Cory Doctorow says the public argument over AI has become an everything argument: labor, public procurement, data centers, customer service, art, medicine, climate, existential risk and paperclip theology, all shoved into the same tiny broadcast segment.
In a June 6 essay on Pluralistic, Doctorow framed that overload as a version of the “Gish Gallop,” the debate tactic named for creationist Duane Gish in which one side throws out more claims than an opponent can answer in the available time. His immediate example was a radio producer who asked him to discuss “AI governance” in a 13-minute slot, then treated a long list of unrelated policy questions as one topic.
Doctorow said the same compression followed him into print, when a newspaper editor asked for an 800-word column that would also address claims that AI is the next Industrial Revolution. His complaint is not that AI raises no policy issues. It is that the industry’s claims, and some criticism of those claims, are too sprawling to evaluate honestly at that speed.
The business question comes first
Doctorow’s main objection is economic. He argues that debates about future AI capability should start with whether the companies building and running the systems can keep paying for the servers, power, chips and staffing required to operate them.
He describes current AI services as heavily subsidized by their providers, with customers liking the product while prices stay far below the cost of delivering it. In his telling, when providers try to raise prices, some customers discover that the tools no longer make financial sense for them.
That is Doctorow’s assessment, and it is doing a lot of work. He points readers to earlier writing of his on AI unit economics and to a market report about S&P Global’s fast-entry proposal as SpaceX listing speculation loomed. He also argues that Elon Musk’s abandoned plan involving Grok and Twitter represented an attempt to shift losses onto a broader class of investors.
The mechanism he wants readers to focus on is plain: if an AI company spends more to produce outputs than customers will pay for them, grand predictions about future capability depend on somebody continuing to absorb the losses. That somebody could be venture investors, public markets, corporate customers, governments or, in the worst version of the story, retirement savers and taxpayers.
Doctorow warns against accepting hype as a premise
Doctorow also criticizes what he calls “criti-hype,” a style of AI criticism that accepts promotional claims as likely and then panics about the consequences. Examples he gives include arguments built around AI taking all jobs, replacing teachers, doctors or governments, or bypassing human judgment at scale.
His counterargument is that many of those scenarios require assuming both technical progress and durable funding. He applies the same test to claims that AI will cure cancer or solve climate change: how much money, energy and carbon emissions would be consumed before any promised breakthrough, and how does that compare with funding other research that already lacks money?
Doctorow connects the essay to his forthcoming book, The Reverse Centaur’s Guide to Life After AI, which Macmillan lists for June 2026. He says he wrote it after repeated demands to answer broad AI claims in formats too short to separate procurement policy from labor markets, data-center siting and speculative machine superintelligence.
He is also taking the book on the road, with listed events in Kansas City, Los Angeles, Menlo Park, New York, Philadelphia, Chicago, Edinburgh and South Bend. The tour pitch is literary. The argument underneath is regulatory and financial: before asking what AI might become, Doctorow wants institutions to ask who is paying the bill, for how long, and under what claims.
This story draws on original reporting from Pluralistic.
