Why would any company release open weights once the investment money stops ?
Releasing open weights have been basically a PR move, the moment those companies need to actually make money they will cut it out as that reduces their client base.
They DO NOT want you to run AI. They want you to pay them to do it
Minimax just released a new model yesterday. You're conflating one company with a countries entire industry. There's more than just Qwen coming out of China.
z.ai did go public on the HK exchange. They are under pressures similar to other public companies.
I know that China models are increasingly being trained and run using Huawei chips instead of Nvidia. I know China has a surplus of electricity from renewables (wind, solar, hydro).
Two years ago a lot of people thought GPT-4o was usable for software development. I didn’t really find that to be the case in general but certainly it could do a lot of useful things. And now Qwen3.5-8B is just as capable and runs fine on an M2 MacBook Air.
QWEN3.5 coder next runs to ~84k context before it poops out on AMD395+ w/128GB. Most of what it's good at is boilerplate find/replace/copy/paste; but being able to scaffold things out and touch up 20-30% of the code is pretty sweet.
People keep repeating this without any real thought behind it because of the high profile resignations on the Qwen team. Meanwhile the Minimax team just released a new open weights version of their 229B model yesterday. So much for that narrative.
The AI landscape in China is larger than just Qwen and Alibaba.
The statement was that China was giving up on open weights, they didn't say anything about licensing. Licensing on these models has always been hit or miss depending on which lab and which release.
but context of the statement is discussion about corps do grab and rent strategy. My understanding is that referenced Chinese model can't be argument in this context, and there is no recent 200B+ params Chinese models with friendly license.
That license is more like business source license vs open source license.
Of course, but for how long? Do you think that companies will keep giving away valuable assets for free forever, or do you think that in the near future there's going to be an open weights model that's so good that people keep using it indefinitely instead of going back to frontier model providers?
The first one is just incredibly naive, the second might be true for some people, for some tasks, but it's not going to capture the majority who're chasing the latest and greatest to "keep up".
> Do you think that companies will keep giving away valuable assets for free forever
If China is forced to choose between giving the entire AI market to the US or releasing free models, they'll be releasing free models as long as it's necessary.
Every time you release the models you even the playing field out for the competition, which ruins a lot of the advantage your bigger competitors had. It also lets smaller players work on the latest tech and then you can make deals with them.
do you think that in the near future there's going to be an open weights model that's so good that people keep using it indefinitely instead of going back to frontier model providers?
We are almost at that point now, where the harnesses and tools are more important drivers of functionality and performance than the model weights themselves. We'll get there.
z.ai models are open weights. GLM-5.1 is very close to Opus with obvious exception of session length.
Only academic models will be true open source as companies can't legally afford to disclose learning inputs.
In regards to "They want to train models on our engineering to replace us". Some software engineers in China can run circles around some of the best teams in Silicon Valley. Days of U.S. hegemony are over. I recommend you make peace and make friends.
I've been using z.ai and codex latest models since last September.
Each release has been an improvement.
codex handles longer sessions but the quality seems to decline and it tends to over engineer and lose focus. It will happily add slop on top of slop...which may pass immediate tests of "code works" but doesn't pass my criteria of "code as craft"
I'm using z.ai GLM with opencode. It's obvious when GLM loses its mind when the session gets too long.
I've been using AI to support programming for around 3 years now. The models have gotten amazing. However, unless there is a significant breakthrough I have determined that it's best for me to focus on short sessions.
I a) organize my work, b) improve my AGENTS.md, ensure source has appropriate comments to guide the models to the patterns and separation of concerns c) use shorter sessions d) review and test without AI. This approach means I still own my code. The AI is just an assistant.
With this approach GLM-5.1 is an excellent model. I never run out of token allotment on z.ai or codex plans. At this point, I only keep my OpenAI subscription as the ChatGPT desktop app is excellent at long web research tasks and I get codex with it.
AST of what? Will it read my clojure code's forms as such? What if my source file has a paran balancing error? I feel I'm thinking of this at the wrong level/angle.
I cannot remember a case, in the last 10 years at least, when I committed code that does not compile. Why should I share that? Also, tree-sitter sort of handles that.
> code that does not compile. Why should I share that?
If you collect test cases for compilers, for example.
> tree-sitter sort of handles that
My worry is that stability of committed ASTs would depend on tree-sitter being stable, and it might be difficult to guarantee that for languages are still in flux. Even most well established languages gain new grammar once every few years, sometimes in backward incompatible ways.
Maybe you meant tree-sitter itself will also be versioned inside this repository?
Also, there is an option to pick a codec for a particular file. Might use tree-sitter-C, might use general-text. The only issue here, you can't change the codec and keep nice diffs.
I usually want the codex approach for code/product "shaping" iteratively with the ai.
Once things are shaped and common "scaling patterns" are well established, then for things like adding a front end (which is constantly changing, more views) then letting the autonomous approach run wild can *sometimes* be useful.
I have found that codex is better at remembering when I ask to not get carried away...whereas claude requires constant reminders.
Depends on your app cache needs. If it's moderate, I'd start with postgres...ie. not have operate another piece of infra and the extra code. If you are doing the shared-nothing app server approach (rails, django) where the app server remembers nothing after each request Redis can be a handy choice. I often go with having a fat long lived server process (jvm) where it also acts for my live caching needs. #tradeoffs
I've started using a container (podman) which is just for the AI tools. I start it up for Codex etc and let it access to the appropriate code directory outside the container.
Anyone else using this approach? Ideas on improvements?
Don't mark the folder as trusted when you open in VsCode. The number of other hooks that may exist is going to be hard to track down (especially because each addon may add their own).
This may only provide a flalse sense of security. Afaik, there is no way to disable workspace settings taking priority over user settings, so a malious repo can easily override them and reenable automatic tasks.
On macOS systems, this results in the execution of a background shell command that uses nohup bash -c in combination with curl -s to retrieve a JavaScript payload remotely
Unrestricted outbound connections, specially from curl/wget/bash
I think that's a bit ungenerous: there is a push and pull between security and seamless user experience and it's never obvious where the line should be set. You really only figure out which way to move it after someone complains.
I'm not sure about the other ones, but I know that helix supports language servers by default and it does not have a workspace trust system like vscode, so LSPs can automatically execute code when you enter a directory
If your hypothetical happens, yes.
China has been working hard to turn domestic investment away from housing. A trustworthy domestic stock market is key.
* The Shanghai and Hong Kong stock market seems to have improved regulatory enforcement. I have no way of measuring this...just stories from others.
* Over the past 10 years the China gov pressed on with building more housing in part to dilute value. Each year they have warned that houses are for living, not speculation. Last year, they dumped a huge amount of cheap lending into the market to provide movement...warning this is the last step...a month ago the 2026 gov priorities list removed protecting the housing market...first time in modern history. Expectation is the next two years will see realized losses in property. It would be a huge mistake if the gov hasn't ensured regulatory enforcement of other segments have not reached maturity for the retail investor. We'll see...
* As for civil courts, over the past 20 years I've run into quite a few stories from friends and business colleagues that needed to go to China court. The stories are similar to what you may hear in the US. No one suggested the court/process itself was dodgy/unfair.
> No one suggested the court/process itself was dodgy/unfair.
for civil disputes, i am sure they are.
For disputes between the gov't and you, i highly doubt it. Is there a single instance of the gov't being sued for a policy that was meant to be political in nature affecting the supplicant?
Even someone like jack ma is unable to use the courts to obtain any justice - his Ant Financials IPO was shut down for political reasons, and he was reeducated. There's no such thing as due process in china.
Name me a single country where a rich person goes against the government and wins? You just don’t see it happen much in the US because the government is run by rich capitalists, but pretty much every country is the same.
It happens all the time here. Wealth isn't even a precondition, but indeed, one needs time and/or money. It helps being organized with other people to share the burden. Over here we have also got the ombudsman.
It is a matter of degrees. The harder it is for a poor individual to be done justice against the government, the weaker the rule of law. On a tangent, parties that play the horn about "law and order" usually mean "rules for thee but not for me".
Not sure where here is for you. But anyway, even if you can win a battle you can’t win a war. If a government wants to do something, it will regardless of how any individual person, rich or not, feels about it.
It just so happens that most western “democracies” are run by rich people, so they can avoid all that unpleasant business by just running the government in the first place.
I would say that’s an extension of the idea that rich people run the US government, which runs global organizations such as the world bank, which runs these ISDS courts.
China is just big enough to be able to ignore these global orgs.
Now try to win a case against the interests or connections of a high ranking official in China.
"Law and order" is not equal to "the rule of law". Both China and the US ascribe to the idea of "might makes right", which is in essence an organized form of lawless state. It is conceptually the same as in criminal gangs, only with vastly better optics. That is why anyone not in power should strive for a rules-based order, for their own best interest at least.
Not sure why you think China is Orwellian and the US isn’t when ICE is literally kidnapping people off the streets. Wake up mate, the Orwellian is coming from inside the house.
European governments regularly lose cases brought by individuals in both domestic and European courts; below are some well-known examples across different countries and legal issues.
E.g.
Broniowski v. Poland (ECtHR, 2004)
Doğan and Others v. Turkey (ECtHR, 2004)
Hirst v. United Kingdom (No. 2) (ECtHR, 2005)
Scoppola v. Italy (No. 3) (ECtHR, 2012)
KlimaSeniorinnen v. Switzerland (ECtHR, 2024)
These judgments show that individuals and civil society groups can hold European states accountable for violations involving property, voting, asylum, climate, and broader rule‑of‑law issues.
They often lead to legislative change, financial compensation, or policy reversal, and many are used as precedent by lawyers and activists in new cases across Europe.
I’m not a lawyer, and many of these cases are not famous enough to be reported on in a format easily understandable to a layperson. I’m also not going to read through case resolutions to respond to a hacker news comment. I did take a cursory look at the examples you wrote though.
I will admit that my original statement lacks nuance, which makes it easy to nitpick.
Having read some of your cases though, a pattern emerged: it’s usually supra national organizations adjudicating these cases, and the nations are not bound by the rulings.
For example, in Hirst vs UK it was ruled that it’s a violation of human rights to deny prisoners the vote, and yet the UK government deliberately ignored that ruling and as a result prisoners still can’t vote in the UK. Not to mention that when this case was brought up in a UK court it was dismissed.
Hirst v United Kingdom (No 2) (2005) ECHR 681 is a European Court of Human Rights case, where the court ruled that a blanket ban on British prisoners exercising the right to vote is contrary to the European Convention on Human Rights. The court did not state that all prisoners should be given voting rights. Rather, it held that if the franchise was to be removed, then the measure needed to be compatible with Article 3 of the First Protocol, thus putting the onus upon the UK to justify its departure from the principle of universal suffrage.
There are numerous examples of citizens winning court cases against the government.
Take the just the uk, three examples:
Anti‑protest regulations (Liberty v Home Secretary)
Air pollution litigation (ClientEarth v UK Government)
Rwanda asylum plan (AAA & Others v Secretary of State
Windrush - Members of the Windrush have repeatedly challenged the Home Office over wrongful detention, removal, and denial of rights, leading to government admissions of unlawfulness and an official apology in 2018.
Your claim is false. The world is not the same the world over, civil liberties are better in some places than in others.
Prisoners still can’t vote, people are getting arrested for peacefully protesting holding signs, and the Rwanda ruling was overruled by parliament and the only reason the plan was stopped is because the PM changed.
> Each year they have warned that houses are for living, not speculation. Last year, they dumped a huge amount of cheap lending into the market to provide movement...warning this is the last step...a month ago the 2026 gov priorities list removed protecting the housing market...first time in modern history.
Perhaps one of a few genuinely positive policies which only China can do. Meanwhile western countries will rather stab their economies to death than accept even just stagnating real estate prices.
>> No one suggested the court/process itself was dodgy/unfair.
Not sure where this is coming from. The EU recently just won a WTO dispute[1] against China that prohibited patent holders (often EU companies) from pursuing or enforcing patent infringement cases in non-Chinese courts -- it violated several provisions of the WTO's Agreement on Trade-Related Aspects of Intellectual Property Rights (TRIPS Agreement), including Articles 1.1, 28.1, and 28.2.
Foreigners generally view the Chinese court system with significant skepticism, primarily due to a perceived lack of judicial independence from the ruling Communist Party (CCP), opacity, and the use of the judiciary to serve political goals.
1. DS611: China – Enforcement of Intellectual Property Rights
I've been using GLM for over 6 months and pretty happy.
reply