Live data from Hacker News

DeepSeek-v3.2: Pushing the frontier of open large language models [pdf]

huggingface.co

311–320 of 485 posts

Re: DeepSeek-v3.2: Pushing the frontier of open large language models [pdf]

#311
post #261

Earlier quoted context omitted.

Ah, so exactly like Uber, Netflix, Microsoft, Amazon, Facebook and so on have done to the rest of the world over the last few decades then? Where do you think they learnt this trick? Years lurking on HN and this post's comment section wins #1 on the American Hypocrisy chart. Unbelievable that even in the current US people can't recognize when they're looking in the mirror. But I guess you're disincentivized to do so…

Except domestic alternatives to the tech companies you listed were not driven out by them, they still exist today with substantial market share. American tech dominance elsewhere has more to do a lack of competition, and when competition does exist they're more often than not held at a disadvantage by domestic governments. So your counter narrative is false here.

> American tech dominance elsewhere has more to do a lack of competition,

Do you believe the lack of competition is purely because the products are superior?

US tech is now sort of like the dollar. People/countries outside the US need and want alternatives to hedge against in the event of political uncertainity but cannot do it completely for various reasons including arm twisting by the US govt.

One example is some govts and universities in the EU are trying to get rid of MS products for decades but they are unable to.

Re: DeepSeek-v3.2: Pushing the frontier of open large language models [pdf]

#312

Earlier quoted context omitted.

> Infrastructure owners with access to the cheapest energy will be the long run winners in AI. For a sufficiently low cost to orbit that may well be found in space, giving Musk a rather large lead. By his posts he's currently obsessed with building AI satellite factories on the moon, the better to climb the Kardashev scale.

The performance bottleneck for space based computers is heat dissipation. Earth based computers benefit from the existence of an atmosphere to pull cold air in from and send hot air out to. A space data center would need to entirely rely on city sized heat sink fins.

And the presence of humans. Like with a lot of robotics, the devil is probably in the details. Very difficult to debug your robot factory while it's in orbit.

That was fun to write but also I am generally on board with humanity pushing robotics further into space.

I don't think an orbital AI datacentre makes much sense as your chips will be obsolete so quickly that the capex getting it all up there will be better spent on buying the next chips to deploy on earth.

Re: DeepSeek-v3.2: Pushing the frontier of open large language models [pdf]

#313

Earlier quoted context omitted.

>winning on cost-effectiveness Nobody is winning in this area until these things run in full on single graphics cards. Which is sufficient compute to run even most of the complex tasks.

Why does that matter? They wont be making at home graphics cards anymore. Why would you do that when you can be pre-sold $40k servers for years into the future

I didn't say winning business, I said winning on cost effectiveness.

Re: DeepSeek-v3.2: Pushing the frontier of open large language models [pdf]

#314

Earlier quoted context omitted.

Because Moore's law marches on. We're around 35-40 orders of magnitude from computers now to computronium. We'll need 10-15 years before handheld devices can run a couple terabytes of ram, 64-128 terabytes of storage, and 80+ TFLOPS. That's enough to run any current state of the art AI at around 50 tokens per second, but in 10 years, we're probably going to have seen lots of improvements, so I'd guess conservatively…

I appreciate your rabid optimism, but considering that Moores Law has ceased to be true for multiple years now I am not sure a handwave about being able to scale to infinity is a reasonable way to look at things. Plenty of things have slowed down in progress in our current age, for example airplanes.

Someone always crawls out of the woodwork to repeat this supposed "fact" which hasn't been true for the entire half-century it's been repeated. Jim Keller (designer of most of the great CPUs of the last couple decades) gave a convincing presentation several years ago about just how not-true it is: https://www.youtube.com/watch?v=oIG9ztQw2Gc Everything he says in it still applies today.

Intel struggled for a decade, and folks think that means Moore's law died. But TSMC and Samsung just kept iterating. And hopefully Intel's 18a process will see them back in the game.

Re: DeepSeek-v3.2: Pushing the frontier of open large language models [pdf]

#315
post #229

Earlier quoted context omitted.

Any car company. Uber. All tech companies offering free services.

What American car company competes overseas on price?

All the American cars (Ford, Chevrolet, GM...) are much cheaper in Europe than eg. German cars from their trifecta (and other Europe-made high end vehicles from eg Sweden, Italy or UK), and on par with mid-priced vehicles from the likes of Hyundai, Kia, Mazda...

Obviously, some US brands do not compete on price, but other than maybe Jeep and Tesla, those have a small market penetration.

Re: DeepSeek-v3.2: Pushing the frontier of open large language models [pdf]

#316
post #16

Well props to them for continuing to improve, winning on cost-effectiveness, and continuing to publicly share their improvements. Hard not to root for them as a force to prevent an AI corporate monopoly/duopoly.

As much I agree with your sentiment, but I doubt the intention is singular.

It's like AMD open-sourcing FSR or Meta open-sourcing Llama. It's good for us, but it's nothing more than a situational and temporary alignment of self-interest with the public good. When the tables turn (they become the best instead of 4th best, or AMD develops the best upscaler, etc), the decision that aligns with self-interest will change, and people will start complaining that they've lost their moral compass.

Re: DeepSeek-v3.2: Pushing the frontier of open large language models [pdf]

#317

Earlier quoted context omitted.

80% of the parts in iPhones are manufactured in China, and they have completely and utterly dominated in Enterprise (Ever heard of someone using a Blackberry in 2025? Me neither.) so there’s one example.

The software is made by Apple. Hardware can’t magically intercept communications and the manufacturing is done mostly in Taiwan. If Apple doesn’t have a process to protect its operating system from supply chain attacks, it would be derelict

Hardware can do any "magic" software can, which should be obvious since software runs on it. It's just not as cost-effective to modify it after shipping, which is why the tech sector is moving to more sw less hw (simplified, ofc, there are other reasons).

Re: DeepSeek-v3.2: Pushing the frontier of open large language models [pdf]

#318

Earlier quoted context omitted.

>winning on cost-effectiveness Nobody is winning in this area until these things run in full on single graphics cards. Which is sufficient compute to run even most of the complex tasks.

Nobody is winning until cars are the size of a pack of cards. Which is big enough to transport even the largest cargo.

Lol its kinda suprising that the level of understanding around LLMs is so little.

You already have agents, that can do a lot of "thinking", which is just generating guided context, then using that context to do tasks.

You already have Vector Databases that are used as context stores with information retrieval.

Fundamentally, you can have the same exact performance on a lot of task whether all the information exists in the model, or you use a smaller model with a bunch of context around it for guidance.

So instead of wasting energy and time encoding the knowledge information into the model, making the size large, you could have an "agent-first" model along with just files of vector databases, and the model can fit in a single graphics cards, take the question, decide which vector db it wants to load, and then essentially answer the question in the same way. At $50 per TB from SSD not only do you gain massive cost efficiency, but you also gain the ability to run a lot more inference cheaper, which can be used for refining things, background processing, and so on.

Re: DeepSeek-v3.2: Pushing the frontier of open large language models [pdf]

#319

Earlier quoted context omitted.

As much I agree with your sentiment, but I doubt the intention is singular.

It's like AMD open-sourcing FSR or Meta open-sourcing Llama. It's good for us, but it's nothing more than a situational and temporary alignment of self-interest with the public good. When the tables turn (they become the best instead of 4th best, or AMD develops the best upscaler, etc), the decision that aligns with self-interest will change, and people will start complaining that they've lost their moral compass.

It's not. This isn't about competition in a company sense but sanctions and wider macro issues.

Re: DeepSeek-v3.2: Pushing the frontier of open large language models [pdf]

#320
post #308

Earlier quoted context omitted.

Well consumers care about the cost to them, and those we know. And deepseek is destroying everything in that department.

Yes. Though we don't know for sure whether that's because they actually have lower costs, or whether it's just the Chinese taxpayer being forced to serve us a treat.

Third party providers are still cheap though. The closed models are the ones where you can't see the real cost to running them.
Post reply on HN