The Reverse Centaur’s Guide to Life After AI

Cory Doctorow · Farrar, Straus and Giroux · 2026

Good book about the current state of AI in our culture and the impact of AI. Cory is understandably cynical about the hype.

This is a book any technologist should read. It’s generally the next chapter of the enshittification of everything. Everything also means GenAI. He uses the reverse centaur as an effective visual metaphor for what’s happening, or what the promise of AI fails to deliver. A centaur is a human head on a horse, the human mind using the superior power to drive human wants using the machine to increase human power. The machine helps the human. Bikes, cars, chainsaws, calculators, even your (not ai assisted) phone makes you a centaur. A reverse centaur is where the human is conscripted to do the machine’s bidding, at the speed of the machine. He gives the example of an ‘I Love Lucy’ episode where Lucy and Ethel are supposed to wrap chocolates by hand coming down the conveyor belt. The belt speeds up, making for a great laugh for us the viewers. But Lucy and Ethel have no control over the machine, and they have to meet the speed of the machine. They are the wet-ware extension of the machine, they are the reverse centaur.

Cory asserts that this is where we get it wrong, and all of AI is really about labor rights, or lack of them. Good news is he thinks this part will fail. Bad news is we’re not getting Star Trek Utopia any time soon.

The internet has taken over all corners of our lives. It’s in everything we do, and access to reliable, resilient internet is required for minimally participating in our world. (think about this next time you see a person experiencing homelessness or extreme poverty with a smartphone - you cannot exist without access to the internet.) Now, AI seems to have taken over every corner of the internet and how we use it. Cory does a great job summarizing it, and boiling it down to a simple universal truth, hucksters gonna huck. Is this an AI bubble? Maybe. Is it huge? Maybe. Will we all lose our jobs and become batteries? Probably not. But what is true is AI is being forced on us in everything, and it doesn’t always make the experience better. It does make more money for the tech bro oligarchs.

The tech-bro-oligarchs (I’ll call them TBOs from now) don’t actually care if it works, or if it makes it better. They’re selling the hype, which sells the promise of stock growth. This kinda sucks, since if the TBOs were honest and let AI (and ML more broadly) focus on real problems, we could solve some pretty big things. But the TBOs are not selling AI, or chips or anything but stocks, and stocks are sold and valued on the promise of growth. Growth forecasting is a made-up equation based on hype. One recommendation he gives is don’t help them with the hype machine. He calls it criti-hype, a term coined by Lee Vinsel : critics of AI inadvertently fuel the industry’s inflated valuations by exaggerating what AI can actually do into doomsday scenarios.

Taking press releases from startups and covering them with hellscapes. Lee Vinsel, quoted in Pluralistic

In other words, when you talk about the super scary hacker models, you’re only helping the TBOs sell the hype machine and inflate the growth.

I think a key lesson he wants us to understand with any technology, including AI is the following.

What your technology does is less important than who it does it for, and who it does it to. Cory Doctorow

Read the book, @ me on bluesky or email me. I’d love to chat about it and technology overall.

ps. As I was writing this, Mythos 5.1 and Fable 5.1 were announced. Be open, honest and grounded as you review and read the reviews of the models. Some of it may be true, and some might be hype. ¯\_(ツ)_/¯