Hacker News new | past | comments | ask | show | jobs | submit
Every single day, three things are becoming more and more clear:

    (1) OpenAI & Anthropic are absolutely cooked; it's obvious they have no moat
    (2) Local/private inference is the future of AI
    (3) There's *still* no killer product yet (so get to work!)
This has got to be bait..

1) OpenAI and Anthropic are killing it, and continue to do so, their coding tools are unmatched for professionals.

2) Local models don't hold a candle to SOTA models and there's nothing on the horizon that indicates that consumers will be able to run anything close to what you can get in a data center.

3) Coding is a killer product, OpenAI and Anthropic are raking in the cash. The top 3 apps are apps in the app store are AI. Everyone who knows anything is using AI, every day, across the economy.

The grandparent is definitely wrong on (3). Yes, coding is a killer product, I agree with you.

On (2), I agree with you for local models. BUT, there are also the open source Chinese models accessible via open-router. Your argument ("don't hold a candle to SOTA models") does not hold if the comparison is between those.

On (1), I agree more with the grandparent than with your assessment. Yes, OpenAI and Anthropic are killing it for now, but the time horizon is very short. I use codex and claude daily, but it's also clear to me that open source is catching up quickly, both w.r.t. the models and the agentic harnesses.

>BUT, there are also the open source Chinese models accessible via open-router.

I thought so myself, but after burning a lot of money on OpenRouter in a few days I just subscribed to Z.ai's Coding Pro plan and using the subscription is much, much friendlier with my wallet.

> the open source Chinese models accessible via open-router

And? They aren't as good as SOTA models. Even the SOTA model provider's small models aren't worth using for many of my coding tasks.

loading story #47685402
I don't want to respond to 100 comments about the same thing, and this one happens to be on top, so, in my humble opinion:

(1): You don't have to be an Ed Zitron disciple to infer that OpenAI and Anthropic are likely overvalued and that Nvidia is selling everyone shovels in a gold rush. AI is a game-changing technology, but a shitty chat interface does not a company make. OpenAI and Anthropic need to recoup astronomical costs used in training these models. Models that are now being distilled[1] and are quickly becoming commoditized. (And frankly, models that were trained by torrenting copyrighted data[2], anyway.) Many have been calling this out for years: the model cannot be your product. And to be clear, OpenAI/Anthropic most definitely know this: that's why they've been aquihiring like crazy, trying to find that one team that will make the thing.

(2): Token prices are significantly subsidized and anyone that does any serious work with AI can tell you this. Go use an almost-SOTA model (a big Deepseek or Qwen model) offered by many bare-metal providers and you'll see what "true" token prices should look like. The end-state here is likely some models running locally and some running in the cloud. But the current state of OpenClaw token-vomit on top of Claude is fiscally untenable (in fact, this is why Anthropic shut it down).

(3): This is typical Dropbox HN snark[3], of which I am also often guilty of. I really don't think AI coding is a killer product and this seems very myopic—engineers are an extreme minority. Imo, the closest we've seen to something revolutionary is OpenClaw, but it's janky, hard to set up, full of vulnerabilities, and you need to buy a separate computer. But there's certainly a spark there. (And that's personally the vertical I'm focusing on.)

[1] https://www.anthropic.com/news/detecting-and-preventing-dist...

[2] https://media.npr.org/assets/artslife/arts/2025/complaint.pd...

[3] https://news.ycombinator.com/item?id=9224

loading story #47688870
> And to be clear, OpenAI/Anthropic most definitely know this: that's why they've been aquihiring like crazy, trying to find that one team that will make the thing.

Anthropic is up to $30B annual recurring revenue. I wish I had failing business models like that.

> Token prices are significantly subsidized and anyone that does any serious work with AI can tell you this. Go use an almost-SOTA model (a big Deepseek or Qwen model) offered by many bare-metal providers and you'll see what "true" token prices should look like.

I'm not sure what think you are saying here, but if you look at the providers for both "almost-SOTA model (a big Deepseek or Qwen model)" or at the price for Claude on AWS Bedrock, Azure or on GCP you will quickly see inference is very profitable.

loading story #47687606
loading story #47688262
No killer product? Coding assistants and LLM's in general are the single most awe-inspiring achievement of humanity in my lifetime, technological or otherwise. They've already massively improved my and others' lives and they're only going to get better. If pre and post industrial revolution used to be the major binary delineation of our history, I'm fairly confident it will soon be seen as pre and post AI instead.
I know right? 8-year-old me dreamed of being able to articulate software to a computer without having to write code. It (along with the original Stable Diffusion) are Definitely one of the coolest inventions to ever come along in my lifetime
Coding assistants are currently quite hard to run locally with anything like SOTA abilities. Support in the most popular local inference frameworks is still extremely half-baked (e.g. no seamless offload for larger-than-RAM models; no support for tensor-parallel inference across multiple GPUs, or multiple interconnected machines) and until that improves reliably it's hard to propose spending money on uber-expensive hardware one might be unable to use effectively.
This is an argument against the grandparent's points (1) and (2), not their point (3).
loading story #47684063
Computers get better and cheaper. That’s not a forever problem.
loading story #47685283
No killer products... just robots that can do vulnerability analysis at the level of a decent security engineer and write code without tiring.
I've also been using the LLM in Posthog and it has been impressive. I need to check if I can also plug a MCP/Skill to my actual claude code so that I can cross reference the data from my other data source (stripe, local database, access logs etc.) for in depth analysis
This might be up your alley - have Posthog and a ton of other SaaS tools connected so you can run analysis across quant/qualitative data sources: https://dialog.tools
> Coding assistants and LLM's in general are the single most awe-inspiring achievement of humanity in my lifetime

Landing a man on the moon is way more impressive. Finding several vaccines for a once in a century pandemic within a year of its outbreak is and achievement that in its impact and importance dwarfs what the entire LLM industry put together has achieved. The near-complete eradication of polio, once again, way more important and impactful.

Those are all good things, but with the current AI boom we've invented something with the potential to invent those kinds of things on its own, if not now then in the near future. It's far more important and impactful to invent a digital mind that can invent an arbitrary number of vaccines than to just invent one vaccine, no matter how hard it was to invent the vaccine by hand.
yeah, painting yourself into a corner at 10x speed is hardly the most awe-inspiring achievement of humanity.
I don't see how its possible to think this. AI coding assistants are some of the most useful technologies ever created, and model quality is by far the most important thing, so I doesn't make sense why local inference would be the path forward unless something fundamentally changes about hardware.
The hardware will change. We know that.
> no moat

I'd like to think the superior product wins. But Windows still thrives despite widespread Linux availability. I think sometimes we can underestimate the resilience of the tech oligopolies, particularly when they're VC-funded.

VC can spend all the money in the world and it won't matter if the cost of switching providers is effectively zero.

If I want to switch from Windows to Linux, I have to reconsider a whole variety of applications, learn a different UX, migrate data, all sorts of annoyances.

When I switch between Codex and Claude Code, there is literally no difference in how I interact with them. They and a number of other competitors are drop in replacements for each other.

>I'd like to think the superior product wins. But Windows still thrives despite widespread Linux availability.

That's because by most metrics Linux is inferior is Windows.

What benefit is there to dropping $50k on GPUs to run this personally besides being a cool enthusiast project?
Why would anyone need more than 640Kb of memory?
Exactly the point though. In the 640KB days there was no subscription to ever increasing compute resources as an alternative.
Well, there kinda was - most computing then was done on mainframes. Personal / Micro computers were seen as a hobby or toy that didn't need any "serious" amounts of memory. And then they ate the world and mainframes became sidelined into a specific niche only used by large institutions because legacy.

I can totally see the same happening here; on-device LLMs are a toy, and then they eat the world and everyone has their own personal LLM running on their own device and the cloud LLMs are a niche used by large institutions.

loading story #47684441
Intel has just released a high VRAM card which allows you to have 128GB of VRAM for $4k. The prices are dropping rapidly. The local models aren't adapted to work on this setup yet, so performance is disappointing. But highly capable local models are becoming increasingly realistic. https://www.youtube.com/watch?v=RcIWhm16ouQ
That's 4 32GB GPUs with 600GB/s bw each. This model is not running on that scale GPUs. I think something like 96GB RTX PRO 6000 Blackwells would be the minimum to run a model of this size with performance in the range of subscription models.
> I think something like 96GB RTX PRO 6000 Blackwells would be the minimum to run a model of this size with performance in the range of subscription models.

GLM 5.1 has 754B parameters tho. And you still need RAM for context too. You'll want much more than 96GB ram.

Is it so hard to project out a couple product cycles? Computers get better. We’ve gone from $50k workstation to commodity hardware before several times
Subscription services get all the same benefits from computer hardware getting better. But actually due to scale, batching, resource utilization, they'll always be able to take more advantage of that.
It will run exactly the same tomorrow, and the next day, and the day after that, and 10 years from now. It will be just as smart as the day you downloaded the weights. It won't stop working, exhaust your token quota, or get any worse.

That's a valuable guarantee. So valuable, in fact, that you won't get it from Anthropic, OpenAI, or Google at any price.

That's why we all still use our e machines its never obsolete PCs. Works just the same it did 20 years ago, though probably not because I've never heard of hardware that's guaranteed not to fail.
Agree directionally but you don't need $50k. $5k is plenty, $2-3k arguably the sweet spot.
as a local LLM novice, do you have any recommended reading to bootstrap me on selecting hardware? It has been quite confusing bring a latecomer to this game. Googling yields me a lot of outdated info.
First answer: If you haven't, give it a shot on whatever you already have. MoE models like Qwen3 and GPT-OSS are good on low-end hardware. My RTX 4060 can run qwen3:30b at a comfortable reading pace even though 2/3 of it spills over into system RAM. Even on an 8-year-old tiny PC with 32gb it's still usable.

Second answer: ask an AI, but prices have risen dramatically since their training cutoff, so be sure to get them to check current prices.

Third answer: I'm not an expert by a long shot, but I like building my own PCs. If I were to upgrade, I would buy one of these:

Framework desktop with 128gb for $3k or mainboard-only for $2700 (could just swap it into my gaming PC.) Or any other Strix Halo (ryzen AI 385 and above) mini PC with 64/96/128gb; more is better of course. Most integrated GPUs are constrained by memory bandwidth. Strix Halo has a wider memory bus and so it's a good way to get lots of high-bandwidth shared system/video RAM for relatively cheap. 380=40%; 385=80%; 395=100% GPU power.

I was also considering doing a much hackier build with 2x Tesla P100s (16gb HBM2 each for about $90 each) in a precision 5820 (cheap with lots of space and power for GPUs.) Total about $500 for 32gb HBM2+32gb system RAM but it's all 10-year-old used parts, need to DIY fan setup for the GPUs, and software support is very spotty. Definitely a tinker project; here there be dragons.

loading story #47686339
The 4-bit quants are 350GB, what hardware are you talking about?
qwen3:0.6b is 523mb, what model are you talking about? You seem to have a specific one in mind but the parent comment doesn't mention any.

For a hobby/enthusiast product, and even for some useful local tasks, MoE models run fine on gaming PCs or even older midrange PCs. For dedicated AI hardware I was thinking of Strix Halo - with 128gb is currently $2-3k. None of this will replace a Claude subscription.

Google doesn't release Gemma 4 if Gemini is similiar good.

We probably talk abuot a year of progress diffeerence.

Its also still quite expensive for an avg person to consume any of it. Either due to hardware invest, energy cost or API cost.

Also professionally I don't think anyone will really spend a little bit less money of having the 3th quality model running if they can run the best model.

I'm happy that we reach levels were this becomes an alternative if you value open and control though.

(1) is absolutely not true if you actually use these models on a regular basis and include Google in here too. The difference in reliability beyond basic tasks is night and day. Their reward function is just so much better, and there are many nuanced reasons for this.

(2) is probably true but with caveats. Top-tier models will never run on desktop machines, but companies should (and do) host their own models. The future is open-weight though, that much is for sure.

(3) This is so ignorant that others have already responded to it. Look outside of your own bubble, please.

> Top-tier models will never run on desktop machines

Sorry, but you don't know that

I mean it's not hard to understand that if good model can run on consumer hardware, even better models can run in data centers
If we get to the point where a local model can reliably do the coding for a good majority of cases, then the economic landscape changes significantly. And we are not that far from having big open weight models that can do that, which is a first step
Larger, yes, absolutely. Better? Right now it seems that bigger is better, but if we are thinking about long term future, it's not obvious that there isn't a point of diminishing returns with regards to size. I can also imagine a breakthrough, where models become much smaller, with the same or better capabilities as the current, very large ones.
loading story #47686203
I was trying to use Claude.ai today to learn how to do hexagonal geometry.

Every time I asked a question it generated an interactive geometry graph on the fly in Javascript. Sometimes it spent minutes compiling and testing code on the server so it could make sure it was correct. I was really impressed.

Anyway I couldn't really learn anything since when the code didn't work I wasn't sure if I had ported it wrong or the AI did it wrong, so I ended up learning how to calculate SDF and pixel to hex grid from tutorials I found on google instead.

This is also my exact experience
Posted this after mythos came out? The hutzpah
No moat: yes. Cooked: no. It's a race. Why assume they're going to lose? It relies on (2) which is only true if AI usefulness plateaus at some level of compute. That's a huge claim to be making at this stage. (3) AI has lots of killer products already. The big one is filling in moats. Unrealized potential though for sure.
>(1) OpenAI & Anthropic are absolutely cooked; it's obvious they have no moat

I think big corporations will continue to use them no matter how cheap and good other models are. There's a saying: nobody was fired for buying IBM.

How good would open source models be if they couldn't distill higher quality private models?
(3) is simply a lie spread by engineers who have no other context. I manage some real estate (mid-term rentals) and everyone I know has switched over to AI robo-handlers to do the contact at this point. It's almost a passive investment at this point. Some can even handle interfacing with contractors and service requests for you. Revolutionized the field in my opinion.
The model is the killer product