I’ve said this before. The Chinese models are significantly better and will outcompete the models from the US, it was just a matter of people realizing that.
My other prediction, being that they will lobby for tariffs or banning Chinese models outright also seems to be coming true.
A Dark-Money Campaign Is Paying Influencers to Frame Chinese AI as a Threat
Build American AI, a nonprofit linked to a super PAC bankrolled by executives at OpenAI and Andreessen Horowitz, is funding a campaign to spread pro-AI messaging and stoke fears about China.
I do wonder how Europe is going to react. Will they just focus on their home grown Mistral or will they consider Chinese open weight models? I feel like the EU is quite wary of anything Chinese and that many people won’t fully comprehend the actual security risk and that they will initially dismiss are avoid them, but they can’t ignore it forever. Qwen 3.6 35B which can be ran at home is already leaving Mistral’s latest models in the dust.
Anthropic banned me for no reason last week, was merely using it for some SaaS project. Appealed, nothing. No reason given other than usage policy blabla.
Google Gemini, unusable, timed out 9 out of 10 times.
Very expensive, and unsustainable in my opinion. All the big boys lack capacity and I doubt they will be able to get that sorted.
Now running Qwen3.6:27B, opensourced by Ali Baba, locally. Works really well actually.
It’s honestly shocking how good Qwen 3.6:27b is, it actually outperforms Qwen 3.5:397b which itself was released only in February. I’m convinced that local models are the future. In a year or two, we’ll probably get to the point where local models are as good as Claude is today, and at that point it kind of doesn’t even really matter if frontier keeps getting better. It’s going to be good enough for vast majority of use cases. On top of that, you can already do a lot with tooling to make the model perform better. This paper is a great example. And I think this is very much an underexplored area. Current agentic harnesses are very primitive in nature, they just give the model some tools to play with, but do little in a way of guidance. ATLAS is a really interesting project in that’s attempting to make a smarter harness, and their results are pretty impressive. If you’re already running Qwen locally, I recommend checking it out.
Personallly, I think qwen3.6 is already pretty close to Claude.
Did not know about ATLAS, will definitely try that out.
It’s really getting there, it’s the first local model that really does feel like frontier quality.
that’s what happens when the western alternatives are 10x more expensive
and closed source
It’s like China has a few smart people left in their government!
They’ve fucked up plenty, their asset bubble is going to be crippling just like everybody else, but overall they are in a vastly better position than the US to advance their imperialistic goals.
I don’t see any reason why their asset bubble will be crippling. The reason it’s crippling in capitalist economies of the west is cause the government is expected to bail out the investors. In China, they just let these companies fail and then nationalize their assets if needed. If you look at the whole real estate bubble, there was never any crash there precisely because it was a handful of rich investors who ended up eating the loss instead of it being socialized.
As I understand it, housing speculation has filled a similar role to 401(k)s in the US, the relatively illiquid asset that “always goes up”.
People will feel a lot poorer when their apartment loses 70% of its value, and I don’t think that’s exactly the same as, say, a big leveraged builder going out of business.
You’re right of course the US won’t let one rich man go broke so it is worse there.
Chinese household savings hit a record high in 2024 https://www.wsj.com/livecoverage/stock-market-today-dow-jones-bank-earnings-01-12-2024/card/chinese-household-savings-hit-another-record-high-xqyky00IsIe357rtJb4j
90% of families in the country own their home giving China one of the highest home ownership rates in the world. What’s more is that 80% of these homes are owned outright, without mortgages or any other leans. https://www.forbes.com/sites/wadeshepard/2016/03/30/how-people-in-china-afford-their-outrageously-expensive-homes
Student debt in China is virtually non-existent. https://www.forbes.com/sites/jlim/2016/08/29/why-china-doesnt-have-a-student-debt-problem/
It’s pretty clear that there is very little negative impact from housing prices falling on the vast majority of the population. And it’s great news for young people who are moving from countryside to the cities and can now afford cheap housing.
The key reason why 2008 was a clusterfuck was because Obama decided to bail out the bankers and fuck the working class in the process.
I used to own a house free and clear. If the value dropped by half, I would still feel a lot poorer.
When the price of assets held as a store of value drops, the holders of those assets are impoverished. That’s already bad.
Now, when a fall in the price of assets held as a store of value drops, and that triggers a series of defaults that ripples across the financial system, that’s even worse.
I can’t see from the outside the degree to which Chinese society has been financialized, but I’m betting it’s more than the CCP wants to let on.
I mean, literally the first link in my reply is saying that household savings in China are at record high levels. 🤷
When the cost of hosing, the basic necessity for living, drops that is in fact a very good thing. And the record high savings number clearly demonstrates that housing is not a primary investment vehicle for majority of the population. As Xi put it, hosing is for living. This was an intentional policy choice and a correct one.
If you think that Chinese society has been financialized then you’re utterly clueless on the subject and have no business discussing it.
Can’t watch the video what ai do the Chinese offer? Is it safe?
It’s a lot cheaper to use hosted models and it’s open source so you can run them in house and customize them any way you like. And in terms of quality, there’s no significant difference. So, everybody is naturally using cheap and open models.
Among others, the Qwen Coder and Instruct models (by Alibaba) are generally well regarded; MIT Technology Review had a good analysis of Chinese open-weight models in February 2026 – https://www.technologyreview.com/2026/02/12/1132811/whats-next-for-chinese-open-source-ai/
