Oh you sweet summer child...
- Posts
- 3
- Comments
- 62
- Joined
- 2 yr. ago
- Posts
- 3
- Comments
- 62
- Joined
- 2 yr. ago
Excellent news, awful comparison in the title. Also, the script of the video definitely extensively used LLMs, there are so many phrases there that are the epitome of AI slop, but I'm glad to have learned about it nonetheless.
Stasi State or Socialist Paradise is the book I've read on it. It focuses mostly on the good aspects (and especially makes very good comparisons to other countries at the time) but also mentions and to some extent contextualized the bad aspects too. I think it would probably be a good starting point despite being more editorialized than a strict history book, and it has citations for further reading.
Side note: I read that book just before visiting Berlin, and of course I had to visit the DDR Museum there. It was a real shock seeing the propaganda there after reading the book. It almost felt like satire of anti-DDR propaganda. At one point the museum mentioned that people in the East had more sex on average than people in the West, and the museum tried to explain it as "there wasn't much else to do in the DDR." That and the fact that people repaired things ("a DIYers heaven") were the only good things the museum had to say about the country. They even said that its revolutionary child care systems were just communist indoctrination, and had nothing to say about women's rights or anything like that.
Honestly I have not been super impressed with Kimi K2. Maybe the thinking model is better, but in my experience GLM has been much better. I'll still give it a shot though.
I saw a Hacker News article of someone running Deepseek R1 for $6k, although still too expensive IMO
Do you remember what their setup was? My guess would be CPU inference with a metric fuckton of RAM if they were running it at the full quantization, which could work but would be pretty slow. But for $6k it'd be impossible to buy enough VRAM to run it at full quant on GPUs.
Thanks comrade!
Fair enough, I must say I haven't tried local models (tfw no GPU ;_;). I guess my take is that if it costs a tenth of a cent on OpenRouter to use a SOTA open source model, I might as well do that, but I can see the appeal of local models for easier queries.
I mean, it's a good editor without those features too, but imo they have a really good implementation of the LLM stuff
I've been using and enjoying Zed. It's like Cursor but open source and written from scratch in Rust instead of a VS Code fork. Highly recommend checking it out for seeing what coding agents are capable of. Zed recommends using Claude as the model, and in my experience Claude does work the best, but other models work quite well as well.
Zed, in my opinion, is the right way to do coding agents. It's part of your chat interface, and it can whip out an entire project from the ground up, but it really feels like you're in the driver's seat still, and it will show you diffs of what it changed and you can approve or reject them easily. It feels natural to use and also the best way to actually be a software engineer using a tool.
I don't really like DeepSeek as a coding model (though I do like it in a no-tools chat context), but GLM 4.6 (another open source Chinese model, made by a company called z.ai) has been very good imo and it works great in Zed. You can use it through OpenRouter or the official z.ai API. If you just buy tokens directly, it costs pennies and you only pay for what you use. Also, z.ai has a very cheap monthly plan for like $3 USD with a lot of usage.
Though, imo OpenRouter makes the most sense for trying out LLMs because you can just put a few dollars into their service and use it for GLM 4.6 or DeepSeek-V3.2 or Qwen-Coder or Claude or any other model and see which ones you like the most. Also, new models come out of China very regularly and are often quite good (Minimax-M2 just came out recently and seems super promising), and if your money is in OpenRouter you can just try it out easily. Plus, if you ever, at any point, put $10 into an OpenRouter account, you get 1000 free messages/day across their free models forever, which is the most generous "free" tier I've found. And then you can use those credits to play around with LLMs as coding agents. I put $10 in my account a few months ago and still haven't run out with reasonable usage as a coding agent.
One caveat about OpenRouter is that you should set it up to prioritize the first party APIs under the hood (e.g. if you want to use GLM 4.6, you can configure OpenRouter to use the official z.ai servers). This is because many providers quantize the models, which makes them less good, and also then it's a 3rd party getting your data rather than the (possibly CPC-affililated) company that actually makes the model.
IMO, local models don't make sense for the average person. A computer capable of running DeepSeek-V3.2 at full precision would cost well over $50k (iirc). Of course, you can't be sure your data isn't being mined without running it locally, but I'm just writing open source software for fun, so I don't really care that much.
Please feel free to ask any questions if you want more info! I have strong opinions about this stuff and I'm happy to share.
I'm aware of that, most folks on here really respect Mao but aren't Maoist. (I recognize you are probably just helping with a distinction, not trying to be snarky)
Yeah I figured, but just in case I wanted to make the clarification
the Maoists I've briefly talked to about this make the case that a lot of the talking points lobbed at Shining Path are ones that anti-communists lob at all socialist revolutions, and so that would be a reason to be critical of the sweeping assumption that Shining Path = bad.
Yeah this is basically what I meant when I said I wasn't well read enough on it.
The whole China debate is something I also struggle with. Like, China is a superpower at this point, why would they feel compelled to trade with the Phillipines? Why did they, for a time, side with Cambodia (iirc). I cannot think of reasons why China could not make a more moral choice in these instances.
I think the sibling comment from purpleworm has a good explanation for the Philippines. China seems to be trying very hard to not export revolution, for better or worse, and appear "politically neutral" at a world stage. I think this is something I'm personally disappointed in, but it also seems they're playing the long game and that the world, and the future possibility of leftist movements, would be much worse off if China were to cease to exist.
As for Cambodia, are you referring to China backing the Khmer Rouge (and in particular, against the Communist Party of Vietnam)? IIRC that was still under Mao, and I think a lot of the poor foreign policy decisions of that time are a direct result of the Sino-Soviet Split, which ended up manifesting in very weird ways (the US and Mao's China both backing the Khmer Rouge against Vietnam, which was backed by the Soviets, for example).
At the same time, I have heard compelling arguments that justify China's cobalt mines in terms of there being more reciprocal benefit as opposed to colonial extraction, and it does seem to operate differently from other forms of colonialism. So they do seem to operate in a unique capacity compared to the US.
Would you happen to have any resources about the cobalt mines? I would like to read more about that if you have it handy, plus it would be nice to keep it in my back pocket.
This is so fucking funny omg
Hard mode: those rules, plus
- Not X. Not Y. Just Z. (or, "But Z.")
- Not with X, but with Y.
- utterly
- —
- visceral
Here's my (not super well read, mostly based off of discussions with comrades) take:
- There's a difference between enjoying Mao and the contributions he made to the theory and being a Maoist
- Maoism was synthesized by Chairman Gonzalo of the Shining Path/Communist Party of Peru, which is a whole thing. Many communists denounce the Shining Path as being communist in name only, similar to the Khmer Rouge. I don't want to get into the weeds of this discussion because I am not well read enough on the matter, but it is certainly baggage that exists.
- Maoists unarguably do a lot of good in the world. The communist insurgency in the Philippines is a great example of this. The guerillas there have set up a parallel state to the bourgeois one and are resisting and expanding, and materially making things better for people in the rural Indigenous communities, and they're a real thorn in the "official" Filipino state.
- Maoists tend to be anti-China, both for ideological reasons (they think that socialism with Chinese characteristics is revisionism, and that China fell off/liberalized when Deng Xiaoping became the leader), and material ones, i.e. China selling arms to the Filipino bourgeois government, which it uses to oppress the revolution. This is because China trades with everyone, and is not the Soviet Union. In fact, many/most Maoists think that there are no actually existing socialist states. Thus, Maoists are often grouped in with ultra-leftists and Trots, who also don't think there are any actually existing socialist states, and criticize AES states from a left perspective.
I think @ColombianLenin was calling Maoists out particularly over their criticisms of China, which, given the context, seems like a relevant point of discussion.
It seems that EqualyzAI, the main company mentioned in the article, isn't even using a hosted version of DeepSeek. They're running the model themselves on their own hardware, so nobody but EqualyzAI is seeing this data, not even DeepSeek or Huawei.
(Edit: Also EqualyzAI only has to pay for the server costs of running the model, they don't have to pay DeepSeek anything to be able to self host it.)
From your second quote:
Like all companies that build on DeepSeek, they can choose to either host their products locally and pay for computing and storage infrastructure, or go through providers like Huawei. EqualyzAI does the former.
So that means that DeepSeek is not getting a cent from this company. It's open-weight, meaning if one has sufficiently powerful hardware they can just run DeepSeek, unlike OpenAI state of the art models, which can only be run by companies that contract with OpenAI to get the weights (as far as I know, this is basically just Google (Vertex) and Amazon (Bedrock)).
But... even considering that DeepSeek is a more lightweight/efficient programme and China overall is rapidly expanding their electricity output... it still seems hard to imagine any profit is actually happening
I think DeepSeek is absolutely burning money. Right now, almost all Chinese models are all open-weight. I've seen numerous hypotheses for why this is the case, but I think the one that convinces me the most, at least for DeepSeek, is that they're doing it as advertising/recruiting. But the revenue that DeepSeek has is only from charging per token on their API as described in your first quote, and they're competing with every other GPU provider for these prices, so it's an aggressive race to the bottom. It's possible that DeepSeek is even running this at a loss to get more training data from people using their API.
In any case, DeepSeek has made a lot of innovations relating to doing more training with less power, because they are currently relatively GPU-poor. NVIDIA chips are hard to come by in China and so DeepSeek can't really buy any more of the top tier models than they already have. Some of these are used for running the inference for the API, and some are used for the training. But even with all of these optimizations, it costs a lot of money to train an LLM, and it's hard to imagine that with how often they're releasing models, they're actually breaking even, given that at best they have small margins on their API.
Democratic People's Republic of Chattanooga
Any chance you could post an (anonymized) sample of the writing style? think have a pretty good vibes-based intuition about what's Al. Also,
(just as the state of the world)I can deal with the sideways ones but the trains in my city have backwards seats :(
The material conditions of the world would have to be so different from what they currently are in order for this to happen, so it's hard to speculate. Tensions currently are slightly increasing between ROK and US (it seems) but the ROK bourgeoisie would certainly not willingly give up their class position, and I'm sure that the US would go to bat for them if push came to shove. So I don't think it would be able to happen until the US's influence at the world stage diminishes sharply. Currently it seems to be on the decline but obviously still very relevant.
My prediction of how it would ultimately happen would be that the proles in ROK would gain more class consciousness as a result of labour struggles, and the labour movement would get powerful enough to threaten revolution, in which case a post-recolution society could consider/discuss reunification.
I think the current free travel status is very much dependent on the US's role in geopolitics, so if that sharply declining is a precondition for reunification, I think that it would be a much different situation than it is now.
I heard that before going on strike, Canadian posties were escalating labour action incrementally, where first they refused overtime work, then they threatened to stop delivering flyers, which apparently makes up a substantial amount of Canada Post revenue.
Choosing to interpret "N/A" as "we don't use soap here"