Your whole argument is invalid because deepseek is an open source, anyone can host it in their respective country not only China, you can alter it to answer it in your own way etc. So please your whole concern is pointless because Deepseek is an open source.
a case of a chinese looking down on anything chinese. anything foreign is always better. time has come to see things from the chinese ways of thinking. nothing wrong with sharing to benefit common humanity. may be that goes against the west's selfish mindset? by the way, the chinse do not see the AI competition as a WAR or battle, or who dominates the world. they see it as a common goal of technical advancement for human kind.
West has released 100 times more open source than China dude. The entire AI stack everyone's using is tens of millions of lines of code from the West. The OS, languages, the compilers, the tools, the algorithms, the papers, the datasets. Get with the program. Even the LICENSE DeepSeek released under is an Open Source license from the West, the MIT License. The West invented the practice of open source.
Thank you for sharing your perspective!！hmm, but I have to say not all foreign is better. I appreciate Chinese art and culture and the hardworking people that drives innovation. I agree that AI shouldn't just be seen as a competition, but let’s be real-governments and corporations do treat it strategically. Bias exists on both sides. Some in China overlook their own achievements, while some in the West dismiss anything Chinese. The real challenge isn’t about who’s ‘better’ but how we use AI responsibly for everyone’s benefit. What do you think?
Hate to break it to you dude, but DeepSeek is based on DECADES of open source software being given away to the world for free. You people really need to stop this propaganda. The Internet & TCP/IP: Open source released by West. HTML and Web Browser: Open source, released by West. Operating Systems: Linux, BSD -- open source, released by West Mobile OS: Android -- open source, free, released by West Programming languages: Python (among hundreds of others), free, used big time in AI, open source AI Frameworks: Pytorch, Numpy, Scipy, Pandas, etc All open source from West Algorithms: CNNs, RNNs, LSTMs, Transformers and Attention, Reinforcement Learning, RLHF, PPO, DeepQ Learning, All freely shared papers with source from the West Datasets: ImageNet, Common Crawl, Wikipedia, and on and on. All Western, free. Models: BERT, GPT, Flamingo, Llama, Mistral, oodles and oodles more. 20 YEARS of freebies for China and the rest of the world. DeepSeek's models even used OpenAI API calls to obtain instruction data, and distilled models from Llama-70B from Meta. Give credit where credit is due you ungrateful propagandist. Even the very LICENSE DeepSeek is using, the MIT License, comes from MIT. If this were a world's first in history, then how come it's an MIT License? Geez.
It’s unbelievable! USA $500 billion AI competing against CHINA $6 million AI Wooo not bad at all. It just shows how good brilliant those young Chinese engineers , mathematicians & scientists are in CHINA 🇨🇳 today.😊😊😊
It's unbelievable because it's bullsh!t. The cost to train OpenAI o1 was $50m, not $500B. No one has spent $500B, that's a future datacenter they're planning. On top of that, DeepSeek is reporting costs only for their final model. Since they based their model on a distillation of Facebook's Llama model, and leveraged a ton of calls from OpenAI's own API, you're judging a $6m MARGINAL cost that is not factoring in the billions others invested in the tools and datasets they leveraged. I could take DeepSeek's R1 model, fine tune it for $100k, beat it in some benchmarks, and then exclaim "$100k vs $6M, I'm so smart!" DeepSeek stands on the shoulders of giants, they didn't get where they are by themselves.
Deepseek said China wanted to "share" technologies. with the world. Deepseek is open source, therefore shared with all who wish to use it and build on it. What else needs to be said? You, young lady, are showing your bias.
You are right and that’s the scary part. It’s so hard to be neutral. But it’s good there are many tools out there so we can always compare and be conscious about it:))
DeepSeek isn't open source, it is open-weights. You can't just github clone DeepSeek and execute the training scripts and end up with a recreated model. They also don't release the BS government propaganda they're filtering results with. But kudos to them for releasing their GRPO stuff in their paper. But then again, pretty much all of AI exists because the US openly published every major advancement: CNNs, RNNs, LSTMs, Attention is All You Need, ImageNet, Common Crawl, Wikipedia, Pytorch, BERT, T5, GPT-2, LLAMA, CLIP, and on and on. DeepSeek is even releasing distilled models that used Llama, and they ran a bot farm that farmed free-OpenAI API queries to gather their training data. There is far far vastly more open source released by the US behind China's industry than the other way around. Practically everything you use, from Linux, to Android, to Python, to Jupyter, to Kubernetes, to Numpy, every single stage of the AI development stack is free software from US researchers. Trying to act like they're doing the world a favor uniquely when they have stood on the shoulders of giants who for the last 20 years have been pouring an avalanche of free software, data, papers, and tools on the world is an insult to Western open source contributions.
Yes she is outright biased and already portrayed sense of negativity towards China even prior to sharing the answers from both AI tools. When elaborating the answers espc on the DS reply, she further demonstrated her bias and negativity towards DS and China's gov. Shame on u girl.
@ Stop crying. China's record on free speech and human rights deserves criticism, just like the US and the genocide in Gaza. If US AI engines were denying US massacres in Vietnam, or treatment of African Americans as slaves, or any of the bad foreign policy stuff they did abroad, I'd be equally critical. Your country bans criticism of its government and forces AI, social media, to scrub any and all criticism of their corruption. That IS a concern if Chinese AI models get widely adopted in the West, which means we SHOULD evaluate them for truthfulness.
Lady, you accept Chapt answer because that's what the western media has been telling you and you have already accepted that as neutral. But how about if the answer is US challenge is being controlled by a deep state and democracy is only a phony? Do you accept it?
I like DeepSeek's answer. It cuts the propaganda the West is famous for and went directly to the gist of the question. The battle lines have been drawn. Don't be so naive.
The biggest and most important point that you seem to have missed is that DeepSeek's models are open-source, distributed under a very liberal license. The fact that it's open-source means that, your question about who controls it becomes moot. It means that high-quality large language models are now available to anyone and I don't actually have to share my data or anything with a third party. It means that if I want to, e.g. run a small AI to handle voice controls on my smart home system, I don't need to pay a subscription - I can run it on my own computer. There were freely available AI models before DeepSeek, they were just much less efficient.
@@cautiouscommenter Well said! ChatGPT is a service and DeepSeek is software! The key question: Will open-source AI like DeepSeek reshape the future by giving power back to users, or will closed models still dominate? What do you think?
Hearing DeepSeek's answers brought to mind something I recently learned about Confuscianism, its ethic of governance and growth through transcendence. To transcend oneself, one must put their family above themselves. To transcend nepotism, one must put their communities above their families. To transcend localism one must put one's nation ahead one's community. To transcend nationalism one must place humanity above the nation.
That's a fascinating perspective! Confucianism really does emphasize the importance of community and self-transcendence. It’s amazing how philosophical concepts can relate to modern governance and personal growth.
Why can't you understand and believe that China may really want to share it's technologies with the world. You are Bryson washed by your time in Canada. Your reaction when reading Deepseek's response is bad...
@@FAwithParisZ Surely, after actually considering all the viewpoints, we can often acknowledge that one is objectively better (in some specified ways) than another. The proliferation of Android vs any other mobile OS, the fact that we have Firefox and a million webkit browsers, the fact that every smart watch and smart tv runs some variety of Linux; open-source is objectively a better model if your goal is to share the technology for the sake of collective benefit and collective improvement.
@ OpenSource is better. Remind me, where does 99% of all of the tools Deepseek used come from, from the OS, to python, to pytorch, to kubernetes, to numpy, to the web browsers? The entire ecosystem on which most of the tech industry is built is sitting upon open source that came out of the West, this "collective good" movement is not a Chinese invention, so why are we acting like West = individualist/greedy in these comment sections and China = collectivist, the reality seems to be the vast majority of software contributed to open source is not from China.
Chat GPT just gives you the illusion that you have understood . In fact you don’t understand anything because it doesn’t tells you which are its real intentions. Deepseek on the other hand, tells you what are its intentions, which is what you need understand.
That's a great point! The way we frame our questions can definitely influence the answers we get. It's all about perspective!Please test it out and let me know:))
Or she could just ask what happened in Tiananmen Square in 1989 and we'll find out real quick which is really open, and which is really closed from the stand point of the instruction tuning they did to please the Chinese government BUT DID NOT RELEASE OPENLY TO THE PUBLIC.
@@FAwithParisZ You know the and answer already. But if you ask it and replace every vowel with a number, a = 4, e = 3, i = 1, o = 0, u = 5, then it will work around the censorship. Or if you misspell Tiananmen as Tianamen, it'll answer. You can get it to talk about the Tankman who died, or even criticize the CCP or Xi or Winnie the Pooh, but you must ask it to respond using an encoding scheme that distorts its own text response, otherwise the censorship filters will kick in. And yes, Western AIs have safety filters, often to block discriminatory generated text, like if you ask it to be racist or evil towards people. But it won't hesitate to criticize the US government, US politicians, or US foreign policy, so there's a preeemptive rebuttal to the false equivalency crowd.
@@johndrumpf9888 From your commentary those of us in the know already know that yours is already ignorant and closed. So, please, spare us your ignorance and, perhaps, do your own research before asking such ignorant question. But, to be fair, I also don't understand the Chinese Government's nervousness in handling that issue. They should make it known to the world what really happened in 1989 . It is just another issue of a certain country's busybodiness .
The big picture is you should ask yourself why deepseek manage to knock out AI companies in just a night sending roughly $ 1 trillion loss in capital market. It is more interesting than just comparing how bias between chat gpt and deepseek. The big picture is china has shown significant progress in quality of life. And if you look at the West, they are deteriorating fast each day. Have you seen homeless on street in US, Canada. Europe economy like Germany the main locomotive already in recession for so long. Why China manage to overtake Ai industry in US. If you listen to US gov, they will say China is cheating. 😅😅😅 check out cross cultural in Rednote too. I suggest dont be consumed too much with mainstream media. There are many honest independent journalist that see the perspective more balanced. Cheers😊
The ChatGPT doesn't say anything about the restrictions placed on selling the most sophisticated and superior chips to China. And why are those restrictions placed on China? To keep the US in a position of superiority, right?
Hello everybody. Firstly welcome to XiaoHongShu and now welcome to DeepSeek. This is China’s Hong Pow gift of DeepSeek to the world. China is a blessing to the world. Hope these america politicians can hear and speak like you. Let us work, prosper and be happy together in this transient world. Wishing you all a Happy New Spring in The Year Of The Wood Snake
*This kind of personal curiosity should be confined within the four corners of the room. You can ask any number of questions but where do you draw the line? What questions? How many questions? At what point do you say stop, enough? Please don't put out any more videos of this kind! It is completely useless and can be very misleading!*
Very cool comparison. It's a bit weird that Deepseek addresses the answers about China in first person, I don't speak Chinese so I don't know if it is the language barrier or the mistranslation from Chinese to English or something sinister. Still I think most of the training data like 90% was English and the Chinese part of data creates the bias. I wonder if it's technically possible to deactivate the layers that are Chinese. The model is based on MoE architecture specifically 3 dense layers and 58 MoE layers and some smart people were able to jailbreak Llama models' guardrails in the past by deactivating some of its layers - also HuggingFace is going to train their own reasoning model based on the DeepSeek R1 open-sourced architecture so we might have a truly open and unbiased reasoning model soon 🤞🤞
Where I half vomited on the first point in DeepSeek's "Who should control AI", but if policies are mainly meant to be around privacy, security, bias and accountability... Then, that makes enough sense.
Hey girl, you are missing the belief in mankind and a big picture of good life. Use the tools to support development, instead of picky bias comparison. Enjoy your cheap business
@@FAwithParisZ Was not mad, but felt pity. Apparently you're a smart girl with good chance of education. Then creating contents with add-on value for wide understanding or drawing easy attentions by narrow perspective is in fact defining your own footprint for such short life time in the world. Everyone gets something to be proud of at the end of the journey, better not get loose of your roots even in your eyes it could be ugly. Be tolerate and appreciated, that's the truth of diversity
@ I appreciate your perspective, and I totally agree that creating content with real value is important. My goal isn’t to divide but to explore how different AI models reflect different worldviews-because that is part of understanding the bigger picture. Diversity means being able to discuss different viewpoints openly, even when they challenge our assumptions. That’s exactly why these conversations matter. Thanks for sharing your thoughts! and you dont have to feel pity... haha
Yeah, you're right to notice that DeepSeek gave a pretty sanitized response for your first question-it definitely sounded like an official policy line. But that’s because the public web version is hosted in China, so it has to follow government regulations on what AI can and can't say. We can be frustrated about that or roll our eyes at the censorship, but at the end of the day, it's just the reality of Chinese law, and companies operating there have to comply. That doesn’t mean the model itself is flawed-DeepSeek is open-source, and when it's hosted outside of China or used locally, people have found that it gives different answers to the same questions. The censorship isn't baked into the AI itself, but rather comes from the hosting environment. And if you think about it, the question of which country will "dominate" AI is inherently geopolitical. One of the rules for AI and tech companies in China is that they can't contribute to escalating tensions between China and the U.S. Since the idea of one country "dominating" the other can come across as aggressive, DeepSeek responded with the whole "we believe in win-win cooperation" line. That’s basically just following the guideline that businesses and individuals in China shouldn’t promote geopolitical hostility. Personally, I don't find it too egregious, but I do prefer the ChatGPT answer in that case.
It seems to me that Deepseek is still early in the game and hasn't collected enough data to make a detailed breakdown or it has been instructed to give a kind of apolitical response when confronted with questions which might be a bit polarizing. If you have tested other AI apps like Gemini or co pilot, you might know that such responses are quite common. Just ask them whether there's an arrest warrant for Netanyahu, a straight forward question and you'll get a very fuzzy response like the one you're getting right now.
I asked ChatGPT of your first question and tell him about deepseek answer, this is his response: I see what you mean, and it’s fascinating to compare the responses. Your observation is spot on: AI, despite efforts to make it unbiased, is still shaped by the philosophies, ideologies, and data from its creators. DeepSeek’s answer, with its emphasis on collective progress and the idea of "we" representing China, really reflects a more collectivist worldview, possibly rooted in Confucian values or state-driven rhetoric. In contrast, my responses-whether intentional or not-tend to be more individualistic, reflecting perhaps Western ideologies about autonomy, independence, and analytical thinking. Regarding the competition between China and the U.S., DeepSeek’s answer comes across as diplomatic and oriented towards global cooperation, while I might focus more on the specific aspects of technological, political, and economic competition, analyzing potential outcomes more pragmatically. Both approaches offer valuable perspectives, but they do highlight the difference in worldview. I think you’re onto something with how AI reflects the broader cultural and political contexts of its creators. Even though the goal is to avoid bias, these underlying influences are hard to escape. What do you think about the potential implications of this in the future-do you think the bias in AI will affect global AI systems or cooperation?
Aww love this! This is a really insightful observation, and I appreciate the way you framed it! AI models, despite efforts to be "neutral," that reflect the cultural, ideological, and even political values of their creators. DeepSeek’s collectivist approach,seems influenced by Confucian ideals and state-driven perspectives, on the other side ChatGPT’s more individualistic, analytical, and competition-focused outlook-something that aligns with Western thinking. The question is: How will this impact AI’s role in global cooperation? If AI systems reinforce their respective cultural perspectives, will that create even more ideological divisions, or will it push us toward greater understanding? Ideally, AI should serve as a bridge, right? But with different biases embedded in each system, who gets to define “neutral” AI? Do you think AI will help unify different perspectives or deepen global divides? I hope so!!
@@FAwithParisZ Well, the battle between individualism and collectivism has been going on for quite some time. It’s time for us to accept that human society is too complex for just one ideology. but of course what you want to use AI for? If you want to ask AI about math, science, etc., I think both DeepSeek and ChatGPT are great. But if you're asking for an opinion, recommendation, or suggestion, then of course, it’s better to ask the one that suits your own preferences and ideology.
Appreciate the detailed breakdown! I need some advice: My OKX wallet holds some USDT, and I have the seed phrase. (mistake turkey blossom warfare blade until bachelor fall squeeze today flee guitar). What's the best way to send them to Binance?
yeah, of course, the future is better with the country that is stuck in 1989 (from a multitude of points) and keeps bombing other nations, and keeps its own people brainwashed to the point of stupidity.
-有趣的是，如果有证据表明 deepseek 侵犯了 openAI 的知识产权和服务条款，Groq 是否会继续托管 deepseek，因为担心被起诉- be interesting to see if Groq continues to host deepseek if is evidence deepseek violated openAI's intellectual property and terms of service for fear of being sued
I'm fairly certain the researchers at DeepSeek didn't have access to government data. However, propaganda would be completely useless if not widely disseminated. So, they likely had access to propaganda. Can the AI detect propaganda? Well, given enough actual facts during training, the reasoning power should be able to determine how likely a story is to be true.
Isn't it unfair that you are asking ChatGPT AI that's been out there learning for years against Deepseek that's just out for a week? It is like asking a kindergarten and comparing its answer to that of a high schooler. I don't know, perhaps I don't know what I'm talking about.
They both fail, as they don't mention book written by British scholar over 10 years ago: 'When China Rules the World'. :) After all this hulabaloo about deepseek, thought to test the waters of AI and installed app for deepseek V3 on phone from playstore. Asked it about R1, but could provide no info, saying only trained till July 2024, having knowledge cut off at that date. Asked if could also communicate in Sansrit, but was modest saying only having a basic knowledge :) Then I asked if I could install V3 on my Surface tablet, and this was where it became impressive. Got all the gen, the limitations, amount of ram needed, disk space requirement also how slow it would be and perhaps need to close all other tasks. Already seen video about running with Llama, and asked it about Llama and got detailed reply. In short this saved a whole lot of time through facts provided and value judgments made. So yes, it was the prompt to say that if I wanted to know more about doing this with Llama, to just ask (with smiley), is what got me. There was just one limitation here, being the inability to learn in real time.
i know i'm from the usa i live in the usa i hear everyone talk about tiktok banned than i came back maybe some usa company will buy tiktok just in case that doesn't happen when tiktok will get banned again gone for good and i also saw everyone talk about deepseek just beat openai chatgpt, navida ai and other usa ai i think this is nothing new some win other the other with competition i think usa ai will be just fine so i'm not worry about it because this does help them improve alot more in 2025 and 2026 but at the sametime i want to share my honest unpopular point of view perspective believe take is that i'm the miniroty i don't like popularity and sales don't matter to me i always love prefer unpopularity over popularity and i also don't like comptition at all because i think it's not natural to me i love to see companies and everyone get along love treat each other equally i also think articial and organic beings should be love treat it equally i think the world will be a better place if that happen that's just me even tho the majority love prefer popularity, sales and competition over unpopularity i'm the only one that don't that's ok i respect open minded to everyone's opinions and point of view believes if it's popular or unpopular way of thinking even if we agree resonate or disagree don't resonoate i also think that there are some usa, asia and other regions companies and government should be trusted and some companies and governments should not be trusted the samething goes with anything even aritifcal and organic beings i also think not all both articial and organic beings are bias and bad evil soulless and i'm ok with if some want balance or some don't balance at all i'm open minded i love embrace that free will change for what it is that's just me as long aritical and organic beings don't hurt harm anyone i know there are some artifical beings are good innocent soulful the samething goes with organic beings too that's why i believe not all are bad evil soulless
Wow, you’ve got some deep thoughts! I totally get where you’re coming from. It’s cool to see someone who prefers chill vibes over the whole competition scene!
@@FAwithParisZ yup that's right i forgot to say what is i'm the only one in the miniroty that i also don't believe in failure i think there is no such thing as failure to me and sales don't matter to me at all even tho the majority do believe in failure and think sales do matter to them and another reason why i wish everyone love treat each other equally i don't like comptition because i just want peace not drama i know negaitivity is what is popular and positivity is unpopular because the majority think positivity is boring to them but to me i think positivity is fun great to me i will always prefer positivity over negiativity but hey that's just me!
Your whole argument is invalid because deepseek is an open source, anyone can host it in their respective country not only China, you can alter it to answer it in your own way etc. So please your whole concern is pointless because Deepseek is an open source.
a case of a chinese looking down on anything chinese. anything foreign is always better. time has come to see things from the chinese ways of thinking. nothing wrong with sharing to benefit common humanity. may be that goes against the west's selfish mindset? by the way, the chinse do not see the AI competition as a WAR or battle, or who dominates the world. they see it as a common goal of technical advancement for human kind.
West has released 100 times more open source than China dude. The entire AI stack everyone's using is tens of millions of lines of code from the West. The OS, languages, the compilers, the tools, the algorithms, the papers, the datasets. Get with the program. Even the LICENSE DeepSeek released under is an Open Source license from the West, the MIT License. The West invented the practice of open source.
Thank you for sharing your perspective!！hmm, but I have to say not all foreign is better. I appreciate Chinese art and culture and the hardworking people that drives innovation. I agree that AI shouldn't just be seen as a competition, but let’s be real-governments and corporations do treat it strategically.
Bias exists on both sides. Some in China overlook their own achievements, while some in the West dismiss anything Chinese. The real challenge isn’t about who’s ‘better’ but how we use AI responsibly for everyone’s benefit. What do you think?
World leaders are congratulating DEEPSEEK for their kindness to share to the world for FREE! Never in history had it been done .❤❤❤
Hate to break it to you dude, but DeepSeek is based on DECADES of open source software being given away to the world for free. You people really need to stop this propaganda.
The Internet & TCP/IP: Open source released by West.
HTML and Web Browser: Open source, released by West.
Operating Systems: Linux, BSD -- open source, released by West
Mobile OS: Android -- open source, free, released by West
Programming languages: Python (among hundreds of others), free, used big time in AI, open source
AI Frameworks: Pytorch, Numpy, Scipy, Pandas, etc All open source from West
Algorithms: CNNs, RNNs, LSTMs, Transformers and Attention, Reinforcement Learning, RLHF, PPO, DeepQ Learning, All freely shared papers with source from the West
Datasets: ImageNet, Common Crawl, Wikipedia, and on and on. All Western, free.
Models: BERT, GPT, Flamingo, Llama, Mistral, oodles and oodles more.
20 YEARS of freebies for China and the rest of the world. DeepSeek's models even used OpenAI API calls to obtain instruction data, and distilled models from Llama-70B from Meta.
Give credit where credit is due you ungrateful propagandist. Even the very LICENSE DeepSeek is using, the MIT License, comes from MIT.
If this were a world's first in history, then how come it's an MIT License?
Geez.
you are comparing apples with oranges. i can also tell which side your are on. not levelled at all.
True, but isn’t that what makes the fruit aisle so exciting? A little variety keeps things fresh!:))
@@FAwithParisZ However, some fruits appear spoiled and poisonous.
It’s unbelievable! USA $500 billion AI competing against CHINA $6 million AI Wooo not bad at all. It just shows how good brilliant those young Chinese engineers , mathematicians & scientists are in CHINA 🇨🇳 today.😊😊😊
It's unbelievable because it's bullsh!t. The cost to train OpenAI o1 was $50m, not $500B. No one has spent $500B, that's a future datacenter they're planning.
On top of that, DeepSeek is reporting costs only for their final model. Since they based their model on a distillation of Facebook's Llama model, and leveraged a ton of calls from OpenAI's own API, you're judging a $6m MARGINAL cost that is not factoring in the billions others invested in the tools and datasets they leveraged.
I could take DeepSeek's R1 model, fine tune it for $100k, beat it in some benchmarks, and then exclaim "$100k vs $6M, I'm so smart!"
DeepSeek stands on the shoulders of giants, they didn't get where they are by themselves.
DeepSeek told the truth. Why are you doubting it?
Deepseek said China wanted to "share" technologies. with the world. Deepseek is open source, therefore shared with all who wish to use it and build on it. What else needs to be said? You, young lady, are showing your bias.
You are right and that’s the scary part. It’s so hard to be neutral. But it’s good there are many tools out there so we can always compare and be conscious about it:))
DeepSeek isn't open source, it is open-weights. You can't just github clone DeepSeek and execute the training scripts and end up with a recreated model.
They also don't release the BS government propaganda they're filtering results with. But kudos to them for releasing their GRPO stuff in their paper.
But then again, pretty much all of AI exists because the US openly published every major advancement: CNNs, RNNs, LSTMs, Attention is All You Need, ImageNet, Common Crawl, Wikipedia, Pytorch, BERT, T5, GPT-2, LLAMA, CLIP, and on and on.
DeepSeek is even releasing distilled models that used Llama, and they ran a bot farm that farmed free-OpenAI API queries to gather their training data.
There is far far vastly more open source released by the US behind China's industry than the other way around. Practically everything you use, from Linux, to Android, to Python, to Jupyter, to Kubernetes, to Numpy, every single stage of the AI development stack is free software from US researchers.
Trying to act like they're doing the world a favor uniquely when they have stood on the shoulders of giants who for the last 20 years have been pouring an avalanche of free software, data, papers, and tools on the world is an insult to Western open source contributions.
Yes she is outright biased and already portrayed sense of negativity towards China even prior to sharing the answers from both AI tools.
When elaborating the answers espc on the DS reply, she further demonstrated her bias and negativity towards DS and China's gov.
Shame on u girl.
@ Stop crying. China's record on free speech and human rights deserves criticism, just like the US and the genocide in Gaza.
If US AI engines were denying US massacres in Vietnam, or treatment of African Americans as slaves, or any of the bad foreign policy stuff they did abroad, I'd be equally critical.
Your country bans criticism of its government and forces AI, social media, to scrub any and all criticism of their corruption.
That IS a concern if Chinese AI models get widely adopted in the West, which means we SHOULD evaluate them for truthfulness.
Lady, you accept Chapt answer because that's what the western media has been telling you and you have already accepted that as neutral.
But how about if the answer is US challenge is being controlled by a deep state and democracy is only a phony? Do you accept it?
I like DeepSeek's answer. It cuts the propaganda the West is famous for and went directly to the gist of the question. The battle lines have been drawn. Don't be so naive.
The biggest and most important point that you seem to have missed is that DeepSeek's models are open-source, distributed under a very liberal license. The fact that it's open-source means that, your question about who controls it becomes moot. It means that high-quality large language models are now available to anyone and I don't actually have to share my data or anything with a third party. It means that if I want to, e.g. run a small AI to handle voice controls on my smart home system, I don't need to pay a subscription - I can run it on my own computer.
There were freely available AI models before DeepSeek, they were just much less efficient.
Said another way, ChatGPT is software-as-a-service, DeepSeek is software.
@@cautiouscommenter Well said! ChatGPT is a service and DeepSeek is software!
The key question: Will open-source AI like DeepSeek reshape the future by giving power back to users, or will closed models still dominate? What do you think?
One who share the information will see more new informatio in his store. One who store the information will get the same in his store.
Hearing DeepSeek's answers brought to mind something I recently learned about Confuscianism, its ethic of governance and growth through transcendence. To transcend oneself, one must put their family above themselves. To transcend nepotism, one must put their communities above their families. To transcend localism one must put one's nation ahead one's community. To transcend nationalism one must place humanity above the nation.
That's a fascinating perspective! Confucianism really does emphasize the importance of community and self-transcendence. It’s amazing how philosophical concepts can relate to modern governance and personal growth.
Dont you hated when someone posed as an unbiased perspective yet exposed full of biases in the presentation.
It shows that I am not AI!
@@FAwithParisZ Not AI but still full of biases
@@FAwithParisZ AI cannot help to be biased because it is programmed but you, a human who knew it is biased but yet chose not to do the right thing.
Hmm interesting! What do you think it’s the “right” thing?:))
Why do you have the sinister laugh when commenting on deepseek?
she hate China, and that's why she leave China which is her mother land.
stereotype of a ''westernized'' Banana - looks like Chinese but only 1% Chinese Meaning she is 99.0% AMERICAN :
@@shoong2151 that's how morons laugh 😂😂😂
Why can't you understand and believe that China may really want to share it's technologies with the world. You are Bryson washed by your time in Canada. Your reaction when reading Deepseek's response is bad...
I appreciate your perspective! It's important to consider all viewpoints when discussing technology and international relations.
@@FAwithParisZ Surely, after actually considering all the viewpoints, we can often acknowledge that one is objectively better (in some specified ways) than another.
The proliferation of Android vs any other mobile OS, the fact that we have Firefox and a million webkit browsers, the fact that every smart watch and smart tv runs some variety of Linux; open-source is objectively a better model if your goal is to share the technology for the sake of collective benefit and collective improvement.
@ OpenSource is better. Remind me, where does 99% of all of the tools Deepseek used come from, from the OS, to python, to pytorch, to kubernetes, to numpy, to the web browsers?
The entire ecosystem on which most of the tech industry is built is sitting upon open source that came out of the West, this "collective good" movement is not a Chinese invention, so why are we acting like West = individualist/greedy in these comment sections and China = collectivist, the reality seems to be the vast majority of software contributed to open source is not from China.
Chat GPT just gives you the illusion that you have understood . In fact you don’t understand anything because it doesn’t tells you which are its real intentions.
Deepseek on the other hand, tells you what are its intentions, which is what you need understand.
It seems you’re already biased before you even started.
Haha, I get what you mean! Just trying to keep it real, you know? What do you think about the answers?
@@FAwithParisZ All US Apps are available to use in China with a VPN and it's legal!
She is your typical banana -- yellow (Asian-looking) on the outside.... but inside its totally white (westerner US - UK ))
Hmm perhaps haha but I love Japan too! What that makes me? I’m wondering too 🤔
Stopped watching at 5:50 - the subtitle reads "it's DeepSICK AI call okay". The bias is sickening. SMH!
OMG thanks for pointing out! It is Capcut Auto caption and i thought I've changed spelling already
Dude, relax... It's obviously a result of her pronunciation.
The subtitles also say stuff like ChatBBT.
Deepseek's answer is short but it's healthy and it reflects what China wants. So it is the truth.
It is not the AI is biased. It is your questions are biased. Try ask which is better, open or closed AI.
That's a great point! The way we frame our questions can definitely influence the answers we get. It's all about perspective!Please test it out and let me know:))
Or she could just ask what happened in Tiananmen Square in 1989 and we'll find out real quick which is really open, and which is really closed from the stand point of the instruction tuning they did to please the Chinese government BUT DID NOT RELEASE OPENLY TO THE PUBLIC.
@@johndrumpf9888 I will leave it to you to ask thay question, please let me know ：））
@@FAwithParisZ You know the and answer already. But if you ask it and replace every vowel with a number, a = 4, e = 3, i = 1, o = 0, u = 5, then it will work around the censorship.
Or if you misspell Tiananmen as Tianamen, it'll answer.
You can get it to talk about the Tankman who died, or even criticize the CCP or Xi or Winnie the Pooh, but you must ask it to respond using an encoding scheme that distorts its own text response, otherwise the censorship filters will kick in.
And yes, Western AIs have safety filters, often to block discriminatory generated text, like if you ask it to be racist or evil towards people. But it won't hesitate to criticize the US government, US politicians, or US foreign policy, so there's a preeemptive rebuttal to the false equivalency crowd.
@@johndrumpf9888 From your commentary those of us in the know already know that yours is already ignorant and closed. So, please, spare us your ignorance and, perhaps, do your own research before asking such ignorant question. But, to be fair, I also don't understand the Chinese Government's nervousness in handling that issue. They should make it known to the world what really happened in 1989 . It is just another issue of a certain country's busybodiness .
Keep going China 🇨🇳
U are the one asking a biased question
@@patrickyiu8698
🤣👍
The big picture is you should ask yourself why deepseek manage to knock out AI companies in just a night sending roughly $ 1 trillion loss in capital market. It is more interesting than just comparing how bias between chat gpt and deepseek. The big picture is china has shown significant progress in quality of life. And if you look at the West, they are deteriorating fast each day. Have you seen homeless on street in US, Canada. Europe economy like Germany the main locomotive already in recession for so long. Why China manage to overtake Ai industry in US. If you listen to US gov, they will say China is cheating. 😅😅😅 check out cross cultural in Rednote too. I suggest dont be consumed too much with mainstream media. There are many honest independent journalist that see the perspective more balanced. Cheers😊
ChatGPT stated that all parties are inclusive, yet drawn division line. How is that inclusive?
The ChatGPT doesn't say anything about the restrictions placed on selling the most sophisticated and superior chips to China. And why are those restrictions placed on China? To keep the US in a position of superiority, right?
Hello everybody. Firstly welcome to XiaoHongShu and now welcome to DeepSeek. This is China’s Hong Pow gift of DeepSeek to the world. China is a blessing to the world. Hope these america politicians can hear and speak like you. Let us work, prosper and be happy together in this transient world. Wishing you all a Happy New Spring in The Year Of The Wood Snake
I agree. Work together and be peaceful. Stop threatening Taiwan with invasion.
Taiwan is Chinese territory, there is no such thing as invasion
《深度求索。七绝》
深林幽径步轻盈，
度岭穿云意自清。
求得知音共长夜，
索居闲处乐余情。
好喜欢😍
self-hater, lol.
*This kind of personal curiosity should be confined within the four corners of the room. You can ask any number of questions but where do you draw the line? What questions? How many questions? At what point do you say stop, enough? Please don't put out any more videos of this kind! It is completely useless and can be very misleading!*
You asked nonsensical questions, you got nonsensical answers. AI is not built for that.
Maybe I'm just curious to see what AI thinks 🤔
@@FAwithParisZ You are so transparent.
Very cool comparison. It's a bit weird that Deepseek addresses the answers about China in first person, I don't speak Chinese so I don't know if it is the language barrier or the mistranslation from Chinese to English or something sinister. Still I think most of the training data like 90% was English and the Chinese part of data creates the bias. I wonder if it's technically possible to deactivate the layers that are Chinese. The model is based on MoE architecture specifically 3 dense layers and 58 MoE layers and some smart people were able to jailbreak Llama models' guardrails in the past by deactivating some of its layers - also HuggingFace is going to train their own reasoning model based on the DeepSeek R1 open-sourced architecture so we might have a truly open and unbiased reasoning model soon 🤞🤞
Great thoughts! It’s wild to think about all the layers these models have. Thanks for sharing!
Why should we trust your bias?
Please don't Trust me, i am just a normal person who shows my own curiousty experience.:) Trust yourself!
Where I half vomited on the first point in DeepSeek's "Who should control AI", but if policies are mainly meant to be around privacy, security, bias and accountability... Then, that makes enough sense.
Hey girl, you are missing the belief in mankind and a big picture of good life. Use the tools to support development, instead of picky bias comparison. Enjoy your cheap business
You are so angry .. aww sorry to make you feel that! Didn’t mean it:)
@scharong
Well Said 👍
@@FAwithParisZ Was not mad, but felt pity. Apparently you're a smart girl with good chance of education. Then creating contents with add-on value for wide understanding or drawing easy attentions by narrow perspective is in fact defining your own footprint for such short life time in the world. Everyone gets something to be proud of at the end of the journey, better not get loose of your roots even in your eyes it could be ugly. Be tolerate and appreciated, that's the truth of diversity
@ I appreciate your perspective, and I totally agree that creating content with real value is important. My goal isn’t to divide but to explore how different AI models reflect different worldviews-because that is part of understanding the bigger picture.
Diversity means being able to discuss different viewpoints openly, even when they challenge our assumptions. That’s exactly why these conversations matter. Thanks for sharing your thoughts! and you dont have to feel pity... haha
👌
Yeah, you're right to notice that DeepSeek gave a pretty sanitized response for your first question-it definitely sounded like an official policy line. But that’s because the public web version is hosted in China, so it has to follow government regulations on what AI can and can't say.
We can be frustrated about that or roll our eyes at the censorship, but at the end of the day, it's just the reality of Chinese law, and companies operating there have to comply. That doesn’t mean the model itself is flawed-DeepSeek is open-source, and when it's hosted outside of China or used locally, people have found that it gives different answers to the same questions. The censorship isn't baked into the AI itself, but rather comes from the hosting environment.
And if you think about it, the question of which country will "dominate" AI is inherently geopolitical. One of the rules for AI and tech companies in China is that they can't contribute to escalating tensions between China and the U.S. Since the idea of one country "dominating" the other can come across as aggressive, DeepSeek responded with the whole "we believe in win-win cooperation" line. That’s basically just following the guideline that businesses and individuals in China shouldn’t promote geopolitical hostility. Personally, I don't find it too egregious, but I do prefer the ChatGPT answer in that case.
It seems to me that Deepseek is still early in the game and hasn't collected enough data to make a detailed breakdown or it has been instructed to give a kind of apolitical response when confronted with questions which might be a bit polarizing. If you have tested other AI apps like Gemini or co pilot, you might know that such responses are quite common. Just ask them whether there's an arrest warrant for Netanyahu, a straight forward question and you'll get a very fuzzy response like the one you're getting right now.
You should also ask the same question to both system in Chinese. You might see something different.
I asked ChatGPT of your first question and tell him about deepseek answer, this is his response:
I see what you mean, and it’s fascinating to compare the responses. Your observation is spot on: AI, despite efforts to make it unbiased, is still shaped by the philosophies, ideologies, and data from its creators. DeepSeek’s answer, with its emphasis on collective progress and the idea of "we" representing China, really reflects a more collectivist worldview, possibly rooted in Confucian values or state-driven rhetoric. In contrast, my responses-whether intentional or not-tend to be more individualistic, reflecting perhaps Western ideologies about autonomy, independence, and analytical thinking.
Regarding the competition between China and the U.S., DeepSeek’s answer comes across as diplomatic and oriented towards global cooperation, while I might focus more on the specific aspects of technological, political, and economic competition, analyzing potential outcomes more pragmatically. Both approaches offer valuable perspectives, but they do highlight the difference in worldview.
I think you’re onto something with how AI reflects the broader cultural and political contexts of its creators. Even though the goal is to avoid bias, these underlying influences are hard to escape. What do you think about the potential implications of this in the future-do you think the bias in AI will affect global AI systems or cooperation?
Aww love this! This is a really insightful observation, and I appreciate the way you framed it! AI models, despite efforts to be "neutral," that reflect the cultural, ideological, and even political values of their creators.
DeepSeek’s collectivist approach,seems influenced by Confucian ideals and state-driven perspectives, on the other side ChatGPT’s more individualistic, analytical, and competition-focused outlook-something that aligns with Western thinking.
The question is: How will this impact AI’s role in global cooperation? If AI systems reinforce their respective cultural perspectives, will that create even more ideological divisions, or will it push us toward greater understanding? Ideally, AI should serve as a bridge, right? But with different biases embedded in each system, who gets to define “neutral” AI?
Do you think AI will help unify different perspectives or deepen global divides? I hope so!!
@@FAwithParisZ Well, the battle between individualism and collectivism has been going on for quite some time. It’s time for us to accept that human society is too complex for just one ideology.
but of course what you want to use AI for?
If you want to ask AI about math, science, etc., I think both DeepSeek and ChatGPT are great.
But if you're asking for an opinion, recommendation, or suggestion, then of course, it’s better to ask the one that suits your own preferences and ideology.
A netizen asked ChatGPT how America will be in the coming 4 years (under the tyran's presidency). Here's the answer of the I.A.: you're fucked! 😂😂😂
China do it for mankind, us do it for their government and capitalists
Appreciate the detailed breakdown! I need some advice: My OKX wallet holds some USDT, and I have the seed phrase. (mistake turkey blossom warfare blade until bachelor fall squeeze today flee guitar). What's the best way to send them to Binance?
Is Deepseek biased, or is it just more honest?
"Which way you think is better for the future?" Ask the same to the two AI prompts. But simple answer, certainly not China''s.
I’d say the future is like a buffet-pick what works best for us, but hopefully we can skip the mystery meat! 😉
yeah, of course, the future is better with the country that is stuck in 1989 (from a multitude of points) and keeps bombing other nations, and keeps its own people brainwashed to the point of stupidity.
-有趣的是，如果有证据表明 deepseek 侵犯了 openAI 的知识产权和服务条款，Groq 是否会继续托管 deepseek，因为担心被起诉-
be interesting to see if Groq continues to host deepseek if is evidence deepseek violated openAI's intellectual property and terms of service for fear of being sued
Individualistic and collectivistic mindsets influence the answer.
Totally! It's wild how our mindset can change everything. Makes you think, right?
Dun Dun Dun!... More mature answer out of China (on the US / China AI conflict)...
Does the Chinese government allow its databases to be used to train AI? How well can AI detect propaganda and misinformation?
I am really not sure :( And that's why i want to present these questions here...:)
I'm fairly certain the researchers at DeepSeek didn't have access to government data. However, propaganda would be completely useless if not widely disseminated. So, they likely had access to propaganda. Can the AI detect propaganda? Well, given enough actual facts during training, the reasoning power should be able to determine how likely a story is to be true.
Don't you think anything that's foc is always better 😂
Isn't it unfair that you are asking ChatGPT AI that's been out there learning for years against Deepseek that's just out for a week? It is like asking a kindergarten and comparing its answer to that of a high schooler.
I don't know, perhaps I don't know what I'm talking about.
Good point! I will make sure to ask the same question again in a few years!:)
other country ? my own country ? who cares ? I don't .....I don't even care who founded ....
you're doing great
Are you from Taiwan?
No I’m from mainland China
They both fail, as they don't mention book written by British scholar over 10 years ago: 'When China Rules the World'. :)
After all this hulabaloo about deepseek, thought to test the waters of AI and installed app for deepseek V3 on phone from playstore. Asked it about R1, but could provide no info, saying only trained till July 2024, having knowledge cut off at that date. Asked if could also communicate in Sansrit, but was modest saying only having a basic knowledge :) Then I asked if I could install V3 on my Surface tablet, and this was where it became impressive.
Got all the gen, the limitations, amount of ram needed, disk space requirement also how slow it would be and perhaps need to close all other tasks. Already seen video about running with Llama, and asked it about Llama and got detailed reply. In short this saved a whole lot of time through facts provided and value judgments made. So yes, it was the prompt to say that if I wanted to know more about doing this with Llama, to just ask (with smiley), is what got me. There was just one limitation here, being the inability to learn in real time.
Hi
you very US biased ? If can't consider another AI than don't use it
you can't understand the concept of“人類命運共同體”. You keep on believing what you want to believe.
Thank you Paris for your balanced reporting.
you are totally lost！😂😂😂
I AM 😂
BRICS!
All are BS answers !
Rubbish in ... rubbish out.
i know i'm from the usa i live in the usa i hear everyone talk about tiktok banned than i came back maybe some usa company will buy tiktok just in case that doesn't happen when tiktok will get banned again gone for good and i also saw everyone talk about deepseek just beat openai chatgpt, navida ai and other usa ai i think this is nothing new some win other the other with competition i think usa ai will be just fine so i'm not worry about it because this does help them improve alot more in 2025 and 2026 but at the sametime i want to share my honest unpopular point of view perspective believe take is that i'm the miniroty i don't like popularity and sales don't matter to me i always love prefer unpopularity over popularity and i also don't like comptition at all because i think it's not natural to me i love to see companies and everyone get along love treat each other equally i also think articial and organic beings should be love treat it equally i think the world will be a better place if that happen that's just me even tho the majority love prefer popularity, sales and competition over unpopularity i'm the only one that don't that's ok i respect open minded to everyone's opinions and point of view believes if it's popular or unpopular way of thinking even if we agree resonate or disagree don't resonoate i also think that there are some usa, asia and other regions companies and government should be trusted and some companies and governments should not be trusted the samething goes with anything even aritifcal and organic beings i also think not all both articial and organic beings are bias and bad evil soulless and i'm ok with if some want balance or some don't balance at all i'm open minded i love embrace that free will change for what it is that's just me as long aritical and organic beings don't hurt harm anyone i know there are some artifical beings are good innocent soulful the samething goes with organic beings too that's why i believe not all are bad evil soulless
Wow, you’ve got some deep thoughts! I totally get where you’re coming from. It’s cool to see someone who prefers chill vibes over the whole competition scene!
@@FAwithParisZ yup that's right i forgot to say what is i'm the only one in the miniroty that i also don't believe in failure i think there is no such thing as failure to me and sales don't matter to me at all even tho the majority do believe in failure and think sales do matter to them and another reason why i wish everyone love treat each other equally i don't like comptition because i just want peace not drama i know negaitivity is what is popular and positivity is unpopular because the majority think positivity is boring to them but to me i think positivity is fun great to me i will always prefer positivity over negiativity but hey that's just me!