Working Column

AI is a normal technology, and in the absence of the bubble, we'd call this collection of technically interesting, sometimes useful tools "plug-ins":

https://pluralistic.net/2026/02/19/now-we-are-six/#stock-buyback

A chatbot can't do your job, but an AI salesman can absolutely convince your boss to fire you and replace you with a chatbot that can't do your job:

https://pluralistic.net/2025/03/18/asbestos-in-the-walls/#government-by-spicy-autocomplete

Despite the fact that the AI can't do your job, there are many ways that AI can be used to erode your wages and working conditions:

https://pluralistic.net/2026/04/06/empiricism-washing/#veena-dubal

The workers who say that their jobs are worse and the things they produce are much worse as a result of AI are correct; but the workers who say their work is much better thanks to AI are also correct. This only seems like a riddle until you understand that the most important fact about any technology (including AI) isn't what it does, but who it does it for and who it does it to:

https://pluralistic.net/2025/09/11/vulgar-thatcherism/#there-is-an-alternative

When a boss fires a worker and gives their jobs to an AI, it usually means that they don't care if that job is done well, which is why customer service jobs are being handed over to AI:

https://pluralistic.net/2025/08/06/unmerchantable-substitute-goods/#customer-disservice

Bosses also love firing coders and replacing them with AI – first, because bosses are really angry about the decades when tech workers were in short supply and bosses had to pretend to like them, and second, because if you're selling AI as a way to replace workers, what better way to convince a potential customer than to fire the workers your own company depends upon? (All that said, the coders who are excited about their new AI coding tools have a point – when a worker is in charge of their work and thus when and how they use a tool, we should defer to their own experience):

https://pluralistic.net/2025/08/05/ex-princes-of-labor/#hyper-criti-hype

Artists are also a favorite target of AI bosses, which is weird, because the wages of creative workers add up to a total that rounds to zero when compared with the unimaginably large sums AI companies will have to take in if they are to pay back the trillions they've spent to date (let alone the trillions more they're proposing to spend in the near term). All of this raises a foundational question: can AI "art" ever be good? (Spoiler: probably not):

https://pluralistic.net/2025/03/25/communicative-intent/#diluted

Media companies say they have the answer to the AI art question: they'll create (or assert) a copyright that lets them control AI training. This is an incredibly transparent ruse: media companies are artists' class enemies, and if we get a new right to control AI training, our bosses will demand that we sign it away to them as part of their non-negotiable, one-sided standard contracts:

https://pluralistic.net/2024/11/18/rights-without-power/#careful-what-you-wish-for

For creative workers, the answer to these new would-be tech bosses isn't asserting a new right that will be expropriated by the old media bosses who've been ripping us off forever. Our salvation lies in leaning into the US Copyright Office's interpretation that holds that AI-generated works can't be copyrighted, because copyright is only for human creations. That means that the only way our bosses can get a copyright over the things they want to sell is to pay us to make them:

https://pluralistic.net/2026/03/03/its-a-trap-2/#inheres-at-the-moment-of-fixation

Many of the seemingly urgent AI questions that people won't shut up about are distractions, because they assume that AI will lastingly infiltrate every part of our society. In reality, the AI companies are losing unimaginable amounts and have no path to profitability:

https://pluralistic.net/2025/06/30/accounting-gaffs/#artificial-income

The only jobs that AI can do better than humans are jobs that shouldn't exist, like figuring out how to maximize undetectable wage-theft:

https://pluralistic.net/2024/12/18/loose-flapping-ends/#luigi-has-a-point

AI is also really good at figuring out how to do individualized price-gouging, another thing that shouldn't exist:

https://pluralistic.net/2026/01/21/cod-marxism/#wannamaker-slain

Despite AI's manifest unsuitability to do jobs that should exist, bosses keep firing people and replacing them with chatbots that do their jobs very badly. This allows bosses to indulge their solipsistic fantasy of a world without people, in which customers, workers and suppliers are statistical artifacts and bosses are unitary geniuses who simply imagine a product or service and then it is delivered, without any ego-shattering confrontations with people who know how to do things:

https://pluralistic.net/2026/01/05/fisher-price-steering-wheel/#billionaire-solipsism

This is catastrophic, and not just for the parties involved today. The AI bubble will pop, and when it does, the chatbots that do these jobs (badly) will be switched off. Meanwhile, the workers those chatbots replaced will have retrained, retired, or become "discouraged." No one will be around to do those (necessary) jobs. AI is the asbestos we are shoveling into the walls of our civilization and our descendants will be digging it out for generations:

https://pluralistic.net/2025/09/27/econopocalypse/#subprime-intelligence

The real existential AI threat isn't that we'll accidentally teach the word-guessing program so many words that it awakens and becomes a vengeful god. The real risk is that when the bubble bursts we'll indulge the ruling class's reflex to austerity, and that this will continue the decades of mass economic traumatization that makes people into easy marks for fascists:

https://pluralistic.net/2026/04/12/always-great/#our-nhs

But when the AI bubble pops, that won't be the end of AI – it will be the end of the bubble. When the AI bubble pops, we'll have mountains of GPUs at fire-sale prices, skilled workers liberated from the imperative to help their bosses promote their stock swindle, and open source models that will yield tremendous dividends to anyone who sets out to optimize them:

https://pluralistic.net/2025/10/16/post-ai-ai/#productive-residue

As you can see from the links above, I developed The Reverse Centaur's Guide to Life After AI in the same way that I developed Enshittification: in public, through a series of essays, which I periodically synthesized into major, widely shared speeches:

https://pluralistic.net/2025/12/05/pop-that-bubble/#u-washington

Making my working notes public is a hugely effective way of producing and refining critical work, and it's been my method for 25 years now:

https://pluralistic.net/2021/05/09/the-memex-method/

It's a method that's let me produce a string of international bestsellers, published by some of the largest publishers in the world. Nevertheless, Amazon refuses to carry my audiobooks:

https://pluralistic.net/2022/07/25/can-you-hear-me-now/#acx-ripoff

That's because I have an iron-clad requirement that my work be sold in open formats, without the "digital rights management" that blocks you from moving the books you bought on Amazon to someone else's apps. Digital rights management (DRM) enjoys bizarre legal protections so that it's a felony for me to give you the tools you need to move the books I wrote out of an Amazon app and into a competitor's app:

https://pluralistic.net/2026/01/14/sole-and-despotic/#world-turned-upside-down

What's more, these outrageous legal rights extend around the world, because the US Trade Representative spent decades bullying America's trading partners into passing laws that criminalize the act of fixing the defects in America's tech exports, which is why farmers can't fix their John Deere tractors, hospitals can't fix their Medtronic ventilators, and no one can sell you an app that stops Apple and Google from spying on your phone:

https://pluralistic.net/2026/01/01/39c3/#the-new-coalition

Amazon's Audible controls 90% (!) of the audiobook market, and they will not sell any book unless they can permanently lock it to their platform. That means that every time a writer sells you an audiobook on Audible, they create a "switching cost" that stops you from leaving Audible for a competitor. Not only is this fundamentally unjust, it's also terrible for creators: if our audiences can't leave Amazon, then we can't leave Amazon either, which means Amazon can (and does!) steal millions of dollars from writers without losing our business:

https://pluralistic.net/2022/09/07/audible-exclusive/#audiblegate

Which is where these Kickstarter campaigns come in. Whenever I sell a new book to a publisher, I arrange to make my own independent audiobook for it, which I sell everywhere except the platforms that have mandatory DRM: Audible, Apple and Audiobooks.com. There are some very good DRM-free audiobook stores, notably Libro.fm and Downpour.com (Google Play also sells audiobooks without DRM). But most people have never heard of these, so it wasn't until I started pre-selling my audiobooks on Kickstarter that I was able to make my stubborn refusal to sell out to Audible into a paying proposition. My agent tells me that if I'd sold out to Audible, I'd have paid off my mortgage and I'd be able to give my kid a full ride through a fancy US college. I don't make that kind of money from these Kickstarters, but they do very well nevertheless, and they're a critical part of my family's finances.

The Kickstarter is live for the next three weeks:

https://www.kickstarter.com/projects/doctorow/the-reverse-centaurs-guide-to-life-after-ai

You can pre-order print copies of Reverse Centaur, as well as DRM-free ebooks and audiobooks (narrated by me!) for Reverse Centaur and Enshittification. Normally, I offer custom-signed copies of the print books, but Enshittification was so successful that I haven't stopped touring it and I'm in a new city every couple of days, so there's no way I can reliably get into a warehouse to sign the latest batch of orders. Instead, I'll be posting the contact details for every bookstore that's hosting me on my tours (US in June, UK in September) and you can order signed copies from them, which I'll personalize after my events there so they can ship them to you.

I've also decided to raise money for the Electronic Frontier Foundation (eff.org), the nonprofit I've worked at for nearly 25 years. EFF is the oldest, best and most effective tech rights organization in the world, and its mission has only gotten more important over the years. EFF's outreach folks are offering a special membership package for backers of the Kickstarter, which includes an EFF hat and stickers, as well as an Enshittification pin and two Enshittification stickers:

https://pluralistic.net/2026/04/24/poop-emoji-plus-plus/#devin-washburn

The audiobook is fully recorded and finalized and you can listen to the first hour of it here:

https://archive.org/details/reverse-centaur-audio-sample

It came out great (as always!), thanks to the terrific direction of Gabrielle De Cuir of Skyboat Media and editing from Wryneck Studios' John Taylor Williams. Gabrielle's directed all my audiobooks since 2017, and John's been mastering my podcasts since 2006 (!!), so we constitute a very well-oiled machine.

Working out my ideas in public allows me to produce my Pluralistic newsletter, and with it, a large volume of free, high-quality work that's licensed under a generous Creative Commons license that lets anyone reproduce, translate, redistribute and even sell my articles. If you've enjoyed that work, I hope you'll consider backing the campaign! Selling books is how I pay the bills and keep the lights on, and as ever, this is the only way you can get a major publisher's ebooks and audiobooks with no DRM and no "terms of service." These are truly ebooks and audiobooks that you own. You can sell them, give them away, or lend them out – so long as you don't violate copyright law, we're all cool:

https://www.kickstarter.com/projects/doctorow/the-reverse-centaurs-guide-to-life-after-ai

If you'd like an essay-formatted version of this post to read or share, here's a link to it on pluralistic.net, my surveillance-free, ad-free, tracker-free blog: