Let’s Pop That Bubble

A High School English Teacher’s Notes on Cory Doctorow’s The Reverse Centaur’s Guide to Life After AI

Reverse centaur. A phrase I never, in a million years, could have thought of myself grounds Cory Doctorow’s brilliant little book, The Reverse Centaur’s Guide to Life After AI. The image of the centaur is a powerful one in mythology: body of a horse, head and heart of a human. In Doctorow’s use of the metaphor, technology—AI specifically—is the horsy part. His argument runs something like this: There’s nothing wrong with AI per se. In fact, for tech-savvy folks like Doctorow, there are a great number of pretty nifty uses—transcription and image manipulation among the uses Doctorow acknowledges frankly for the creation of his book. The issue is that for many—if not most—of the folks using AI, it is not the person that is using the tech. Rather, that is reversed. The tech is using the person, as in the cover for this book: a human bottom with a horsy head. And this reversal is not good for human beings.

The second part of Doctorow’s title is worth pondering, too: life after AI. Unlike so many folks writing breathlessly about the topic, conjuring dream/nightmare scenarios where robots are alternatively saving/destroying the planet, Doctorow zeroes in on the fantastically inflated investment in AI by the so-called “Magnificent Seven,” Nvidia, Apple, Microsoft, Google, Amazon, Meta, and Tesla, companies which, as he memorably terms it, are currently passing around a “$100 billion dollar IOU.” The only inevitable thing with regard to AI for Doctorow is that the bubble these companies have created will pop. There will be an “after AI” moment.

There’s good challenge too in this book, for folks like me who sometimes heap scorn on the tech itself, in addition to the companies hyping it. Passages like the following from his excellent concluding chapter, “After the AI Bubble” are worth closer study, especially Doctorow’s distinction between being “anti-AI” and “anti-AI bubble”:

“It’s fine to be ‘anti-AI bubble’ but it’s pretty silly to be ‘anti-statistical analysis’ and ‘anti-machine learning’ and ‘anti-automated inference.’ Long after the bubble is gone, many of these tools will recede to the status of boring utilities, nurtured by open-source weirdos and a few ambitious startups.”

Like Karen Hao does at the end of her excellent Empire of AI, Doctorow closes his book with a genuinely socially beneficial use of AI—the work of the Human Rights Data Analysis Group (HRDAG), which collects and analyzes “data about genocides, war crimes, and mass scale human rights abuses, using rigorous statistical methods and expert scientific communication skills to inform truth and reconciliation proceedings, war crimes tribunals, and other high-stakes forums.”

Maybe some day, HRDAG can help inform, through use of an LLM, a truth and reconciliation proceeding on the role of the “Magnificent Seven” in imperiling the planet, and facilitating a global recession? Up to this moment, though, the HRDAG have focused on accountability of governments and governmental agencies. For instance, in the US, they conducted the firs-ever national census of police killing of civilians. In Colombia, they examined and analyzed “Fifty years worth of records of extrajudicial killings during the country’s brutal civil war.” Both of these examples necessitated the use of LLMs to sort the data in helpful ways; the work wouldn’t have been possible without AI. At the same time, neither of these use cases are helped by AI hype or by the massive AI bubble the Magnificent Seven and their ilk have constructed in these past years.

Again, like Karen Hao, Doctorow underscores that whatever “AI Future” is on the horizon, it is not in any way “inevitable.” Doctorow writes in the concluding lines to the book:

“How we use AI is up to us. Whether we use AI is up to us. The future can be ours, if we never stop remembering that the most important fact about a technology isn’t what it does, it’s who it does it for, and who it does it to.”

When I consider these closing lines of Doctorow’s book from my position as an American high school English teacher, I want to be sure to think carefully about that last phrase: “who it does it for, and who it does it to.” Currently, GenAI in schools and classrooms is doing a number on trust between teachers and students, as well as trust between teachers and administrators. (See my own “Having the (Awkward AI) Conversation” and Matt Brady “AI: The Trust-Breaker” for on-the-ground writing on this topic from high school teachers.) This loss of trust is a huge, unquantifiable cost of school leaders’ unreflective embrace of a new technology that has not demonstrated its pedagogical worth. This loss of trust is also something that, aside from a few places online, or in pockets of conversations with friends and colleagues in between classes, is not really being addressed by schools. How many AI-centered faculty/staff meetings are dedicated to restoring trust between students and teachers? Or between teachers and administration? This is, in my mind, essential work for this moment. I don’t see or hear it being taken up.

Rather, as is customary for education discussions in this country, responsibility is being shifted to individual teachers for use of this technology that many teachers did not want in their classrooms in the first place. Some classic reverse centaur stuff going on. Consider Doctorow’s accounting of how AI is being marketed in the medical field:

“The market’s bet on AI is that an AI salesman will visit the CEO of Kaiser and make this pitch: ‘Look, you fire 9/10s of your radiologists, saving $20m/year, you give us $10m/year, and you net $10m/year, and the remaining radiologists’ job will be to oversee the diagnoses the AI makes at superhuman speed, and somehow remain vigilant as they do so, despite the fact that the AI is usually right, except when it’s catastrophically wrong.

‘And if the AI misses a tumor, this will be the human radiologist’s fault, because they are the ‘human in the loop.’ It’s their signature on the diagnosis.’

This is a reverse centaur, and it’s a specific kind of reverse-centaur: it’s what Dan Davies calls an ‘accountability sink.’ The radiologist’s job isn’t really to oversee the AI’s work, it’s to take the blame for the AI’s mistakes.”

This sounds a lot like the way Anthropic is marketing its new “Claude for Teachers” AI product, a product that teachers are on the hook for making FERPA compliant, by, as Benjamin Riley puts it in his excellent and scathing review of the product, making individual teachers “take on the work of anonymizing their student data each and every time they enter it into Claude.” Part of critiquing the AI Bubble so it pops sooner rather than later is quickly and accurately noting moments when you see you and your colleagues are being pitched as “accountability sinks.” Thanks, Benjamin Riley for showing one such sink. Administrators out there reading this post, please just say no to “Claude for Teachers.”

Another Doctorow-inspired question that doesn’t get a lot of discussion in schools is what are the tangible benefits as a worker that teachers like me are receiving for schools embracing these technologies? What about the benefits students (who are also workers) are receiving? I hear “more time” and “preparation for a changed economy” time and again, respectively, but many administrators seem to think these “benefits” speak for themselves. They don’t, and many teachers and students would dispute this account: Teachers, by and large, do not experience AI in the classroom as a time-saving technology, and it is not clear how AI in the classroom is “preparing students for a changed economy.” Can we have a conversation about this disconnect?

Can administrators be transparent and clear about the amount of money they are spending to implement AI in their schools? And share how that spending compares with how they are compensating their teachers? Are salaries for teachers keeping pace with inflation? Are staff and faculty being cut as AI investment is increasing? These are the pertinent and sometimes awkward labor questions to task when AI is on the table for discussion at school meetings—not simply “how is AI to be implemented in your class.”

Speaking of money, how about conversation around our retirement portfolios, and the funds which schools and we, its employees, contribute to? If, you accept, as I do, Doctorow’s analysis, that the Magnificent Seven has inflated an AI Bubble to dangerous proportions, perhaps it is time now to take a closer look at the holdings of our school’s default 401K and 403b funds? What percentage of those holdings are tied up in the Magnificent Seven / Big Tech / AI investment?

Last summer at this time, inspired and provoked by the calls for divestment in funds by college students protesting the US’s support for the genocide against Palestinians, I took a closer look at my own 401K, and realized the default fund on offer, BlackRock, in addition to having numerous companies funding the genocide in Gaza, and extensive holdings in the “Magnificent Seven,” was also invested in CoreCivic and Geo, private prison companies that contract with ICE. It was not an impossible task to “rebalance” my investments (with the help of www.prisonfreefunds.com) so I was not investing in those companies. (And yes, rebalancing in this way meant much less profitability for me the investor.) Other employees were grateful to know this information when I shared it, especially in light of the federal ICE occupation that descended on our city, Minneapolis, this past year. From an ethical and a practical point of view, heavy investment into big tech on the part of educational institutions should not be a given, and I think should be up for discussion and debate—discussion and debate that I think should involve teachers, students, and parents.

Solidarity amongst workers is key to ‘popping the AI bubble’ and Doctorow’s words in a speech that contained the germ of this book, one he made in December, 2025, are worth holding onto as we continue to navigate this strange and tumultuous moment:

“To pop the bubble, we have to hammer on the forces that created the bubble: the myth that AI can do your job, especially if you get high wages that your boss can claw back; the understanding that growth companies need a succession of ever-more-outlandish bubbles to stay alive; the fact that workers and the public they serve are on one side of this fight, and bosses and their investors are on the other side.”

Amen, Cory Doctorow. Thanks for talking back to the many myths and lies of this moment. Thank you too for offering cogent analysis of these ‘growth’ companies, and for clearly seeing the fight we are in for what it is. I recommend this book wholeheartedly, with all my (human) heart. And my centaurian “hooves” are digging into dirt with glee, too, as I type the final words of this review into a Microsoft Word document before pasting it into Substack.

---

NOTES:

-If you don’t have the funds/time to buy/read Cory’s entire book (which is worth both your time and money), you can find the excellent talk the book is based on here.

-I referenced Karen Hao’s excellent Empire of AI in this review (and elsewhere) as well as Benjamin Riley and Matt Brady. Other recommendations of AI-Bubble critics who I read and admire (and a non-exhaustive list) include: John Warner, Brian Merchant, Emily M. Bender, Matt Seybold, Alex Hanna, and Astra Taylor. Please do, add your own in the comments below, reader.