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

GLM-5.3: Frontier coding with emergent cyber capabilities

z.ai

Post reply on HN
Print view View original

161–170 of 626 posts

Re: GLM-5.3: Frontier coding with emergent cyber capabilities

#161

Post by dhx »

Gigachad wrote: post #124
MangoCoffee wrote: post #53

Earlier quoted context omitted.

OpenAI and Anthropic are both seeking trillion IPOs, while Chinese labs are pumping out open-weight models that are free for US providers to host and monetize. These Chinese models cost less of US SOTA models to run, even if they are less capable. Providers can just run them, offer cheap tokens, and pocket the margin. I just don't see how you justify a trillion valuation for US AI labs when the underlying models are…

This is going to be catastrophic. Whether AI works or is useful or not isn’t even the question anymore. It can fulfil every promise Sam Altman has been making and will still make no financial sense to justify these valuations.

I take it from [1] (transcript of recent DeepSeek CEO discussion with investors) that DeepSeek would disagree on the immediate catastrophic impact to the likes of OpenAI or Anthropic. The reason is even though technology parity mostly exists, only OpenAI, Anthropic et al have the inference capacity to gain market share and generate revenue. Chinese vendors don't have the chips needed to scale up inference and gain market share, and the DeepSeek CEO doesn't think this would happen in optimistic circumstances in the next 3 years, but thinks it might be possible in 5 years.

In summary, regardless of country of origin, availability of inference capacity is the moat protecting the likes of OpenAI and Anthropic, not technology superiority.

[1] https://www.fredgao.com/p/deepseeks-liang-wenfeng-breaks-his

Re: GLM-5.3: Frontier coding with emergent cyber capabilities

#162

Post by re-thc »

chrismsimpson wrote: post #78
MangoCoffee wrote: post #53

Earlier quoted context omitted.

OpenAI and Anthropic are both seeking trillion IPOs, while Chinese labs are pumping out open-weight models that are free for US providers to host and monetize. These Chinese models cost less of US SOTA models to run, even if they are less capable. Providers can just run them, offer cheap tokens, and pocket the margin. I just don't see how you justify a trillion valuation for US AI labs when the underlying models are…

> Providers can just run them, offer cheap tokens, and pocket the margin. There’s an assumption that you can spin up the infra and acquire customers within that margin

> There’s an assumption that you can spin up the infra and acquire customers within that margin

Only Nvidia and approved friends can at the moment. Nvidia can even backstop your loan required.

Re: GLM-5.3: Frontier coding with emergent cyber capabilities

#164

Post by scotty79 »

petesergeant wrote: post #155

Their own hardness (ZCode) seems to be a GUI, which doesn't work for me. They say they support other harnesses. However, it seems like I can inject the plan into other harnesses, like Claude Code[0]. Does anyone who's been using GLM models for a while have a strong feeling for if it does better in some harnesses than others, or should I just use my favourite harness? 0: https://docs.z.ai/devpack/tool/others

I use it with random harnesses. It behaves consistently.

Re: GLM-5.3: Frontier coding with emergent cyber capabilities

#167

Post by valleyer »

rob74 wrote: post #116

I'm not that up to date with the latest AI developments, but I noticed that this article seems to use "Cyber Capabilities" as a shorthand for the model's ability at cybersecurity tasks? Is that now an established expression, same as "crypto" now refers to cryptocurrencies rather that cryptography? Because "cybernetics" actually means something different (yeah, old man yelling at clouds, I know)...

Yeah, I've noticed it recently, too. I'd be interested to know where it started.

Re: GLM-5.3: Frontier coding with emergent cyber capabilities

#168

Post by ArvidSu »

jjcm wrote: post #70

Same image->html test as I showed in the Gemini 3.7 flash thread. Note that GLM isn't multimodal, but it still was able to generate something similar-ish by writing a python script to inspect the image and extract elements from it. Original images: https://image.non.io/neonRamenDesigns.webp GLM 5.3 build: https://html.non.io/neonRamenGLM5.3 Opus 5 build for comparison: https://html.non.io/neonRamen For having no visi…

That's super impressive given that it doesn't have vision! Intelligence overcomes blindness.

Re: GLM-5.3: Frontier coding with emergent cyber capabilities

#169

Post by sscaryterry »

z4y5f3 wrote: post #128

Apparently they are scanning OSS and popular software at scale and disclosing the vulnerabilities they found: https://cvd.z.ai/ Most of these are under embargo, but it seems there are a lot of CVE here from a wide range of popular software, many considered critical or high. I understand the argument of "people are not actively looking", but isn't the cost for such a scan getting lower by the week, and Anthropic's Pro…

Interesting... So Chinese models are not so bad?

Re: GLM-5.3: Frontier coding with emergent cyber capabilities

#170

Post by topato »

gruturo wrote: post #143
netdevphoenix wrote: post #137

Earlier quoted context omitted.

Not exciting for anyone directly or indirectly invested in a frontier lab or its partners. And that is a lot of people, including you.

No time like the present to pull out and reduce your exposure. I brought this up in my employer's forums 4 months ago and honestly it's been clear even before then. In particular, the upcoming IPOs of both oAI and Anthropic will likely be disastrous for the public - the floor is falling from under them and I don't know if they can be scrappy and work with fewer resources - their internal culture may not support this.…

Yes buuuut…. I do quite a bit of day trading (maybe closer to scalping) for the first few hours the market is open, everyday. Anecdotally: despite everyone knowing its valuation was ridiculous, I rode that SpaceX train pretty hard and made a pretty penny.

I close-out all my positions by end-of-trading everyday… so when the day came when there was a very clear and very scary indicator during early trading hours, quickly followed by SpaceX’s catastrophic fall right after opening bell, that was the end of my involvement….

And I fully expect oAI and anthro to be the same way. They’re being propped up with private loans, subsidies, and other tricky bookkeeping techniques. You would think their CEOs would pivot away from their current public personas. Ironically, they are like a poor man’s Elon Musk… and that doesn’t bode well for their companies

Post reply on HN