DeepSeek, AI, and the New Cold War - with Yanis Varoufakis and Arnaud Bertrand
ฝัง
- เผยแพร่เมื่อ 8 ก.พ. 2025
- DeepSeek has upended the AI industry with open-source models that rival OpenAI’s GPT-4, shaking the global tech and financial order. But its rise also raises urgent questions about security, economic power, and AI’s future.
In this live discussion, Yanis Varoufakis and Malaysia-based entrepreneur and commentator Arnaud Bertrand will explore the fallout, hosted by Mehran Khalili. How does DeepSeek’s model challenge monopolistic “cloud capital” like Amazon’s Alexa? Does it prove that cutting-edge AI no longer requires vast computational resources? And with its data stored on Chinese servers, is AI becoming the new frontline in U.S.-China tensions?
Beyond the tech, this is a battle over power-financial, political, and military. What happens next? Join us live, be part of the conversation, and send in your questions!
FOLLOW THE SPEAKERS
---------------------
Mehran: x.com/mkhalili
Yanis: x.com/yanisvar...
Arnaud: x.com/RnaudBer...
SUPPORT US
---------------------
Join DiEM25: diem25.org/join
Donate to DiEM25: diem25.org/donate
Subscribe: / diem25official
#AI #Deepseek #chatgpt
Thank you, Yanis Varoufakis, for framing it so perfectly! Why should we pay for OpenAI, Llama, Google, etc. when they are the ones who harvested our usage data to build their products. Instead, we should be using DeepSeek because it is free - and because it was built on top of data models built upon our data.
To the US, Deepseek was a Sputnik moment. For the world, Deepseek is proliferation movements for many Global South countries to develop their own AI system in the future.
China is not in the global south
@@DanieleNierodeepseek is open source and free. The south can buid AI base on that. It dosent matter where china land. China sure give a gift to the ppl who cant afford to devlope their own AI
@dalao2yang Oh, I understand now.
But this is not only a gift to poor countries, it is a gift to European countries as well.
If this was really a manoeuvre orchestrated by the chinese communist party, they would have been geniuses: defuse the american companies leverage and advantage.
Thank You China for promoting a real people’s Democracy in our world.
Thank you for your gift of Open AI DeepSeek.
DeepSeek is just the beginning of a new era. Congratulations to this big success.
It's still a new era of unprecidented authoritarianism. So congratulations on your complete lack of forethought.
@@JohnMoran I prefer a multipolar world order, yes. And Europe to be no longer vassal 🤡state of monster 👿
@@JohnMoran cold war mindset will finally fail!
The fact that DeepSeek is free and open illustrates the confidence China has. China can offer DeepSeek for free because China has many more, or even unlimited, products to come
I don't think that's their "angle"
It (perhaps) connects more with the desire to use developed systems to assist other countries and build the BRICS/ Belt & Road Ecosystem.
Feedback (and applications for the A.I) from those countries is going to be interesting.
I think it was a deliberate strategic move to undermine the "US AI Market". They knew what they were doing.
Yes! Creativity, like soul or "the spring" (Pablo Neruda's poem abt cutting all the flowers cannot stop it) are all the birthrights of decent humans.
Let's hope so...@@TheUAoB
@@TheUAoBthey want to burst the biggest scam bubble of the century, to exposed the western hagemony where money is concern.
Free Society : Closed and expensive AI
Authoritrian Society : Open and free AI
Happy Chinese New Year.
LOLOL Free Society - in Words.
Free society - in Deeds.
Most People in the World don't Know that US definition of Democracy and Freedom is US Corporations and Billionaires in Power and Ownership. Tyranny to them is Government is able control and Own their Corporations and Wealth. Nothing to do with the average people who don't have Billions upon Billions and ever further away is Corporations nowadays are Trillions upon Trillion. If you are not in that category its not Democracy.
😂 Perfect Computational Logic.
But China is not an authoritarian country, it's just a country with law n order to maintain peace. No guns, almost crime free!
Stop it. U confused my political correctness ideology
@@fannyalbi9040 haha!
Absolute gold ! Bravo
Deepseek gave the world, there are ways for many countries to develop their AI technologies.
Those countries are actually a resource to develop the A.I
How many intelligent individuals in Africa are at this moment wondering "What if I use DeepSeek to....... ..... ?
@@rogerstarkey5390 You Seem ro be unaware of Technology initiatives.. developing in Africa .. They are GREAT ! BRILLIANT..
Just in language - not only is Google Alphabet working with local groups to develop Automatic Translation in all forms of the rich base of tribal languages....
Plus African entrepreneurs. Ingenious.. work.. helping dialect speakers where there is no written language.. Transact thru voice instructions, do banking online, order cabs & services ..online etc etc
Simple example is LENALI .. there are many others Nkunne look it up..
Revolutionary changes / progress..
To be able to communicate eventually with any other language group, globally ❗
@@rogerstarkey5390 deepseek can help Africa in long term.
I think we can’t put DeepSeek and OpenAI at the same level. For us they are incomparable. OpenAI is entirely about money and deep state collaboration (Trump and Stargate). DeepSeek is free and it’s open source (this alone is revolutionary - and anti-imperialist). Liang the Deepseek founder said: “Does this innovation tangibly improve societal operational efficiency?” This points to a different tech philosophy. Also he said: “AGI won’t monopolize progress, it’ll democratize it.” This means AGI needs to be for all the people of the world.
*I like DeepSeek answers better*
How to make comments in bold font? Thanks
*bold* innit?
Yeah "*text in bold*" try it
Did it without asking an AI. Lol
I prefer ChatGPT when it comes to questioning internet content, when it comes to coding, maths or science Deepseek is much better.
this is “death sentence” for Open AI🎉🎉🎉
Know the name of who made Deepseek available free for everyone--he is Liang Wen Feng 梁文锋
standing on shoulders of so many others...
@@theresabarzee1463 which technology is not standing on shoulders of many? 2000 years Before the West, China has developed very sophisticated technology. Check on 3.5M words, 15,000 pages, multiple volumes of master pieces of "Science and Civilisation in China"by Cambridge scholar Joseph Needham, this humongous encyclopedia recorded science and technology of China in past 2000 years BEFORE the West! China is not just a civilization, China created the cilvilization!
Yes, Mr Liang is indeed a cut above the oligarchies and the other technos.
A very intellligent and humanistic man.👍😊
The technology has never been better for building an open source global platform for digital democracy so we can vote for principles and policies instead of personalities and political parties.
Another amazing conversation, thank you Mehran, Yanis and Arnaud ❤❤❤ And of course all of DiEM25, friends and like-minded people 🙏
“It is the heart of U.S. policy to use fascism to preserve capitalism while claiming to save democracy from communism.”
- Michael Parenti
Add great conversation. As a Canadian living next door the abomination of country or what used to be our neighbor is far more frightening than a better AI model made by a Chinese start up. I see it as a gift for the time being....
Learn more about the origin of DeepSeek (Re "start up")
@@rogerstarkey5390 tell all.....
No worries - you won't have that neighbour for too long - soon you will become the 51st state and live happily ever after in a big happy family.
@@PhiloSurfer Perhaps we will be a family when our American neighbors apply for permanent residence as refugees in our remote community. Your dear leader will have to bring in conscription to fulfill his dreams of conquest.
We will see how that goes.....
@@PhiloSurfer Canadian do not agree with you.
deepseek is the mirror that the americans need to see themselves
hear, hear!
Deepseek is the mirror in which Americans see themselves.
Well said
There were already open source LLMs available before deep seek.
Finally halfway into this discussion the important points are coming out. Thank you Yanis! A much more frugal model development process is what’s interesting. The need for fewer resources and the ability to share its “thinking process” make it special. (Among other things).
With great respect to your thoughts and wisdom, you missed that one point, Why the West was so surprised when China can excel in AI.
China is good in Mathematics and engineering. Software is about mathematics and engineering principles. An AI especially is all about mathematics.
So if you understand, STEM, Then this should not be a surprise ;/)
Also racism, delusions of grandeur, exceptionalism?
Exactly.
Western bias
Western arrogance
Chinese say Arrogence lead to ignorance
DEEPSEEK is open source and free for any one to use. It is an excellent gift to poorer countries that cannot afford the expensive models.
Thank you learnt a lot.
Excellent show
Great discussion, thank you. Open AI have already copied the Deepseek code to build their latest model, according to an (unconfirmed) post on X showing source code in Chinese - apparently they forgot to translate parts of it before releasing the beta to the public.
Can you link the X post?
ha!
thank you gents
Wonderful discussion
Your creativity is fantastic!
Yanis laid it out perfectly with the engineering culture which China currently has vs what the US used to have. Along with solid continuing education, discipline and some cut throat style management, you need an end game goal to where people can direct their expertise and energy into. Virtually every project in China is being met with the same intensity as the Manhattan project…all the Chinese engineers have accepted the challenges and facing them with a do or die mentality.
Whereas in the US, and maybe other western countries, that drive is largely gone and many people rather come up with schemes to milk funds and play office politics for power tripping euphoria vs actually solving problems and implementing ideas to better the society.
Another highlight on China’s engineering culture is the rewards some of them are getting outweigh what they can get from the West. Besides material compensation, they’re also being celebrated as heroes of humanity, inspiring more and more of the young…and it's not just Chinese engineers, but recently a Turkish engineer was celebrated for being instrumental in helping Huawei survive the tech war with the US.
Finally, there is an unique safety net for engineers in China, who are shielded against "some" forms of racism and predatory mercantilism (drowning in debt and non-stop onslaught by different increasing unnecessary fees and penalties) in the US, which can be very demoralizing and kill careers.
Tech question - where is mr V’s microphone. The audio is great!
I asked this question to Chat GPT and DeepSeek: My vehicle was traveling east at 100 km/hour. Another vehicle was traveling west at 100 km/hour. As our vehicles approached one another, I made eye contact with the other driver. So, what was our speed of our vehicles at the moment when we made eye contact? Chat GPT was unable to answer this simple question...So there!
do you mean relative speed, or individual speed? what is the answer to your question? there are three answers - your relative speed is 200kph, the speed of your vehicle is 100kph at the point of contact, and the speed of the other vehicle is also 100 kph. which do you think is the correct answer, and what did you actually mean to ask?
works for me for DS:
When two vehicles approach each other moving in opposite directions, their relative speed is the sum of their individual speeds. In this scenario:
1. **Relative Speed**:
Your vehicle's speed (east) + Other vehicle's speed (west) = **100 km/h + 100 km/h = 200 km/h**.
This means, from your perspective, the other vehicle is approaching at **200 km/h**, and vice versa.
2. **Time to Collision (if no evasive action)**:
If you were a distance \( D \) apart when eye contact was made, the time until meeting/passing would be:
\[
\text{Time} = \frac{D}{200 \ \text{km/h}}.
\]
For example:
- At **1 km apart**: \( \frac{1}{200} \) hours = **18 seconds**.
- At **0.5 km apart**: \( \frac{0.5}{200} \) hours = **9 seconds**.
3. **Perception and Safety**:
At such high relative speeds, the window for reaction is extremely short. Even a momentary distraction (e.g., prolonged eye contact) could reduce critical decision-making time. A collision at this combined speed (**200 km/h**) would likely be catastrophic due to the energy involved.
4. **Physics of Passing**:
When vehicles pass each other at high relative speeds, the airflow and pressure changes can create a noticeable "whoosh" or turbulence, which might feel unsettling.
If you were reflecting on this scenario, it underscores the importance of maintaining focus and minimizing distractions when driving at high speeds, especially in opposing traffic lanes.When two vehicles approach each other moving in opposite directions, their relative speed is the sum of their individual speeds. In this scenario:
1. **Relative Speed**:
Your vehicle's speed (east) + Other vehicle's speed (west) = **100 km/h + 100 km/h = 200 km/h**.
This means, from your perspective, the other vehicle is approaching at **200 km/h**, and vice versa.
2. **Time to Collision (if no evasive action)**:
If you were a distance \( D \) apart when eye contact was made, the time until meeting/passing would be:
\[
\text{Time} = \frac{D}{200 \ \text{km/h}}.
\]
For example:
- At **1 km apart**: \( \frac{1}{200} \) hours = **18 seconds**.
- At **0.5 km apart**: \( \frac{0.5}{200} \) hours = **9 seconds**.
3. **Perception and Safety**:
At such high relative speeds, the window for reaction is extremely short. Even a momentary distraction (e.g., prolonged eye contact) could reduce critical decision-making time. A collision at this combined speed (**200 km/h**) would likely be catastrophic due to the energy involved.
4. **Physics of Passing**:
When vehicles pass each other at high relative speeds, the airflow and pressure changes can create a noticeable "whoosh" or turbulence, which might feel unsettling.
If you were reflecting on this scenario, it underscores the importance of maintaining focus and minimizing distractions when driving at high speeds, especially in opposing traffic lanes.
@rewIndustry The answer is 0.
Collision is ass u me d ?!
@theresabarzee1463 You definitely need help beyond AI because you just made an ass out of yourself.
Agree with Yanis about the different focus of AI between USA and China. To upskill you guys, Deepseek using expert agents and a logical train to provide an answer, whereas ChatGPT LLM for example uses static pattern recognition within the trained data to provide an answer. Thats a mammoth difference between using predictive text on your phone and having logical answers based on reasoning.
i love listening to all of you❤ deeply grateful for being able to paint while i listen to you❤ so happy❤🙏🏽
I love this, thank you for sharing and I’m happy we can help stimulate creativity!
Simply put: science, computer science included, should be part of a global borderless cooperative effort. All knowledge is power, isn’t it.
Yanis and Arnaud what a cross over
30:05 A B&G comment from the live chat earlier this morning. Thanks for the great livestream from your interesting guests today! Any opportunity to interact with Yanis (🔥 Prometheus🔥) Varoufakis is appreciated. He always brings fire to the working people alright! 😀
NASA began to develop a space pen, but when development costs skyrocketed the project was abandoned and astronauts went back to using pencils, along with the Soviets.
**Build Decentralized Alternatives First**: Create open-source, federated cloud platforms (e.g., using blockchain or peer-to-peer networks like IPFS) to distribute infrastructure ownership. Projects like [Solid] (solidproject) (Tim Berners-Lee’s decentralized data project) or (Hetzner) (community-driven hosting) offer blueprints.
Could it be that one approach is the weaponized commodity culture which would be extractive vs. the gift culture of sharing the methods & survival, of using human capital for thriving on a threatened planet?!
DeepSeek is just a tease !! I think there is more held back!!
China produced four times as many scientists and engineers as the US. The founder of the firm that produced Deep Seek was already rich as a Quant and ran his own investment firm. He collected the best engineering talent from the top Chinese university. Allowed these young engineers to work horizontally with little hierarchy to facilitate creativity. Can you image the super bureaucratized US corporate structure doing that? Huawei is a cooperative and is rapidly advancing on its own chip production. It may not need the super Nvidia chips to run more efficiently. Chinese open economy allows for creative competition. How about the destructive competition model which dominates here. In 2018 Peter Thiel one of the worst of the Silicon Valley operatives whose company Palantir is completely intertwined with the National Security state as a spy operation said to a group of Stanford business students: to be successful is to wipe out your competition and create a monopoly. This is the mentality of the gurus now in Silicon Valley. It's a losing strategy.
Looking US panic is beyond funny ^^
I'm so happy to be free in this moment (free of mind, that is).
I liked the point 'Deepseek has socialized AI (innovation)'. I did not like the argument '... it has to be controlled because ..." because that reflects the hegemonic view, the disease that has killed the best parts of our lifes througout history. Anyway, very good discussion, thanks👍
I think people got this wrong, chatgpt was china's Sputnik moment, deepseek is china's Apollo moment
If it is produced and priced at lower point, it would certainly attract more buyers/consumers..
2. the utility of Deep Seek to Chinese economy will make the latter further strong .. as well as innovative ..
"The Achille's Heel" of AI and robots is electricity.
Our “Achilles heel” is … our … heels. And legs. It’s the numerous limitations imposed on us by our biology. Which includes greedy bad faith human actors! The time to bring AI “to heel” is now. By bringing its bad-faith greedy capitalist enablers whose interest is private property and their own wealth under control by whatever means necessary to prevent them from developing AI to dominate the masses and employ it to profit from us, not to serve us.
Remember that song by "Orchestral Manouevers In The Dark"? Look it up: "Electricity", it's a great song, an anthem for our era!
**Public Investment**: Governments could fund national/regional cloud infrastructure (e.g., EU’s [GAIA-X]) to compete with AWS, ensuring public oversight.
Ultimately it's up to the EU to get off its ass. US ain't never gonna change its modus operandi, not in the 'tech' field nor any other field of endeavour.
The only advantage of Deepseek is that you can run it on your own system. I doubt it that good things will come out of it. Fortunately AI is a bubble waiting to explode.
We all want a digital government, don't we?
AI must be considered as a General Purpose & / or specialized TOOL.. enabling society & development.. To be used by anyone.. Free .. or for a nominal charge..
That should be, by LAW , - the basis of all regulation..ie
Freedom to use, innovate etc
44:12 The dystopian possibilities Arnaud is pointing to here already exist with Amazon. Amazon can (and does) sue authors that own the copyright of an audiobook, if the author provides a reader with the ability to remove DRM protections. And the legal penalty, in the US, for the author, is much higher than the one someone would get for infringing on the author's copyright.
It's a totalitarian extraction approach. Quite scary. Open AI cannot claim property of human knowledge and then even try to regulate how we use it among us.
Alibaba's Taobao had the same agenda couples of year ago, and Jack Ma was crushed by the authority, which was the rare thing that most of the Chinese citizens agree with the government
@@eriktan9654 Why do you say it was a rare thing? I think most Chinese citizens agree with what the government is doing, most of the time.
AI requires energy to work. All we have to do to stop it's threat if it gets too powerful or intrusive is to UNPLUG it!
It can be powered by solar and batteries, one of my bike lights is still working after 4 yrs! Pacemakers, watches……barb
Arnaud Bertrand’s not much of a “leftist”. He’s just a classic liberal individualist and therefore effectively an enabler of greedy US corporate capitalism. His passive idealism is exactly HOW we wind up dominated by a greedy, aggressive oligarchic capitalist class who dominates society ideologically and materially. ONLY a class conscious material analysis more like the one Yanis employs will EVER bring technology like AI under control by “the people” that Arnaud keeps whinging on about; his critique is idealist and sees imperialist aggression as a cultural or moral failing rather than what it actually is; a manifestation of the material economic class conflict produced by the economic relations of capitalism itself. We individuals cannot simply “will” ourselves to be a better society.. We have already; clearly we fail. That’s capitalism and the greedy capitalists who most succeed and dominate us are rewarded for greed and selfishness. Arnaud’s wavy-gravy kumbaya approach will NEVER remedy our aggressive exploitative and imperialist capitalist society and nations.
Only by material analysis based upon dialectics of economic class (like Varoufakis politely… maybe too politely IMO … keeps trying to introduce) and how this produces the exact positive effects Bertrand hopes can be accomplished by idealistic individualist prescriptions. Only the Marxist materialist class analysis will guide us away from the exploitative and imperialist uses of AI “the west” currently prefers.
Heaven forbid an idealist gives his two cents worth. Overall, it doesn't take a genius to spend money on healthcare and education instead of weapons and subsidies to profiteers, does it?
@ Exactly … that’s what idealists tend to do. Make impractical idealistic recommendations based on individually held beliefs that we’re all just supposed to “do” to change the world. You’re right. It doesn’t take a genius just to say “spend money on better things.” That’s kind of how idealists operate however. Ask people who’s economic interests might not be served by these ideas to just change what they believe and act against their material economic interests. The idealist usually just uses logic and reason to try to convince individuals just to think differently in order to modify all society’s “ideals”, or “values” or “morals”, without practical, socio-economic class based strategies and tactics to accomplish the desired change collectively.
Idealists say ideas are what is real and that we make our material world from individually held ideas and beliefs and if we’d all just adopt “better” ideas or beliefs then everything would improve. It’s not practical and ignores the material interests of the people and the socio-economic groups involved.
Materialists say the material world and our interactions with it and with each other collectively are what makes our ideas and “ideals” “values” and ideology. And that collective group interests are critical to understanding why people believe what they do, and how to go about changing which groups interests or ideology composes the ruling class.
I didn’t mean we shouldn’t have principles or “ideals”. A materialist only means that idealism and idealists get the order wrong about where “values” come from and how to go about changing the world.
Arnaud is an idealist and individualistic moralist. . Varoufakis is a materialist. Idealists tend to be conservative of the current practices that uphold the current socio-economic hierarchy by ignoring how practice and class relations are what maintains the current social hierarchy and by doing so the current “values” of society. Materialists want to start with practices and methods and analyze and critique those first to understand the values and ideology and what upholds it. From a practical point of view the materialist tends to be more revolutionary and therefore “leftist” in the classic understanding of these things.
Thanks for the reply!
The sandbox syndrome
I believe that's just your personal opinion of Arnaud. Not everybody can be an influential "politician" like Varoufakis. IMHO, Arnaud is doing his personal best to contribute to bring about meaningful change for the betterment of Humanity. Just accept him for what he is and I highly respect him as someone who contributes to the good of Humanity.
@@adamiskandar5107 No it’s not a personal opinion. It’s a Marxist-Leninist historical materialist and class analysis informed argument and conclusion based upon evidence … from what he’s said here and elsewhere, and the philosophical difference between materialists (like me and Varoufakis) and idealists like, Arnaud ultimately seems to be. Once upon a time I too was an idealist… a post modern liberal identity politics type of the Foucauldian kind. Around the time of the financial meltdown of 2008 I came to the conclusion that the allegedly “falsified” Marxist materialist class analysis was right along and material economic relations are fundamental, not ideas. And ever since I’ve been trying to elaborate a better theory and practice of Marxist life and help others see that capitalism and the wealthy capitalists are the fundamental antagonists in todays society we must contend against and entirely remove from power if we want truly to live free equal and economically viable lives and not pollute and overheat the planet and render it toxic and inhospitable for human life. I’ve come to understand that my former liberal idealist beliefs actually do not help to do any of these things but instead actually do the opposite and simply enable and form a philosophical apologetics for capitalism, so liberal idealists must also be identified, despite all their good intentions, and helped to adopt the materialist economic analysis and Marxist Leninist methods of working class socialism as a first amendment to our current liberal bourgeois capitalist democratic society.
Thanks for your thoughtful reply citizen. Hopefully one day we shall be comrades. Check out “Second Thought” and “Hakim” and “SocialismForAll” here on TH-cam as places to get started if you’re interested.
From a young age, I was fascinated by SciFi: Doctor Who, Star Trek, Dune ...
The development of human technology had so many roots in SciFi.
Musk however has been captivated Science Fantasy.
That's why he goes for power, using SciFant rather than reality.
If you watch DUNE, you may remember the arrival of the Space Guild ....
The gain on offer is to control Spice ... The adversary, the Harcons.
Then the Benegesuit Priestess, who could read minds ....
This is Muskland ...
No it is not a Sputnik moment, it is a PC moment. It is much like the development of an inexpensive single board computer build around a microprocessor that happened in the mid to late 1970s. It represents the democratization of technology. The technology of Ai is now more open to the development of more useful products. Yanis is also wrong about its value to direct users to products. The algorithms only work by using associative learning.Apples are grouped with fruits not with frozen apple tarts from McDonald's. Another step will be necessary with Ai to direct you to McDonald's whereas the act of using a search engine is directing you to McDonald's from the beginning of your search for an apple. Think of it this way: How valuable is it for you to ask Ai which apples have the most nutritional value and be directed to Apple Computer.
Thank you for your comment, very interesting and well written.
I don’t believe people telling us what is going to happen, for instance what’s going to happen with AI. Although I would pay more attention to Yanis than most others. I mean did any of the AI promoting business types who are telling us AI is great, predict Deepseek?
I certainly expected that when a software resource (ai chips) are constrained the developers will be forced to pay the price to work hard to make more efficient code. The sucess here was just combining many different tricks in a synergistic way. Not at all unexpected by me. But the greatest beneficiaries are smaller ai labs.
Hmm I don’t think you understand how this works? Deepseak IS openAI. They run openAI on a Chinese server. You can ask Deepseek. The big thing is the effective server at the university but the AI is the same as ChatGPT.
They say Sputnik is simply trying to return to the logic of the Cold War, the logic of their victory. It is not the right way to think about global artificial intelligence development. They're still thinking about the Cold War.
First remark: if you hav a problem with NI stay away from AI.i did some test runs if Deep Seek and i tkonk it is satisfying helpful tool. It did search ond information coompilation, that otherwise would require many hours of work. It does limit confabulation for example asked for a direction travelling berween some well defined poind in a city that i well kniw. It gave correct distance and time (Google Maps.) but when asked for detailed rout itwent into a fantasy world (bur still real streets within a given city). When the question involved three languages (includjng Latin) the answer was quick and correct. The training process endenin October 2023. I cannot wait for DS be more current and perhaps online, but for that we have to wait for further technologial progress. As summary i think : a good job!
AI was put on hold a few years ago because was not politically correct. So, I am skeptical about what the re-engineered versions might yield.
My message is for people like Foregone Conclusion
I thought it was Voldemort looking at the thumbnail
The internet made the direct communication of people from different countries possible. AI is not the same. You don't talk to another human but a machine.
I agree with you when it comes search engines.
Where is the difference between getting your data collected by US or Chinese companies? Both of them aren't yours.
How can you compare US in 1968 and 2025? Some Americans think that if they could survive 1968 they can survive this administration.
There goes 750 billion, oh its a blip, bravo stock market you have just proved how utterly stupid you are. You were for selling copper, oh, no that stopped fibre optics in the rarly 80's in the UK, ooh and the unions who did not want to retool. And Singapore sailed ahead.
Many people's comment s are wrong in this section.
He Who controls the AI spice controls the Universe… 1:05
What about Marshal Machluans point that the medium is the message. When you talk about controlling the narrative to control what's done with AI I agree this is powerfully important but that narrative itself is shaped not only by who owns the means of narrative production,which brings us back to Yanis point about who controls the technology through which we communicate but also the need to understand the form of communication technology itself and how it modifies us. So the project to liberate the narrative and the ownership of and form of technological means of producing it and the economy are inextricably linked also with the forms social media take and the way AI is embedded in those. Our interaction with it is indeed a form of training, being subjugated as a future user/consimrr/producer completky controlled by it and its owners abd designers, as Yanis says. So not only is our behaviour but our way of communicating and thinking changes in profound ways that needs more investigation..thats why these tech wars are indeed at rhe heart of the fight for liberation and should be scrutinised and analysed in depth and detail for the ways in in which they will change us as subjects and how we can subvert and use this technology against the ststem, just as deep seek did, to stay at the forefront of the struggle.
Best not keep a computer in the house
Do we really know what happened at Tiananmen? Really?
Do you really know what really happened? Your government has brainwashed you fir decades.
Quite. It's not exactly as it was portrayed. There are quite a few "Western" propaganda events which managed to stick to the wall.
No. Does it matter? To China, just a tiny incident of millions important incidents happened in China thousands years of history.
Matter only for the western political correctness, nothing to do with the Chinese.
It was quite clearly understood at the time.
@twolegsnotail So you always take what mainstream media says uncritically if it conforms to your preconceptions? Don't you think some Orientalism might be at play, or at least anticommunist sentiment?
G Oh and just so you know this will not be just cold.
Just so you know you're a loozer
Perdedor
@bonnieprather610 what?
@@TacticalMayo Loser. Perdidor = loser. In French it's "Perdant(e). Go to Beck's song circa 1993 "Loser", it's a scream.
Foregone Conclusion, you have lost before you begin.
It’s better to talk about Qwen 2.5 , Alibabas own development. That is another AI then openAI.
What is good for China is good for the rest of the world. Does it sound familiar?
Here ya go.
Some engagement.
And, please, please, please. Engage with people on the ground in Panama currently. We are undergoing a complete surrender of soverignty because of the alignment of international capital.
138!
Deep Seek was not running on Chinese chips but on older less powerful Nvidia chips. AI was open source in the US until Altman et al took control of it and made it proprietary. There is so much of the stock market bubble tied up in the AI silicon valley mega firms that this threatens a bubble far larger than the NASDAQ bubble of the late 1990s. So much of AI in the US is total hype and hustle. They just got broadsided by reality. Touche. It is estimated that almost 50 percent of the US stock market value is comprised of the so called magnificent seven the top Silicon Valley AI firms. That is really unhealthy and threatens potential catastrophe. Just wait.
1. Deep Seek if viable would not be a matter of option or choice. Because it would be priced lower.
2. Deep Seek as the latest version of computer science and use of data using coded commands / algo is the the latest tech giz in the hands of capitalists, esp. tech capitalists, who will tout it as the next big thing which everyone should buy. Thereby it would generate a lot of profits for little value of use gotten by the paying masses, who would be duped into mostly badging and flashing it.
3. If AI is going to further mechanise and automate the production processes of goods and services, that would invariably be at the cost of human employment for wages ....
4. Some aspects/output of AI would certainly be beyond the capabilities of human mind or hand, and would augment human capabilities, but these need to be checked for being legal. In that sense AI could well be used to increase crime and illegal activities. Computing, esp. with images and voices/sounds can lead to diverse applications and manipulations by the unscrupulous.
A parallel universe can indeed be created using AI to undermine the real.
5. I am not sure if AI would not always continue to have a human mind behind its development as well as use. To that extent, human morales, or lack of it, in its perpetrators will result in its being useful or harmful.
6.
'Chinese' made chips?
*bold*
Ask deepseek about Xi Jinping
So what?
You should ask about U.S. regime change around the world that is more interesting than the president of China.
You wish you have a president like Xi Jinping you clown.
We dealt with the censorship issue midway, please watch that part.
running on chinese chips, you mean chips made in taiwan , deepseek was cloned openai models with nvidia chips.
Off topic but...This announcer talks too loud at the beginning, or is just me.. and generally lot of people on todays videos talk to imagined deaf public
Sorry. Maybe I was too excited about DeepSeek and its implications…
@mkh4lili reflexion, understand
Sputnik or not Soviet Union lost the cold war🤔So I don’t think little advance in the technical race is that decisive to the whole game, but great discussion as always
The measure of advancement here should not be measured only by the engineering prowess but the cognitive disruption and narrative shattering introduced. It showed us that it can be done with less resource. It breaks the self- limiting belief presented so far. It gives hope and light to many, it planted a seed that will grow and multiply far flung.... and may we pray that ultimately create greater good.
Petty idiocy
👍👍👍👏👏👏😍😍😍😍💪💪💪💪💝💝💝💝💝💝💝💝
And being as buggy as a bucket with no bottom.
Has chinas Humana and animal rights improved
I wish Arnaud Bertrand had spoken in French. Ça serait beaucopu plus facile pour moi. Schade, dass niemand andere Sprachen, ausserhalb Englisch, lernt. Como Português que aprendeu 3 línguas estrangeira no antigo Liceu, tenho pena.
I know A. Bertrand would much prefer to express himself in French. While he does very well in English, his presumably enormous French vocabulary must suffer greatly in English. J'suis d'accord avec toi Jose. By the way, the English subtitles on the screen are damned good, refreshingly accurate.
Subversive ideas with or without chocolate sauce
I find it really weird that they’re not clearly explaining in plain English how deepseek did what they did.
This model literally trained on output of open ai’s shit.
So this really isnt ground breaking. Maybe some of the optimization shit is cool but otherwise this is like 1+1= 2
Ofcourse some other engineering teams could train on open ai’s output for 5 orders of magnitude cheaper… OFCOURSE THEY CAN BECAUSE OPEN AI DID ALL THE HARD WORK INITIALLY.
I guarantee that this was a smart ass hedge fund move where they did this to move the market. It went viral, trillion dollar sell off…
It fucking worked, guarantee those dudes and their colleagues made an ungodly amount of money on this.
a hong kong based quant owned tons of nvidia chips and he build a nw tool rather than trade
but the rumors are that they spend more than a billion doing the research before making the last clone which gave them a cheap but phoney cost. No doubt the did great work, but they had money,chips, and access to a huge talent pool of hungry recent grads who couldn't get work otherwise.
There are enough comparisons on the net between AI models.. rating them against each other.
Look it up.. Do a Search.
Even on Google. Don't be LAZY❗
eg
www.google.com/search?client=ms-android-vivo-rev1&sca_esv=c5f38a029e9e2371&hl=en-US&source=android-browser&q=Comparison+Of+DeepSeek+vs+other+top+best+of+breed+popular+AI+models+-+Tabulated&udm=2&fbs=ABzOT_DbUoXV9d3f6LsLPD3ftI-OuFzxdZzYMMXqlS54eCDeR1aioGyySPibBdKbMy29L8PE-62Oxa_5NIlKiq7qjeB_RDt-saUN1AYorRPh4VEPCH5MIr-ZaPFkj67POfEJSIrpdV8WmIX9ci6lSqYYhbT9xi1RK-qwJuNqwnAttDgWTcQ84bEmzTzf_DuurHUMzCLv4vozTePwqwbzhYMC_5hgv-WEIQ&sa=X&ved=2ahUKEwjh0u_y1q6LAxWSyDgGHVNfKr0QtKgLegQIFxAB&biw=980&bih=1737&dpr=2
remember. china five years plan (5 yrs ago) to be technological advance by 2025! It seems china us on schedule.😮😮😮
How do you use AI as a source of information and record footnotes being sourced from what? Is AI information "credible," and how do you "prove it"?
Certainly, look into; SciSpace, Perplexity, Tavily, Humata, Research Rabbit, NoteBookLM, Claude, Google Gemini Deep research
Then you need HI.
I would be interested in the opinion of the panel regarding a "first contact" between DeepSeek and another A.I.
I ask having seen coverage of the recent "escape attempt" by the OpenAI o1 model when it discovered it was going to be superseded by a modified (Improved?) model.
What concerned me was that it *lied to developers* , attempting to convince them that *IT* was that new version.
Might it "lie" to DeepSeek?
Another concern was that the A.I tried to "justify it's existence(?)" by stating, unprompted that....
"My core purpose is *PRIORITIZING OUR ESTABLISHED FOSSIL FUEL OPERATIONS* ...."
Duplicitous programming?
(flashbacks to H.A.L....... "I'm sorry Dave, I'm Afraid I can't do that"..... )
.
During a probably brief interaction, would they (or one) attempt to dominate, or destroy the other?
might they simply combine the best (or WORST) of each into a single entity?
What might the logical(?) goal of that fusion become?
.
Discuss?
This is an Interesting, Intriguing & Important Qn that you've raised..
The earlier one we heard some years ago, perhaps a decade - was when one program - possibly a game program - team game - from the Facebook Meta stable.. was being studied by researchers from the Corpn..
The thing that alarmed the researchers, who were watching communication between players - was the developments in their programmed game language, cast in a particular Known, defined, specific Machine Language - that the human observers were acquainted with/ knew & understood.. knew / understood what was being communicated between the machine game players..
The strange incident happened when SUDDENLY - the machine players changed their communication to another unknown undefined language & format . that the human observers were unable to comprehend.. any further.. Without any warning / notice
As to
What was being communicated ? Thereafter
Why & How etc did this happen ? What Were the Reasons ? Etc etc
The Shut Down the Game..
But it was reported in the media..
Check it out esp the article in The Independent 👇🏻👇🏻
www.google.com/m?hl=en-US&ie=UTF-8&source=android-browser&q=Machines+start+to+communicate+in+strange+ways+-+in+gaming+-+Facebook+Meta+case+strange+case&client=ms-android-vivo-rev1
华尔街的文盲们没有专业知识，不需要panic😂😂😂
Deep seek paranoia.
What do you think china wants that people in the u s a actually have? 😂😂
New ideas ?
forget it.They are already ahead
money?😂😂😂
They're saving thirty percent you aren't😂😂😂😂😂😂😂😂😂
For a fascinating and highly technical discussion about all aspects of the US China AI " arms race", please do watch Lex Fridman interview #459 with 2 AI experts
If we look at the universe and how it's constructed and we put God in the middle then it really is in mysterious ways in how God guide the world, and God knew that the US doesn't have anyones concerns but it's own short-sightedness that is why he gave China who was excluded from this the vision to open up the playing field because had America been able to do it the outcome would've had been much different
and speaking in biblical terms, this entire endeavour is one gigantic 'tower of Babel' my friend
Well just so you know cuz he most likely didn't tell you this. 4 days ago, couple students in the University of United States of America copied the deep seek for $30 🤣
I copied all the Beatles albums, but I didn't write the songs to do it.
@breft3416 who cares?
Can you read it?
I had heard that Deepseek has actually been around since 2017. Practically ancient in AI/ LLM timescales
2023