Hagen Blix
LightNews LightNews
banner
hagenblix.bsky.social
Hagen Blix
@hagenblix.bsky.social
2.7K followers 2.1K following 2.2K posts
Linguist, Cognitive Scientist, Occasional AI Researcher, Immigrant in NYC, Co-Author w/ Ingeborg Glimmer of 'Why We Fear AI' - out now: https://bookshop.org/a/114797/9781945335174
Posts Replies Media Videos
hagenblix.bsky.social
See here in the German wikipedia, for example
November 13, 2025 at 9:11 PM
hagenblix.bsky.social
Oh, and of course Eric Schmidt himself makes an appearance a couple of times, too
Another tech billionaire, former Google/Alphabet CEO Eric Schmidt, is similarly dreaming of magical solutions that just so happen to be what he’s selling. The New York Times reports him saying that AI “had too much potential, to let concerns about climate change get in the way.” In light of all that potential (for profit, we presume), Schmidt doesn’t feel that we should focus on reducing emissions too much: “we’re not going to hit the climate goals anyway.” What should we do instead? Well, Eric Schmidt would “rather bet on AI solving the problem.” We should—in the words of another tech marketer—hope that “AI is a problem and a solution.”25 That sure sounds like a salesman’s dream to us
November 11, 2025 at 8:36 PM
hagenblix.bsky.social
I do, of course, have a paragraph from "Why We Fear AI" for this
An LLM is an abstraction over the status quo (or rather, the text of the status quo) that has, at its core, no other aim than producing more text that is just like the text already in existence. One result is that the dominant ideas, the dominant textual patterns of our time, are recast as “intelligence” per se. We might call the complex mass of concepts and ideas, the channels into which ideas flow often and easily, or the boundaries of what is reasonable to debate in so-called “polite society”—the dominant ideology of an age. An age’s dominant ideology tends to presuppose that the power of that age is natural; the hallmark of ideology, in this sense, is that it is easy to recognize only when it’s past its time. We may no longer think that an almighty god appointed a king to rule us, that women are naturally unfit to vote, or that sex outside of a heterosexual marriage is immoral. But at various times, to think otherwise (to think of replacing the monarchy with a republic, to affirm that women are people, or being in favor of sexual or gay liberation) was the far more difficult, even dangerous thought—if not to have, then at the very least to put in writing. Ideology, of course, is never complete, other thoughts are always possible. But ideology is easy to think, it’s what our thoughts are already marinated in. And hence, ideology will be dominant in the textual patterns of an age. Seen in this light, what is called “intelligence” in a language model, is really that—ideology as a thing that can pretend to be a person.
November 11, 2025 at 8:36 PM
hagenblix.bsky.social
At a reading group that had just read our book, someone told them they loved that our book was making left wing theory so accessible, but that they wished we hadn't presupposed that people know the word "neoliberal".
Today, I realized that the Firefox default dictionary doesn't know the word either
The word "neoliberal", with red squiggles that suggest a misspelling
November 11, 2025 at 7:50 PM
hagenblix.bsky.social
Automating away accountability is, in fact, one of the many purposes of AI (and of faux personhood more generally, from that of AI to that of corporations)

"We can't manually update AI-generated fields" even when that's clearly stupid and unethical is part of that agenda
Pages 47f of "Why We Fear AI": Any previous bureaucracy, no matter how intransparent, how incomprehensible, how arbitrary in its exercise of power, was still, in some obvious sense, a structure made up of people—even if it precluded the possibility of communality. Up until today, when one faced a bureaucracy, even a hostile one, one always knew, at least intuitively, that the structure was made up of people. Even if no one person in particular was responsible, people were. Bureaucracies may have already perfected the art of continually deferring responsibility from one operator, bureaucrat, or office to the next, until the caller, customer, applicant, etc. gives up from exhaustion. But even if we knew nobody in particular was ever responsible in practice, it was simultaneously obvious that the deferred responsibility is a tactic of diversion, that it is, in a very real sense, a lie (even if it is usually a lie that one cannot do anything about). If a bureaucrat was practically insulated from responsibility, a model, in contrast, is simply genuinely incapable of it. It does not simply obfuscate the responsibility within the bureaucracy by continually deferring it, but instead it becomes the irrational locus of responsibility—incapable of bearing responsi- bility, yet “happy” to take it on.
November 10, 2025 at 4:27 PM
hagenblix.bsky.social
Fun & weird fact: Schloss Dagstuhl, which today houses the Leipniz Center for Computer Science, features the von der Leyen crest 🤷
November 10, 2025 at 2:35 PM
hagenblix.bsky.social
<Insert apocryphal quote about capitalists selling ropes here>
A metal plate version of a NYPost cover "the red apple" featuring Mamdani
November 7, 2025 at 1:36 AM
hagenblix.bsky.social
Great post, even though I hadn't come across that graph yet! Wondering how that compares internationally?
I don't think the EU job openings track any kind of stock market nearly so neatly, but the job openings also fall around Q3/2022

ec.europa.eu/eurostat/sta...
November 6, 2025 at 12:01 AM
hagenblix.bsky.social
Just remembering this picture of Lina Khan and Steve Bannon (from April)

I feel like Lina Khan always talks about the importance of US's national security interests when she talks about monopolies, too. Not sure my enemy's enemy is necessarily my friend here

xcancel.com/nancyscola/s...
November 5, 2025 at 9:55 PM
hagenblix.bsky.social
lmao, someone in the last reading group I did was literally complaining "wtf does 'a deed of history' even mean??", and I told them the German one says "historical deed"
An excerpt from the blogpost linked in OP: "In between those two examples, we have Fowkes translating Marx as saying that discovering the “manifold uses” of different objects “is the work of history.” The Princeton translation makes that “is a historical act.” There’s no sense in which readers would have a different understanding of the idea Marx is expressing depending on whether they read “is the work of history” or “is a historical act.” But the rhythm of the sentence has become noticeably worse."
November 2, 2025 at 11:43 PM
hagenblix.bsky.social
Why, firefox, why
November 1, 2025 at 2:01 PM
hagenblix.bsky.social
If someone needs a book on this, that links it to the nature of capitalism, class, the role that knowledge and technology play in history, and what power has to do with that... we've got you covered, of course

hagenblix.github.io/wwfai/
A book, called "Why We Fear AI: On The Interpretation of Nightmares"
October 30, 2025 at 5:06 PM
hagenblix.bsky.social
Kevin Roose and the New York Times seem like a match made, if not in heaven, then at least by the @nytpitchbot.bsky.social

"Does it actually work?" - "not really, but the power!"

xcancel.com/kevinroose/s...
October 30, 2025 at 1:44 PM
hagenblix.bsky.social
Wish the essay had some numbers on what part of this is just flowing through these companies vs being paid out as profits or unusually large salaries, but still "raising money to buy lists to raise money" is a standard big nonprofit thing, too, and it really is the opposite of organizing.
October 29, 2025 at 2:39 PM
hagenblix.bsky.social
I don't know if this tells us much about how long until the next downturn, given how frequent these "milestones" are
October 28, 2025 at 6:45 PM
hagenblix.bsky.social
As we write in our book: "If there is anything of interest to say about the billionaires’ [AI] dreams, other than their transparently infantile and self-serving nature, it is perhaps merely that they range from the mundanely apologist to the triumphantly narcissistic."
October 25, 2025 at 2:58 PM
hagenblix.bsky.social
To quote our book: "For capitalism, the worship of technology and anti-intellectualism are two sides of the same coin"
Page 131f of "Why We Fear AI": It is not the purpose of science to speak truth to power; under capitalism and for capital, its purpose is to train specialists for the pyramid, and lay the groundwork for the further development of technology, so as to enable companies, national economies, militaries, or states, to gain competitive and strategic advantages. But in order for science to do that, it has to be able to operate according to its own rules, and make its own purposes. And, at least within certain limits, these rules of science sometimes lead it to speak truth to power anyway. Climate change is an obvious case in point. But while the quest for truth, or understanding, may be the (or at least a) point of the enterprise from the perspective of science and the individual scientists, from the standpoint of capital it is mostly an inconvenient cost of doing business. Capital and capitalists dream of a world that is rational (based on measure- ments, statistics, predictions, etc.), but not comprehensible (where nobody has to be paid special wages for their comprehension, and all understanding resides in ruly prediction machines, evaluated by one metric of success—profit). A world in which only those truths that are good for profit are spoken to anyone. Capitalism loves technological developments, but it despises people having more (economically valuable) knowledge than absolutely necessary (and it resents even that). For capitalism, the worship of technology and anti-intellectualism are two sides of the same coin—quite analogous, of course, to the desire to increase productivity and decrease knowledge through deskilling.
October 23, 2025 at 4:23 PM
hagenblix.bsky.social
Alliances with fascists are our best and only hope? Fuck this shit
The coalition forming around these goals of promoting competition and protecting workers’ and consumers’ rights cuts across partisan lines. Safetyists and skeptics are finding unexpected allies, from progressive Democrats focused on labor equity to right-wing populists suspicious of corporate power. Former White House strategist Steve Bannon helped mobilize opposition to the federal AI moratorium, framing it as a giveaway to Silicon Valley. Sen. Josh Hawley (R-MO) has proposed algorithmic-audit mandates and copyright protections for artists, while Sen. Bernie Sanders (I-VT) has proposed a “robot tax” to redistribute productivity gains from automation to displaced workers. Their convergence suggests a nascent bipartisan agenda with politicians and influential voices on both sides starting to treat AI governance not as a culture war, but as a struggle over economic equity and institutional control. Anxiety about AI, if left unchecked, leads to paralysis. But when channeled into strategic coordination, it can become a force for change. The governance of AI should not be dictated solely by Silicon Valley or an administration intent on winning the AI race at all costs. It should be shaped by civil society, recognizing AI as a technology with transformative potential — both good and bad — that demands democratic mobilization and negotiation, not technological surrender. History shows that unlikely coalitions can redirect runaway systems. In the case of AI, they may not just be possible — they may be our best, and only, hope.
October 18, 2025 at 1:43 PM
hagenblix.bsky.social
Look, it doesn't know how to store anything in memory (0/10), but it's pretty good (4/10) at retrieving the nothing that is stored.
October 17, 2025 at 6:03 PM
hagenblix.bsky.social
We actually make a very similar point in our book - we also talk about land mines, and we link it to a whole series of arguments about faux accountability (from the operating logic of bureaucracies crowding out reality in favor of measurements of reality to citizens united)
From "Why We Fear AI", p. 65 These are, in some ways, continuations of the tendencies to automate killing in war—a landmine, too, makes an automated “decision” to explode and kill. But Lavender, quite unlike landmines, is simultaneously a system that creates an illusion of objectivity and order, as if it could take responsibility. It is itself an automated replication of a bureaucracy, veiling and deferring responsibility, as if the deter- mining factor were the AI, and not the military’s demand for a kill list of a certain length.
October 12, 2025 at 9:55 PM
hagenblix.bsky.social
Haha, you're welcome! Shannon Vallor also recommends it btw!
“Blix and Glimmer take the reader on a deeply engaging, funny, and provocative tour through our deepest anxieties and darkest fantasies of an AI-dominated future. Why We Fear AI offers profound insight into how AI and capitalism today work in mutually reinforcing lockstep to deskill and devalue humans, and what it will take to wake up from our worst AI nightmares and build a better world.” — Shannon Vallor, author of The AI Mirror
October 8, 2025 at 6:42 PM
hagenblix.bsky.social
It's always the car dealership owners, isn't it? We even have that one in our book
The more divided the working class is, the harder it is for them to find a common interest and common ground. This is obvious at the level of wages (where division undermines collective bargaining power, by separating out different interests). But it also applies at the level of politics, which is dominated by shifting alliances that sections of capital make with strata of the work- ing class. Of course, we have only touched upon many of these developments, and viewing the working class as stratified by their access to economically valuable knowledge has many potentially useful applications that we won’t be able to get to in this book. Consider, for instance, the right-wing attempts to channel anger among some working-class people away from collective conflict with their bosses, and into anti-intellectualism. There are ongoing attempts from the right to declare a barista with a college degree and $100k of student debt the “elite,” while the inheritor of a used car dealership, a small capitalist that makes seven figures in profit per year, is declared “working class.”
October 2, 2025 at 7:53 PM
hagenblix.bsky.social
This was the quote:

kalebhorton.ghost.io/2025-so-far/...
I also have advice. Everybody loves reading advice on the computer, so I’ll share it: the best thing you can do right now is log off as hard as you can. Go outside, talk to people in real life where it’s actually kind of rude to talk about the news, try to actually see the friends you usually just text message. Go for a long drive and turn the phone off while you do it. Get back into your hobbies or pick one and learn it for a while. Watch one of those studio movies that reviews called “wildly miscalculated” and you haven’t seen since high school. Play an album you like but find embarrassing. Go to free community events even if they sound stupid. If you take the freeway, try the surface streets. Go to a bad diner and just order some bad coffee because even bad coffee is good coffee. You can’t help anybody when you’re exhausted and keep posting one million college-educated rewordings of “I would love to be dead right now” on the computer. Walk away from the thing and try out some of those normal things you hear about and if you get bored that’s wonderful because we’re not supposed to get bored anymore. It turns out boredom is the Cadillac of feelings.
October 1, 2025 at 10:05 PM
hagenblix.bsky.social
10/10 headline. New Democratic party plan: Be incompetent and unhinged
September 29, 2025 at 3:01 PM
hagenblix.bsky.social
"Can Liberalism Be Saved?" - judging by that Kissinger answer, I'm gonna go with... nah, probably not.

www.newyorker.com/news/q-and-a...
September 23, 2025 at 5:02 PM