July 11, 2026
Half human, half panic
Reverse centaurs are the answer to the AI paradox
People say AI is either a miracle helper or a blame-shifting boss in disguise
TLDR: Doctorow argues that AI feels great when you control it, but awful when bosses use it to dump impossible workloads on workers who then take the blame. Commenters split between praising the idea as a perfect description of modern work and blasting it as hypocritical anti-AI grandstanding.
Cory Doctorow’s big idea is deliciously simple: people love artificial intelligence when it acts like a helpful sidekick, and they hate it when it turns them into the exhausted human buffer for a machine’s mistakes. His example — the now-infamous summer reading guide stuffed with books that didn’t even exist — had commenters nodding along, because the real scandal wasn’t just the fake titles. It was the image of one embarrassed writer being set up as the fall guy for work that used to take a whole team. In internet terms: the chatbot wrote checks, and the human had to cash the humiliation.
The comment section, though, quickly turned this into something much bigger and much spicier. One camp loved the “reverse centaur” label and called it a sharp way to explain why AI feels empowering for some people and miserable for others. Another took the metaphor and sprinted straight into doom: one commenter compared humans to obedient geese desperate to be led, saying we’re basically built for these control systems already. Casual! Others turned the whole thing political, cheering Doctorow’s anti-"there is no alternative" message as a kind of resistance speech against bosses and tech hype.
But not everyone was buying the sermon. The sharpest backlash accused Doctorow of hypocrisy: AI is bad for writers, but magically noble when open-source people use it? Even the lighter comments had edge — one person just wanted a more readable link and another celebrated the refreshing joy of reading something that didn’t sound machine-generated. In other words, the community verdict was messy, dramatic, and very online: AI isn’t just a tool fight, it’s a power fight.
Key Points
- •The article uses the automation theory concepts of “centaur” and “reverse centaur” to explain different experiences with AI.
- •Doctorow cites a Hearst-published summer reading guide containing nonexistent books generated by a chatbot as an example of AI-related editorial failure.
- •Reporting by Jason Koebler found that one freelance writer’s byline appeared across much of a 64-page supplement, suggesting a workload far beyond traditional editorial staffing.
- •The article argues that in such cases the worker acts as a human overseer and “accountability sink” for AI output rather than as a conventional author.
- •Doctorow contrasts that with his own use of OpenAI’s Whisper for transcription, presenting it as a worker-controlled AI tool that improved his workflow.