Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

China’s open-weights AI strategy is winning

werd.io

Post reply on HN
Print view View original

591–600 of 978 posts

Re: China’s open-weights AI strategy is winning

#591

Post by tomtheelder »

SwellJoe wrote: post #562
geophile wrote: post #352

The lesson of the last 50 years of the computer and software marketplace is that free and low-end eventually wins. - PCs destroyed minicomputers. Mainframes survive, but serving a much tinier portion of the market than they used to. - PC office productivity software destroyed expensive professional products. - Windows (low end) and Linux (free) completely destroyed the UNIX marketplace, and again, have taken huge mar…

"I think that in 10-15 years, we are going to have consumer PCs (and phones!) running models doing pretty much anything that frontier models can do right now." I don't think it'll take 10-15 years. Gemma 4 31B in the 4-bit QAT is competitive with the frontier of less than three years ago and runs on any high-end 32GB gaming PC GPU or a large-ish Mac. The question is whether the frontier will continue to get better at…

I think the question is even a bit more nuanced than that. Even if frontier models can maintain a big gap that gap has to actually _matter_. If a local model satisfies my everyday use cases adequately then I may not really care that a frontier model is 5, 10, 100x better at ultra high order reasoning tasks.

I think that reality is probably not all that far off for a huge swath of use cases.

Re: China’s open-weights AI strategy is winning

#592

Post by Razengan »

geophile wrote: post #352

The lesson of the last 50 years of the computer and software marketplace is that free and low-end eventually wins. - PCs destroyed minicomputers. Mainframes survive, but serving a much tinier portion of the market than they used to. - PC office productivity software destroyed expensive professional products. - Windows (low end) and Linux (free) completely destroyed the UNIX marketplace, and again, have taken huge mar…

> The lesson of the last 50 years of the computer and software marketplace is that free and low-end eventually wins.

Except, uhm, for ..you know, that one company that hit a trillion cap

But you're right: Just like how million dollar computers with 1 bit of RAM performing 1 operation a second and taking up a colossal cave were replaced by $1 laptops with a zillion zekabytes running at a trillion hertz (exact values may vary),

the sprawling data centers of today with a quadrillion GPUs powered by black holes will get replaced by breakthroughs in hardware and most importantly, algorithms:

The human brain is proof right here that intelligence doesn't require dinosaur-sized hardware or eat half the sun every second.

I actually wonder if we're seeing the limits of discrete binary logic: Maybe it's high time to give analog ternary and all that funky jazz an honest try :)

Re: China’s open-weights AI strategy is winning

#593

Post by elictronic »

kube-system wrote: post #302
elictronic wrote: post #288

Earlier quoted context omitted.

China wants the US economy to flounder. Building our entire growth model on software that can be copied and taken by a small group of people will have no possible consequences.

China is looking after their own interests, but they absolutely don't want their largest export market to struggle. The global economy is not a zero-sum game, and the idea that it might be is the root of many of our policy issues in the US.

Chinas interests today are a US with reduced power and a global economy more focused on China.

Someone loses power for someone else to gain. It’s literally the definition of a zero sum game? China targets areas it thinks it can win and dominate in the future.

Re: China’s open-weights AI strategy is winning

#594

Post by Danox »

TeMPOraL wrote: post #371
elictronic wrote: post #285

Earlier quoted context omitted.

Nvidia selling more GPU's at the cost of it's datacenter business is pretty close to Kodak selling digital cameras at the cost of film. The data center side is so bloated anything that eats into it is a huge negative. Their data center business brings in 20x the gpu market. Local open weight models will be what pops the bubble and China will do anything in it's power to enable that pop.

> Nvidia selling more GPU's at the cost of it's datacenter business is pretty close to Kodak selling digital cameras at the cost of film. You mean they resisted the idea trying to protect their legacy business, and it ended up all but killing them?

Kodak, Xerox, IBM, Motorola of Schaumburg, Illinois are perfect examples of companies that did not want to upset the apple cart, another example outside tech of tech that did not want to go into the future was US Steel, and while I’m at it to a lesser extent Ford and GM are teetering since 1973…

Re: China’s open-weights AI strategy is winning

#595

Post by comrade1234 »

x313 wrote: post #225
comrade1234 wrote: post #15

Earlier quoted context omitted.

We're using deepseek with the idea that we would switch to something better when more of our customers are using the ai features but it ends up deepseek is awesome for what we're doing and so we probably won't switch because it's so much cheaper. The ai libraries we use let us switch models with just a configuration change.

Similar story here. DS models are absurdly good value for mid-end tasks. I've found DSv4 Flash to be ~10% the cost of GPT-5.4-mini/Claude Haiku at similar performance. We used to pay OpenAI >1m$/month for fraud classification, NER, etc. Sadly the US companies no longer care about non-coding-agent uses. I imagine uptake will continue to increase as the corporate infra improves. Right now it's still bad - for example,…

Yeah we don't use it for coding either. Basically text analysis within specific constraints. It works beyond my expectations.

Re: China’s open-weights AI strategy is winning

#596

Post by calebkaiser »

realusername wrote: post #467
_fizz_buzz_ wrote: post #452

Earlier quoted context omitted.

So, OpenAI and Anthropic say the Chinese models are only as good because they distill their models. How true is that. I am sure it adds something. But is it more like a marginal 1% improvement or something really significant?

I also don't believe it, if it was as easy as that, we would have hundreds of competitors. The truth that Anthropic and OpenAI will not say, is that these Chinese labs have a lot of talented people.

I am strongly in favor of open models, open source ML more broadly, and am pretty critical of the cynical positions adopted by major US labs vis a vis open models.

But this is an insane characterization. Literally every single researcher and executive at OpenAI and Anthropic would say that "these Chinese labs have a lot of talented people." They hire from them (and vice versa). Tencent's chief AI scientist was poached directly from Deepmind, who poached him from Anthropic, etc etc etc. Do you think there are just zero people from China working at US frontier labs?

And even beyond that, the entire ML ecosystem (including people at OpenAI and Anthropic) get excited about research published by Chinese labs. Deepseek's GRPO paper set the ecosystem on fire for a little while.

The contention from OpenAI and Anthropic around distillation has basically been "Labs that distill from us get to bootstrap their model at a much lower price point". Or, in other words, "If we didn't invest in building the teacher model, it wouldn't be possible for these labs to distill their student model." Which I'm not very sympathetic to, but is a far cry from how you're characterizing it.

Re: China’s open-weights AI strategy is winning

#597

Post by JamesBarney »

geophile wrote: post #537
jeffreyrogers wrote: post #479

Earlier quoted context omitted.

> free and low-end eventually wins Not in SaaS which is what LLMs are. You can get VMs for much cheaper than AWS, Microsoft, and Google offer them but large companies (and startups) are happy to pay a premium for the support, reputation, and reliability that they perceive those companies as offering. Same thing for some of the managed database providers who are effectively selling a very heavily marked up version of…

But aren’t the frontier models heavily subsidized? That’s not sustainable.

This really depends what you mean. If you spend $5.00 on anthropic tokens does that cost Anthropic more than $5.00 to serve? No.

If you amortize all of their training and salaries over that $5.00 then yes.

If you only amortize the training costs of that specific model then again we're back to no.

Re: China’s open-weights AI strategy is winning

#598

Post by z3ugma »

Varelion wrote: post #217
z3ugma wrote: post #208

Earlier quoted context omitted.

This is the history of enclosure of the commons since the beginning of capitalism

I'm sorry, is there further reading material on this? I am pretty new to political/economic science.

https://www.progress.org/wiki/enclosure/

https://www.thelandmagazine.org.uk/articles/short-history-en...

Re: China’s open-weights AI strategy is winning

#599

Post by gexla »

I don't think we're far enough into this game to know what winning even looks like. If the US frontier labs are losing, it may have been their own missteps on handling such massive change of being the fastest growing apps ever. Until China can ever show me something new rather than just cheaper, then I just see efforts from that region as a force of commodification. OpenAI will go down in history as the company who brought what many understand to be AI to the first billion people.

Re: China’s open-weights AI strategy is winning

#600

Post by JamesBarney »

awllau wrote: post #552
apparent wrote: post #163

Earlier quoted context omitted.

The full quote: > When entrepreneurs walk into the offices of Andreessen Horowitz (a16z), a big American venture-capital firm, the odds these days are that their startups are using AI models made in China. “I’d say 80% chance [they are] using a Chinese open-source model,” says Martin Casado, a partner at a16z. This is very different from what the author portrays. It may be the case that many pre-funded startups are u…

Casado says he was misquoted and walked this back: https://x.com/martin_casado/status/1990462245541982546 "Well, not quite. I'd say 20-30% use open source. Of those I'd say 80% use Chinese based models. So closer to 16-24%." Reads pretty differently.

This should really be the #1 comment on this post.
Post reply on HN