased efficiency but through labor intensification and free verification of machine errors.
Doctorow links this to the global financial bubble around AI, which he estimates at $1.4 trillion. The seven largest companies make up more than a third of the stock market, but their models rest on mutual expectations. This bubble will inevitably burst, but until that moment, corporations have an incentive to squeeze maximum value out of "inverse centaurs" to maintain the appearance of profitability.
An Analyst's Perspective
Doctorow's concept isn't just a literary metaphor but an important indicator for investors. When I look at reports of productivity growth in companies adopting AI, I always ask: was this growth achieved through technology or through increased strain on personnel? If we see a model where AI doesn't replace routine but merely speeds up the conveyor belt, demanding humans work at the machine's pace—that's not progress but a new round of exploitation. The true value of technology will only emerge when we stop using people as "insurance" for imperfect algorithms and restore their role as the "head" rather than the "body." The future is indeed ours, but only if we consciously choose the "centaur" model rather than its inversion.