Cory Doctorow delivers a blistering correction to the prevailing narrative that artificial intelligence is simply the next inevitable wave of productivity, arguing instead that we are witnessing a top-down imposition of a technology that actively loses money with every use. While the tech press obsesses over capability benchmarks, Doctorow exposes a fundamental economic inversion: unlike the early web, which workers clamored to adopt to get their jobs done, AI is being forced upon a resistant workforce by executives desperate to justify massive capital expenditures. This is not a story of innovation; it is a story of coercion, surveillance, and a bubble built on "dogshit unit economics."
The Brokered Peace of the Early Web
Doctorow anchors his argument in a historical comparison that many in the tech industry have conveniently forgotten. He points to Lotus Notes, the clunky precursor to modern office suites, not as a triumph of user experience, but as a "brokered peace" between IT managers and the workers they tried to control. "No one had to force-feed the web to workers," Doctorow writes, highlighting the organic demand that drove the internet's adoption. Workers historically "smuggled" tools like Hotmail or Visicalc into the enterprise because they needed to bypass arbitrary restrictions to get their work done.
The author reframes the conflict not as a security failure, but as a clash between rigid policy and the reality of labor. He notes that IT departments were asked to do the impossible: enable workers to be effective while simultaneously preventing them from taking any action that could harm the business. "The only way to eliminate the possibility that a worker will take an action that harms the business is to gag that worker and lock them in a dungeon," Doctorow observes. This stark imagery underscores the futility of trying to micromanage human ingenuity. The early web succeeded because it aligned with the worker's desire for agency, whereas the current AI push aligns only with the executive's desire for control.
Workers need flexibility and freedom to achieve business goals, and that flexibility and freedom means that those workers might (deliberately or accidentally) thwart the business's goals.
This historical lens is crucial because it reveals that the current resistance to AI is not "technophobia," as some executives claim, but a rational response to a tool that offers no immediate utility to the user. Critics might argue that some workers do adopt AI voluntarily to automate dull tasks, and Doctorow concedes this point, noting that some see value in offloading "bullshit jobs." However, he quickly pivots to the broader trend: "Virtually every major company now has a program to force workers into using AI." The distinction is vital; voluntary adoption signals a useful tool, while forced adoption signals a political mandate.
The Economics of Coercion
The most damning part of Doctorow's analysis lies in the financial reality of the AI boom. He dismantles the comparison between the dot-com bubble and the current AI frenzy by focusing on unit economics. In the early days of the web, "every new web user brought the web closer to profitability." In contrast, Doctorow cites Ed Zitron's assessment that AI has "dogshit unit economics," where "every new AI user makes AI less profitable."
This economic inversion explains the behavior of the executive branch and corporate leadership. Because the technology is a money-losing endeavor, management must compel usage to generate the data and activity required to sustain the valuation of these companies. "AI is the money-losingest endeavor in human history," Doctorow states, a claim that reframes the entire industry's aggressive marketing as a desperate attempt to hide a loss-leader strategy. The pressure is no longer on the technology to prove its worth to the worker; it is on the worker to prove their worth to the technology.
It's impossible to overstate how important Lotus Notes and similar products were, because workers demanded the right to use the web on their work computers, and they made those demands so forcefully that managers had to completely re-do their IT policies.
Doctorow connects this to the concept of "work-to-rule," a form of labor action where employees follow every rule to the letter, grinding operations to a halt. He warns that if workers are denied agency and forced to use tools they don't trust, they may resort to similar tactics, treating AI mandates as "damage and route around them." The irony is palpable: the administration and corporate boards are trying to solve a productivity problem by introducing a tool that requires surveillance to enforce, potentially triggering the very inefficiencies they seek to eliminate.
The author also touches on the generational divide, noting that while young people were once the vanguard of web adoption, they now "hate AI." This reversal is telling. It suggests that the technology is not solving a problem for the next generation of workers, but is instead being imposed as a mechanism of surveillance and throughput maximization. "Labor-led automation produces improvements in quality, while capital-driven automation increases throughput (often at the expense of quality)," Doctorow writes. This distinction separates tools that empower workers from tools that merely extract more labor for less pay.
The Unnecessariat and the Future of Work
Doctorow concludes by warning of the rise of the "unnecessariat," a class of workers who are becoming superfluous not because they lack skill, but because the technology is being used to replace them rather than assist them. He distinguishes between "centaurs" (workers assisted by technology) and "reverse-centaurs" (workers recruited to serve as peripherals for machines). The current push for AI, he argues, is creating the latter.
What the technology does is nowhere near as important as who the tech does it for and who the tech does it to.
This framing shifts the debate from technical capability to power dynamics. The article suggests that the AI bubble is not a market correction waiting to happen, but a structural failure of the technology's alignment with human needs. While some may argue that AI will eventually become profitable and useful, Doctorow's evidence suggests that the current model is fundamentally broken. The reliance on forced adoption and surveillance indicates a lack of organic value, a sign that the bubble is not just inflated, but hollow.
Bottom Line
Doctorow's argument is a necessary corrective to the hype, successfully demonstrating that the AI boom is driven by capital's need to justify losses rather than labor's need for better tools. The strongest part of his case is the historical contrast with the early web, which proves that truly transformative technology is adopted from the bottom up, not forced from the top down. However, the piece's biggest vulnerability is its reliance on the assumption that management will not eventually find a way to make the economics work, even if it means further eroding worker autonomy. The reader should watch for the inevitable clash between these forced AI mandates and the growing resistance of a workforce that sees the technology as a threat rather than a tool.