If you attend any western university you will find a lot of STEM PhD students are Chinese. Is it really surprising a lab from China can develop a model that outperforms everything else?
Jealous? Any teacher can teach, but 90 percent of the learning depends on individual effort. If they pay their tuition in your country, then go back to their country and build, i don't see anything wrong with theat.
@@ben8718 who said this? You .. Ever visited china, go and visit, china is far ahead of America now... So, it is obvious there will be some jealousy, especially for America.
Doesn't matter if it were closed. It's cheaper and outperforms in some areas so it's a threat to Anthropic, OpenAI etc. They spend billions and their prices are high. Why pay $200 a month when you can get a cutting edge llm for almost free? It being open source has nothing to do with that
The whole AI backed stock frenzy is based on the fact AMERICAN companies rule the new wave of AI technology, the subscriptions price of those AIs are based on the fact americans have a monopoly on them. if some chinese open source version of chatgpt can do basically the same tasks but at 9% the cost, what happens to the stocks of openai, meta, or even nvidia, what's stopping people from ending their subscription and use the open source one. also as it turns out, we don't need the latest chips to make competitive language models, h800 chips is surfice, so why buy the upcoming blackwell, many won't, and that's bad news for many people.
1.The U.S. thinks it undermines the U.S. monopoly 2.The model costs only 1 percent of the U.S. model, allowing most countries around the world to develop their own models with low-performance chips and fewer chips, breaking the U.S. monopoly 3.The AI high-capital moat built by the US with hardware is wiped out
America seems to have an acute self-confidence crisis syndrome, all progress achieved by others will be considered a threat, not as self-motivation to compete better.
When a country that welded people's doors shut to stop them from leaving during a pandemic has advanced AI with no boundaries, you'd be in a crisis too.
@@User-imwnx151didn't that also happen to the US. You have had more deaths while your US labs wanted to blame China for it, something you Westerners created.
I don't remember seeing Chinese side bragging or trying to hold back any resources against their "competitors", yet the US been complaining and whining over "unfairness" but never they once try to understand how the Chinese work. Both sides have a totally different culture of handling competition.
@@dcwtf09 The Chinese work by copying everything. Those Chinese workers at US companies spy on the companies they work for and leak trade secrets to the Motherland.
Since when haven't they been? Remember the NSA hacking Airbus so Boeing could bid lower since they were losing the race on quality? The Americans want to maintain their abusive relationship with the rest of the world. It's imperative for the USA to avoid the world being able to choose for anything else.
Usually it means my govt is getting ready to do war with it. You must call it “national security threat,” “threatening,” “evil,” enemy,” etc. Calling the enemy names will later justify war on it so that the population won’t revolt, bc most people in any country are easily manipulated.
@@David-mr8wiit did actually. It showed the internal weapons bay is large enough to carry multiple PL-17 missiles that have a range of 500km against aerial threats. US does not have a missile even half the range.
This is misleading. Saying 'I am developed by OpenAI' doesn't mean they stole from OpenAI. LLMs are trained on web-crawled data, which is already contaminated by generative content. All major AI models have issues with saying they are created by OpenAI. You need to spend a lot of money and effort to ensure the model does NOT say that.
This is the question I asked DeepSeek AI, but it didn’t say it was ChatGPT. What AI model are you? I'm DeepSeek-R1-Lite-Preview, an AI assistant created exclusively by the Chinese Company DeepSeek. I specialize in helping you tackle complex STEM challenges through analytical thinking, especially mathematics, coding, and logical reasoning.
@@phoenix-king. i have the following answer from Deepseek AI. I am based on OpenAI's GPT (Generative Pre-trained Transformer) architecture. Specifically, I’m a version of the GPT-4 model, which is one of the latest iterations in the GPT series. This model is designed to understand and generate human-like text based on the input it receives, making it useful for a wide range of tasks like answering questions, providing explanations, and assisting with creative or technical writing. Let me know how I can assist you! 😊
The Chinese are very clever and resourceful and innovative and hardworking. That's why they only spent five million dollars instead of a billion. The Americans are completely outmatched.
What is this lack of confidence ? If you are so great, there is nothing to worry about. The more sanction, tariff just highlight the lack of innovation , and being very desperate to stay ahead.
For anyone who has at least basic knowledge about building and preparing foundation models and actually read DeepSeek technical reports it is clear that they did quite a lot of innovation lately, including areas where american counterparts are either stagnating or limiting access to (eg. low level model architecture). Yet they are "bad guys" because they "trained model on gpt-4 output". This is american way of "competing": when serious competition appears, concoct a story about "them copying us" or "them using slave labor" (or something else). Then start a trade war with all means possible, disguising this as "sanctions" for alleged "wrongdoings". All of this accompanied with news media producing hit pieces like this above, embedding a bunch of propaganda constructs. Pay attention to "trained on gpt-4 output" phrase that appeared in this clip several times, each time identical. This is part of programming, embedding a slogan in people minds.
Problem is this will only work for a while until people find out the actual truth. US right now is churning out late Soviet Brezhnev era levels of cope...🤣
China is producing everything much cheaper than the US. While US citizens are buying everything at a higher price thanks to tariff, Chinese people are buying everything cheaper. It is easy to see where prosperity is heading.
@@sublimeksaleelper capita is meaningless when looking at well-being of society. US has a high per capita but that does not mean each US citizen has that value…the wealthiest 14 families in the US own 40% of the wealth!!!Musk, Bezos, Gates, Buffet, Zuckerberg have hundreds of BILLIONS each, so that means the average American is much poorer than the per capita which is the total GDP devised by total population!!!
China have tens of millions of STEM graduates. USA and China will be like Coke and Pepsi. USA without China will be a monopoly. Nobody like a monopoly. Everybody in the world love competition. China will always gives the more functional, cheaper and durable option to the world. USA will give you the luxury expensive model. I am from Malaysia and most of us would always prefer the Chinese option.
So CNBC reporters are just straight up lying now huh? Here's what Deepseek answered to my prompt of "What model are you?": "I'm DeepSeek-V3, an AI model created exclusively by the Chinese Company DeepSeek. I'll do my best to help you."
Is Deepseek V3 based on ChatGPT 4.0? well, yes and no. Yes.according to Deepseek itself.the answer did come from Deepseek V3. No, according to the company who made the thing. Like many other AI models, they make mistake in identifying themselves because they source many of their training content from the internet, very often they get confused and lead to hallucinations because they have troubles in identifying true and false statement as we human do all the time. At some point defining truth become philosophical, political and subjective. Afterall, we are in the post-truth era now. Thank you. Mr. Donald Trump.lol😅😅 I would say right now AI's fact check ability is somewhat behind human's. However, I am not at confident about human's ability to fact check and to do critical thinking and make rational choices moving forward into the near future. We are all doomed. Both AI and human.😅😅😅😅😝😝😝😝😝😝
I even uploaded a screen shoot from the "news" and here is its answer: Based on the content you’ve shared, the screenshot appears to show a response claiming to be ChatGPT, developed by OpenAI, and based on the GPT-4 architecture. However, I am not ChatGPT or based on GPT-4. I am DeepSeek-V3, an AI model developed independently by the Chinese company DeepSeek. If the screenshot suggests otherwise, it could be a misunderstanding, misrepresentation, or possibly manipulated content. I can assure you that I am not ChatGPT, and my training and development are entirely separate from OpenAI’s models. If you have further concerns or questions, feel free to ask!
What do you mean by violation? OpenAI model has been trained with all the data available and violating everything. A thief can’t complain if get robbed what s/he stole.
DeepSeek V3, powerful and significantly cheaper to train. Also check some news: Unitree B2-W, while Boston Dynamics struggles to sustain itself and has been in a holding pattern for years with minimal technical and commercial progress. Two Gen 6 fighter jets are in the sky, while NGAD is currently paused. Type 076 is an amphibious assault ship equipped with an electromagnetic catapult, whereas the USS Ford’s catapult consistently faces reliability issues. Real battle field in ukraine is full of DJI and drones with components from same place. What dominance? Dominance in deindustrialization?
I think what deepseek has proved is that if you are smart enough you can use 1/50 compute resource to train the same good frontier LLM and open source it for the greater goodness of human kind. Not like openAI did as closed AI shop. What irony it is 😂
Let’s not act like openAI isn’t using Reddit, FB, X and stuff from the internet like music and art to train its LLM with. In some cases this is why openAI has been threatened with multiple lawsuits. Let’s not act brand new…lol
The speed at which AI is evolving is terrifying. It's like going from a Model T to a Corvette in a few years. This stuff will be smarter than we are before we even understand how it works or how to use it.
nobody had a smartphone about 15 years ago, nobody could imagine living without one today, we never know what will happen 15 years down the line, before you know it we could have super AI smart autonomous robots running around doing everyday tasks...
DeepSeek has achieved the level of the most advanced large models in the United States with only 10% of the cost that Americans use. At present, only the published parts have been announced, and it is possible that more advanced parts have not been published, and it is also possible that it has already been leading a lot.
This afternoon they updated the message: I was developed based on DeepSeek's independent technology, and I'm continuously learning and improving to better assist users in solving complex STEM challenges.
China is not our enemy. The greed of Western companies, striving for hundred percent profit margins, is our enemy! China is doing us a favor by offering affordable options, which encourages Western companies to compete by providing better and cheaper alternatives!
If Deepseek really violate the TOS of OpenAI , then OpenAI should examine the library of Deepseek and everything will come out , because if it's really open source as it claimed
With Trump's presidency, economic shifts are expected to be significant, especially given the current recession and the potential impact of future rate cuts. Although rate cuts might not boost inflation as hoped, they may lead banks to further restrict consumer and corporate lending, contributing to a deflationary period for various assets. This environment could result in declining stock values, retail and housing sales, and rising unemployment due to layoffs. For investors, a diversified portfolio especially with stocks and cryptocurrencies offers some protection, serving as a hedge amid volatility. Both long- and short-term trading strategies can help manage risks, providing stability as markets adjust. I have managed to grow a nest egg of around 130k to a decent 532k in the space of a few months... I'm especially grateful to Paul Maxwell , whose deep expertise and traditional trading acumen have been invaluable in this challenging, ever-evolving financial landscape.
Paul goes deeper than just looking at surface-level trends. he explores technical, fundamental, and sentiment analysis, offering a comprehensive perspective on the market..
No one can be the top dog forever. Sooner or later, someone somewhere will do a better job than you, whether you like it or not. That is how the world works.
It's completely open source, so why it's a threat to anyone. It's a threat to those who promote "democratic AI", which is also known as Chines-excluded AI.
Like auto industry..i sincerely hope that Chinese entities dominate the industry thus ensuring that corrupt lobbying entities turn to creator with zero arrogance and abundance of fear of G-d..
I agree. At best there will be a US or “western" arm that will do training on the open model. The conga line of billion dollar startups training their own LLMs just ended. VCs likely working on clawing back AI investment as I write.
@@sudoname I think that the demand for compute will continue to increase, and the capabilities of models will just get higher as the end goals for these companies is AGI
@@sudoname that’s like saying NVDA Blackwell gpus will decrease GPU demand because the GPUs will be much faster and more efficient. I think it’s the opposite
American AI companies: "Open source? Maybe we should partner and take advantage of the cost saving. Nah... we'll just push our bloated overpriced model and hope nobody notices"...🤣
Who didn’t see that coming? The speed to market with accuracy as good (or better in some cases) as elite AI leaders and doing so with significantly such lower investment is incredibly impressive.
@@shoe_Bin this is what I got ... I’m DeepSeek-V3, an artificial intelligence model created by DeepSeek. I’m designed to assist with answering questions, generating creative content, and helping with a wide range of tasks. If you have any questions or need assistance, feel free to ask! 😊
@@shoe_Bin I’m DeepSeek-V3, an artificial intelligence model created by DeepSeek. I’m designed to assist with answering questions, generating creative content, and helping with a wide range of tasks. If you have any questions or need assistance, feel free to ask! 😊
here is what I got ... I’m DeepSeek-V3, an artificial intelligence model created by DeepSeek. I’m designed to assist with answering questions, generating creative content, and helping with a wide range of tasks. If you have any questions or need assistance, feel free to ask! 😊
Deekseek developer also stated that they aim to provide an alternative AI source to the market, which is proven to be achievable within a reasonable cost and resource, would eventually benefits the entire CONSUMER end, instead of being forced to pay hundreds or thousands of dollars of subscriptions for a development service.
while DeepSeek and similar companies are inspired by existing technologies, they bring their own contributions to the AI space rather than simply replicating ChatGPT.
it's actually better than GPT 4o and surprisingly good at coding than the king himself claude 3.5 sonnet, it's open source and cost only 2$ for 7 million tokens 😂😂
@@Whatismyhandle So? And ChatGPT used China's BAIDU to train theyre models. whats your point? Chat GPT is open source.. The fact china did this with just $5 million dollars and weaker GPUS just proves to show how superior china is at everything it does
Chinese tech giant Baidu discovered the Scaling Law earlier than OpenAI.The core of large model development lies in the “Scaling Law”-a principle that suggests the larger the training data and model parameters, the stronger the model’s intelligence capabilities. This concept is widely attributed to OpenAI’s 2020 paper, Scaling Laws for Neural Language Models, which has since become a cornerstone of AI research. However, Dario Amodei, a co-author of the OpenAI paper, former VP of research at OpenAI, and founder of Anthropic, revealed in a podcast episode in November that he had already observed the Scaling Law phenomenon in 2014 while researching AI with Andrew Ng at Baidu. Amodei noted that as the amount of data provided to the model increased, along with the model’s size and training duration, the model’s performance improved significantly. This informal observation was later validated in OpenAI’s GPT-1 language model and has since been regarded as the “golden rule” of large model development.
Ai stocks and investing needs a second look and someone better explain why the HIGH PRICES? MASS HYPE? When the output is similar in China with LESS steller tech>>>??????
One’s better is only by competing each other will lead the world . Blocking , sanctions and trade war does help the economy, wake up Americans . Make America great again by Always thinking of and doing it it by getting up everyday and go to work . Don’t be lazy. Work hard , then you can play hard .
US really need to bring in these bright Chinese researchers. Stop the US-China trade and tech war. Open to China so you can attract these top AI talents into America and stay ahead. To win the tech battle aginst China, you will need Chinese. As racist as i may sound..
Anyone with a healthy stable mind uses the word ‘ compete ‘ instead of ‘ threat’.
US people are warlike, not civilized yet, 😂😂
For the business world it’s synonymously the same where domestic or international.
It’s always us vs them for some people
Obviously, the place where garlic is security risk doesn’t have healthy stable mind
@@naomi5418 Did you hear that humming sound from the made-in-China fridge? I think it's spying on us! 😎🤫🤭😂🍌🇺🇸
No one else is allow to compete with US, otherwise it is a threat.
usa s biggest threat is usa
yeah ，even the garlic，hahaha
😂😂😂😂
When Chinese came to Africa, they talked about trade.
When US people came to Africa, they talked about China.
@@amandagrant4331 in the past, they talk about slavery, so, I would say it's an improvement already...lol
If you attend any western university you will find a lot of STEM PhD students are Chinese. Is it really surprising a lab from China can develop a model that outperforms everything else?
Jealous?
Any teacher can teach, but 90 percent of the learning depends on individual effort. If they pay their tuition in your country, then go back to their country and build, i don't see anything wrong with theat.
Americans are discussing LGBT or other issues, science and technology, give up. It's too difficult. Be a couch potato~~~
i think hes just stating facts that its natural chinese can do this, because in US, these programers are mostly chinese as wsll.@@anyetitan8410
How insecure these people are 😂
因为吸血鬼害怕大蒜.
Oh, come on, what do you expect? They are even afraid of garlic!
Why america is so obsessed with china? Mind your own business
Americans are really scared that they cannot keep up with tough competition 🤔🤔🤔😅😅😅
中国美国是在互相紧盯
China likes to make anything american make, but dirt cheap, that hurts their profit.
@@ben8718 who said this? You ..
Ever visited china, go and visit, china is far ahead of America now... So, it is obvious there will be some jealousy, especially for America.
nah, u know, that's just media. U not hype about it then it's not for u. That word is just a click bate XD.
It’s open source, what do you mean by threat?
Doesn't matter if it were closed. It's cheaper and outperforms in some areas so it's a threat to Anthropic, OpenAI etc. They spend billions and their prices are high. Why pay $200 a month when you can get a cutting edge llm for almost free? It being open source has nothing to do with that
The whole AI backed stock frenzy is based on the fact AMERICAN companies rule the new wave of AI technology, the subscriptions price of those AIs are based on the fact americans have a monopoly on them. if some chinese open source version of chatgpt can do basically the same tasks but at 9% the cost, what happens to the stocks of openai, meta, or even nvidia, what's stopping people from ending their subscription and use the open source one. also as it turns out, we don't need the latest chips to make competitive language models, h800 chips is surfice, so why buy the upcoming blackwell, many won't, and that's bad news for many people.
China
Threat to Americans monopolizing and making mooooore money 😂
1.The U.S. thinks it undermines the U.S. monopoly
2.The model costs only 1 percent of the U.S. model, allowing most countries around the world to develop their own models with low-performance chips and fewer chips, breaking the U.S. monopoly
3.The AI high-capital moat built by the US with hardware is wiped out
America seems to have an acute self-confidence crisis syndrome, all progress achieved by others will be considered a threat, not as self-motivation to compete better.
Yeah, the US is psychologically not healthy.
When a country that welded people's doors shut to stop them from leaving during a pandemic has advanced AI with no boundaries, you'd be in a crisis too.
@@User-imwnx151didn't that also happen to the US. You have had more deaths while your US labs wanted to blame China for it, something you Westerners created.
@@User-imwnx151 I hope fewer people in the U.S. share your way of thinking; otherwise, the country may soon find itself on a downward path.
@@User-imwnx151 cool lol
If you can't compete:
- Threat
- Sanction
- Ban
can't ban open-source model that can be downloaded on a whim. cats out of the bag.
They literally fear competition with china 😅
You get a product that is just as good or better at a price 95% lower. China is ahead.
6 of the 10 AI enginners are chinese, both in Chinese or US companies. US must get the point
I don't remember seeing Chinese side bragging or trying to hold back any resources against their "competitors", yet the US been complaining and whining over "unfairness" but never they once try to understand how the Chinese work. Both sides have a totally different culture of handling competition.
no Its Indians , Indians will save USA !
@@dcwtf09 The Chinese work by copying everything. Those Chinese workers at US companies spy on the companies they work for and leak trade secrets to the Motherland.
But the Indians are smarter than the Chinese and the Americans right ?
Since when Americans are afraid of competition?
Since when haven't they been? Remember the NSA hacking Airbus so Boeing could bid lower since they were losing the race on quality?
The Americans want to maintain their abusive relationship with the rest of the world. It's imperative for the USA to avoid the world being able to choose for anything else.
Threat to America? Why everything is a threat to America?
Because, they fear competition 😅
Absolutely, everything is a threat 😂. Have you not heard about garlic from China being a threat to national security ? Ask FL senator Rick Scott. 😂😂😂
Usually it means my govt is getting ready to do war with it. You must call it “national security threat,” “threatening,” “evil,” enemy,” etc.
Calling the enemy names will later justify war on it so that the population won’t revolt, bc most people in any country are easily manipulated.
Because America is a threat to everything.
And they think garlic is more of a threat
You dont dare to report China tested world first 6th generation jet,right?
The flyovers didn’t really show off its capabilities though did they.
@@David-mr8wiit did actually. It showed the internal weapons bay is large enough to carry multiple PL-17 missiles that have a range of 500km against aerial threats.
US does not have a missile even half the range.
@ oh wow I only saw it doing a fly by in the videos I watched.
@@David-mr8wi At least they have a video. Where's the video of the mythical beast called NGAD from the USAF?🤣
@@David-mr8wiit’s made of paper. Don’t worry just a paper dragon
This is misleading. Saying 'I am developed by OpenAI' doesn't mean they stole from OpenAI. LLMs are trained on web-crawled data, which is already contaminated by generative content. All major AI models have issues with saying they are created by OpenAI. You need to spend a lot of money and effort to ensure the model does NOT say that.
这个媒体撒谎了，网上有模型链接，输入问题你可以得到回应，这是假新闻，模型返回的是我不是ChatGPT，跟chatGPT没有任何关系，使用不同的模型和算法，这个媒体一如既往的撒谎。
This is the question I asked DeepSeek AI, but it didn’t say it was ChatGPT.
What AI model are you?
I'm DeepSeek-R1-Lite-Preview, an AI assistant created exclusively by the Chinese Company DeepSeek. I specialize in helping you tackle complex STEM challenges through analytical thinking, especially mathematics, coding, and logical reasoning.
@@phoenix-king.
i have the following answer from Deepseek AI.
I am based on OpenAI's GPT (Generative Pre-trained Transformer) architecture. Specifically, I’m a version of the GPT-4 model, which is one of the latest iterations in the GPT series. This model is designed to understand and generate human-like text based on the input it receives, making it useful for a wide range of tasks like answering questions, providing explanations, and assisting with creative or technical writing. Let me know how I can assist you! 😊
Term and service violations. Well, OpenAI uses copyrighted materials for training.
OpenAI literally violating TOS for their data mining
@@thenewbanker1225 before you say this do you know how many ongoing copy right lawsuit OpenAI has?
@@jase298hkhk And top OpenAI whistleblower conveniently died. Like Boeing ones.
Bingo
OpenAi在2020年的重要论文中抄袭2017年百度的论文《Deep Learning Scaling is Predictable,Empirically》但是不在论文中标明来源，这样真的不太好。
The Chinese are very clever and resourceful and innovative and hardworking. That's why they only spent five million dollars instead of a billion. The Americans are completely outmatched.
What is this lack of confidence ? If you are so great, there is nothing to worry about. The more sanction, tariff just highlight the lack of innovation , and being very desperate to stay ahead.
For anyone who has at least basic knowledge about building and preparing foundation models and actually read DeepSeek technical reports it is clear that they did quite a lot of innovation lately, including areas where american counterparts are either stagnating or limiting access to (eg. low level model architecture). Yet they are "bad guys" because they "trained model on gpt-4 output". This is american way of "competing": when serious competition appears, concoct a story about "them copying us" or "them using slave labor" (or something else). Then start a trade war with all means possible, disguising this as "sanctions" for alleged "wrongdoings". All of this accompanied with news media producing hit pieces like this above, embedding a bunch of propaganda constructs. Pay attention to "trained on gpt-4 output" phrase that appeared in this clip several times, each time identical. This is part of programming, embedding a slogan in people minds.
Problem is this will only work for a while until people find out the actual truth. US right now is churning out late Soviet Brezhnev era levels of cope...🤣
to some ppl, they are bad guys because they are Chinese lmao. look at other comments and u find that out.
They can’t find any think to win over China
我喜欢美国媒体日益增多的歇斯底里😅
人家只是客观报道，啥坏话也没说，能听懂英文吗？
@@1.4billion65 “threatening America’s AI dominance” 这也不算是？
@@enryhen6859 能听懂英文吗？她具体采访里面有哪句说你中国坏话了吗？
@@1.4billion65能把競爭說成威脅就是缺乏安全感
@@bonkujo TH-cam标题不都这么下的吗？别人报道中间有说你中国坏话吗？你们粉红能不能别总是浑身带刺?
China is producing everything much cheaper than the US. While US citizens are buying everything at a higher price thanks to tariff, Chinese people are buying everything cheaper. It is easy to see where prosperity is heading.
yeah chinese per capita income never going to be anywhere near the US atleast till 2100
@@sublimeksaleel😂😂但是可以用100元买美国人价值1000元的东西。
@@sublimeksaleel It does not have to. When you money can buy things that are three times cheaper, your money goes a much longer way.
The proper term is not “cheaper” but “more efficiently and smarter processes “
@@sublimeksaleelper capita is meaningless when looking at well-being of society. US has a high per capita but that does not mean each US citizen has that value…the wealthiest 14 families in the US own 40% of the wealth!!!Musk, Bezos, Gates, Buffet, Zuckerberg have hundreds of BILLIONS each, so that means the average American is much poorer than the per capita which is the total GDP devised by total population!!!
China have tens of millions of STEM graduates.
USA and China will be like Coke and Pepsi.
USA without China will be a monopoly.
Nobody like a monopoly.
Everybody in the world love competition.
China will always gives the more functional, cheaper and durable option to the world.
USA will give you the luxury expensive model.
I am from Malaysia and most of us would always prefer the Chinese option.
DeepSeek: How may I help you?
America: Deep 💩💩💩...panic city 😅😅😅 We need tariffs! 🥊🥊🥊
US fears most is competition.😅
anthing challengs the US dominance is considered as a threat
So CNBC reporters are just straight up lying now huh? Here's what Deepseek answered to my prompt of "What model are you?":
"I'm DeepSeek-V3, an AI model created exclusively by the Chinese Company DeepSeek. I'll do my best to help you."
No one is sTrAigHt Up LyInG. You can find plenty of people reporting the same thing.
Is Deepseek V3 based on ChatGPT 4.0? well, yes and no. Yes.according to Deepseek itself.the answer did come from Deepseek V3. No, according to the company who made the thing. Like many other AI models, they make mistake in identifying themselves because they source many of their training content from the internet, very often they get confused and lead to hallucinations because they have troubles in identifying true and false statement as we human do all the time. At some point defining truth become philosophical, political and subjective. Afterall, we are in the post-truth era now. Thank you. Mr. Donald Trump.lol😅😅
I would say right now AI's fact check ability is somewhat behind human's. However, I am not at confident about human's ability to fact check and to do critical thinking and make rational choices moving forward into the near future. We are all doomed. Both AI and human.😅😅😅😅😝😝😝😝😝😝
I even uploaded a screen shoot from the "news" and here is its answer: Based on the content you’ve shared, the screenshot appears to show a response claiming to be ChatGPT, developed by OpenAI, and based on the GPT-4 architecture. However, I am not ChatGPT or based on GPT-4. I am DeepSeek-V3, an AI model developed independently by the Chinese company DeepSeek.
If the screenshot suggests otherwise, it could be a misunderstanding, misrepresentation, or possibly manipulated content. I can assure you that I am not ChatGPT, and my training and development are entirely separate from OpenAI’s models. If you have further concerns or questions, feel free to ask!
@@experiment9024 Do you seriously think that the Chinese aren't going to program it to say that?? LMAO!
@@OneTequilaTwoTequila Even that, it is still false what the video shows. That is the point.
Yes Chinese models are great and open source :)
and free to use.
US AI is like US eV's. Good, bad or ugly, it's too expensive.
😅
A new model AI for free open . So can you imagine how much money will be saved by this new model on the world???
This is what the United States cannot accept. If anyone can produce a fifth-generation stealth fighter at low cost, how can Americans make money
@ oh yeah
用户只会想着便宜好用，只要满足这两点，就没有人可以禁止它。
how a OPEN SOURCE AI treat US AI? Everyone can use it...
they like to control the 'open' source projects so they can inject attack in them later - check out Snowden's reports
What do you mean by violation? OpenAI model has been trained with all the data available and violating everything.
A thief can’t complain if get robbed what s/he stole.
DeepSeek V3, powerful and significantly cheaper to train.
Also check some news:
Unitree B2-W, while Boston Dynamics struggles to sustain itself and has been in a holding pattern for years with minimal technical and commercial progress.
Two Gen 6 fighter jets are in the sky, while NGAD is currently paused.
Type 076 is an amphibious assault ship equipped with an electromagnetic catapult, whereas the USS Ford’s catapult consistently faces reliability issues.
Real battle field in ukraine is full of DJI and drones with components from same place.
What dominance? Dominance in deindustrialization?
I think what deepseek has proved is that if you are smart enough you can use 1/50 compute resource to train the same good frontier LLM and open source it for the greater goodness of human kind. Not like openAI did as closed AI shop. What irony it is 😂
Warmonger doesn't allow any competitor to challenge them...
Let’s not act like openAI isn’t using Reddit, FB, X and stuff from the internet like music and art to train its LLM with. In some cases this is why openAI has been threatened with multiple lawsuits. Let’s not act brand new…lol
The speed at which AI is evolving is terrifying. It's like going from a Model T to a Corvette in a few years. This stuff will be smarter than we are before we even understand how it works or how to use it.
nobody had a smartphone about 15 years ago, nobody could imagine living without one today, we never know what will happen 15 years down the line, before you know it we could have super AI smart autonomous robots running around doing everyday tasks...
DeepSeek has achieved the level of the most advanced large models in the United States with only 10% of the cost that Americans use. At present, only the published parts have been announced, and it is possible that more advanced parts have not been published, and it is also possible that it has already been leading a lot.
This afternoon they updated the message:
I was developed based on DeepSeek's independent technology, and I'm continuously learning and improving to better assist users in solving complex STEM challenges.
😂 H-1b IP bye bye
lol, I saw that too
Ai overcapacity coming soon
China is not our enemy. The greed of Western companies, striving for hundred percent profit margins, is our enemy! China is doing us a favor by offering affordable options, which encourages Western companies to compete by providing better and cheaper alternatives!
Impossible.. isn't America the greatest tech country in the world..?? So says Trump?
😅😅
They used Chat GPT to train their models and managed to outperform Chat GPT..
这是个开源项目，并未是使用已经被数据污染的ChatGPT来辅助训练，那样毫无意义。你的摸黑和造谣只能让你显得是个无知的小丑罢了。你的摸黑谣言改变不了什么，因为专业人士知道是怎么回事。
Google uses Baidu's big model to train itself, so his Chinese answers all say that they are Baidu's model.
it’s not the main reason that deepseek only spent 6m to train ChatGPT 4 level model.
The same with WeChat. It has both features of Facebook and Twitter, and many more.
2:30 Because the Chinese engineers are sleek and efficient; can do more with less; can do faster with cheaper parts.
That woman knows nothing about AI.
she just read the script In front of her
She’s hotter than you
She is good making sarcastic smile
Because most of the $$$$$ are going to fill the pockets of the bosses... lol
Why the US can't allow other country to compete against them? Why must you think that it is a threat?
If Deepseek really violate the TOS of OpenAI , then OpenAI should examine the library of Deepseek and everything will come out , because if it's really open source as it claimed
China is always a threat. US never a threat😂
Better than you is a threat???
This just confirms what everyone should know: America is One Big Con Job 😂
With Trump's presidency, economic shifts are expected to be significant, especially given the current recession and the potential impact of future rate cuts. Although rate cuts might not boost inflation as hoped, they may lead banks to further restrict consumer and corporate lending, contributing to a deflationary period for various assets. This environment could result in declining stock values, retail and housing sales, and rising unemployment due to layoffs. For investors, a diversified portfolio especially with stocks and cryptocurrencies offers some protection, serving as a hedge amid volatility. Both long- and short-term trading strategies can help manage risks, providing stability as markets adjust. I have managed to grow a nest egg of around 130k to a decent 532k in the space of a few months... I'm especially grateful to Paul Maxwell , whose deep expertise and traditional trading acumen have been invaluable in this challenging, ever-evolving financial landscape.
He mostly interacts on Te le grams, using the user-name paulmaxwel
The clarity and precision in Paul maxwel market predictions are astounding. I'm so grateful to have found his insights here on TH-cam as well.
I have been working with Paul Maxwell for about a year now, his total professionals has delivered solutions well above my expectations, all thanks.
Paul goes deeper than just looking at surface-level trends. he explores technical, fundamental, and sentiment analysis, offering a comprehensive perspective on the market..
Well this is not my first time of seeing his name on social platform. I think he's extremely brilliant and lucrative at the same time.
No one can be the top dog forever. Sooner or later, someone somewhere will do a better job than you, whether you like it or not. That is how the world works.
A Toyota or a Rolls Royce, it does the same thing. It gets you from point A to point B. Your choice.
It's completely open source, so why it's a threat to anyone. It's a threat to those who promote "democratic AI", which is also known as Chines-excluded AI.
Like auto industry..i sincerely hope that Chinese entities dominate the industry thus ensuring that corrupt lobbying entities turn to creator with zero arrogance and abundance of fear of G-d..
The Chinese is building a highway over OpenAi moat, LMAO
Wow challenging USA how dare are they
Is US is unbeatble country forever?😅
Try to find top AI research papers without a Chinese name in it. Not many.
NVDA sales will defined take a hit if deepseek is so capable with 6m, imagine if the Chinese invest 1B
I think the opposite
@@t.w2000 what are your arguments why NVDA can still hold/grow? i think chip demand will slow down.
I agree. At best there will be a US or “western" arm that will do training on the open model. The conga line of billion dollar startups training their own LLMs just ended. VCs likely working on clawing back AI investment as I write.
@@sudoname I think that the demand for compute will continue to increase, and the capabilities of models will just get higher as the end goals for these companies is AGI
@@sudoname that’s like saying NVDA Blackwell gpus will decrease GPU demand because the GPUs will be much faster and more efficient. I think it’s the opposite
China :Thank You, America.Thank you, Biden. Thank you, Gina Raimondo. No sarcasm intended, only smerk and giggles😝😝😝😝😁😁😁😁
Jai Hind. Even a China garlic can be construed as National Security threat
American AI companies: "Open source? Maybe we should partner and take advantage of the cost saving. Nah... we'll just push our bloated overpriced model and hope nobody notices"...🤣
A threat of monopoly is a good threat.
Who didn’t see that coming? The speed to market with accuracy as good (or better in some cases) as elite AI leaders and doing so with significantly such lower investment is incredibly impressive.
media in the states are going nuts everyday
US is dominating AI? Surprised 😂😂😂
It's not threatening, it's dominating, get it?!
Americans can't stand it again!
actually gpt also say itself baidu wenxinyiyan
Do cover Open AI’s whistleblower Suchir Balaji’s murder in silicon valley suspiciously. His parents have gathered evidence against the suicide claims.
I asked DeepSeek the same question and it doesn't identify as chatgpt ... so not sure where she got her answer from.
What propaganda? This is not CCP media.
it showed same output as CNBC.
@@shoe_Bin this is what I got ... I’m DeepSeek-V3, an artificial intelligence model created by DeepSeek. I’m designed to assist with answering questions, generating creative content, and helping with a wide range of tasks. If you have any questions or need assistance, feel free to ask! 😊
@@shoe_Bin I’m DeepSeek-V3, an artificial intelligence model created by DeepSeek. I’m designed to assist with answering questions, generating creative content, and helping with a wide range of tasks. If you have any questions or need assistance, feel free to ask! 😊
here is what I got ... I’m DeepSeek-V3, an artificial intelligence model created by DeepSeek. I’m designed to assist with answering questions, generating creative content, and helping with a wide range of tasks. If you have any questions or need assistance, feel free to ask! 😊
New planes, ships, AI... 😲
Only in China is this legal
Hella yea!!! AI shouldn’t be for profit they are using our data!! I’m up for open models like this one
This is Free and Open Source.
The whistle blower of OpenAI "commited suocide" ;-P , just as Boeing whistle blower. America is great!
Deekseek developer also stated that they aim to provide an alternative AI source to the market, which is proven to be achievable within a reasonable cost and resource, would eventually benefits the entire CONSUMER end, instead of being forced to pay hundreds or thousands of dollars of subscriptions for a development service.
Another national security risk.
while DeepSeek and similar companies are inspired by existing technologies, they bring their own contributions to the AI space rather than simply replicating ChatGPT.
US is too much worried.
I'm about to test deep seek v3 and seeing this news of it being a "threat" is beyond me LOL
it's actually better than GPT 4o and surprisingly good at coding than the king himself claude 3.5 sonnet, it's open source and cost only 2$ for 7 million tokens 😂😂
this is huge wtf- AI malinvestment has been identified - bubble just went pop
fun fact, us never had any ai dominance, just ahead a bit, together with china
China China China
China again😂😂😂
If you don't publish news about china daily, your full day is wasted.😅
; -CNBC 😅
DeepSeek is certainly no deepshit.
Overcapacity
That’s no need to put victim infront to be a drive for advancement.
I ask same question, but give the right answer. Please put video from generating the answer that can be trusted
ai companies gave the models out like free candy and trained it on everyone data, and now people are train on thier models/data shocking
Human rights, freedom and national security are the first issues
you just don’t understand how young of AI is! So AI models can be changed or created all the time
How did they tell their children, it's just a job, or, China is a threat?
Basically Chinese figure a way to reverse engineering ChatGPT in a lower cost.
Go to school bro.
@ knock your head and come out something useful. Apparently they are using ChatGPT to train their model.
@@Whatismyhandle So? And ChatGPT used China's BAIDU to train theyre models. whats your point? Chat GPT is open source.. The fact china did this with just $5 million dollars and weaker GPUS just proves to show how superior china is at everything it does
@ that is what I said they reverse engineered ChatGPT at a lower cost. What is wrong statement with that?
@@Whatismyhandle
Chinese tech giant Baidu discovered the Scaling Law earlier than OpenAI.The core of large model development lies in the “Scaling Law”-a principle that suggests the larger the training data and model parameters, the stronger the model’s intelligence capabilities. This concept is widely attributed to OpenAI’s 2020 paper, Scaling Laws for Neural Language Models, which has since become a cornerstone of AI research.
However, Dario Amodei, a co-author of the OpenAI paper, former VP of research at OpenAI, and founder of Anthropic, revealed in a podcast episode in November that he had already observed the Scaling Law phenomenon in 2014 while researching AI with Andrew Ng at Baidu. Amodei noted that as the amount of data provided to the model increased, along with the model’s size and training duration, the model’s performance improved significantly. This informal observation was later validated in OpenAI’s GPT-1 language model and has since been regarded as the “golden rule” of large model development.
Why is it a threat? Competition is good for everyone. How insecure can they be? These people have confidence crisis.
Ai stocks and investing needs a second look and someone better explain why the HIGH PRICES? MASS HYPE? When the output is similar in China with LESS steller tech>>>??????
Well, they out thought us again!😖
America can only cry in these days, while the china is developing fast.
One’s better is only by competing each other will lead the world . Blocking , sanctions and trade war does help the economy, wake up Americans . Make America great again by Always thinking of and doing it it by getting up everyday and go to work . Don’t be lazy. Work hard , then you can play hard .
😭😭😭everything is threatens. lets watch dump and dumper movie
US really need to bring in these bright Chinese researchers. Stop the US-China trade and tech war. Open to China so you can attract these top AI talents into America and stay ahead. To win the tech battle aginst China, you will need Chinese. As racist as i may sound..
how welll does it do in the ARC AGI test and frontier math against world leading o3?