Transcription
America's uh idea of winning is to create a god, and that god will live in San Francisco, and the god works for America.
EMF Morgan Stanley >> Right now, it's just a money bubble, and it's money for nothing. So I think that's going to end all bubbles. For spelled the internet whistling. [Music]
The AI fantasy bubble. Chinese AI was focused on real-world application, like a uh, like in the physical world. So, uh, it's very much uh oriented towards making things. So, uh, any technology people want to see it improves production. So, uh, the first major AI application came in mining, while the US moved, uh, moved in the direction of a larger language model. You ask a question, it seems like it's talking to like a, like a person.
>> Hey, Chubby BT, how are you doing?
>> I'm doing fantastic, thanks for asking. How about you?
>> Pretty good.
>> What's up? So, my friend, uh, Barrett here, he's been having trouble sleeping lately. And, uh, I want you to tell him a bedtime story about robots and love.
>> Oh, a bedtime story about robots and love. I got you covered.
Human mind is a natural attractor to something like that. Even though we don't know if it makes money, actually, nobody makes money. But the, but, but, but the Chinese, we, uh, thought that they were behind and in something. So naturally, they want to catch up. So that's why, uh, uh, uh, like Deep Seek happened.
>> This is an absolute challenge to the supremacy of Nvidia and the high-end chips. If you can produce an AI model at that scale without the high-end chips that has revolutionized, look at this. Nvidia down 11, 12%. This would be the single biggest market cap wipeout ever. Deepseek, the number one app on the app store.
>> Uh, Deep Seek was not on the radar screen. A bunch of people, uh, who didn't go to Ivy League schools or didn't work for famous technology companies, but, uh, they just got together with very little resources, they made something happen. So, um, it's, uh, it shocked the world and shocked the Chinese government. Deepseek is a chatbot. Okay? It does what OpenAI does. But the issue is that, uh, you know, Chinese still want to make it useful in the physical world. So it has been integrated into, uh, uh, uh, electric EVs. So, uh, the driver can communicate with the car through Deepseek, and Deepseek kind of talks like a real person. So it becomes useful in the physical world.
>> It's the most humbling thing I've ever seen. 70% of all EVs in the world, electric vehicles, are made in China. Uh, they have far superior in-vehicle technology. You get in, you don't have to pair your phone automatically. Your whole digital life comes is mirrored in the car. You have an AI companion that you can talk to, a chat GBT equivalent in China. To your all the automatic payment is already there. You can buy movie tickets. It has facial recognition, so it knows who's in which seat and which, which media you like.
>> Wait, why don't Ford cars have that? Would you need to go with Apple or?
>> Yeah, because Google and Apple decided not to go in the car business. We are in a global competition with China, and if we lose this, we do not have a future. Ford. So I think this is the most important angle of AI, how it can help people in the physical world, and I see that China is doing a really pretty good job in that regard, and in many ways, probably China is leading. I'm not denigrating the virtual world, and, and a lot of people spend so much time, uh, in the virtual world, so they, they think this is the real world, right? So, so when they talk about AI, they, they feel like, wow, oh, this AI can make a picture for me. I think a lot of people enjoy that. So, this is fine. This is also worth something. But Chinese is much more down to earth. Down to earth. No. So, so, uh, how do you make a machine, uh, uh, productive today? In the race for AI, who's winning? America or China? If the answer is America, how close is China to us? And what do we do to make sure the answer remains America will win? Mr. Alman, we'll start with you. [Music]
>> Uh, American models, including some models from our company, OpenAI, and Google, and others, are the best models in the world.
>> American language is very much dominated by all these sport competitions like baseball or or football. So, it's always they put a scorecard on, on, on everything. And while the US has a dream that, uh, uh, we're going to have a super, uh, uh, intelligent AI, so America's idea of winning is to create a god, and that god will live in San Francisco. Okay? And the god works for America.
Chinese approach, these open-source approach is not really about the winning, right? How you going to win if everything is shared, right? So, uh, Deepseek is open source, and anybody can download the model to your own computer, much, much cheaper than, uh, what OpenAI, uh, charges because that Deep model is, uh, is 10 times more efficient, uh, than OpenAI's model. So it uses a lot less energy. So it costs a lot less money to run.
The number one factor that will define whether the United States or China wins this race is whose technology is most broadly adopted in the rest of the world. This is a global market, and it will be defined, as technology markets typically are, by network effects. 18% of the people of the world live in China. 4% live in the United States. 78% live somewhere else. So, uh, for, uh, the global South, where, uh, 7 billion people live, uh, I think that the, uh, uh, the US approach is not, uh, kind of accessible. The US business model is too expensive for them. They're investing hundreds of billions of dollars, and they expect the users to pay a lot more so they can make more money. Right? China is comes from low price. Everything China does is from low price. Then China tries to make the product equally as good as in the West. Okay? So, uh, so it becomes irresistible for emerging economies where people's income is limited. So, uh, so this is where I see that, uh, uh, the West, not just in terms of, uh, uh, AI, in many other areas, it's going to live inside a wall, a wall of a high cost. So increasingly, over the next decade, uh, the global South will become integrated, while the global North, we have our own tie walls around. [Music]
>> Could you sketch a difference between how the AI industry in the United States is being financed and how that is done in China?
>> Well, you, in the United States, like money is, is means nothing, right? So, uh, because there's so much money flowing around. So, uh, uh, uh, it's like a company's worth a trillion dollars, you know, uh, uh, uh, open air gets a hundred billion dollars to burn, you know, even, you know, open air doesn't make money, but it's worth like, uh, hundreds of billions of dollars, kind of. So, so, so the US money means something very different in the United States.
>> But how come that there is so much money going around?
>> Oh, that that goes back to, uh, uh, 2008, when the US started this, uh, quantitative easing.
>> Remember that? 200 billion is a starting point. This is, it's going to take a while to get this program up and going, and, um, and, uh, then it can be expanded and increased over time.
>> Uh, the central bank's balance sheet, uh, began with $900 billion. Then, uh, it added $4 trillion. Enjoy your Thanksgiving turkey and, uh, have a good holiday. Thank you.
>> Then, uh, pandemic happened.
>> So they doubled the, uh, the quantity of easing, another $4 trillion, when the people were worried about staying alive, and the stock market went through the roof. The consequences of, uh, the Federal Reserve's policy are far-reaching. They have distorted the US economy. Everybody is playing with money. Everybody is playing with money. Not, not many people are really doing useful work. You know, you, you go to the United States, you ask people, what do you do for a living? It's, it's, not many people can tell you that they're doing something useful. Everybody is a consultant. Everybody is a lawyer, a social activist, and, uh, something now, but everybody's well-paid. Uh, something like this will never happen in China. Okay? Will never happen in China. But the, uh, uh, money comes very, very, very, uh, hard. Everybody has to work very hard for a little bit of money.
The US has become a fantasy land. I think the US is going to face some difficulties, uh, fairly soon. AI is like a dot-com bubble. In 2000, uh, people like a pay. When they hear AI, they pay. Uh, they give lots of money to these guys. Anybody with a PPT can raise a few billion dollars. And, uh, uh, as an AI company, what does the AI company do? They, they, they buy chips. Nvidia says it's a, it's an AI company. It makes chips. It's, there's no AI in there. It gets Taiwanese company TSMC to make the chip for $1, then resells it for $10. That's, that's the business model. Then this profit is priced at 100 times something like that. So the company is worth like $3 trillion. There's no AI in this company. It's just a chip design company.
This one node here replaces that entire supercomputer. We're building AI factories. This is the XAI Colossus factory. This is Stargate. The computing part of it, these systems are 40, 50 billion dollars of it. The reason why people build factories is because, you know, you know the answer. The more you, the more you buy, say it with me, the more you buy, the more you make. You know, when Deep Seek came out, the media share, uh, dropped a lot. It's, it's come back. You see, nothing has changed. Nvidia still exists. People don't need so many powerful chips anymore. Why N, why, why Nvidia share prices back? Because there are people who are raising money. All these, all these so-called entrepreneurs, uh, they have startups, they come up with, come up with a new idea of training a different AI model, buying chips. It depends on how much money that is going to this fund these AI startups. Then these startups' money becomes Nvidia's revenue, you see? So, so Nvidia's revenue is somebody else's investment. It's all based on a dream, and ultimately, all that investment, would that work out? Actually, the people who invest don't care. They think that, oh, all I need is for this shop to be get listed in the stock market, and they're dumb, they're big fools down the road. So I have, I don't care if AI works or not. All I need is for this thing to stay, uh, on this path for another year or so, I'll get out and rich and become so cynical.
>> Yeah. So that's what's going on. Um, but I understand that there's also a lot of Chinese that invest in the United States.
>> Oh, yeah. Oh, Chinese are suckers for a bubble. Even if you tell them that it's a bubble, they go for it. They say, "I know when to get out." You know, I'm smart. I know when to get out. So, uh, so much money, Chinese money has been pumped into the US bubble.
>> You said that the success of the Chinese AI company DeepS could finally burst the US stock bubble.
>> No, the, the key issue is, you know, bubbles, uh, always happen. You look at the economic history, it happens regularly. Uh, the, the foundation for a bubble is a monetary policy. You have, there's too much money going around. So, uh, what, what happens is that you have a shiny light somewhere, the money will go there, and that's where the bubble is. So when everybody's talking about AI, you know, it's, it's a bubble. It doesn't mean it's not useful. It's a bubble. The issue is there, how, when it's going to burst. So you look at the US now, the US has so many problems. The fiscal deficit is like a, for every 100 days, they run up $1 trillion dollars in fiscal deficit. It's just an unimaginable number, right? So we have 60% of the people who live paycheck to paycheck. So they, they basically, uh, uh, and 20% of the people who buy groceries on debt, they, something service called, uh, buy now, pay later, kind of a loan sharking business. So this is like a, you, you look at the economy like that, you say, hey, this thing can, can like break it very soon, right? It doesn't, it's because that all these big wigs on Wall Street, they figured out how to hold it together. Three fund management companies, Black Rock, State Street, and Vanguard, they control over 80% of the market capitalization, okay? So this market is can be manipulated, okay? So, uh, so when, uh, when people get scared, they sell, the market goes down five or 10%. But these three players will look for an opportunity to make it go up again because it's in their interest, and they have lunches together every week, and they fix the market. They figure out how to keep the bubble going, consolidation. So the United States bubble has become extremely resilient because for structural reasons, for structural reasons. So it takes a lot, a lot to knock this thing off.
So will it never pop? There are two things that will pop this. One is foreigners get scared. They don't buy US debt. They don't buy US treasure bonds. Then they bond, the interest rate will go high, very high. So that will pop up, pop the bubble, right? But the foreigners, like, if they panic, where else they go? Uh, Americans kind of like the, hey, you have nowhere else to go. You see? If you don't like me, that's okay, but you don't have anybody else. You see? So, so that's a kind of a, uh, uh, I may be ugly, but I'm the only one guy around. So there, there's nothing you can do. So this is the action outside force. Uh, that's outside force. Internally, it's inflation. US inflation is very serious. Americans, uh, uh, perspective is still, every problem I have, I will throw more money at it. If I don't have the money, I'll print it. So, so it's still the, this way of thinking, you know. Uh, and they, so far, they've got away with it for decades. So why should they change their mind? Why should change the, the way they of doing things? It's free lunch. [Music] [Music]
So what will happen if it bursts?
>> Well, I don't know. It's this time, uh, uh, uh, really, uh, it may trigger a revolution in the United States because of wealth. You see, this, this money, monetary bubble has uplifted the wealth to, to a very small number of people. 10% of the people in the United States account for 50% of consumption. 10%. Jeff Bezos had this, uh, this wedding at the Venice, uh, for, you know, he spent $50 million just on a wedding. You know,
>> It's not a kind of a fairy tale wedding. These are, these both are very old people. So, so why do you need to show off like that? Because they can. You see.
>> This side, this side. [Applause]
Globally, there's too much money flowing around, and people buying properties, go around the world to buy properties. This marginal demand is pushing up prices. So unaffordable housing has become a global thing everywhere. People talking about unaffordable housing. That is ridiculous. That is housing is a mass, mass market product. How could it become unaffordable to most people? Right? So this is a monetary phenomenon. So this means that, uh, when the US bubble bursts, it means that global, the global part is also going to burst. Now, the good news is that the cost of living crisis in rich countries, that should go away. The cost of living should come down, but on the other side, your asset value is going to be worth less. Life gets easier, but you are not as rich as you thought.
So are you looking forward to the bursting of the bubble?
>> I think that we have to go through it. When this thing bursts, like it will be like the 1920s, the Gatsby, you know. [Music] [Applause]
The longer the bubble lasts, the more dangerous it becomes. You look at what happened in, uh, in, uh, in Germany in the 1930s after the bubble burst. Uh, different countries had different reactions, right? Some blamed somebody else and went extreme, right? Uh, there will be some serious political ramifications. But the, the, the, uh, the scary part is that somebody will blame some, uh, blame another country, right? China is the scapegoat. Why we poor now? Chinese stole our money. So when the time comes, uh, it will be a test for everybody.