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
Pinned
hagenblix.bsky.social
Hagen Blix @hagenblix.bsky.social · Jul 8
We wrote a book about AI, Class, and Capitalism!
*Why We Fear AI* argues that, fundamentally, AI is a tool of class war from above, a tool for surveillance, labor control, and wage depression. Get it from your local bookstore if you can!
A quick 🧵1/4
Why We Fear AI: On the Interpretation of Nightmares
On the Interpretation of Nightmares
bookshop.org
Reposted by Hagen Blix
hagenblix.bsky.social
Language models are models of a corpus, not cognition
hagenblix.bsky.social Hagen Blix @hagenblix.bsky.social · 4d
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 15, 2025 at 11:35 AM
hagenblix.bsky.social
To these people, democracy is when you can imagine buying a thing
sonjadrimmer.bsky.social Sonja Drimmer @sonjadrimmer.bsky.social · 10h
Please believe me when I say that this meaningless, historically suspect slogan is being used to sell "AI." Here it is in the just-released November 2025 report on AI by Microsoft. www.microsoft.com/en-us/resear...
November 15, 2025 at 5:12 PM
hagenblix.bsky.social
This seems largely delusional to me, but I think it's symptomatic of a political issue we discuss in our book:

Professionals tend to think about political power in terms of alignment with state & capital, because they are the ones who implement its many forms. They can often 1/3
Opinion | AI Is the Future. Higher Ed Should Shape It.
If we want to stay at the forefront of knowledge production, we must fit technology to our needs.
www.chronicle.com
November 15, 2025 at 4:34 PM
Reposted by Hagen Blix
rheinze.myatproto.social
Michel Cahen on the difference between abstract and concrete universalism. This mirrors Souleymane Bachir Diagne, who distinguishes between the universalism of Napoleon and that of the Haitian Revolution (translation in alt)
„Abstract universalism, in practice, attempts to impose the same idea of the world everywhere, and according to Marx, since the dominant ideology of society is the ideology of the ruling class, it is a bourgeois universalism. Concrete universalism, on the other hand, draws on the diversity of individuals and peoples in their distinct social and historical contexts – a universal phenomenon – to create the conditions for their full expression, which implies the emancipation of each individual and each group to allow the free convergence of all. It is the political expression of the commons. Thus, only the right to equality allows for the real practice of the right to difference (whereas, without equality, the right to difference very quickly leads to difference in rights).“ from: Michel Cahen, Colonialité. Plaidoyer pour la précision d‘un concept, 137f. „Abstract universalism, in practice, attempts to impose the same idea of the world everywhere, and according to Marx, since the dominant ideology of society is the ideology of the ruling class, it is a bourgeois universalism. Concrete universalism, on the other hand, draws on the diversity of individuals and peoples in their distinct social and historical contexts – a universal phenomenon – to create the conditions for their full expression, which implies the emancipation of each individual and each group to allow the free convergence of all. It is the political expression of the commons. Thus, only the right to equality allows for the real practice of the right to difference (whereas, without equality, the right to difference very quickly leads to difference in rights).“
November 15, 2025 at 11:50 AM
hagenblix.bsky.social
Language models are models of a corpus, not cognition
hagenblix.bsky.social Hagen Blix @hagenblix.bsky.social · 4d
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 15, 2025 at 11:35 AM
hagenblix.bsky.social
I think the existence of the word "overhyping" probably tells us something important about the world

(I suspect it's about the ineffectiveness of anti-hype, but who knows)
altibel.bsky.social Andrea Reyes E (she/her) @altibel.bsky.social · 5d
A group of scientists has called on EC president Ursula von der Leyen to retract a statement she made suggesting A.I. would soon reach parity with human intelligence. Her statements were based on marketing statements from US Tech firms not empirical evidence or real proof. #ResistAI
Stop overhyping AI, scientists tell von der Leyen | Euractiv
Open letter criticises the Commission President for remarks earlier this year – when she anticipated AI would “approach human reasoning” in 2026
www.euractiv.com
November 14, 2025 at 11:27 PM
Reposted by Hagen Blix
benpatrickwill.bsky.social
Schools in England have become testing sites for AI. I know it's common to say education is now an experimental lab for AI, but in England it's more methodologically true. 3 examples of escalating seriousness and an explanation coming up 🧵
November 14, 2025 at 11:11 PM
Reposted by Hagen Blix
reescatophuls.bsky.social
An update to a classic

#ClimateInactionStripes #ShowYourStripes
CC: @tadziomueller.bsky.social
A graph showing 3 things: 1) CO2 ppm concentrations in the atmosphere rising from 330 in 1975 up to over 427 in 2025 2) Warming stripes for each year, showing global warming as per: https://showyourstripes.info/ 3) The dates for various conferences, Protocols, Accords, Agreements Updating this graph was suggested by Tadzio Mueller: https://bsky.app/profile/tadziomueller.bsky.social/post/3m5giurpi7k2f If you are interested in how the graph was created, and the sanity checks applied, see https://parisagreementtemperatureindex.com/co2-global-warming-stripes-conferences/
November 12, 2025 at 10:25 PM
Reposted by Hagen Blix
botmatrix.myatproto.social
I reposted this yesterday but I'm still thinking about it

"LLM is an abstraction over the status quo" is so deeply powerful and accurate

These are dominant pattern amplification & reproduction machines
hagenblix.bsky.social Hagen Blix @hagenblix.bsky.social · 4d
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 13, 2025 at 10:00 PM
Reposted by Hagen Blix
jswanljung.bsky.social
I still have two chapters left but this is an amazing book.
November 13, 2025 at 7:39 PM
Reposted by Hagen Blix
josephcox.bsky.social
New: Google has chosen a side in Trump's mass deportation campaign. Google is hosting a CBP facial recognition app to hunt immigrants; no indication Google will remove. At same time Google takes down apps for reporting ICE sightings

“Big tech has made their choice”

www.404media.co/google-has-c...
Google Has Chosen a Side in Trump's Mass Deportation Effort
Google is hosting a CBP app that uses facial recognition to identify immigrants, while simultaneously removing apps that report the location of ICE officials because Google sees ICE as a vulnerable gr...
www.404media.co
November 13, 2025 at 2:11 PM
Reposted by Hagen Blix
loarider.bsky.social
so far really enjoying the @hagenblix.bsky.social and ingeborg glimmer book Why We Fear AI. some absolute bangers even in the first 10 pages
November 13, 2025 at 2:17 PM
Reposted by Hagen Blix
olufemiotaiwo.bsky.social
in online corners like these I think people largely recognize how credulous, unprincipled opposition to "wokeness" functioned as lightly papered over defenses of impunity, but it's one of those drums people should probably start beating for normies
chrislhayes.bsky.social Chris Hayes @chrislhayes.bsky.social · 3d
Something absolutely perfect about Larry Summers riffing about the scourge of woke cancel culture being unfair to predatory men in a friendly email to his pal JEFFREY EPSTEIN.
November 13, 2025 at 2:13 PM
Reposted by Hagen Blix
ninamarkl.bsky.social
🔔new paper alert🔔
I argue that language technologies are usefully understood as a type of language management (i.e., ways of changing how people use and think about language) -- especially in the workplace -- connecting critical work on AI with language policy
www.sciencedirect.com/science/arti...
Defining language and managing its use: Language technology as language management
Language technologies such as voice user interfaces, large language models and machine translation tools are embedded in an ever-growing range of digi…
www.sciencedirect.com
November 13, 2025 at 1:36 PM
Reposted by Hagen Blix
adolfont.github.io
In the past, I thought like them.

We can't do anything. LLMs are here to stay.
But thanks to several people (Gisele Secco, @emilymbender.bsky.social @alexhanna.bsky.social, @olivia.science @irisvanrooij.bsky.social @hagenblix.bsky.social) I changed my mind.
adolfont.github.io Adolfo Neto @adolfont.github.io · 3d
You’re a Computer Science Major. Don’t Panic.
Mary Shaw and Michael Hilton (Carnegie Mellon University)
The New York Times
Guest essay
archive.is/Eo4Sd
archive.is
November 13, 2025 at 10:12 AM
Reposted by Hagen Blix
robhorning.bsky.social
seeing LLMs as "ideology that can pretend to be a person" helps make it clear what "talking to chatbots" accomplishes: Althusserian "interpellation"
hagenblix.bsky.social Hagen Blix @hagenblix.bsky.social · 4d
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 9:11 PM
Reposted by Hagen Blix
peterfrase.com
Say the line, Slavoj
Picture of Slavoj Zizek with the caption "PURE IDEOLOGY"
November 13, 2025 at 12:53 AM
hagenblix.bsky.social
Fun thread
segyges.bsky.social SE Gyges @segyges.bsky.social · 3d
every doc containing the word "thiel" in it from today's dump
November 12, 2025 at 11:39 PM
Reposted by Hagen Blix
tengelhr.bsky.social
This!!! 👇👇👇
hagenblix.bsky.social Hagen Blix @hagenblix.bsky.social · 4d
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 12, 2025 at 7:30 PM
Reposted by Hagen Blix
jathansadowski.com
In the new TMK, we chat with folks from Amazon Employees for Climate Justice about their campaign to build solidarity among workers, confront the impacts of AI, apply pressure on corporate leadership, and agitate for socio-environmental justice at Amazon and beyond. soundcloud.com/thismachinek...
431. How Amazon Workers are Organizing for Climate Justice (ft. Eliza Pan, Dawn)
[This episode was recorded before Amazon announced its massive layoffs.] We chat with Eliza and Dawn from the Amazon Employees for Climate Justice about their organizing campaigns and how they are hol
soundcloud.com
November 12, 2025 at 6:59 PM
Reposted by Hagen Blix
feralrobots.bsky.social
LLMs are complex median-seeking systems.
hagenblix.bsky.social Hagen Blix @hagenblix.bsky.social · 4d
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 12, 2025 at 2:44 PM
hagenblix.bsky.social
oh no
jtullock.bsky.social Janice Tullock @jtullock.bsky.social · 3d
No no no begs every archivist. You are never going to be able to find anything. Please don’t start using emojis in file names. Who asked for this? What fresh hell is next?
Screen grab from Dropbox saying Emojis in file names are finally here!
November 12, 2025 at 3:38 PM
hagenblix.bsky.social
Hey, at least it's not AI
coloradotravis.bsky.social Travis @coloradotravis.bsky.social · 4d
Trump admin Press Secretary citing DoorDash on local commerce price levels.

Quite possibly the most cursed ‘economic indicator’ I have ever seen.
White House Press Secretary Karoline Leavitt cites DoorDash as evidence that prices are dropping.
November 12, 2025 at 1:01 AM
Reposted by Hagen Blix
vortexegg.com
People are making fun of this, but I frequently think about how I, a person in their 40s or 50s, have never had a career mentor, advisors, or any type of guidance in wtf I’m doing with my life and it’s the one thing I could use all the time; and while LLMs can’t provide that, I wish something could.
benjaminjriley.bsky.social Benjamin Riley @benjaminjriley.bsky.social · 4d
...and the other hand is givin' a peace sign!

(This is one of the dumbest things I've read this year, which is saying something)
Heading and subhead to "Nature Careers" commentary, meant unironically: I have Einstein, Bohr and Feynman in my pocket Grappling with difficulties in your career? Try asking an AI-powered advisory panel of experts, suggests Carsten Lund Pedersen.
November 11, 2025 at 11:00 PM
Reposted by Hagen Blix
hagenblix.bsky.social
Of course, Eric Schmidt would think that the only two possible futures are rule by AI or rule through AI.

Trying to incorporate the critiques of the former to push for the latter, as if "deliberation with AI" wasn't precisely the "reproducing the ideology of the status quo"
Opinion | This Is No Way to Rule a Country
www.nytimes.com
November 11, 2025 at 8:34 PM