Bookpile 12: The Reverse Centaur's Guide to Life After AI

I'm not always in my work bubble – in my work bubble, everyone has a degree of knowledge and criticality about the current state of AI hype. Sometimes I get out, and run into people who feel that what we're currently experiencing is inevitable, that we don't have a choice, and that the tools and products being foisted upon us by companies like OpenAI are just going to happen to us, whether we want them or not. When talking to the people in my life who are less in the AI-critical bubble, I do my best to stress that this doesn't need to be inevitable, that the economics of current applications don't stack up, and that another world is possible.

Now I have a book I can recommend to those people. There are so many brilliant and critical works out there on the assumptions and problems with the AI boom. But none of them are quite as handy and accessible as Cory Doctorow's new book, The Reverse Centaur's Guide to Life After AI. This thing is a wildly convenient arrow to have in your quiver, laying out what's happening at the moment, what we can take from it and what might be better to leave aside, and what's wrong with inevitabilism.

It's a tidy and well-bounded book, providing essential arguments and insights, as well as being pithy and easy to read. As is his habit, Doctorow manages to make complicated things comprehensible. Most importantly, he decouples the technologies that actually exist now from the things that are being promised, making it that much easier to not just accept Large Language Models (LLMs) as one step on the path to inevitable Artificial General Intelligence (AGI). This is one of the points I tend to not see enough in general conversation, not least because the marketing machines of tech companies are heavily invested (literally and figuratively) in making us believe that AGI is just around the corner and will change everything when it comes.

The treatment of economic issues is the other best thing about this book. Clearly and comprehensibly explaining the financial mechanisms that cause bubbles to happen is a major public service. Once again, decoupling the hype from the underlying factors motivating the actions of the hype machine is widely needed, and not happening enough in public discussions of AI.

Read the hell out of this book, even if you already know the arguments. The accessibility and humour of Cory Doctorow's writing always helps, and now he's done it for AI. I've recommended it to three people this week already, and it'll likely continue to be my go-to when I run into others who only hear the hype and doom, and don't have easy access to the nuanced and hopeful arguments.