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

261–270 of 626 posts

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

#261

Post by budsniffer952 »

meowface wrote: post #209
VulgarExigency wrote: post #190

Earlier quoted context omitted.

The CIA ran one of the world's largest cryptography companies, for DECADES[1]. Are you truly so naive that you believe intelligence agencies that have more to gain from stifling the discovery of vulnerabilities they know of and use wouldn't do so? [1] https://www.washingtonpost.com/graphics/2020/world/national-...

I believe it is unlikely. (Not because I do not believe NSA is hoarding 0-days, but for many other reasons.) I'm curious: to any professional vulnerability researchers reading this, what do you think?

[flagged]

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

#262

Post by mcintyre1994 »

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…

If you look at the distribution of their findings in the linked post, most of theirs are issues introduced a long time ago, almost all before 2006.

Complete speculation, but I wonder if they and Anthropic are scanning very different codebases and Anthropic's skew would be in the other direction.

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

#264

Post by mcintyre1994 »

andy_ppp wrote: post #232
zorked wrote: post #181

Earlier quoted context omitted.

There's a chance that the real reason why they want to ban Chinese models is that they are so good at fixing bugs and preventing exploits that intelligence agencies have been using for espionage and surveillance for a long time.

Anyone who knows anything realises banning things is a) impossible and b) your enemies will use them anyway, you are just depriving your own side of the advantages.

In this case depriving US companies would be the point though, so that's not necessarily a disadvantage.

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

#265

Post by matheusmoreira »

MangoCoffee wrote: post #53
aliljet wrote: post #5

This is absolutely still shy of Sol and Fable, but only just by a hair. Ridiculous results. There's still not a compelling economic reason to drop OpenAI courtesy of the ludicrous reset addiction that's taken place, but it feels like we're on the precipice. How are you all toying with running this kind of thing in a mega quantized way locally? Two weeks out from released weights, but this is still just GLM 5.2 with p…

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…

> I just don't see how you justify a trillion valuation for US AI labs

Market is irrational.

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

#266

Post by eru »

cromka wrote: post #214
Gigachad wrote: post #124

Earlier quoted context omitted.

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 seriously need to start considering the scenario in which this leads to next global financial crisis.

These crises are manufactured by the central banks.

Compare and contrast how the dot-com bust did _not_ lead to global financial crises. Nor did Black Monday, nor the recent string of bank failures in the US.

('Manufactured' above means that central banks are responsible. I make no judgement on intent here. Around 2008 it was incompetence by the Fed and ECB as far as I can tell. The Fed started paying interest on excess reserves and the ECB even increased rates. Twice. Amongst quite a few other missteps.)

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

#267

Post by budsniffer952 »

numpad0 wrote: post #222
sscaryterry wrote: post #169

Earlier quoted context omitted.

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

They've always been good enough for double digit less money. Always. Anyone thinking "Chinese models fake models built using dirty distillation scam" don't know what they're talking about. Distillation is just forcing the model to use an exam prep workbook for training instead of generic publicly available textbooks. The models themselves has to be smart enough for that to work. It's the exact same thing as Asian tig…

Their best models are getting more and more expensive, and still aren't SOTA.

It's almost like there's an actual cost to developing these models, and the Chinese don't have magic dirt that allows them to do it at a fraction of the cost.

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

#268

Post by andsoitis »

layer8 wrote: post #225
andsoitis wrote: post #204

Earlier quoted context omitted.

> Just hosting it in the EU and promising not to retain / sell the data let's you charge a healthy extra and compete in many areas other players can't. It's been a few years. Has anyone done this successfully yet?

There are a over a dozen EU open-weight providers. I’m not sure if they are even charging that much of an extra. EU-based clients have little reason to use non-EU inference providers.

> EU-based clients have little reason to use non-EU inference providers.

Which models are most popular in Europe?

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

#269

Post by alienbaby »

One htought I had; if The chinese allow unfettered access to cyber capabilties while th US does it's best to neuter it's model releases, from China's point of view they have the US all tied up in knots dealing with problems they don't give people the tools to solve. China giggles as it watches the US under threat from people using it's models. The US is restricting citizens from owning this particular kind of weapon, while China is handing it out to the wrolds citizens freely. It feels like the US would only come out worse overall?

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

#270

Post by KronisLV »

ljosifov wrote: post #193
KronisLV wrote: post #87

Their coding plan switched to credits, didn’t it? What are the rate limits like, compared to Anthropic or Kimi K3? I remember trying their Coding Plan out before the change and the 5 hour limits felt too restrictive then even for light/medium work, especially cause of the whole peak and off-peak thing: https://blog.kronis.dev/blog/z-ai-s-glm-5-2-is-a-great-model... Nowadays, I’d probably go with their Max plan if the…

Wdym "sadly they don’t support using Claude Code"? For the longest time that's all Zai supported - Claude code. I'd run it via export ZAI_ANTHROPIC_BASE_URL="https://api.z.ai/api/anthropic" export ZAI_ANTHROPIC_AUTH_TOKEN="$ZAI_API_KEY" claude-zai() { { local -; set -x; } 2>/dev/null ANTHROPIC_BASE_URL="$ZAI_ANTHROPIC_BASE_URL" ANTHROPIC_AUTH_TOKEN="$ZAI_ANTHROPIC_AUTH_TOKEN" claude "$@" } $ claude-zai I liked Claude…

I am not talking about GLM models being served through an Anthropic compatible API, that part is perfectly fine and I'm glad they support it!

I am talking about ZCode, the program, being unable to delegate to other harnesses, like using Claude Code (or even OpenCode) within their UI, so that an Anthropic subscription can be used, because Anthropic don't let you use 3rd party harnesses directly.

It's basically what Paseo: https://paseo.sh/ and Kepler https://www.gitkraken.com/kepler and Zed https://zed.dev/ support doing.

ZCode doesn't seem to work at that level, it instead feels comparable to OpenCode or Codex or Claude Code directly, while also being desktop oriented - you just make API calls directly within it.

It's okay if it's not a goal of theirs, it's just that their UI is really really nice and that would be a cool direction for them to also go in some day.

Post reply on HN