Disclosure - Fireworks kindly gave me some free credit to experiment with GLM to help write this article.
needs to be higher up
GLM 5.2 and the coming AI margin collapse
401–410 of 495 posts
Re: GLM 5.2 and the coming AI margin collapse
#402Earlier quoted context omitted.
So Deepmind/Meta/etc... would all have to cut off their EU offices even though they hold incredible talent? I'm just not finding this scenario likely.
Does meta has an AI training data center in EU? They can just operate and provide normal access to their services, just block AI access. This is already happening, apple would not release new Siri in EU (granted it's due to a regulation clash) but this would be a testing point. If Europeans are still paying the same prices for sub par services/products to their American counter parts, it's win-win situation for those…
Meta's FAIR has several R&D offices in the EU, yes. So you are saying their labs can conduct R&D on models in the EU, potentially even train them there, they just can't have production LLM inference serving or release the model weights? I'm just not seeing it.
A not-insignificant portion of the AI/ML research community is in the EU.
Re: GLM 5.2 and the coming AI margin collapse
#403Earlier quoted context omitted.
Yeah, Linux-the-kernel does have a stable ABI indeed, but this is not relevant for most ISV desktop software out there. In my comment above I was referring to Linux-the-OS (aka GNU/Linux). The userspace libs don't have a stable ABI at all, and this is a widely discussed problem. Other operating systems built on top of Linux-the-kernel don't have this problem, Android has a really stable ABI.
> The userspace libs don't have a stable ABI at all, and this is a widely discussed problem. And DLL hell isn’t? Or the shambolic mix of 32 and 64 bit libraries on Windows? Anyway, desktop binaries are increasingly rare for business software.
I dont think dll hell is a problem anymore
Re: GLM 5.2 and the coming AI margin collapse
#404Earlier quoted context omitted.
Bedrock is really out of date with the models it offers, to the extent that I'm not sure they even have plans to update what's on there now they have the deal with Anthropic. They're still offering Qwen 3 , not even 3.5 and certainly not 3.6. GLM 5 is the newest z.AI model they have, when it's 5.2 that would be the one to worry Sonnet. There are some ok models on there (Qwen 3 Coder Next is usable and fast, for insta…
Because these models are not going to stand up legally
That's a stretch but it doesn't change the outcome.
Re: GLM 5.2 and the coming AI margin collapse
#405Earlier quoted context omitted.
Multiple providers (who need to make a profit) offer the same 4.40 rate for glm-5.2. It's not subsidized. Deepseek's 0.86 or whatever is likely subsidized but alternate providers offer it for a price comparable to glm-5.2.
According to deepseek themselves, their current rates are NOT subsidised. They have published tons of articles dedicated to performance and efficiency engineering. Feel free to have a look...
Why is no other inference provider offering similar prices then?
Re: GLM 5.2 and the coming AI margin collapse
#406Regarding the lack of vision part, if you are using Claude or opencode, I've made a skill[1] that let's you talk with any models in Claude/opencode mid-session. You ask "Have claude opus to look at this PDF for a second opinion" during a session of claude with GLM5.2 or opencode with GLM5.2 It doesn't need to pass whole conversation history as context (unlike /model), you can ask follow up to that forked model (which…
FWIW I've seen subagents remain open for followups on latest Claude.
Will give it a try thanks for the heads up. Mind if I ask if you are referring to SendMessage?
I was testing on Claude code 2.1.196. SendMessage was not available. Skimming through their change log didn’t seem to have anything related to SendMessage. There is “ Fixed SendMessage silently misrouting when a re-spawned agent reuses a previous agent’s name — the tool now detects the mismatch and asks the caller to retarget” from 2.1.199. Not sure if they are related
Re: GLM 5.2 and the coming AI margin collapse
#407People have been making claims about the commoditization of llms since chatGPT, and they've been wrong every time as quality and prices and differentiation have increased.
Re: GLM 5.2 and the coming AI margin collapse
#408Earlier quoted context omitted.
> the world's dominant power by virtue of having roughly half the total world population If human+ level AI takes off one would expect to see a great decoupling of power from population.
Sure, but until then, proportional to each of [population, education/educated workers, capital, instantaneous industrial base, energy supply]. Asia's diverse, but I'd say they seem to be doing pretty well with rapid improvements across all fronts. In comparison, the US's weaker (not weak-weak, just weaker) areas currently seem to be educated workers, instantaneous industrial base, and energy supply (relative to rapid…
> People talk about China having demographic issues too, but they're a dictatorship, they can make it shift if they care to.
China has a significantly bigger problem with demographics than the EU does, it is just on a slightly longer fuse, compare:
https://ourworldindata.org/grapher/children-born-per-woman?f...
The big drop in Chinese fertility is going to be very disruptive in the near future, because it is much less gradual than European trends and the retiree/workers ratio is going to spike much harder because of that.
Having full authoritarian control is not gonna change anything now because it is already much too late (action would have been required like 35 years ago).
Best they can do is get through it somewhat smoothly.
edit: This is an even better visualization (projected working age population fraction)
https://ourworldindata.org/grapher/population-young-working-...
Re: GLM 5.2 and the coming AI margin collapse
#409Meanwhile: > China’s Ministry of Commerce has led meetings over the past month with major AI companies, including Alibaba, ByteDance, and http://z.ai/ , to discuss measures that would restrict overseas access to cutting-edge AI models, including models that have not yet been released. > The discussions reportedly include not only closed-source models but also open-weight models. > Future regulations could take the fo…
which is a result of the United States and other western forces (Israel) weaponizing AI to create propaganda
Re: GLM 5.2 and the coming AI margin collapse
#410Meanwhile: > China’s Ministry of Commerce has led meetings over the past month with major AI companies, including Alibaba, ByteDance, and http://z.ai/ , to discuss measures that would restrict overseas access to cutting-edge AI models, including models that have not yet been released. > The discussions reportedly include not only closed-source models but also open-weight models. > Future regulations could take the fo…
It absolutely blows my mind that everyone is abundantly aware that money is not shared freely, and yet they are already popping champagne that the CCP are surely going to share power freely.
I never thought the chinese were giving away valuable technology out of kindness. I thought having them be free and open weight was part of a strategic move to undermine US tech dominance. I suppose if that was the case something has changed or maybe they simply didn't care until recently