July 26, 2026, (Inside AI) — When the AI investment frenzy finally collapses, executives who rushed to replace human workers with algorithms will face a harsh reckoning: lost institutional knowledge that takes years to rebuild. That is the central warning from journalist and science fiction author Cory Doctorow, who is bringing his post-bubble survival guide to Australia.
Doctorow, known for coining the term "enshittification" to describe platform decay, argues the question is not if the AI bubble will burst, but when. In his new book, The Reverse Centaur's Guide to Life After AI, he maps out what happens after the hype cycle ends.
"The thing about bubbles is that after they burst, we look back and we say 'well, obviously that was the weak link'," Doctorow said. The current boom, he notes, floats on massive investments from Gulf sovereign wealth funds and billionaires into datacenters. A collapse in those deals could pop the bubble, as could the circular investment loop between chip makers, AI firms, and tech giants. Delayed IPOs from companies like Anthropic and OpenAI add further fragility.
"We can see in retrospect, when we don't have to continue pretending that this thing is going to go on for ever, that it was always very, very brittle. It was always very fraught," he said.
Skills Lost to the 'Four Winds' Won't Return Easily
For businesses that bought the promise of cheap AI labor, the aftermath will be brutal. Fired, retired, or retrained workers take with them deep process knowledge that cannot be simply reactivated. "It's going to be a bit a little like Fury Road where there's just going to be stuff you have, but we don't know how to do again," Doctorow warned. "And it's going to take a really long time to figure out how to do it again."
This echoes research on organizational forgetting, where lost tacit knowledge cripples productivity for years. A 2023 study in Organization Science found that firms replacing experienced staff with automated systems saw a 23% drop in problem-solving capability during disruptions. Doctorow's narrative turns that data into a stark human story.
Copyright Laws Won't Save Creative Workers
Doctorow's visit comes just a week after Australian Prime Minister Anthony Albanese promised "plain as day" copyright laws to force AI companies to pay for training data. The creative sector cheered, but Doctorow is skeptical. He says relying on copyright "won't solve anyone's problems"—a stance that pits him against both AI firms and large media conglomerates.
"It's not like the media companies suing the AI companies want to pay media professionals, right? Like, no one has ever accused Rupert Murdoch of feeling that journalists don't get enough money," he said. "The only other constituency, apart from a few creative workers, for a new copyright is our bosses."
Instead, Doctorow advocates for stronger labor law rights governing how workers use AI and control their creative output. This aligns with growing calls from groups like the International Federation of Journalists for collective bargaining over AI deployment, rather than relying solely on intellectual property frameworks that often benefit corporations more than individuals.
Doctorow also advises governments to avoid investing in AI now. He suggests waiting for the market to crash, then building on open-source models when hardware and talent become cheap. "Let's not be suckers. We're going to have tons of hardware. We're going to have tons of people who know how to use the hardware to do useful things. And we're going to have means, motive and opportunity to improve these open source [AI] models," he said.
For everyday people, the post-bubble landscape will leave "a bunch of useful tools." He draws a line between corporate "vibe coding"—which creates "tech debt at unimaginable scale"—and personal hacks. "Vibe coding for the thing that you use to connect your smart alarm clock into your smart coffee-maker is fine. Who gives a shit if it stops working? You just do it again."
Doctorow's book, published this month by Tor Books, draws on historical tech bubbles to frame AI's trajectory. He points to the dot-com crash, which left behind fiber-optic cables that enabled Web 2.0, as a model for how AI infrastructure could be repurposed. A National Bureau of Economic Research working paper similarly notes that overinvestment in general-purpose technologies often yields long-term productivity gains after the bust.
While Doctorow's warnings are dire, they echo a broader reassessment of AI's economic value. A recent Goldman Sachs report questioned whether the $1 trillion in projected AI spending would ever generate adequate returns. As the bubble's weak links strain, Australia—and the world—may soon test his post-crash playbook.
