Should add lofi beats to chill and burn flags to lol

Comrade the same could have been said of coal back in Marx’s day, or even its use today in the global south. And refusing to think about it or entertain contemporary questions doesn’t solve them, it’s putting our heads in the sand (the ostrich technique) and hoping the rolling storm passes over us unscathed. LLMs are means of production, many professionals are using them already in many different ways. Whether they lower the socially necessary labor time and rate of profit is not entirely calculated yet, but this isn’t necessary to count as MoP. And AI has entered our relations of productions already.

I have an essay here on the question https://en.prolewiki.org/wiki/Essay:Intellectual_property_in_the_times_of_AI.

None of these claims say anywhere that Stalin was responsible for them

Deng: [He laughs.] Listen, they can call me anything they like in the West, but I know Khrushchev well; I dealt with him personally for ten years, and I can assure you that comparing me to Khrushchev is insulting.

This is partly why I say I love Deng so much. It makes western ‘marxists’ seethe and look at that, he even calls them out for being ineffective and irrelevant.

In truth Deng continued a lot of what Mao had started or planned to start.

The conclusion is these 2000s new atheist figures are actually very dumb individuals and should fade into irrelevance but instead we let them poison the discourse for an entire generation instead

the communists discussing Stalin have managed mass production of tanks meanwhile the anarchists are still sitting on the ground (presumably because they didn’t find anybody who likes building 100 chairs in their spare time). MLism undefeated 😎

lol this is so true. I love talking about medieval Europe.

We should be careful not to extol the virtues of the imperial examination system too much. I also used to believe this but when I did my research to write on the history of China I had to face the fact that it was a deeply flawed system back in the imperial times, though in its time it had a reason to exist and remain an institution, it’s not quite meritocracy as we understand it to be today.

There is first of all a problem with pure meritocracy - not everyone comes from the same background. This is why in China today (but they’re not the only country doing it), ethnic minorities have easier access to universities among. A poor farmer coming from a rural village will simply not have the same outcomes available to them than a coastal business owner’s son. Over generations, these starting positions do get equalized, but it’s a long process with unequal development of course.

This was also the case historically with the examination system. At its core this system, which had different levels for every promotion, allowed anyone to take it. But not many families could not only afford to pay a tutor for their son, but also let their son study for several years instead of helping on the farm. For some “not too well-off” families, this could be a way into money, so you would see it at the landowning level (feudalism in China was different from European feudalism and while I don’t know too much about it to say definitively, there definitely were landowners who were not nobles, and land was openly traded between families and the state, and this is how there came to be these landlords that got purged during the revolution.)

In fact, even at its largest in history, comparatively few officials came from the imperial examination compared to aristocratic families that just got a job because of their background. I don’t have the numbers right now but it was much lower than you’d expect from how much this system is said to be important. The third mechanism for recruitment was receiving a recommendation from someone already in the administration.

Materially speaking, the system got started in the early Han but saw widespread use in the Song dynasty. This was because when the Tang came to an end following countless rebellions and wars, many of the aristocratic families that had a place in government died off or vanished, so there needed to be a new ‘strata’ of advisors trained for the needs of this new state. They had unified China through conquest and urgently needed civil servants to now administrate this territory.

And even during the Song expansion of the system, only about 50% of the population was eligible for the exams in the first place, and far fewer could even afford to study for it. Women were excluded by default, as well as merchants (because confucianism considered merchants to be a low class as they did not produce anything themselves lol). In the end, only about 100 people were allowed to pass the exam each year, which is certainly a choice, but also seems very low for a large country such as China - not only geographically, but also economically and demographically, with an estimated 175 million people. And 100 was just for the “entry-level” exam, there were further ones you could take to be promoted further up with even lower passing rates.

Libertarians figuring out they do not live outside of society as they previously thought

chat help with my macroeconomy I’m not very good at this. Is it good to lock myself out of a 1.5 billion people strong market or nah?

I would have said the same thing as the author if I’d written that, counting ‘beating’ as being exceedingly cheaper while delivering comparable results, and doing it under sanctions. Deepseek makes a lot of sense for hobby projects because of the price, though I’m hearing about professional devs ditching Claude for V4-pro - but before deepseek, there was no reasonable solution for agentic at home, and you were stuck on debugging on the web interface.

Mind you speaking of benchmarks I have no idea what these things are actually supposed to represent lol. I found v4 good at recall and memory, but when talking to it (doing research, clarifying questions etc as opposed to just having it code), I found its overall output pretty diminished, like an old GPT 3.5 “you’re so right - and here’s why you are”. You can gloss over it but they had found a great mix by late 3.2 imo.

The 4-bit quantized GGUF for granite 4.1 is sub 5GB, so it’s probably going to run on any modern machine even if it’s not particularly built for Vram… 6 gigs is what I had on my old 1080 gpu.

https://huggingface.co/unsloth/granite-4.1-8b-GGUF/tree/main

Deepseek v4 pro! Top up your credit as you go and they’re having a sale until May 31st, but even without the sale 1M output tokens is “only” 3.48. Flash is only 0.28 per 1M output.

lmao I never saw the second one before

It’s pretty much the only thing I still use Gemini for, vision and image creation when I need a dumb meme like this:

but google hates when people use VPNs and fingerprint-scrubbing extensions so half the time it doesn’t work. Since deepseek upgraded to agentic search shortly before v4 I don’t even use perplexity anymore. I also noticed you can run web code (html css and js) in a sandbox on the web interface directly if it writes you some code, but it’s no replacement for proper agentic software imo.

Incidentally gemini’s vision was very good, so I’m excited to try deepseek’s when it comes out for me. It understood AI-generated abstract paintings just fine, AI-generated being the keyword here because it’s never seen it before, since I had just generated it with an SD model (unless there’s super deep cross-contamination where it somehow intuitively understands the output of another neural network).

And of course deepseek is completely free and has no paid tier or throttling whatsoever.

Example found on twitter:

that’s pretty impressive tbh.

Only some people were invited to test it, I wasn’t 😢

if they can integrate it wholly with the base model so that you can use it in agentic this would be huge. If you think about it, how many times in a week do you tell people “I have this problem” and send them a screenshot without further context? That’s what vision would do in agentic, it could see the graphical result of its code and instantly know what it needs to fix, it’s like adding a whole new dimension to solve problems with basically.

It’s identity politics (yes I know the word is charged but this is also literally what it means). Living in a place does not confer someone instant total knowledge about the conditions of their experience. I live in capitalism and what do I know about capitalism or running a company? Has everyone living in a neoliberal hellhole read Mises, Hayek and the other so-called ‘real’ economists that championed the misery we experience?

Sure I can say that it’s pretty miserable, that wage is not keeping up with cost of living etc. None of this confers me the LTV, the theory of surplus value, etc. Just start with where you live and reframe from there imo. In capitalism there are also dissidents (we are proof), regime bootlickers and others “I just want to grill” apolitical types. There are also liars who tell you their new law will be good for workers when it’s the complete opposite (like France graciously “allowing” employees in bakeries and flower shops to work on May 1st at increased wages, the only reason people “agree” to work on May 1st is because wages are too small and don’t cover cost of living so people feel compelled to take the day).

This is why no one person is an authority on the subject of “this is how living here is like”, no matter how much zeal they might put in their speech. Most people don’t even fully understand their political system even if they have civics class in school, or just did the bare minimum to get through school and move on.

If you like AI China is doing a lot for local inference.

If you don’t like AI China is accelerating the death of western tech companies.

Basically thank China and Mr. President Xi Jinping.

Thanks for the link, I’ll have to check it out. I don’t see your addition as a point of criticism but precisely as additional sources of information; the strategy is part of a broader context that is often missing in the west as well, but I am also interested in how they use this technology on the ground and in the day-to-day, like what does it do and how does it work in this context? It’s an entire field of research and application that’s missing in the west (partly because we don’t produce our own food anymore lol). Outside of these circles if you start talking about smart agriculture you’ll get met with blank stares, we have no idea this is even possible.