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

GLM 5.2 and the coming AI margin collapse

martinalderson.com

Post reply on HN
Print view View original

261–270 of 495 posts

Re: GLM 5.2 and the coming AI margin collapse

#261

Post by port3000 »

jmyeet wrote: post #105

I think OpenAI, Anthropic and SpaceX are going to envy the dinosaurs because there's not asteroid coming for them, there's three: 1. There will be no moat around frontier AI models in the future. China is going to make sure that happens. It's a national security interest for them. DeepSeek was the first shot across the bow for that but it won't end with them. There are other labs and there are non-Chinese actors too.…

Interesting, so point 2 means that a lot of the hardware being installed now won't be able to run the frontier models of 2029? How does that change the demand for compute/models in the future, I can imagine that even if OpenAI/Anthropic have a moat in 2029 there will be so much older hardware and such a hangover from that investment boom that there will be very little installed capacity that can run it

Re: GLM 5.2 and the coming AI margin collapse

#262

Post by nr378 »

typ wrote: post #236

Unlike the belief that frontier AI is expensive due to a high margin, and going to be expensive if there is no competition. My understanding is that, under certain circumstances (which is most likely true), the price will be driven down just because of profit seeking. The frontier LLM labs run on a huge fixed cost and very low marginal cost. They need the economies of scale to make sense of the business (an incentive…

> The frontier LLM labs run on a huge fixed cost and very low marginal cost.

> Imagine that you want to buy a few B300s to run GLM 5.2 and rent the service out to other people. How could this business be viable and sustainable in the first place?

My understanding is the frontier labs have huge fixed costs and relatively low marginal costs because they have to bear the cost of training the model/R&D, and then amortise that cost over their userbase.

By contrast, if I buy a few B300s and run GLM5.2 and rent the service out to other people, I can be profitable at a comparatively very small scale because I got the model for free.

Re: GLM 5.2 and the coming AI margin collapse

#264

Post by DrScientist »

fny wrote: post #5

I'm not convinced raw costs matter: 1. Compute costs collapsed since the advent of Cloud and yet hyperscalers still have fat margins. 2. Many open source office suites exist yet none compete with the ubiquity of gsuite or office. GitHub, Slack are similar examples. 3. Both Windows and macOS dominate the home desktop space despite free alternatives existing for a long time. 4. Many formerly open source infrastructure…

1,2,3 are dominated by platform stickness or even active lock-in.

Can't say I see the same advantages to stop you switching the model you use.

> It seems that enterprises will pay top dollar for service guarantees, integration, and someone they can sue.

Sure. Though it does depend on whether you need regular updates. If you want the model to be aware of the latest research - then fine. However it already does the job, you might prioritize stability over constant change.

> It's nobody gets fired for buying IBM all over again.

Except they when they did when IBM was no longer good value for money.

> but I don't see any historical analogues

None at all? You mentioned IBM - who is using AIX on IBM hardware in 2026? Who is using Solaris on Sun hardware? It's pretty much all gone to linux on commodity hardware.

Remember Netscape - thew browser company? Killed by Microsoft bundling of IE. How hard would it be for Apple to bundle GLM based services?

Re: GLM 5.2 and the coming AI margin collapse

#265

Post by Aissen »

fny wrote: post #5

I'm not convinced raw costs matter: 1. Compute costs collapsed since the advent of Cloud and yet hyperscalers still have fat margins. 2. Many open source office suites exist yet none compete with the ubiquity of gsuite or office. GitHub, Slack are similar examples. 3. Both Windows and macOS dominate the home desktop space despite free alternatives existing for a long time. 4. Many formerly open source infrastructure…

> 1. Compute costs collapsed since the advent of Cloud and yet hyperscalers still have fat margins.

No, compute costs collapsed (before mid-2025) because of normal technological progress on all fronts of compute.

Re: GLM 5.2 and the coming AI margin collapse

#266

Post by dannyw »

ChadNauseam wrote: post #50
tarpitt wrote: post #35

Earlier quoted context omitted.

GLM-5.2 is not as good as Opus, it's better. I can abliterate GLM-5.2 and have it work on projects that Opus refuses.

I'd guess opus refusals are not an issue for 95%+ of people. Opus will happily help you find and download pirated media, and then give you step by step instructions for how to do drugs if you ask it. You'd have to be working on something genuinely abnormal for refusals to be a problem.

Eh. Refusals for security related tasks seem to be constantly increasing.

Re: GLM 5.2 and the coming AI margin collapse

#268

Post by throwthrowuknow »

I don’t see it. GLM 5.2 seems noticeably worse than Opus and especially GPT 5.5, the poor vision capabilities are also a massive strike against it since these are a huge improvement in the frontier models that can make all the difference when working on anything visual. Running it locally is its biggest advantage but for a lot of use cases that isn’t needed and is a burden to set up and maintain.

Re: GLM 5.2 and the coming AI margin collapse

#269

Post by saberience »

peab wrote: post #107
pants2 wrote: post #86

Earlier quoted context omitted.

If you're developing on top of LLM APIs directly, this is definitely not true. There are differences in how context caching works, in what's available through native harnesses, the types of tools you're fine-tuned on (GPT uses apply_patch while Claude uses edit, with different formats), the API surface (Agents SDK, Responses API, Managed Agents), cost structures, and best-practice guidance all around. Not to mention…

It really depends on what you're doing, but most LLM usage and agentic runs are pretty interchangeable in my experience, and it's usually trivial to switch. If anything, you're better off supporting multiple LLMs as backup because most model providers have been so inconsistent with working all the time

Dude it’s not trivial to switch because the behaviors are different!

You’re clearly not building a product based on an LLM.

I’m still using various old Anthropic and OpenAI models for products I’ve built and released because I can’t risk the behavior changing in unpredictable ways and the users being pissed.

It’s much easier to switch out some deterministic software than an LLM which you’ve spent a ton of time on testing and benchmarking and understanding its nuances. Changing it is like replacing an employee who’s critical to the business.

Re: GLM 5.2 and the coming AI margin collapse

#270

Post by WhereIsTheTruth »

fny wrote: post #5

I'm not convinced raw costs matter: 1. Compute costs collapsed since the advent of Cloud and yet hyperscalers still have fat margins. 2. Many open source office suites exist yet none compete with the ubiquity of gsuite or office. GitHub, Slack are similar examples. 3. Both Windows and macOS dominate the home desktop space despite free alternatives existing for a long time. 4. Many formerly open source infrastructure…

Cost is a measure of scarcity

Once something is abundant, it's hard to justify extracting big margins from it

Which is why so much effort goes into manufacturing scarcity instead

Post reply on HN