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

Deepseek: The quiet giant leading China’s AI race

chinatalk.media

Post reply on HN
Print view View original

161–170 of 476 posts

Re: Deepseek: The quiet giant leading China’s AI race

#161

Post by jurli »

waldrews wrote: post #156

To this day, asking Deepseek "what model are you" typically gives the answer "I'm an AI language model called ChatGPT, created by OpenAI. Specifically, I'm based on the GPT-4 architecture, which is designed to understand and generate human-like text based on the input I receive. My training data includes a wide range of information up until October 2023, and I can assist with answering questions, generating text, and…

This is a common "gotcha" comment from people who don't understand LLMs very well. Occasionally if you ask Gemini it'll say this as well. It has everything to do with the fact that ChatGPT is the most talked about AI model rather than data being trained on it

Re: Deepseek: The quiet giant leading China’s AI race

#162

Post by jurli »

nsoonhui wrote: post #13

I find that the gushing around deepseek is fascinating to watch. To me there are a few structural and fundamental reasons why deepseek can never outperform other models by a wide margin. On par maybe--as we reach the diminishing returns with our investment in the models, but not win by a wide margin. 1. The US trade war with china which will place deepseek compute availability at disadvantages, eventually, if we ever…

[flagged]

Re: Deepseek: The quiet giant leading China’s AI race

#163

Post by jurli »

lomkju wrote: post #6

I feel the GPU restrictions created an environment for Chinese Devs to be more innovative and do more with less. Kudos to the deepseek team!

Makes sense. When you restrict hardware, you have to spend all your energy on optimizing software that everyone else ignores

Imagine if they were forced to use IE7 as the only browser. The frontend frameworks would be blazing fast and we would never have bloatware like React or Angular or npm

Re: Deepseek: The quiet giant leading China’s AI race

#164

Post by jurli »

Etheryte wrote: post #22
wodderam wrote: post #12

Earlier quoted context omitted.

Kai-Fu Lee describes the culture so well in AI Superpowers. The roots are well before GPU restriction. Absolute cut throat competition. Imagine Sam Altman throwing a chair out a window in a meeting lol. The message of AI Superpowers is that China will lag the US at first but once things stabilize this will happen because China has a lot more engineers and a lot more data. Anyone who hasn't read AI Superpowers should…

I don't know, I've been hearing the story that China is about to upend the US as the leading global superpower ever since I was a kid. There's always a new vogue and novel twist put on the rationale and how it's gonna happen, but so far it's like fusion, always a few years away.

It's literally happening lol. When you were a kid China was making shoes and their GDP is 10% of the US. Now they're making drones / evs / high end electornics and it's 80%. This is why people's perception is so unreliable because it's impossible to notice things when they happen over a lifetime

Re: Deepseek: The quiet giant leading China’s AI race

#165

Post by bugglebeetle »

kayewiggin wrote: post #133
solaarphunk wrote: post #67

Earlier quoted context omitted.

Economic superpower perhaps - just take a look at their relative GDP over time.

China has 900M people making less than $400/month, and 600M people making less than $100/month. relative GDP is a joke, go to China to see what most of them are eating (hint: it's unsafe food filled with chemicals, or its mostly carbs) and where these people are living (hint: it's shoddy constructed condos or run down farm houses)

It’s unclear to me why what you’re describing is specific to China and not also what Americans euphemistically refer to as “fly over states.”

Re: Deepseek: The quiet giant leading China’s AI race

#166

Post by sangnoir »

daveguy wrote: post #149
bpodgursky wrote: post #128

Earlier quoted context omitted.

Believe whatever you want. Most of these bases are co-located with NATO or other allies for good reason, the US doesn't have to do everything itself wrt air defense, locating an airlift wing with a fighter wing. But then it's a lower bar than people imagine, for China to buy similar friendship.

You literally called the logistics hubs of the US military -- the bases that move more of the most powerful weapons and military personnel in the world -- indefensible. So you either don't know what indefensible means, or you are a piss poor propagandist.

Yes or No: Can the US singlehandedly defend all those bases without the help of host country? If the answer is "Yes", then China has a long way to go to achieve that capability. If answer is "No", then the bar is much lower, and gp's point is that China can "buy" similar arrangements without too much effort. More directly, is the bottleneck on funding, personnel/matiriel, or diplomacy?

China is building up a lot of soft power with infrastructure projects all over the world - most of them are aimed at improving trade - ports, rail lines and the like. In the next decade or 2, they can reasonably make requests to place a few PLA/PLAA personnel and equipment on bases in strategic places, bases they may have been built using Chinese money.

Re: Deepseek: The quiet giant leading China’s AI race

#167

Post by bpodgursky »

daveguy wrote: post #149
bpodgursky wrote: post #128

Earlier quoted context omitted.

Believe whatever you want. Most of these bases are co-located with NATO or other allies for good reason, the US doesn't have to do everything itself wrt air defense, locating an airlift wing with a fighter wing. But then it's a lower bar than people imagine, for China to buy similar friendship.

You literally called the logistics hubs of the US military -- the bases that move more of the most powerful weapons and military personnel in the world -- indefensible. So you either don't know what indefensible means, or you are a piss poor propagandist.

Yes, it is very easy for a logistic hub that only has an airlift wing to be indefensible in a war against a peer adversary if for example there are no THAAD or Patriot batteries there. It's hub, not a hardened facility.

Many of the US military bases are communication centers or barracks on training bases. They serve important roles but are not "defensible" in many contexts.

Who am I even propagandizing for in this context?

Re: Deepseek: The quiet giant leading China’s AI race

#168

Post by waldrews »

amelius wrote: post #158
waldrews wrote: post #156

To this day, asking Deepseek "what model are you" typically gives the answer "I'm an AI language model called ChatGPT, created by OpenAI. Specifically, I'm based on the GPT-4 architecture, which is designed to understand and generate human-like text based on the input I receive. My training data includes a wide range of information up until October 2023, and I can assist with answering questions, generating text, and…

I'm sure OpenAI breaches copyright just as well. They are just a little bit better at hiding it.

It also tells us the genie is out of the bottle not just in the form of open weights being widely available, but in the form of the text corpuses coming from the existing model. The claimed low cost of Deepseek's training is partly enabled by the availability of all that synthetic data created by the first generation models trained and developed at much higher cost. When the Soviets got hold of the nuke plans, they greatly reduced their development costs by primarily by not having to redo all the experiments that led to dead ends. What's amazing is that time it's different; nobody needs OpenAI's secret sauce anymore, just enough data - some of it happily supplied by ChatGPT itself, and they can experiment with different architectures and either get tolerable results with an architecture already in textbooks, or greatly improve efficiency by innovating.

Re: Deepseek: The quiet giant leading China’s AI race

#169

Post by nimbius »

nsoonhui wrote: post #13

I find that the gushing around deepseek is fascinating to watch. To me there are a few structural and fundamental reasons why deepseek can never outperform other models by a wide margin. On par maybe--as we reach the diminishing returns with our investment in the models, but not win by a wide margin. 1. The US trade war with china which will place deepseek compute availability at disadvantages, eventually, if we ever…

1. China already has a domestic 3nm process and competitive video card industry that openly and actively seeks independence from sanction. Huawei is evidence that sanctions are not as effective as foreign policy leaders may think.

2. Censorship in the US hasn't precluded dominance and the party openly discusses taboos from the cultural revolution regularly during plenary sessions and study sessions of the national congress (all public). Output censorship isn't the same as input.

3. Redhats llm and ai efforts are all open source as well. Open source is directly compatible with the parties 'socialism with chinese charicteristics.'

Re: Deepseek: The quiet giant leading China’s AI race

#170

Post by cess11 »

yoavm wrote: post #146
Bilal_io wrote: post #46

Earlier quoted context omitted.

Western LLMs have a bias when it comes to Israel and Palestine issue. Out of curiously, what part of the Quran do you consider controversial?

Not the OP, but here's one I feel quite uncomfortable with: https://quran.com/en/an-nisa/155/tafsirs - "The Hour will not start, until after the Muslims fight the Jews and the Muslims kill them. The Jew will hide behind a stone or tree, and the tree will say, `O Muslim! O servant of Allah! This is a Jew behind me, come and kill him". Other examples from https://en.wikipedia.org/wiki/An-Nisa include "Men are the prote…

The New Testament has similar passages. One of the most well known has Jesus attacking pilgrims and money changers in the temple. John is rather obviously antijewish. "I have not come with peace" is another well known, not very palatable one.
Post reply on HN