☆ Yσɠƚԋσʂ ☆

  • 578 Posts
  • 341 Comments
Joined 6 years ago
cake
Cake day: January 18th, 2020

help-circle

  • It’s going to be interesting to watch because the usual rationale that China is spying on you doesn’t work for open models people can run locally. And we now see that even in the US, majority of the startups are using Chinese models already. So, there 's a lot of capital invested into continuing to have access to these models. On the other hand, we have these massive AI corps in the US whose entire business model is threatened, and who will spend colossal amounts of money lobbying to ban the competition.








  • 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.



  • 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.















  • I think it’s safe to say that by 2025 there was compromise on the 8xxx or 9xxx Kirin chips out there at that time.

    No, it’s not to safe that at all. You’re making a really wild jump here. It’s like saying that if old version of chips iPhone uses were vulnerable we could say that current ones are. You are firmly into speculation territory here. I don’t know why it’s so hard for you to just admit that your thesis is not supported by evidence.

    Remember though that I’m not claiming these leaks represent the extent of le capability today, but the extent of their capability then. Over time we can expect (and can see based n the expansion of their claims and the capabilities asserted in their leaks over time!) that they would get access to new methods of compromising phones, we just can’t know the exact extent until something leaks.

    Except this equally applies to American phones, and in addition to that, there is a risk of intentional backdoors. So, to reiterate for the tenth time now, American phones are just as likely to be vulnerable to malicious attackers, and on top of that they are produced by companies directly working with US and Israel making it likely they would have intentional backdoors. That’s a strictly worse scenario.

    Again, the evidence speaks directly against your analysis. You’re trying to contort the evidence here to fit your narrative instead of looking at it objectively.

    Phone security with burgerland characterisitcs lol.



  • So, to recap, you have not provided any source that suggests modern Chinese devices that are in production today are vulnerable, or any less secure than ios or android. The brute force support is not for the new devices according to your link. It’s for the older devices. I specifically screenshotted the relevant text in the last reply. The fact that a old device could be brute forced does not extend to the argument that new device can. There is no logical reason to suggest that.

    Your assertion that pixels, graphene and iphones are safest against the hardware/software tools the police have is not backed by facts. The capabilities of these toolkits DO NOT back up your assertion because they do not show that CURRENT devices are more vulnerable.

    There is nothing that would convince me of your point because your point is not based on facts available to us. It’s based on your assertion that android and ios are more secure which is not supported by evidence available. If you were to provide concrete proof that these devices that are currently on sale are vulnerable, then we could have a discussion about that https://consumer.huawei.com/en/phones/

    And given that there is no evidence available to suggest that Chinese devices are more vulnerable, then the next question to ask is which vendor is more likely to be compromised. The answer there is obviously that it would be the American vendor.