Hagen Blix

Might be quite a nice case for "dialectics" in the sense of dynamically mutually defined categories that absolutely resist a simple causal model, imho

I'd like to read that whole essay please

Fwiw, I have never seen squirrels as well fed as around UMich. Look at this adorably chubby little fella

Marx's Ethical Vision Abstract. “The communists do not preach morality at all”; this line from The Communist Manifesto might seem to settle the question of whether Marxism has a

I also recommend Marx’s Ethical Vision by @vcwills.bsky.social academic.oup.com/book/57415?l...

Acknowledging this crucial fact is actually essential to figuring out what might be needed to change it

Why We Fear AI: On the Interpretation of Nightmares — Common Notions Press Industry insiders Hagen Blix and Ingeborg Glimmer dive into the dark, twisted world of AI to demystify the many nightmares we have about it. They combine expertise in cognitive science and machine lea...

We've got a chapter on AI and neoliberalism, and how they're connected all the way to the OG 1958 Perceptron paper that introduced neural networks, which cites Hayek himself as an inspiration.

Might be interesting to you!

Creepy thing is that GPT didn’t invent this model—this is essentially the apotheosis of neoliberal ed reformists’ vision for schooling

I actually think that opposition of "it can do everything" vs "it's all hype" isn't the best way of thinking about this either.

Wrote a little thing about why here:

I will also say that it’s a bit wild to me that no press freedom organization or professional journalistic association has tried to contact me or even put out a statement here.

I mean Jesus Christ. A little solidarity from institutions would be nice here!

I think a lot of stuff that goes under the label critique is actually self-help for people who wouldn't want to buy a self-help book, and occasionally, it is a self-help manual for condescension

This is a terrible - but also terribly revealing - oped from Raimondo that shows why the Democratic Party in its current incarnation will never regenerate public higher education around a humanistic vision of knowledge. Public higher ed for them is job training for the working/lower middle class.

At any rate, agreed that that's an important way of looking at it, glad to see it coming up in many contexts, including your writing!

Why We Fear AI — Common Notions Press Fears about AI tell us more about capitalism today than the technology of the future.

You might like our book, then! We discuss the attempt at industrializing language production, including a critical discussion of productivity vs deskilling (and from what perspective they might look indistinguishable). Might be relevant for you!

between managers and concrete knowledge of concrete tasks. More technical and/or managerial mediation layers might limit the degree to which you can draw from workers, engineers, etc to build a managerial structure. I suspect there's a real contradiction here, rooted in conflicts over knowledge

I've been wondering for a while what the drivers of that change were. There is the shareholder value revolution ala Friedman (which Acemoglu and Johnson would attribute this to), and then there's the recursive nature of the conjoined up- and deskilling moves, which would also increase the distance

one of the big problems with the "we're shifting the nature of the work to specifying outcomes" is that it presumes you already understand the problem in sufficient depth to specify an outcome (in natural language, no less, which is by its nature imprecise)

Yeah, once again it's becoming painfully obvious that the substitution of a business school managerial class for the engineering-management pipeline of old is producing it's own manifest contradictions. Because "just specify the outcome" without subject knowledge is exactly the way of the manager

Opinion | America Cannot Withstand the Economic Shock That’s Coming

Venture Capitalist and Biden's Secretary of Commerce, Gina Raimondo, is taking AI as an opportunity to further subsume education under the needs of capital and embrace deskilling.

Make no mistake, this is about shifting income from labor to capital, it's about wage depression

Had a blast talking to Justin, Sadie, and Jay from @librarypunk.bsky.social about anything from libraries, AI and the economy of knowledge to organizing and the movie Pride!

The industrialization of language production is destroying the artisanal production of text.

We should be clear about what this is: The real subsumption of life under the logic of capital

Sure, there are delusions among the ruling class, and yes, some of those are driven by the need to find new markets, but nah, the main driver isn't religion or mass delusion, it's class war, and if anything AI represents the opening of a new front, and a new weapon. We miss that at our peril.

Anthropic’s AI tool Claude central to U.S. campaign in Iran, amid a bitter feud Anthropic’s AI tool Claude is playing a key role in the U.S. military’s campaign in Iran, amid a bitter fight with the Pentagon over the terms of its use in war.

I think it's actually based on WaPo reporting, that DNYUZ looks like it's just another site stealing content or something

No, the US government murdered those children.

Don't let them use AI to obfuscate responsibility.

Right, that's what it seems like, thanks for sharing!
Posting this here for the two sources in case someone else is looking for them

Right, I think these are the two relevant sources on which this is based, but it's not clear if there's a link between them

Satellite images show Iran school strike hit more buildings than earlier reported The images suggest that precision munitions struck other buildings, including a clinic that was also inside the complex.

So, the school may well have been about outdated intelligence
www.npr.org/2026/03/04/n...

And Anthropic's models were used for rapid target selection
www.washingtonpost.com/technology/2...

But it's unclear what the specific connection is (though perhaps reasonable to suspect that there was one)

Thanks! So we do know about Anthropic's systems being used in target selection through Project Maven, but we don't know anything specific about the attack on the school?

Also, is there a source for the claim about Claude? The thing that's referenced in the skeet you quote is itself just a screenshot of a Claude output, so might be totally hallucinated

Not to disagree with the political point at all (we actually wrote about Lavender in our book, too), but a) I'm not sure we know what kind of AI system Lavender is, I suspect it's not (or not primarily) a language model, but probably runs on metadata (location tracking, call and text data etc)

I'm talking about Jacobin magazine ;)