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

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

huggingface.co

Post reply on HN
Print view View original

201–210 of 485 posts

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

#201

Post by littlestymaar »

raw_anon_1111 wrote: post #98
echelon wrote: post #85

Earlier quoted context omitted.

I don't care if this kills Google and OpenAI. I hope it does, though I'm doubtful because distribution is important. You can't beat "ChatGPT" as a brand in laypeople's minds (unless perhaps you give them a massive "Temu: Shop Like A Billionaire" commercial campaign). Closed source AI is almost by design morphing into an industrial, infrastructure-heavy rocket science that commoners can't keep up with. The companies p…

I can’t think of a single company I’ve worked with as a consultant that I could convince to use DeepSeek because of its ties with China even if I explained that it was hosted on AWS and none of the information would go to China. Even when the technical people understood that, it would be too much of a political quagmire within their company when it became known to the higher ups. It just isn’t worth the political cap…

> I can’t think of a single company I’ve worked with as a consultant that I could convince to use DeepSeek because of its ties with China even if I explained that it was hosted on AWS and none of the information would go to China.

Well for non-American companies, you have the choice between Chinese models that don't send data home, and American ones that do, with both countries being more or less equally threatening.

I think if Mistral can just stay close enough to the race it will win many customers by not doing anything.

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

#202

Post by observationist »

beefnugs wrote: post #176
ActorNightly wrote: post #162

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

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 you're going to be able to see 4-5x performance per parameter, possibly much more, so at that point, you'll have the equivalent of a model with 10T parameters today.

If we just keep scaling and there are no breakthroughs, Moore's law gets us through another century of incredible progress. My default assumption is that there are going to be lots of breakthroughs, and that they're coming faster, and eventually we'll reach a saturation of research and implementation; more, better ideas will be coming out than we can possibly implement over time, so our information processing will have to scale, and it'll create automation and AI development pressures, and things will be unfathomably weird and exotic for individuals with meat brains.

Even so, in only 10 years and steady progress we're going to have fantastical devices at hand. Imagine the enthusiast desktop - could locally host the equivalent of a 100T parameter AI, or run personal training of AI that currently costs frontier labs hundreds of millions in infrastructure and payroll and expertise.

Even without AGI that's a pretty incredible idea. If we do get to AGI (2029 according to Kurzweil) and it's open, then we're going to see truly magical, fantastical things.

What if you had the equivalent of a frontier lab in your pocket? What's that do to the economy?

NVIDIA will be churning out chips like crazy, and we'll start seeing the solar system measured in terms of average cognitive FLOPS per gram, and be well on the way toward system scale computronium matrioshka brains and the like.

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

#204

Post by Havoc »

wosined wrote: post #130

Remember: If it is not peer-reviewed, then it is an ad.

Good general approach, but deepseek has thus far always delivered. And not just delivered, but under open license too. "Ad" as starting assumption seems overly harsh

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

#205

Post by hnfong »

ecshafer wrote: post #104
isamuel wrote: post #18

Earlier quoted context omitted.

There is a great deal of orientalism --- it is genuinely unthinkable to a lot of American tech dullards that the Chinese could be better at anything requiring what they think of as "intelligence." Aren't they Communist? Backward? Don't they eat weird stuff at wet markets? It reminds me, in an encouraging way, of the way that German military planners regarded the Soviet Union in the lead-up to Operation Barbarossa. Th…

I don't think that anyone, much less someone working in tech or engineering in 2025, could still hold beliefs about Chinese not being capable scientists or engineers. I could maybe give (the naive) pass to someone in 1990 thinking China will never build more than junk. But in 2025 their product capacity, scientific advancement, and just the amount of us who have worked with extremely talented Chinese colleagues shoul…

I don't think anyone seriously believes that the Chinese aren't capable, it's more like people believe no matter what happens, USA will still dominate in "high tech" fields. A variant of "American Exceptionalism" so to speak.

This is kinda reflected in the stock market, where the AI stocks are surging to new heights every day, yet their Chinese equivalents are relatively lagging behind in stock price, which suggests that investors are betting heavily on the US companies to "win" this "AI race" (if there's any gains to be made by winning).

Also, in the past couple years (or maybe a couple decades), there had also been a lot of crap talk about how China has to democratize and free up their markets in order to be competitive with the other first world countries, together with a bunch of "doomsday" predictions for authoritarianism in China. This narrative has completely lost any credibility, but the sentiment dies slowly...

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

#206

Post by raw_anon_1111 »

coliveira wrote: post #165
StealthyStart wrote: post #116

Earlier quoted context omitted.

This is the real cause. At the enterprise level, trust outweighs cost. My company hires agencies and consultants who provide the same advice as our internal team; this is not to imply that our internal team is incorrect; rather, there is credibility that if something goes wrong, the decision consequences can be shifted, and there is a reason why companies continue to hire the same four consulting firms. It's trust, w…

So much worse for American companies. This only means that they will be uncompetitive with similar companies that use models with realistic costs.

I can’t think of a single major US company that is big internationally that is competing on price.

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

#207

Post by register »

raw_anon_1111 wrote: post #98
echelon wrote: post #85

Earlier quoted context omitted.

I don't care if this kills Google and OpenAI. I hope it does, though I'm doubtful because distribution is important. You can't beat "ChatGPT" as a brand in laypeople's minds (unless perhaps you give them a massive "Temu: Shop Like A Billionaire" commercial campaign). Closed source AI is almost by design morphing into an industrial, infrastructure-heavy rocket science that commoners can't keep up with. The companies p…

I can’t think of a single company I’ve worked with as a consultant that I could convince to use DeepSeek because of its ties with China even if I explained that it was hosted on AWS and none of the information would go to China. Even when the technical people understood that, it would be too much of a political quagmire within their company when it became known to the higher ups. It just isn’t worth the political cap…

That might be the perspective of a US based company. But there is also Europe and basically it's a choice between Trump and China.

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

#208

Post by Havoc »

nimchimpsky wrote: post #2

Pretty amazing that a relatively small Chinese hedge fund can build AI better than almost anyone.

Yeah they've consistently delivered. At the same time there are persistent whispers that they're not all that small and scruffy as portrayed either.

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

#209

Post by hnfong »

spullara wrote: post #62

I hate that their model ids don't change as they change the underlying model. I'm not sure how you can build on that. % curl https://api.deepseek.com/models \ -H "Authorization: Bearer ${DEEPSEEK_API_KEY}" {"object":"list","data":[{"id":"deepseek-chat","object":"model","owned_by":"deepseek"},{"id":"deepseek-reasoner","object":"model","owned_by":"deepseek"}]}

Agree that having datestamps on model ids is a good idea, but it's open source, you can download the weights and build on those. In the long run, this is better than the alternative of calling API of a proprietary model and hoping it doesn't get deprecated.

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

#210

Post by tokioyoyo »

hhh wrote: post #180
tokioyoyo wrote: post #128

Earlier quoted context omitted.

If the Chinese model becomes better than competitors, these worries will suddenly disappear. Also, there are plenty startups and enterprises that are running fine-tuned versions of different OS models.

No… Nobody I work for will touch these models. The fear is real that they have been poisoned or have some underlying bomb. Plus y’know, they’re produced by China, so they would never make it past a review board in most mega enterprises IME.

People say that, but everyone, including enterprises, are constantly buying Chinese tech one way or another because of cost/quality ratio. There’s a tipping point in any excel file where risks don’t make sense, if the cost is 20x for the same quality.

Of course you’ll always have exceptions (government, military and etc.), but for private, winner will take it all.

Post reply on HN