AI startup DeepSeek rivals OpenAI models using far fewer resources, shocks AI industry

S

Skye Jacobs

Posts: 380   +10
Staff
A hot potato: A new player is shaking up the AI landscape. DeepSeek, a China-based AI company, released an open version of its R1 reasoning model on January 20, 2025, and it's already making waves. The model reportedly matches or even surpasses OpenAI's O1 on certain AI benchmarks while using far less computing resources. This has sparked a frenzy of discussion across the tech world. Within days, the DeepSeek app soared to the #1 spot on the App Store, surpassing ChatGPT and underscoring the growing rivalry between Chinese and American tech giants in the race for AI dominance.

Prominent venture capitalist Marc Andreessen was one of those impressed by the feat, writing on X that DeepSeek's model was "one of the most amazing and impressive breakthroughs I've ever seen."

DeepSeek's accomplishment is particularly noteworthy given the company's claim to have trained a model with 671 billion parameters using just 2,048 Nvidia H800s and $5.6 million, a fraction of the resources typically required by industry giants like OpenAI and Google. This cost-effectiveness is even more remarkable considering the U.S. sanctions that restrict the sale of advanced chips to Chinese companies.

Commentators said that for these reasons, the model also has geopolitical implications. "The impressive performance of DeepSeek's distilled models [...] means that very capable reasoners will continue to proliferate widely and be runnable on local hardware, far from the eyes of any top-down control regime," Dean Ball, an AI researcher at George Mason University, wrote.

Some observers believe that DeepSeek's success could potentially benefit the entire AI industry. "If training models get cheaper faster and easier, the demand for inference (actual real world use of AI) will grow and accelerate even faster, which assures the supply of compute will be used," Garry Tan, CEO of Y Combinator, wrote on X.

However, not all reactions have been uniformly positive. Neal Khosla, CEO of Curai, expressed skepticism, suggesting that the company might be a "ccp state psyop" aimed at undermining U.S. AI competitiveness. However, this claim has been challenged for lack of evidence.

DeepSeek-R1 is a reasoning model that employs a step-by-step approach to problem-solving, making it particularly adept at tasks in physics, science, and mathematics. The model contains 671 billion parameters, which contribute to its problem-solving capabilities.

DeepSeek has also released smaller "distilled" versions of R1, ranging from 1.5 billion to 70 billion parameters, with the smallest capable of running on a laptop.

R1 is available under an MIT license, allowing for commercial use without restrictions. According to DeepSeek, the model outperforms OpenAI's o1 on benchmarks such as AIME, MATH-500, and SWE-bench Verified. These assess various aspects of AI performance, including mathematical problem-solving and programming tasks.

One notable limitation of R1 is its adherence to Chinese regulatory requirements. As a Chinese model, it's subject to benchmarking by China's internet regulator to ensure compliance with "core socialist values." Consequently, R1 refrains from answering questions about sensitive topics such as Tiananmen Square or Taiwan's autonomy.

Despite these constraints, DeepSeek's achievement has sparked significant interest. As of Sunday afternoon, DeepSeek's AI assistant has become the top free app in the Apple App Store, surpassing even ChatGPT.

The success of DeepSeek has catapulted its creator Liang Wenfeng into the national spotlight. Recently, he was the sole AI industry representative invited to a high-profile meeting with Li Qiang, China's Premier and second-most powerful leader.

Liang, a Chinese entrepreneur and hedge fund manager, began his journey to AI prominence in the world of quantitative finance. In 2015, Liang founded High-Flyer, a quantitative hedge fund that quickly rose to one of China's "Big Four" quantitative private funds. Under Liang's leadership, High-Flyer pioneered the integration of AI-driven strategies in quantitative investment, transitioning to a fully AI-based approach by 2017.

Liang's foray into AI development began in earnest in 2021 when he started acquiring thousands of Nvidia GPUs for what was initially perceived as an eccentric side project. This prescient move laid the groundwork for DeepSeek, which Liang founded in 2023 with the ambitious goal of developing human-level AI.

Liang's unconventional background has proven to be a unique advantage in the AI field. His team's experience in utilizing Nvidia chips for stock trading has translated well into the challenges posed by U.S. export restrictions on advanced AI chips to China. This adaptability has allowed DeepSeek to innovate in the face of limited access to cutting-edge hardware.

Permalink to story:

AI startup DeepSeek rivals OpenAI models using far fewer resources, shocks AI industry

 
  • Like
Reactions: TempleOrion
kira setsu said:
Not surprising, U.S. pretty much dared China to make a more optimized model by antagonizing them and putting the highest end hardware out of reach.
Click to expand...

Oh look, if it isn't the unintended consequences of our own actions.

But yeah, China is more then willing to throw crazy money (literally hundreds of billions per year for a decade) to grow its own industry, its not shocking how rapidly they can replace US goods. So instead of being constrained by our supply, they're now supplanting it. Then exporting for a fraction of the price.
 
  • Like
Reactions: riev90, gdavid65, Q Wales and 6 others
This could be a turning point for AI accessibility. Distilled models running on laptops mean we’re not far from a world where advanced AI tools are as common as smartphones, completely reshaping who gets to use cutting-edge tech.
 
  • Like
Reactions: TempleOrion
It is only a question of time before China overtakes the West in technology.
I remember saying this for the first time around 2000-2001 in a motherboard forum which no longer exists. People made fun of me. They still are when I repeat this anywhere.
BTW, have you seen the S&P 500 today?
www.bnnbloomberg.ca

Nvidia’s $465 Billion DeepSeek Plunge Is Largest in Market History

Nvidia Corp.’s plunge, fueled by investor concern about Chinese artificial-intelligence startup DeepSeek, erased a record amount of stock-market value from the world’s largest company.
www.bnnbloomberg.ca www.bnnbloomberg.ca
 
Last edited:
  • Like
Reactions: riev90, Ephebus, TempleOrion and 4 others
Makes D. Trump's Stargate initiative seem more important now. I'm guessing Donald had gotten wind about this Chinese startup before the press did. And perhaps the competition will produce better, more efficient products.
 
From China with Lies! Something stinks, China cannot be trusted. I heard a few minutes ago they will now be limiting users allowed to use it. Running out of hardware resources? NEVER trust the CCP!
 
  • Like
Reactions: Tran Bronstein
airbornex said:
This could be a turning point for AI accessibility. Distilled models running on laptops mean we’re not far from a world where advanced AI tools are as common as smartphones, completely reshaping who gets to use cutting-edge tech.
Click to expand...

I believe there are already models that you can download and use far less resources and pretty close to stds already

Probably in future download models you want - eg mathematics , language , visual etc
 
  • Like
Reactions: billc and TempleOrion
Credit where it is due, the technological achievement here is impressive. More proof of the old adage that necessity is the true mother of invention. This said, it does bring up the concern of bias and censorship in AI models, making them the permanent digital repository of not just knowledge but also biases, prejudices and misinformation. You know, just like a regular human.
 
  • Like
Reactions: TempleOrion and BogdanR
BogdanR said:
It is only a question of time before China overtakes the West in technology.
I remember saying this for the first time around 2000-2001 in a motherboard forum which no longer exists. People made fun of me. They still are when I repeat this anywhere.
BTW, have you seen the S&P 500 today?
www.bnnbloomberg.ca

Nvidia’s $465 Billion DeepSeek Plunge Is Largest in Market History

Nvidia Corp.’s plunge, fueled by investor concern about Chinese artificial-intelligence startup DeepSeek, erased a record amount of stock-market value from the world’s largest company.
www.bnnbloomberg.ca www.bnnbloomberg.ca
Click to expand...

Maybe people are laughing at you because you said it 25 years ago and it's still nowhere near being true. Let it go. People used to say the same for Japan in the 1980s and it hit a recession that it still hasn't recovered from.

China is in a deep recession btw, and it might very well stay in it for the next 4 decades as well.
 
Tran Bronstein said:
Credit where it is due, the technological achievement here is impressive. More proof of the old adage that necessity is the true mother of invention. This said, it does bring up the concern of bias and censorship in AI models, making them the permanent digital repository of not just knowledge but also biases, prejudices and misinformation. You know, just like a regular human.
Click to expand...

yes but that $5.6 million quote is not verifiable by anyone because it's a Chinese company. No one has access to its financials except the Chinese government. They've created shell companies and listed them in the NYSE in the 2010s and then dumped them. Wouldn't trust anything out of China.
 
Interesting how within a week of Tiktok's near-removal from America, another Chinese-owned app appears and goes straight to the top of everyone's wishlist, ready to collect as much wilfully-donated data as possible from users.
 
  • Like
Reactions: BogdanR
poohbear said:
Maybe people are laughing at you because you said it 25 years ago and it's still nowhere near being true. Let it go. People used to say the same for Japan in the 1980s and it hit a recession that it still hasn't recovered from.

China is in a deep recession btw, and it might very well stay in it for the next 4 decades as well.
Click to expand...
He who laughs last. We’ll see who that will be.
Irrespective of your opinions, China has come a long way technologically in the last 25 years. It may take another 25 or perhaps “4 decades” longer, I won’t speculate on that, but it will happen.
You see, there’s a huge difference between ‘80s Japan and current China in many ways, particularly where it counts, the latter has a national drive to become Numero Uno, which Japan doesn’t.
 
Last edited:
  • Like
Reactions: TempleOrion
poohbear said:
Maybe people are laughing at you because you said it 25 years ago and it's still nowhere near being true. Let it go. People used to say the same for Japan in the 1980s and it hit a recession that it still hasn't recovered from.

China is in a deep recession btw, and it might very well stay in it for the next 4 decades as well.
Click to expand...

Making stuff up doesn't make it true. That's the problem with right whingers, they live in a fantasy bubble world based purely on emotions and not logic or reason. Therefore the dumb belief in Big Daddies (either invisible ones in the sky, or senile, sociopathic ones on Earth) that'll take care of everything for them so they don't have to think too much 🤣

https://www.statista.com/statistics/1070632/gross-domestic-product-gdp-china-us/

https://worldostats.com/gdp-ppp-by-country-2025-cia-data/

Under Trumpf of course the US decline will be even faster. This humbling of the hubristic tech giants is just the beginning.

And India's growth rate is double China's... 😉
 
  • Like
Reactions: nnguy2 and BogdanR
The Talking Tech said:
Ai will be the death of all of us....These companies are in such a rush in creating something far smarter than humans beings can ever be BUT they don't care about how much power they are giving it....they just care about who is first and whose is smarter.
Click to expand...

Agreed, it’s like all of a sudden we forgot about world poverty, the poor, and the disabled. Now it’s all about buying or conquering entire continents, being “first” in whatever incl. DEI, while feeding us the biggest dose of propaganda there is.

They're trying to mix iron and clay, something the Bible specifically forbids. It’ll be the end of civilization if the AI enterprise comes to pass.
 
Absolutely impressive work by DeepSeek! Their ability to rival top-tier AI models using significantly fewer resources is a game-changer. This breakthrough not only enhances efficiency but also makes advanced AI more accessible and sustainable for various industries. Excited to see the positive impact they'll bring!
 
  • Like
Reactions: TempleOrion
poohbear said:
yes but that $5.6 million quote is not verifiable by anyone because it's a Chinese company. No one has access to its financials except the Chinese government. They've created shell companies and listed them in the NYSE in the 2010s and then dumped them. Wouldn't trust anything out of China.
Click to expand...
But you trust Musk who has been lying about Fake Self Driving for almost ten years; or Altman who has been talking about a singularity, when it was an idea he got out of a comic book!
 
  • Like
Reactions: TempleOrion, nnguy2 and BogdanR
You must log in or register to reply here.

Similar threads

S
OpenAI's newest AI model is switching languages to Chinese and others while reasoning, puzzling users and experts
Replies
4
Views
65
TooQik
T
Shawn Knight
Atari teases handheld game system with a trackball and a numpad
Replies
8
Views
133
Xelions
X
Shawn Knight
VLC celebrates 6 billion downloads, teases AI generated subtitles and translation
Replies
4
Views
95
GoldenGoat
GoldenGoat

Latest posts

Back