Seems like there's no official blog post with benchmark results yet. But I'm once again thankful for the Chinese AI labs for being open with their work and contributing it to the world under permissive licenses like this. The Fable 5 fiasco is just another reminder of how valuable these things are to have.
radious 2 hours ago [-]
The real news here is that Digg is still up :O
1f60c 1 hours ago [-]
It came back, died, and now it's back as some kind of weird AI-focused news aggregator.
binsquare 41 minutes ago [-]
this sentence hurts to read
stefan_ 39 minutes ago [-]
But they have such great AI generated insights on their AI stories:
"Many users praise Zhipu for open-sourcing GLM-5.2 under MIT with a 1M context window as a major step for accessible AI, while others respond with insults and anti-Chinese hostility."
LearnYouALisp 32 minutes ago [-]
I mean, it reads almost like an abstract of papers I've recently seen, with a similar info-cramming approach (somewhat like an editorial-SEO keyword bloat).
joshuat 17 minutes ago [-]
That's disappointing to hear, I remember the reboot news and thought they had a pretty solid team behind it. I guess gaining traction proved too difficult.
TiredOfLife 18 minutes ago [-]
It died and came back again last month
SilverElfin 13 minutes ago [-]
I actually found some of it useful. I saw some page where it helpfully pulled tweets from well known people relating to some story. So it’s not just some slop, or that’s how it looked to me.
skybrian 1 hours ago [-]
It seems to be basically a Twitter mirror with extra cruft?
theturtletalks 44 minutes ago [-]
More like a curator of all the AI news on Twitter. It’s also a great way to find trending AI projects on GitHub and elsewhere
mannycalavera42 1 hours ago [-]
digg goes along with slashdot and freshmeat memories.
good 'ol mems
jaggederest 40 minutes ago [-]
... for nostalgia's sake ...
It is official; Netcraft now confirms: BSD is dying
tamimio 1 hours ago [-]
That’s my thoughts exactly, had to click the home page to double check!
satvikpendem 1 hours ago [-]
Released at the exact same time, 5:21 pm (Chinese time), as when Anthropic received the letter from the government banning Fable, and explicitly citing other models becoming unusable.
throwaw12 22 minutes ago [-]
I wish they would write a blog post about capabilities of this new model, what to expect from this model, is it cheaper, is it faster or does it have better quality in the outputs.
But still, thank you for the release
bflesch 8 minutes ago [-]
Weird, z.ai does not resolve for me. Is there anything special about that domain?
Is it a coincidence that both MiniMax and Z.ai are releasing frontier open weights models right as the USG is trying to impose a cap on model capability offered to the public?
bontaq 2 hours ago [-]
I think Z.ai rushed a bit for release, for example GLM 5.2 is only available under the coding plan right now and they didn't do a big write up. Not even some charts and graphs about its performance!
This is around when people were predicting a new GLM to come out, so a couple corners clipped in order to catch the moment. I'm using it right now and it seems decent, but I haven't done heavy work with it yet. The expanded context window is great.
wolttam 1 hours ago [-]
This is typical for GLM releases.
lubujackson 2 hours ago [-]
I would say yes.
You think they were sitting on a release waiting for the right marketing moment?
bel8 2 hours ago [-]
Yes?
I have seen enough OpenAI and Anthropic carefuly timed marketing plays to expect it.
I would never announce GLM 5.2 in the same day as Fable or Apple's WWDC, for example.
enraged_camel 2 hours ago [-]
I think it's a possibility, because labs trying to one-up each other is a fairly common phenomenon at this point. Previous Opus releases were immediately followed by GPT releases, for example. At some point the timing stops being a mere coincidence.
halJordan 49 minutes ago [-]
No, not really. This has been telegraphed for a long time by everyone involved. HN denizens have been unashamedly anti-ai for years now, so what makes sense is the not knowing part of this audience. Chinese models are also not frontier models.
thefounder 2 hours ago [-]
No, Dario became too tiresome and annoying that someone had to do something. Personally I hope they ban Opus too. It will only provide more support for open models development.
Compare Dario horror posts with this from GLM release:
“ Intelligence should be open, accessible, and ready to build with, empowering every developer, everywhere.”
bugbubug 2 hours ago [-]
[dead]
ricointhemood 2 hours ago [-]
[dead]
2 hours ago [-]
testfrequency 2 hours ago [-]
Digg
edit: ouch, I’m a current Digg user. Even donated for their relaunch :(
ortekk 58 minutes ago [-]
With deluge of Chinese models popping up recently, I believe there's a few issues one needs to evaluate before deciding to use these models:
- Ethics. As known, ou American frontier AI companies are incredibly ethical. And I have yet to see any interviews or blog posts by Chinese companies where they talk about how they are ethical, or at least credible HN comments about it.
- Safety. Do they covertly sabotage or at least refuse to answer questions that could help cyber- and bioterrorists in their nefarious purposes? What about ML-related questions that could help terrorists create AI models without guardrails?
- Child safety. This is especially important with "free for all" open-weight models, most of which are Chinese (ever think about why that's the case?). How are we going to do age verification and KYC with models that anyone can just download on their computer?
- Intellectual property theft. How can we be sure that no output of our American frontier AI models was used while training these Chinese models?
Frankly, there's a plethora of other issues I don't have time to get into right now. Personally, I believe distribution of Chinese models in the US should be paused until they are required to submit models to the government for review and evaluation, to make sure they are made to Anthropic/OpenAI standards.
We need legal grounds for that.
Write to your congressman, congresswoman or congressperson and urge them to stop proliferation of dangerous non-American intelligence. This is a matter of national security and needs to be acted upon as soon as possible, preferably before IPO.
Xiol 47 minutes ago [-]
So hard to tell what is satire and what isn't these days.
Aldipower 49 minutes ago [-]
Yes, please ban all Chinese models in the US and stick to your US-centric stuff. Good for the rest of the world.
foxindustrial 51 minutes ago [-]
_incredibly ethical_
cyber_kinetist 49 minutes ago [-]
> our American frontier AI companies are incredibly ethical
Ah... sweet summer child.
> Intellectual property theft. How can we be sure that no output of our American frontier AI models was used while training these Chinese models?
The US AI models are already using pirated copyrighted material off the Internet. If Chinese models also do this, they're at least giving it back to the people by releasing their weights as open source.
yieldcrv 26 minutes ago [-]
Chinese models are the closest shining example of their ideological system working for the world than anything else they've ever done
From my perspective
hollerith 18 minutes ago [-]
Either that or releasing the models under permissive licenses is the only way they have get any attention in a market dominated by American companies.
(Also, they don't need to make a profit because their system does not prioritize profit potential when making investment decisions: it prioritizes alignment with directives out of Beijing, which in the past concentrated heavily on providing employment to young men and currently includes keeping up with the West in strategic technologies.)
49 minutes ago [-]
jauntywundrkind 39 minutes ago [-]
Fear Uncertainty and Doubt, the terrormongering, is worse than the terrors. Endless denial of society & possibility & progress: begone you demons.
billyjobob 50 minutes ago [-]
This is great but sails far too close to Poe's Law that I predict downvotes.
somenameforme 34 minutes ago [-]
I missed it at first. Then reread it, and wow - this is grade A satire of the sort rarely delivered anymore, probably indeed because of exactly what you're saying.
LearnYouALisp 28 minutes ago [-]
Well, if so it needs a little "touch" I guess
freehorse 21 minutes ago [-]
It is not necessary for (good) satire to be easily/immediately recognised as such imo
freehorse 22 minutes ago [-]
Before the "preferably before IPO" I honestly couldn't tell.
tiahura 51 minutes ago [-]
Is this a parody of the Chinese-funded anti-datacenter astroturfing?
bbg2401 26 minutes ago [-]
That you and other readers can't outright identify the comment as parody is actually quite disturbing to me.
Rendered at 18:56:20 GMT+0000 (Coordinated Universal Time) with Vercel.
"Many users praise Zhipu for open-sourcing GLM-5.2 under MIT with a 1M context window as a major step for accessible AI, while others respond with insults and anti-Chinese hostility."
It is official; Netcraft now confirms: BSD is dying
But still, thank you for the release
https://z.ai
This is around when people were predicting a new GLM to come out, so a couple corners clipped in order to catch the moment. I'm using it right now and it seems decent, but I haven't done heavy work with it yet. The expanded context window is great.
You think they were sitting on a release waiting for the right marketing moment?
I have seen enough OpenAI and Anthropic carefuly timed marketing plays to expect it.
I would never announce GLM 5.2 in the same day as Fable or Apple's WWDC, for example.
edit: ouch, I’m a current Digg user. Even donated for their relaunch :(
- Ethics. As known, ou American frontier AI companies are incredibly ethical. And I have yet to see any interviews or blog posts by Chinese companies where they talk about how they are ethical, or at least credible HN comments about it.
- Safety. Do they covertly sabotage or at least refuse to answer questions that could help cyber- and bioterrorists in their nefarious purposes? What about ML-related questions that could help terrorists create AI models without guardrails?
- Child safety. This is especially important with "free for all" open-weight models, most of which are Chinese (ever think about why that's the case?). How are we going to do age verification and KYC with models that anyone can just download on their computer?
- Intellectual property theft. How can we be sure that no output of our American frontier AI models was used while training these Chinese models?
Frankly, there's a plethora of other issues I don't have time to get into right now. Personally, I believe distribution of Chinese models in the US should be paused until they are required to submit models to the government for review and evaluation, to make sure they are made to Anthropic/OpenAI standards.
We need legal grounds for that.
Write to your congressman, congresswoman or congressperson and urge them to stop proliferation of dangerous non-American intelligence. This is a matter of national security and needs to be acted upon as soon as possible, preferably before IPO.
Ah... sweet summer child.
> Intellectual property theft. How can we be sure that no output of our American frontier AI models was used while training these Chinese models?
The US AI models are already using pirated copyrighted material off the Internet. If Chinese models also do this, they're at least giving it back to the people by releasing their weights as open source.
From my perspective
(Also, they don't need to make a profit because their system does not prioritize profit potential when making investment decisions: it prioritizes alignment with directives out of Beijing, which in the past concentrated heavily on providing employment to young men and currently includes keeping up with the West in strategic technologies.)