1. “Smug Delusions Won’t Save Us” — Matthew McManus, Liberal Currents
This is actually a lengthy review of “The Revolutionary Center” by Adrian Wooldridge, who is a Bloomberg and former Economist columnist. It’s fair to say that Matthew McManus doesn’t like the book, and based on this review I suspect I won’t either for very similar reasons. To be fair, he lost me at the now axiomatic tendency to combine “centre” with a word that sounds more dynamic: see also the “Radical Centre” beloved of Nick Clegg and now Tony Blair.
I thought the piece was particularly interesting in the differing interpretations of what John Stuart Mill represents. Like McManus, I’m far more comfortable with Mill the liberal socialist than Mill the darling of the reactionary centre right.
At least Hayek had the decency to refute Mill and declare him as the point where liberalism took a wrong turn. These days neoliberals love to do their best to co-opt him, notably the Orange Bookers (and yes, this is probably a prompt to myself to read the signed copy of Richard Reeves’ biography of Mill that has been sitting on my book shelf for nearly 20 years).
2. The Truth About AI Nobody Wants to Hear — Rutger Bregman
This video intensely irritated me but I think I’d like to see a smarter person than me fully rebut it.
To be fair, I’m just not Bregman’s target audience. There’s something about his bro-ey, smarter-than-you style that I tend to bounce off of. But I can’t help that he’s doing an awful lot of straw man arguments and cherry picking in this video.
For one thing, who are these leftwing AI delinalists that he’s based his whole argument on? I’ve read a lot of books and articles about the tech industry over the last 12 months, many from a leftwing perspective, but I’ve yet to come across one asserting that the technology isn’t real and isn’t going to fundamentally change the world. He claims that “their “central conviction that this whole Silicon Valley AI project would hit a wall any minute now. That conviction has completely collapsed,” but I just don’t see who this “they” is. Are we perhaps confusing nuance with denialism?
In fact, far from the smooth line of progress that Bregman outlines here, the AI industry has frequently failed to meet its own targets. In fact, I’ve seen the tech sector shifting its goal posts far more rapidly than its critics. Just as fusion electricity is the technology that is permanently 20 years away, AI superintelligence has, for the past 5 years, been permanently 18 months away.
The most telling red flag in this essay is that he heaps an awful lot of his argument on the phrase “stochastic parrot” coined in the famous paper by Emily M Bender and Timnit Gebru. But that is literally just a pithy description of how the technology works. No, Rutger, they don’t reason. That isn’t denialism, that’s just how computers work.
Bregman himself admits that he has never written a single line of code, and then goes on to describe all the magical apps he’s used AI to build for himself. But that’s literally just taking existing code and adapting it to his own specifics.
No one as far as I’ve come across, is seriously disputing that LLMs are very good, and getting better, at writing code. Some of the more contentious claims made in this video amount to little more than reading off a bunch of press releases, with no more critical analysis than that.
Ultimately, this video comes to some very similar conclusions that I’ve been wrestling with myself, about the left’s need to develop a viable vision of what our future needs to look like. But I worry that by buying into an awful lot of the tech industry’s hype, Bregman is either arguing himself out of the conversion or finds himself on the wrong side of it. Cory Doctorow, someone who knows a lot more about tech than Bregman, seems to have a much better grasp on the problems and opportunities the tech sector is presenting us with (I look forward to getting the finished version of his latest book later this month).
Leave a Reply