the forward centaur's guide to complex problems
Tech critics have a lot to say about how the generative bubble will burst. They're far more mum on what to do while we wait to see if it does.
Sci-fi author and pundit Cory Doctorow has become somewhat of a rock star of the anti-AI movement, by which I mean people who see chatbots and image generators as an existential threat to human dignity and future. In his new book, The Reverse Centaur's Life After AI, he tries to lay out the argument that AI in general is vastly overhyped and the $16 trillion poured into it is mostly empty promises, which is why it's pitched at us with as much zeal as it is, and we should totally team up and do... something to stop its imposition on the workforce.
What is that something? Join unions, employee co-ops, and #resist. Good advice if you don't know about or ignore union busting efforts, are in a field with no active unions, can land a job at one of the employee co-ops that make up between 0.01% and 0.02% of all businesses, or have a boss who is more worried about your future than their do-or-die Q whatever KPIs.
In other words, this is going to work for a few thousand people, out of tens of millions of workers in the U.S. Europe has a far better shot since unions for pretty much every profession are firmly stablished there, and laws are far more labor friendly. If you don't live in Europe and don't have a strong union, you're almost certainly doomed and are at the whims of your workplace. For now, Doctorow has little more to offer you, which made some of his fans shrug in confusion.
It's kind of like an expanded version of his argument that code is a liability, so an AI that generates code can only be a liability. But you need to have code to even have software in the first place, so code still has to be written and sure, there's a lot of bad code out there, but the issue is the process that fosters bad code and rushing to send updates to users who don't actually need many of those updates. Both the problem and solution are somewhat nebulous.
writing how to guides, minus the how
And that in itself poses an even bigger dilemma. Yes, certainly there are problems out there in dire need of solving. Many of them are thorny, pervasive, and felt by millions of people, even if they're not sure what the problem is or how to articulate it. A pundit like Doctorow can absolutely give their frustration a voice, but can't really offer a simple and workable solution outside of aggressive enforcement of anti-trust laws to break up all the oligopolies strangling innovation, stifling competition, and exploiting consumers, which, yeah, obviously.
Okay, great, we broke up all the Metas and Amazons, some of us joined co-ops, we tried to get unions off the ground, we posted a lot on social media, but LLMs are still being used because, even as Doctorow admits, there are valid use cases for the technology and it can be helpful for well defined but annoying tasks, as well basic rubber ducking and brainstorming. It's also not possible to forbid a business to use any generative AI in favor of a human. Okay, yes, it's possible, but not exactly practical. You may not like the results, but if the company is fine with them, well, that's their choice.
Despite the title of the book, there is no life after AI, as Doctorow readily admits. We are not going to un-invent it, and while it will fail to live up to the hype of its boosters, who at this point sound like they're either in the midst of a religious fervor, or have given up on taking change of their own lives, we are still going to have to figure out where it fits in our society and economy.
At the root of the anger at LLMs and image barf robots isn't a searing hate for AI and new technology, but the relentless pressure to use it for every task, the constant, daily admonishments against exercising your own judgement, the loud, obvious disrespect of basic human agency, much less caring what users actually want like in the heyday of Silicon Valley, and the haughty disdain for the livelihoods of the tens of millions whose very existence is tied to the 9 to 5 jobs they brag they want to make obsolete with these services, their eyes lit up with almost psychopathic glee.
the trap of individual solutions to systemic problems
If the system was built around making sure that your average John and Jane Q. Public was going to be fed, housed, and have access to medical care during huge, disruptive economic and technological mood swings, I'm willing to bet that the rage and anxiety we see would be turned down a few dozen notches.
Doctorow can soothe his fans that the very best generative AI can do is the equivalent of a very mediocre humans' effort, that the AI bubble will burst, and ultimately, it will all be okay in the end. He can't fix the fact that their bosses don't care and are laying them off by the tens of thousands with zero recourse, no help, and little hope. He can't feed them, make sure they have a roof over their heads, and can get their meds.
The system is supposed to see to that, but the system does not care. It demands that you either work or starve, and if it does not have work for you to do, then that's a you problem. AI is an existential threat to workers not because it can do a better job but because it's good enough to justify removing them from the payroll and cast them all aside, to get lost in a system that would much rather they disappear never to be heard from again, stuck in a Kafkaesque dystopia of red tape.
Real solutions would involve fixing a system in which a trendy tool can become a threat to one's life, and hype by technobabbling oligarchs can cost you your only way to make ends meet, something that I'm sure Doctorow knows he cannot provide, but all those hanging on to his and Ed Zitron's regular infusions of copium know full well. And this is why his proposed solutions feel so hollow, so vague, so unsatisfying. Because they are by virtue of avoiding the conversation we really need to have but refuse to because it's taboo after a century of red scares and 45 years of parasitic hyper-capitalism...
