Transcription
Hello, everyone, good noon. I'm very happy to invite Indigo to join us today for a conversation about AI, which is currently the hottest topic. Indigo is my friend. He is very active in exploring and investing in AI and has achieved great success. He also has a deep understanding of it. So, I invited him to kick off our community's 2026 events. I think this is a very important start because AI has indeed become the biggest variable in the world, and it seems to be just the beginning. So, we will use this as our opening. I can also give a brief preview here: this is our second session. This year, we will have a total of twelve such in-depth live conversations. I think it might be a little different from most live streams. It's actually a creative process. We're not using this live stream to tell stories or anything, but rather, the live stream itself is a space for thinking and collision. Hello, hello, Indigo, hello. Actually, I didn't tell you before, but I don't know if it's feasible. You might have more research on this. If our recordings and conversations are compiled, can they be made into podcasts? Pure audio podcasts. No problem. So, we should have quite high requirements for both of our recordings. Shall I use one microphone? It's okay, it's okay. No, no, I think because I'm also using headphones, you can't be too much better than me. I think your headphones should be fine. Haha, okay. Yes, I've already briefly introduced myself at the beginning. Indigo is my friend, and we live just a few minutes apart in Vancouver. So, Indigo's fans, seeing his beautiful house, it's even more beautiful than what you see live. Yes, yes. Indigo himself is also a founding member of Weibo. And what impressed me most is Indigo's exploration and research in AI in recent years. It's quite interesting, actually. Two years ago, Indigo, do you remember? We met in downtown Vancouver. At that time, AI was already booming, but many people were still hesitant, right? I remember that time, and Zhou Hang was there, right? Zhou Hang. And our classmates in the live stream room probably also know him, the famous investor. At that time, Zhou Hang and I were both very optimistic about the future of AI, but we were also a little uncertain about the development and investment in AI. For example, it feels like this has already been taken over by a few big giants. What opportunities are there for individuals to invest? But Indigo, I think at that time, he seemed more certain, right? He said, "Hey, this..." And I think this is something that left a deep impression on me. Today, we'll start with this. We already have quite a few people in our live stream room, and this is a relatively small-scale live stream for our community members. So, it will be... I have a list of 20 questions here. We'll go through them one by one. So, let's start and get straight to the point. Okay, let me see. I have one here. Okay, no problem. You don't necessarily have to look at it because I might not follow it completely. The general framework is like this. So, we have a reality, and we don't need to say that AI has become the biggest variable in the world. This variable, in my opinion, actually started with the steam engine. This is a steep curve of human transformation, becoming increasingly steeper. I remember Indigo, you also said in your live stream the day before yesterday that humans, the greatest breakthroughs in human destiny come from science. The greatest breakthroughs in human destiny come from science because science is a process of continuous verification. It constantly believes in something and then dares to overturn what it believes without hesitation. Because of this, humanity has achieved a very steep curve through technological progress. This steep curve, in the past many years of human civilization, is actually the first time. This is also a rather magical thing. We are very fortunate that we, whether we are born in the 70s or 80s, are living in this generation that has encountered a stage that is almost steeper than all previous stages combined. That is, at the eve of a great transformation, with everyone's expectation of AGI, or Artificial General Intelligence, or what seems to be happening. But in the meantime, there are all sorts of strange things. On the one hand, people think AI is very powerful, but on the other hand, they think, "What does it have to do with me?" On the one hand, we think AI has become the best teacher, but on the other hand, our children at home are still reciting the Three Character Classic. All sorts of strange things. It is, in my opinion, like the surprising calm before a tsunami. So, today, we will talk freely. And because this is a closed live stream, we can, of course, comply with policies. From imagination to various aspects, I also want to be able to collide Indigo's wisdom and insights in this process. The first thing is, Indigo, you said that what actually determines history is not the superficial bubble, but the deep ocean currents. Can you describe what you think these surging undercurrents in the depths of the real world are? Yes, this theory is not mine. It's by a French scholar, and it's been around for almost eighty or ninety years. So, his era was probably the First and Second World Wars, right? That era. He said that everything that observes the world is determined by geography. Geopolitics, geography, climate, and population. These three, right? Certainly, looking at it now, it's the same. But the impact of technology at that time, because it was the late stage of the Second Industrial Revolution, right? The Second Industrial Revolution had just begun with electricity. What actually determines history is divided into several categories: what are forces, what are trends, and what are individuals, right? The entire historical impact is long-term, medium-term, and short-term. I think looking at it now, technology, since the 20th century, after the Industrial Revolution, has actually become a very, very important force. It's no longer a trend; it has transformed from a trend into a force, into an ocean current. Yes, it's at the bottom of the ocean. So, every time our social structure changes, our productivity changes, our social structure changes, it's all driven by technological change. Since the Industrial Revolution, it has been like this. The GDP of human society has been rising like this because of technology. And we have experienced, if you've heard me share before, the Kondratiev wave cycle. We are currently in the sixth cycle. The fifth cycle is the information age, from the 1970s to 2020, right? This information age, about fifty or sixty years, personal computers, then the internet, mobile internet, and then the mobile internet ended, followed by artificial intelligence. Yes, it's divided like this. The Kondratiev cycle is divided like this. So, this is the underlying operation of the current ocean current. Because this change will cause changes in the upper structures, economic structures, political structures, all structures, educational structures, and growth structures. Because some things won't change. For example, a US presidential election is only four years. It won't change your educational structure or social structure. It's impossible, right? But it only happens after long-term accumulation. So, the significance of this technological revolution is much greater than that of the mobile internet, at least. I'm in the mobile internet business. Right? I've worked in both the internet and mobile internet. I used to be at Sina. Both times, at most, the mobile internet was just a revolution in information transmission and reception, right? It was just information. It was just information transmission. I could easily access information. I could access information 24 hours a day. After accessing information, you could watch videos, order takeout, call for a ride, right? All of this could be done on a mobile phone. This has indeed brought about great changes. But I think the mobile internet has not brought about significant changes in our productivity, right? The change in productivity is still less than the greater changes that should come from the PC era, personal computers, Microsoft operating systems, cloud computing, and such productivity changes. But this time, the AI revolution is directly starting from productivity. Right? You mentioned earlier that... You said that some people are deeply impressed by AI, while others are not. It's jagged, like a gear. Some parts are very powerful, capable of winning gold medals in math Olympiads, right? And in other parts, it can't even count the letters in a word. It's... it's stupid. It's like it can't even count the letters in a word. So, it leads to many people saying, "Hey, in daily use, the answers to my questions are not very good." Right? It lacks emotional intelligence. Right? Sometimes it feels bad. But many professionals, such as programmers, are the most affected. Then, besides programmers, customer service, online customer service, and many automated marketing companies, sales, content creation, automated content creation. These things are being used very quickly. And when I shared, this year it should be in scientific research, the acceleration of scientific research, the science and research part. So, I think everyone's experience might be different, depending on your job. If most of your job, right, doesn't require much support from third-party intelligence, then you might not feel the benefits of AI. You might think it's just another chatbot that chats with you. Yes, there will be such a change. Actually, yes. In my opinion, when we talk about technological progress, we used to say steam engine, then electricity, then computers, then the internet, and then artificial intelligence. People use this kind of narrative. But this narrative can easily lead people to underestimate AI. In my opinion, AI will reverse and re-invent steam engines, electricity, machinery, and even agricultural civilization, even the origin of life. It will overturn everything and re-invent these things. So, it's not a parallel relationship, but we can't put it on a single curve. Hey, and we've reached this point. I think the disruptive nature is particularly strong. But I feel that Indigo, you are actually an optimist, especially a technological optimist. That's undeniable, right? So, there's something very interesting happening now. I think last year, some things left a deep impression on me, and the views of some very powerful people have undergone earth-shattering changes. For example, one of the most typical is Musk. Elon Musk, we know he, he was quite busy in the first half of last year, right? He was busy with the US Department of Efficiency, right? He thought the US economy was not doing well, the deficit was too high, and he had to step forward, lay off people, and fix Twitter. But he actually underwent a very, very big transformation. This transformation actually came from his rival and friend, Peter Thiel, who instilled a concept in him. Peter Thiel. Peter Thiel has a concept: "Old Ma, don't worry about these things at all. You can't finish worrying about them. Anyway, stupid people can't be saved." These are all my additions. No one. Essentially, AI can solve all these problems. The 10x, 100x, or 10^n power release brought by AI will make all your worries disappear. So, actually, after that, Musk underwent a big change. One is that he returned to Tesla and continued to work, not sleeping on the floor in the White House. And another is that he has actually bet heavily on AI, including his own Grok AI. It has now become the second most used by me. The first is Gemini, and the second is Elon Musk's AI. It's getting better and better, and the progress is visibly huge. So, I want to ask you a question: do you believe that AI can cure all diseases? That is, can it solve everything? Even the US's complex, huge, and seemingly unsolvable fiscal problem can be solved. What about you? To what extent will you be optimistic? For example, as optimistic as Cathie Wood, as optimistic as Peter Thiel, as optimistic as Elon Musk, or do you have your own position? I think I'll be as optimistic as Elon Musk. And there's a quote. I paid attention to a company before, DeepMind, right? Later acquired by Google. Right? Hey, when DeepMind was founded, people watched the documentary that came out at the end of last year. DeepMind was actually invested in by Peter Thiel when it was founded, right? Peter Thiel invested in it. And Peter Thiel invited Elon Musk to invest, but Elon Musk didn't seem interested, or something like that. Then he recommended it to Larry Page, right? Right? And then Demis said, "This person only cares about Mars, not our thing." Right? Yes. It was a joke at the time. Okay. Actually, when the company was founded, it had a vision: to solve the problem of intelligence first, and then all human problems. Right? This is its vision. It wants to achieve artificial general intelligence. First, solve this problem. So, now, coming back to your point, the quote you mentioned is very good. The current AI revolution is surpassing any previous technological revolution, right? We are in such an era. At least, I believe so. Because it improves from all dimensions. It might be, at least from the perspective of AI pioneers, comparable to the electrical revolution. But it might even surpass electricity. Electricity essentially provides us with energy and power, while AI provides us with intelligence. These two cannot be compared. Of course, for AI to work well, it also relies on electricity. Now, these two are closely linked. So, I am definitely an optimist. I believe that we humans will always have problems that cannot be solved. This is what David Deutsch, the British theoretical physicist, said. Problems are inevitable, but problems can be solved. This is optimism, right? We will always have problems that cannot be solved, and we will always be able to solve them. Otherwise, society would be static. Static society cannot progress, right? Right? This is my attitude. I think Edmund Maser is also like this. And you mentioned it very well. After being tossed around, he also reflected. He reflected in his podcast, in the All-In podcast. He said those people are so stupid that they are beyond saving. I don't need to do this. From now on, I'll just... Haha, he said it himself, right? He said it directly. And then he said, I think he feels that the US's sky-high debt and various social problems can only be solved by developing AI. Okay. Yes, yes. Sometimes I also have these thoughts, like, "Either we just hand it all over to AI." Haha, yes. Of course, I myself would think this way. I think regarding our first topic, Indigo's "deep ocean currents," we can almost draw a conclusion: the world has been completely changed. I think this is a topic that no longer needs to be debated. That's the first point. Second, I think we can also draw a conclusion. Even now, because there has always been a lingering question: how far can we go with the "big model, big miracle" approach? Will the curve flatten? Will the marginal effect become weaker? I even remember that around 2023, Demis Hassabis of DeepMind actually expressed doubts about this. He said this thing is quite useful, but it's not elegant enough. I remember he said that very clearly. But in fact, a few months ago, his view has completely changed. Not only is it feasible, but he even said that this thing doesn't seem to be showing signs of decay, right? It's still useful. Right? It seems that continuing to expand the scale and computing power is still useful. So, this is a very, very... So, I think for everyone, first of all, regardless of whether we... It's like we're making a decision, we're understanding the world. We should treat this event as a certainty. Don't doubt anymore, saying those people are a bubble, and after the bubble bursts, AI will be a mess, and we'll go back to the old world. That won't happen. Even at the current level of AI, I think it's enough to increase the world's productivity to a very, very deep and very high level. I feel it very deeply myself. Last year, we did a small-scale test. We did an online "Life Compounding Action Camp." It's actually a very labor-intensive task because we did a lot of interactive things and some one-on-one, very in-depth assessments. I spent about two weekends using complex prompts and a bit of Vibe Coding to complete this work. At that time, I was deeply touched. Why? Because the things I did, I'm an outsider. The things I did, if it were in the past, finding a technical team of twenty or thirty people might take half a year to a year to complete, and it would only turn out like this. From the results, I think it's still far from ideal. So, this is really a very, very big touch. Of course, regarding the first topic, there's a very specific question: what should ordinary people do? It's too fast, and it feels like the big companies are taking all the benefits, right? Indeed, like Indigo, you've done some early-stage investments, but these are too difficult. For ordinary people, from an investment perspective, for example, I would tell my classmates or others, "You must at least get some of these leading companies. You have to be their shareholders." Not in the sense of stock trading, but you have to be their shareholders. Otherwise, this is actually a very risky thing. What do you think? Yes, you just asked about ordinary people. Actually, I think there's a point I have that's probably the same as yours. I think AI is very powerful now, but AI's main impact is on social structure. Its penetration is very slow. A technology might develop very quickly in a research lab or a company. Then, in society, only a small number of people, those who are particularly curious and eager to learn, right? These people will probably chase AI first, and then it will slowly penetrate. There's something called new product penetration rate. There's always a group of trendsetters at the forefront, then it expands to a wider audience, and then it declines. This time is a physical effect. This is a problem of the social structure itself. We have the internet now, and things have sped up, right? But there's still a social structure. You have billions of people, and penetration is not a matter of one or two years, but perhaps five, ten, fifteen, twenty years. But I think if we pause AI research right now, the current model's capabilities are enough for us to digest for five years. I agree. I strongly agree with this. And we also have very thoughtful classmates in our live stream room who say that there are still many problems that AI cannot solve. This reminds me of a story I particularly like. Hinton was in a very difficult period in his early years when deep neural networks were in a very bad state. You should also know that he had a hard time finding a job. I remember he eventually went to Canada, right? He was taken in by Canada. Canada, such a foolish country, took him in. At first, because the mainstream academic community believed that deep learning was not feasible, they even gave it a new name, "deep learning," because neural networks sounded like a scam. So, I remember one thing very clearly: Hinton went to a university to look for a job, roughly like this. I've adapted it slightly. Then, he found the head of the department. He said he only needed half the salary. The head said, "Hey, you're okay." And Hinton was also honest. He told him, "I have to admit one thing: I don't understand computers very well." This is very funny. You're doing AI research, and you don't understand computers? Then the head said, "It's okay, we have plenty of computer experts here." After a while, Hinton said, "I also have to admit one thing: my math is not very good either." This is absurd. You're doing AI research, you don't understand computers, and your math isn't great either. But we also know that Hinton's math problems were often solved by his undergraduates and later his doctoral students. But he had very good thinking. He was actually a deep thinker. So, it was very absurd at the time. The department head eventually decided to hire him. Why? Because he thought this was a direction with no prospects at all, and this guy seemed very unreliable. But this gentleman was also a very wise person. He said a thought popped into his mind: "What if this crazy person is right?" This is a very interesting thought. For a front-line explorer, even a small spark, if it doesn't catch fire, it's nothing. But if it catches fire, it's completely disruptive. So, he made a very smart decision. He hired Hinton. And so, I think at this time, for individuals, while it's okay to be skeptical and doubt, when making personal decisions, financial decisions, and career decisions, we must assume, as Indigo just said, that even if all research and development companies stop now, the world might need five years to digest it. Even after five years, this arrow might still hit your chest. So, I think this is the first topic. Let's move on to the second topic. The second topic is also a very interesting statistic from the past two years. I've also paid close attention to this statistic because I used to be in real estate. And I also saw this chart. It's... to put it more directly, the development of human real estate, in... should be in the US, right? Office buildings, buildings for humans, construction has fallen behind buildings for robots. Or, to put it this way, the construction of office buildings for robots is greater than that for carbon-based life forms. Right? We're putting them side by side now, which is simpler. There are two types of beings in the world now: humans, as carbon-based life forms, and silicon-based beings. And everyone needs a place to work, a place to create and produce. For humans, factories, office buildings are built, and at night they go back to their apartments, to their houses. They are all storage. This real estate is essentially about storage. But in fact, this turning point has arrived, right? Investment in data centers has already surpassed that in office buildings. So, this is a very important change. For example, our understanding of assets might need to be re-evaluated. Indigo, can you share with us the shift in asset logic? Yes, I've shared this chart before. There's also an intersection point on this chart. In 2018, the construction of warehouses and logistics warehouses surpassed that of shopping malls. Right? Yes, this is also very interesting. This is a very interesting phenomenon. In China, everyone can feel this, right? People always say that the offline physical economy is not doing well, and then everyone goes into e-commerce. It's the same in the US. It also happened in 2018, right? This mobile internet, Amazon in the US, maybe not in Japan or Hong Kong, because Japan is a very densely populated country with a very developed offline sector. But in the US and China, it's very similar, right? E-commerce is actually quite good. This is a trend. This is a trend. In these two places, on two continents, in two countries, it's a trend. This happened first, and then the pandemic directly pushed it. The pandemic directly promoted more developed logistics, right? And then COVID-19, it added fuel to the fire. And then AI came along and completely rewrote this. And some people discovered, "Okay, the work that people used to do in those cubicles in office buildings, now I can just plug in GPUs." GPUs can do the work of people, right? So, I don't need cubicles anymore, right? The cloud automatically does this work. It becomes an agent, right? Like Anthropic, this company, they say they want to replace all white-collar jobs. So, they replace all shoes, right? The novel. Right? Today, Elon Musk is very, very funny. I just forwarded a podcast. An employee came out and leaked something. He talked about Microsoft, not Microhard, but Microsoft. Half a year ago, they mocked Microsoft, saying that the AI that Microsoft is doing can be done by many people. Microsoft doesn't need to exist. Then he took a new project name, "Microhard," meaning "super hard," and put it on the roof of the data center building. This is a real project. He wants to use numbers to replace human work. Right? Then someone came out and leaked it. And this person was just fired. Right? Haha. It's a funny story. So, I mentioned these two changes. This is called, let's look ahead. You started by talking about what is an ocean current. This is at least above the ocean current, a flow, right? A kind of change. So, people's understanding of assets will change. Often, people from different eras have different perceptions of assets. Young people now, I think people born after 2000 or 2010, they don't have much recognition of real estate. So, in our surroundings, in Vancouver, where you lived, the young people in the digital generation, they might feel better about crypto. Right? Yes, they think these things belong more to their era. DJ, right? They call themselves the "lost generation." You say, your previous generation, the baby boomers, you bought houses so cheaply, $500 or $10,000 to buy a house. Now it's $5 million, right? I can't afford it at all. You say a large amount of assets are hoarded by the old world. People in the new world need something that represents their assets. Exactly. So, they all move towards new digital assets and other places. This is also a kind of dissatisfaction with the old world. For young people, dissatisfaction with the old world. Why did Occupy Wall Street happen in the US before? Right? Right? And this is one direction. So, I think people's recognition of assets is changing. It's changing. And regarding real estate, personally, I think, personally, my personal opinion, land on this planet is definitely scarce. And usable land, and land in core urban areas, is scarce. It's like gold or something. It's scarce. This scarce land has value. It's a scarce asset in the physical world. Just like the digital world, you see data centers now, they won't be built in New York City, right? They'll be built in places like Tennessee. Right? They'll look for land or places closer to people to build. So, this is changing. A lot of what we used to consider apartments, or some properties, office buildings, were very valuable. You invest in them, you invest in shopping malls. I think this era has passed. Hey, speaking of this, I can make a small conclusion for this topic. Because I used to be in real estate, and I have a definition of real estate, especially residential real estate. I think most people misunderstand real estate. For example, what is residential? What is residential? What is it for? Why is residential expensive? I once said a point, and I think it reinforces this point even more. I believe that, so far, the best definition of residential is a bit self-congratulatory, but because it's so correct, I have to say it. Right? So, for example, let's start with what you just said. The reason why the entire asset logic has shifted from its original meaning of satisfying carbon-based life forms to, for example, data centers and computing power services, is essentially following productivity. As Indigo said at the beginning, for example, the mobile internet, what significant contribution has it made to real productivity? There must be some, but not earth-shattering. For example, it was predatory at the beginning. Your newspaper ads became my Weibo ads, right? It was a bit predatory. But AI is creative. It has truly become a productivity tool. So, you'll find that money follows the productivity tool. So, since productivity... Money will definitely follow where the productivity tool is strongest. So, what is residential? I have a definition for it, but this definition is a bit utilitarian, but I don't mean it that way. It's just like this: residential is essentially a warehouse for storing productivity tools. Right? For example, Indigo, you are a productivity tool. Right? Because you just built a house, and you spent a lot of money to build this house, it's because it's worth it, because it needs to store you, right? It must be proportional to your productivity. So, okay, we can have something interesting. Does residential have no future? Or how do we judge its value? There's a very interesting way to think about it, very simplified. Using the sentence I just said: residential is a warehouse for storing productivity tools. So, similarly, for example, people like Indigo, where are they? Right? For example, in Silicon Valley or somewhere, we see that their real estate is relatively firm. Okay. This is the second topic. Let's move on to the third topic. Let's talk about something specific: investment in the AI era. Let's talk broadly about asset allocation for ordinary families. Last year, 2025, was a crazy year in all aspects. For example, the tech giants continued to rise, and this rise resulted in a situation where they accounted for almost half of the market's gains. But in fact, this used to be called the 80/20 rule. 20% of companies got 80% of the profits. But now it has become 0.2% of companies taking 98% of the profits and growth. It has become very, very unbalanced. On the other hand, we also know that the best-performing asset in 2025 was gold. For example, the best-performing gold mines in the US stock market rose by 150%. How much did they rise? Right? And I checked the spot gold price today, it's already approaching $4700 per ounce. It's still very, very crazy. So, at this time, and the A-share market is also in a bull market. So, I think for most people, there's still some anxiety. If they make money, they feel they didn't make enough. If they don't make money, they feel like they missed something. So, from an allocation perspective, since not everyone has the ability to pick stocks or invest in companies, from an asset allocation perspective, can you give some advice to ordinary people and families? Yes, I had a view before. I still hold this view, even in China. I might have missed out on China's real estate. I always thought houses are for living, not for investment. I still hold this view. I don't invest much in real estate. I might buy apartments. Buy one in each city, that's enough. Right? I'll buy land and build it myself. I think you're right. Using it yourself is the most important thing. So, it has utility value, right? Real estate. You have to use it. And you live comfortably. Because this thing, for self-occupation, so it also has asset preservation value. This is how I see it. After I build the house, I'll mortgage it to the bank and borrow money to invest. It's the same, right? I think this is better. Since you've created an appreciating asset, mortgage it. I think most people might not do this. Many people are eager to pay off their loans, but there's no need to pay off the loans, right? After you build the house, just let it... This is a good leverage effect. This is wealth, right? So, even though you pay a little interest every month, you can earn more than double the interest in other markets, like gold, crypto, or stocks, right? So, you make money. This is a strategy. This is a practical strategy for many people. For those who own houses, right? And you refinance it, you can do this. At least in North America, it's easy to do. We... Yes. Second, personally, in the US, because the household ownership of stocks is very high, it should be 70% of households invest in US stocks. Through ETFs, a large amount is bought through ETFs. So, I personally think that if you want to keep up with technology, because every revolution... The stock market is always the vanguard. Money is the smartest. It looks ahead and behind. Because it's too smart, it creates bubbles, right? So, they fantasize. So, you need to be ahead of time. So, at this time, you can't just watch when everyone else is investing. It's impossible. So, for ETFs, look at good assets. Of course, gold. What is gold? I also look at it. I'm okay. Maybe I'm not very fond of it. At least, I and Lao Yu are opposed. Lao Yu is more physical, doing real estate. And then, but I think from the perspective of current international politics and geopolitics, gold is also very good. Right? Because once the world is volatile, gold prices will rise. It's always like this, right? So, this is a resistance to uncertainty in the world, right? This is how gold is. It's a good asset for resisting uncertainty. But it might be slow. Look at last year, it was particularly good, it rose a lot. But it might not be like this every year, right? And it might stagnate for ten years, or even ten years. Right? Ten years. This is possible. And another thing, I think, so personally, I think good assets are the assets of companies that best represent this era, or I have a metaphor. In June 2024, Nvidia's stock surpassed Apple for the first time to become the global number one. Look at the stock prices of US stocks over the past 30 years. Which company ranked first? It will continue for a period, and there might be some ups and downs, but it will last for about ten years. When this company becomes number one, it represents that the industry will be dominant for the next ten years. Ten years later, Apple in 2010, or 20... 2010, two or three years after the iPhone, it became the world's number one, at least for a while. And from 2010 to 2020, or 2024, for more than ten years, mobile internet was the world, right? Apple represented this. Before that, it was Microsoft. Then Microsoft rose again because of cloud computing. Right? Cloud. Microsoft was represented by Windows operating system and Office. So, this is a very important observation. So, Nvidia has risen, and it's very likely that Google will also rise in 2026. Maybe the top position will be Nvidia and Google. So, these two companies are the best AI companies in the world. One makes hardware, the other is developing models. Right? Right? So, it's very clear where you should put your money. For the next 10 years, right? Right? At least it can maintain, at least it won't double every year, it won't grow like this. But when you seek stability, its annual rate of return, at least it keeps up with the era, right? Its company profits. This is my logic. And third, if you feel secure, you have many options. Bonds are a supplement. Many people buy bonds, like putting money in the bank. And another method is also investment. You can also invest in ETFs that focus on specific vertical areas, which also has opportunities, right? Generally, if you think someone is popular, or you think something can be speculated on, what is it in this era? You can choose ETFs listed in the US and around the world. So, for many people, you don't have to invest in a single company. This seems simpler. You are in the AI concept. What is being speculated on quickly? It's popularity, it's metals, right? Now choose metals, choose semiconductors. You can buy semiconductor ETFs. These are all options. Right? Then, let's see if you're asking about China. I think Chinese AI companies, leading Chinese companies. I know Alibaba's models are very well done. I'm just talking about the top companies. If we refer to the US's "Magnificent Seven," then China also has several giants, right? ByteDance hasn't listed yet, but it should be soon. It will be spun off and listed. Tencent and Alibaba, right? These are the ones at the forefront. So, I think companies like Meituan are different. I'm more optimistic about China's Tencent and Alibaba. They are two digital companies, digitally native. Although Alibaba also does e-commerce, Alibaba at least has an absolute monopolistic position in AI and cloud computing. These two. Tencent is purely a digital investor. Tencent has no physical business. And I have another view. I think someone asked me earlier. I'll answer the question from a classmate earlier. Regarding investment recommendations, this is how I see it. When AI liberates productivity, your working hours might decrease, right? Because AI automates work. So, where is the meaning of life? I think the meaning of life is to pursue your previous meaning. It's my work, my existence, right? My work, I have a profession. You use your profession as a label. I'm doing this, I'm doing that. But I think AI will find it difficult to replace manual laborers in the short term. This is still a good direction. So, I saw a research report from AN Dominion yesterday, the Fourth Economic Report. It says that at least its models have replaced high-end intellectual labor. This is what we call de-professionalization. It helps humans de-professionalize. For example, in many industries, such as travel agencies, or what they call education, things that require a high level of education are replaced by AI. And the remaining people, like tour guides, receptionists, ticket sellers, manual laborers. People who tell stories in classrooms are more popular. They don't need much education, maybe 12 years of education is enough. But it replaces people who have 16-18 years of education. This is called de-professionalization. So, our professions will change like this very quickly. The very top-level people are still there. But the top part of the middle class cannot be replaced because AI is still jagged, it's not perfect. The top experts, the experts among experts, cannot be replaced. You are a human genius, the best computer scientist. But all the intermediate knowledge work is replaced. So, in the future, there will only be the top intellectual laborers and ordinary humans who provide various kinds of manual support, emotional value, and communication value. When people rely more on each other, if you go somewhere, "Oh, this school has a living person teaching me," I'll definitely like this school. Right? If it's all AI, it looks like... People are still people. Do they have empathy? Empathy is a connection with peers. You can't have an emotional connection with a machine, right? The brain doesn't have this circuit, unless we modify our brains. I'm not sure about this. Maybe in 50 years, we'll modify it. Possibly. At least not within 20 years. Haha, yes. So, I think this is a realistic issue. We've answered those previous questions. Let's summarize briefly. In plain terms, from the perspective of asset allocation and investment, first of all, in the past one or two hundred years of history, what has truly created value are companies. And companies, according to Indigo's approach, we must follow productivity. The companies with the highest productivity efficiency. Right? For example, AI, as mentioned earlier, Chinese companies like Tencent and Alibaba. Of course, this does not constitute a stock purchase recommendation. But from an individual perspective, we can allocate through ETF configurations to some equity-oriented ETFs. Of course, we can. And I think perhaps one is to follow productivity, and another is moderate diversification, which might be better. If possible, you can have some S&P 500, you can have some Nasdaq index. You can go further here, semiconductor index, or specific leading companies. This is one. Another is, whether it's gold or cryptocurrency, I think the most important point from an asset allocation perspective is the proportion. It will be important. You must control it within a certain proportion. Generally, a more stable approach is to control it between 5% and 15%. As for what you consider it to be, that depends on individual beliefs and judgments. And then, as mentioned, houses. Houses, I think, I think I agree with Indigo on this point. Houses are essentially a service for productivity tools. We shouldn't treat them as a way to make money from mortgages. If it makes you comfortable, makes you happy, and makes you happier at work, that value is greater than anything else. Right? So, do what you can to make yourself comfortable, but don't overdo it, because it's not worth it. Right? So, it's the opposite. So, let's move on to the next topic. The next interesting topic is about people. I think this is also quite interesting. Indigo, actually, when I communicate with you, I don't feel like you are... Actually, you are quite technically oriented. But I think your social skills are good. Right? I think your social skills are quite good. And I think you are good at judging people or projects or these things. You've also invested in Li Jiaqi's operational formula, right? So, let's talk about this from a human perspective. You predict that in 2026, business models will shift from operating channels to owning AI studios. So, from your perspective, let's talk broadly about what this means for ordinary people, from an individual perspective, whether it's about current studios, the artist economy, or what I can do. What does it mean? What does it value? What are the capabilities, opportunities, and challenges? I'll share two interesting stories. I saw a friend asking me. I forgot to mention this point earlier. Let's talk about Lao Yu's question first. He's asking about the entertainment industry. So, I think AI is... This is also... I saw Elon Musk share this in a recent podcast. He said, we've done such high-level intellectual work with AI. What do humans do? Humans provide emotional value. Because humans are social animals. As you just said, we are doing this. You said earlier, we did this. When we first started the company, Meituan, actually, live streaming was just emerging. KOLs existed, right? There was a company in 2015. Right? And then slowly, there was someone doing live streaming here. China's live streaming developed very rapidly. There was no such phenomenon abroad. Now, YouTube also has YouTube influencers, but it's not the same. Each network has different characteristics. But I think these are very representative. So, with the emergence of social network platforms, the social connections between people have become faster. So, at this time, many nodes will emerge. And these nodes provide the value of human connection. For example, you've built a community now, right? You've built your public account. You're actually providing this service to many people, right? But in the past, you might have just written every day. Now, with more automated AI skills helping you, you can do many things. As mentioned earlier, many codes can also help you. This will make you stronger, right? This is a very important point. I think the AI era is about enhancing these network nodes. If you are a network node, and I am a network node, we both have them. Your community is larger than mine. My community is very small. And then, we are both serving, we are both providing this "connection" value to our communities. That is, they are consuming the content we share, or we are doing these things for them. Actually, can a pure AI do this? A pure AI can do it, but the personality of people, their taste, AI cannot replace this. So, in this regard, AI can amplify this. Why do I say... I will transform from a fan economy into a company. When your community matures, AI agents can help you build many things in this community. You couldn't do it alone before. You bought various network services. Now, you can use AI to help you generate videos, provide trading, provide various agents to help you do many things, help you organize the people in your community. It's like a hybrid of an agent network and a human network. This thing, this community, looks like an organization, like an organization. Right? You can have everyone produce content, everyone can sell things here. This is what I remember Balaji, the former CTO of Coinbase, wrote in a book called "Balaji's Prophecy." Balaji's Prophecy predicts that future communities and media will be like this, right? But he's from the crypto side. He says everyone will have their own tokens, and the community will have this kind of belief. This is what was speculated about as DAOs a few years ago. I think in five or ten years, when AI becomes more advanced, it will come true. Because they are self-organizing. Right? Right? This is an automated operating corporate entity. And communities will evolve like this. Because as a creator and organizer of a community, you can direct more AI skills to help you. All community members who join can be served by AI skills. Everyone can connect and create here. It's like a company. And now there's a characteristic. Looking at new startups in Silicon Valley, many new startups are themselves communities. For example, Midjourney, have you used it to draw pictures? Right? Midjourney has a very strong sense of community. It was originally a company that existed within Discord. It only used Discord. It didn't even have an interface. Nothing. It only had a Discord channel. And people drove the AI skills within it to help them draw pictures. Then, they divided into many creator channels and created things together using its model. It naturally formed a community. It's a model developed by the community, a company developed by the community. Right? This is at least common in Silicon Valley now. Many companies, including early-stage companies, are like this. A small group of people, working together. And they watch this company grow from zero. So, you're creating content here, and we're both creating content. Right? Some people are creating software here, and some are creating podcasts here. They are all products. Remember the philosopher from Silicon Valley? Yuval Noah Harari, right? What is the biggest leverage in life? Code and writing. Code and writing. Reasonable and reasonable creation of articles. They are the two biggest leverages in life. Because this thing, code, and what you write, or courses, can earn you money while you sleep. These are the two biggest leverages. Community is what makes this thing operate continuously and spread. It's a network. So, I think it will form companies in the future. And now software companies are themselves communities. Right? This is my answer to your first question. Second, regarding leisure time, which someone asked earlier, I'll connect it to the previous point. Leisure time, and you want to see... I think the biggest use of AI generation technology in our generation, like video generation, image generation, is in games. When you put on VR glasses, I've played games like Meta's Quest and Horizon Worlds. It's tiring to build 3D worlds. My two friends got myopia from it. They did it for a month. And this virtual world is suitable for AI to generate. Humans only need to think. There's a director who asks me to plan here, and AI keeps generating worlds for me. We enter and play together. So, previously, creators asked how creators generate income. You say, if I can create a world, now I can generate income. That's Roblox, right? So, every child can create their own world. They sell tickets, they sell things. If humans can create such interactive worlds, or if my community directly evolves into a 3D world, and when I put on VR glasses or virtual glasses, I can enter it. In the future, people listening to us live won't see it like this. We'll open a virtual world and sit in front of me to watch. This thing has already been realized in Meta's products, but the performance is too poor. It's called Venue. It can have 200 people online, and everyone can enter and watch. You just need to wear a helmet. But this thing hasn't become popular because the effect is too poor. So, in about five years, with the performance, Apple's Apple Watch Pro 3, 4, and a few more generations of development, and with AI enhancing 3D world generation, these won't be problems. I can freely create parallel worlds and create games in them. And consumption in the virtual world can also generate GDP. I can listen to you talk in there, or sell you what I create. It's the same world model. This indeed, the world model can promote the development of the virtual world. Right? Isn't it that humans reduce movement in the physical world and create more GDP in the virtual world? It's the same. So, this... I think games will be very prosperous in a few years, in 5 years. The technology is not good enough now. Wait another 5 years. Because...
Elon Musk 公司他其实招了 big boss，他其实藏了很多人才。他 XAI 里面，他把暴雪最牛逼的游戏策划招过去了，然后以前做那个底特律变人的那个，那个法国的那个双生那公司，游戏策划招过去了。我不知道 XAI 他可能在这个 world model 里面要干这个东西，他应该是在今年底会有一些产品出来的。因为老板戴上 VR 眼镜吧，最后我们迟早都会这个 side over。
对，哈哈，对对。首先他自己也是，首先他自己也是一个重度用户。如果说到这儿其实挺好的，就说 OK，我觉得特别好，就是 Indigo 就讲的话题都特别特别具体，他没有很宽泛的讲。他其实顺着这个从个体来说，他讲了一个，顺着这个社群的这个线索，但讲的讲的非常非常深度。其实我觉得我很喜欢这种就某一点钻的比较深。
那我们就不必拘泥于这个话题，我们稍微往外跳一下。假如是从 OK，就是从游戏的角度来说，那那有几个悬念。就一个悬念呢，就是我们知道这个 Meta，Meta 现在这个他的他的眼镜相对而言做的是最好的，因为算是出了个爆款，是吧？就是算是。虽然，但但我觉得他还没有到那种就完全被验证的，就是它的一个量级还没有到，就是智能手机的这个层面，还隔得甚至远，甚至可能连 Apple Watch 的这个量级都还远没有到。但它似乎对，但它似乎开了个头。那你会怎么看这个 Meta 的潜力？就 Meta 包括他的他的眼镜，还有 Meta 这家公司自身的潜力？
对，这个要聊公司了。我觉得 Meta 现在至少在 AI 上面伤了元气，真的伤了元气。因为这个之前在内部组的，这个他两个团，他有个研究团队，这个以前那个雷坤负责的是吧？闫雷坤。然后呢，还有一个产品团队，拉玛拉玛那个产品团队是真的不行。对，后来就被裁掉了。对，所以说他大模型上直接掉队了。然后这个呢，他把之前很多公司的精锐都用在那个，以前的那个叫做 Meta，那个那个 Reality Lab 里面去了。那些做那个，所以说你看 Meta 的在那个研究上面来看，他在现实世界解分解，3D 生成，那什么东西做的都挺好的。他的那个那个 Segment Word 就是把画面直接识别出来，3D 物体读出来，他这个确实做的非常好。他就是语言模型很落后，就之前公司重心不在这。对，他大量的研究经历，成员都在这个，都在一个叫做叫做 Metaverse 这个上面去了。但是现在呢，因为突然一下语言模型杀出来了，然后他掉头没，就是及时的，没有没控制好这个平衡。所以这个也是这个，这个怎么说呢？公司是有钱，对，公司社交网络非常赚钱，还有钱烧，有钱玩，有钱玩。然后 Metaverse 这一波概念炒太凶了，这个是他两个决策失误。Metaverse 走太快了，然后语言模型走太慢了。然后现在他重新组了团队，组了那个 Alexander Wang 是吧？那个 Scale AI，对，是那个创始人把他并过去了。我看吧，这个要看到今年第二，第一季度和第二季度能出什么产品呢？我不好评价，但是我不很乐观。对，他像 AI 模型的，他能够赶上 Grok，我就谢天谢地了。我就是赶不上，对，赶不上短期的。对，我觉得还是有点挑战，因为 Grok 是在一个在 Musk 的那种高压下的文化下的。
对对，那必须创始人那种死压着。其实这点也有点奇怪，像扎克伯格，他应该也是一个技术背景，但但他这一次表现的在 AI 面前，这件事有点有点手足无措。我我有两个这个观察，然后也听内部人说的，就是包括谷歌人说的。就 Meta 这个公司，它是做社交网络的，它现在的所有的 Cloud 部署，它没有像 Google、微软和 Amazon 这样子的把云计算部署给第三方使用的经验。这个是和它自己使用是完全不一样的，所以 AI 的这个技术 Infra 能力是非常强的选择能力，我要把它部署架构出来。所以他第一个是这个上面有些问题。
然后第二个，小扎是技术人，他是非常好的一个程序员，但是他有点像，他的这个技术 taste 有点像微软，哈哈，你知道我说的什么意思了？哈哈，这个 taste 不够好，哈哈。他选型，选技术总选错。这个之前我记过，我暂时在微博我是非常了解的。小扎我们他见过的，他来过我们这，我们开过会的，他也给我们，小扎特别会说话，特别圆滑的一个人。就他见你，就他天表扬你，然后背后捅刀子是吧？哈哈，对。然后呢，当时是他把他的所有的这个移动设备起用在我们当时叫做用网页来支持，那所谓移动开发，就是他支持这种网无这种手机页面技术。这很明显，据他跳他想跳过这个，他一直想跳过操作系统控制，是对是。但是他就跳不过去，他整个血型选择了，导致所有的 APP 开发全部失败。还好他的收购做得很好，对，还走了点弯路。这块我记得。
对对，okay。那我想继续，再对再那。那我们再跳到游戏，因为你刚你刚才其实游戏这是一个很有趣的一个话题。就我们知道这个微软，其实不是微软，其实腾讯，腾讯其实是一个本质上是主要是靠游戏来赚钱的，而且他在这两年，他的游戏还依然保持比较好的增速，就包括国际市场和国内。这个好像这个 Elon Musk 还专门夸过，是吧？他比较有 taste 的，至少从他收购公司，收购海外游戏公司还是比较有 taste 的。对吧？对，这是一个很很高度的一个评价。那么你会怎么看这个腾讯在游戏这一块的未来会怎么样？他会比如说，举个例子，这是一个会在比如像那个 Elon Musk 他作为一个重度玩家，然后他结合 AI，尤其是结合我们知道，其实游戏的相当多的部分，对于深层次 AI 来说，是用绝大的，绝对的用武之地。有没有可能会被突然一下子，就是就就被被，有一个特别大的一个什么给颠覆掉？或者是哦，对。
这个要看两个路线。一个是本身游戏本身。对，因为我们都玩游，我是玩游戏的，我是老玩家了。Elon Musk 玩的游戏我都玩。对，我对游戏非常了解。现在的游戏呢，有一些游戏我们叫做 Web，就是在线上那种，就是我们可能叫联网式的游戏嘛。还有一些呢，就是提供真实体验的游戏，那些游戏在主机游戏嘛。然后对不对？这两种游戏。然后腾讯的，腾讯的布局，他对网络游戏的收购是做得非常好的，就是能够联网玩的，包括主机游戏的控制，他也做得很好。世界上最好的游戏引擎，他控制了 40% 的股份，在 30 多到 40 左右，就是 Unreal 这个公司是吧？然后而且呢，3D 的这个下一代，就以前我记得这个 Jensen Nvidia CEO Jensen Huang 他们说过了，未来至少在他的 VDA 显卡上面，未来你看到的每一个像素都是深层的，不是渲染的。对不对？这也预示着 generative AI 深层式 AI 未来最快的速度进到游戏里面来。至少现在，刚前面有说游戏里面大量的素材嘛。对，大家说这些素材都是降本增效了，所以是反馈。他比影视剧还直接。而且下一代呢，我们可能会有一个制作观念的转变。游戏的互动形式会全部变的，就是我能够及时生成叙事。那我对这个这个创作方法就不会像以前，因为我要设计好新的剧本，我要定好剧情，然后你这样才能够渲染吧？那你想我 AI 就可以动态的帮你生成，生成 NPC，生成世界。那整个游戏就是一个全动态的了，在你的这个主创人员的这个事件驱动下，变成全动态的了。这个东西，我觉得会解锁一些全新的玩法，而且这是颠覆性的。这个我估计 XI 就想干这个事情，因为他也在拼命的招募，他没有收购公司，他只是在在招募最优秀公司的游戏策划人员，按照 Elon Musk 的意思去重新改造游戏。哈哈，而且他还脑机接口嘛。哈哈，还脑机接口。
是是，这个这个，我觉得是一个我我个人是很看好这个游戏产业下一个颠覆的。而且在这个游戏，全世界游戏公司里面最强大的公司，其实就是腾讯。如果算游戏，如果把腾讯看成一家游戏公司，他太牛逼的公司。他在世界上各地收购和投资控股，都是很厉害。所以说不管哪个地方产生变革，腾讯肯定是第一个要享受到的。是，所以我觉得可能这一点，就是腾讯在游戏上受益于 AI 这件事，其实我觉得有一点被低估了。
okay。那我们继续再往下，就是那第五个问题。我们有点慢，但没关系，后面我们可以稍微快一点。但是因为慢，但是我觉得就我们谈的内容，就是 Indigo 还是给我们分享了很多特别，就信息量非常非常大。但我这稍微做个小广告，就是我们在这儿呢，就会完了之后会整理一些文本出来，然后到时候大家可以加我们的那个，我们有一个微信的客服，然后大家加我们的那个微信的客服，然后可以就我们的社群小伙伴们，就是就可以会有一些分享的一些相关的资料。就是哎，我没有看到在那个，对，大家就扫这个二维码就行了。扫这个二维码，然后待会我也会把那个让我们的助理，把那个 Indigo 的在那个 X 上，还有 YouTube 上都都有自己的频道。待会我们也会把链接发出来，然后大家直接直接加上就行了。
那我们再继续再下一个问题，就是在这个时候呢，就是一个特别，就我们具体点，就我相信我们的这个房间里都是大家很多的年轻人，很多都是好学的人。然后在这个时候呢，学习变成了一个很奇怪的事情了。为啥变成一个很奇怪的事情？因为最好的智慧是来自于一线的，就好像比如说举个例子，都是这些这些大厂的人，这些创始人，然后他们的他们的对这个世界的理解，就就坦率说，就 Indigo，就是我我这两年我也见了一些老一代的科学家，就教授而且就是教神经网络的，我还真见过两个，就是去请教，但还是学了很多东西，但是也会让我有一点失望。就真的就算了，我还不如去听播客，有一些他真的是会出现了这样。所以其实对个体而言，就出现了一个特别，这就是我们这个问题，就特别具体，就对个体来学习这件事，然后现在又有太多的这种这个这个知识网红，是吧？又不太懂，然后呢就胡说八道。那你能不能给一个，就你也提到了一个，要建一个高信噪比的一个信息摄入系统。那这个时候呢，你能够去把你这能不能给到我们的小伙伴们一些具体的一些，哎，一个框架？
对，这个信息系统因人而异。因为我的工作和我的职业原因，因为我做投资嘛，做投资呢，和做早期，反正我什么阶段投资我都做。所以说我会对科技的这些信号很关心，在科技趋势变化的。刚才我讲的这，我其实讲的我都有例子，我都是看过的这个例子的。对，然后然后呢，所以这个东西呢，我是有可靠的新闻源，而且新闻源能够做的比较挖的深一点的，这种深度报道。对，这相关英文，英文世界里面的这个还是可以的。有的新闻源。
然后第二个，你刚强调的，就是老教授聊天还不如听播客。我说实话，我也是听播客学习，但是播客学习肯定是不系统的。但但是现在呢，你想系统性的学，有 AI 在，你想发现这个事情的兴奋点和亮点，欸，这个东西新趋势是什么？那你就听播客，因为 AI 不会告诉你新趋势的，AI 是个被动的东西。然后播客就会你去了，今天最头部公司的播客，最头部的机构的播客。你就像美国，我觉得美国现在的 VC 都是媒体 VC is media。他们这个 A16z，红杉，他们都是媒体做得很好，而且他会邀请他们投资公司里面最最好的这研究人员，CEO 来分享。但是投资公司分享，他还是有些商业目的的。最纯粹的就是邀请研究员来分享的这播客，或者直接邀请挖人木的员工来分享播客。所以说我觉得之前有一个人，有一个博主写了，选播客也是要有品位的。对，选播客。所以大家不要天听那些什么企业 CEO 的播客了，那个东西大部分都是 CEO，他也不方便乱说，因为说错了，还股价波动。对吧？他不敢乱说，他有 PR 控制。然后另外一个，很多 CEO 出来是做 PR，对吧？这是他两个需求。那，你还不如就去听研究员的，或者听这个领域里面的学者，专家的播客。英文世界非常多这种播客，你只要好好找都找得到。有一些专门采访这些播客，比如说我，我很喜欢听人工智能，就是阿尔伯塔大学的播客。阿尔伯塔大学，哈，阿尔伯塔有一个大神嘛，就是有一个，他叫这个 Rachel Sad 是吧？对，三三巨头，三巨头之一，对不对？他，他不算三巨头，他是新新一代巨头，新一代巨头。okay，三巨头，对，其实三巨头有两个都在加拿大嘛，再加，他这三个都在加拿大。然后，然后这个，这个里面是有比较好的这个信息的。
还有一个是研究员的，还有一些那种采访公司里面的项目组的。挺好的，有些在在在北美这边也挺多的。就是采访，不是，有一些他项目产品发布了，比如说 OpenAI 有一些产品，比如说像可乐 out，有一些产子产品发布了。那么他们会做一些，比如 Jimmy 奶，Jimmy 奶的团队我听过，他们是最敢分享的，就好像 Google 对这个 PR 控制的不严，你知道他们什么都敢讲，哈哈，对，他们什么都分享。挺有趣的。所以说这个东西你可以找的很。然后你有了这些信线，信息线索之后呢，你再把这些记下来，你再用 AI 去做深度 deep research，然后再 AI 给你编一个学习教程出来，再来学。懂了吧？对，这个是我觉得比较好的方法。
等一下，你继续。okay，好，我继续说。你弄你的，对，那我就其实我们就做一个小结。其实现在最好的学习的方式，就是直奔最前沿的，这个最厉害的这些人，他们往会是目前来看呢，基本上会是通过播客或者在 X 上其实都比较强。一个比较好的学习方式呢，比如说我们可以让 AI 来列出来，哪些是符合这些标准的，然后让各个 AI 互相拷问一下。其实刚才 Indigo 也介绍了一个学习的方式，其实也是我自己比较喜欢的，就是我自己比较喜欢的，我把它称作叫所谓的三频法。所谓的三频法呢，就是其中的就我面前有三个屏幕。一个屏幕是人类就最厉害的这些人，不管是他，他的 book，还是他的书。然后呢，还有一个屏呢，是 AI，凡是讲的这些有些概念或者有些东西，哎呦，他的因为我毕竟不是专业人士，有些概念我让 AI 解释给我听，或者让 AI 帮我把知识点补齐。中间的屏是我自己的，我自己的学习，我自己不管是通过写作，或者通过什么样的方式把它输出。所以我觉得这三屏会是一个比较有趣的方式。okay，让我们继续在下一个。
那么我们再说一下一代。那么这下一代也包括我们，也可以说我们自己。就是关于学习和教育这块，我们就快速的把，我把它几个问题稍微合并一下。就是就 Indigo 里说了一个关于老虎的一个概念，因为就我们在一个碎片化的时代，其实我们完全变成了一个及时满足的一个动物。就我自己现在就包括也是，我就是就比如说，我们任何一个人，我那就在我们社群里，我说让大家做个实验，你闭着眼睛，让自己啥也不干，待 60 秒。因为我们觉得一分钟是一眨眼就过去了，60 秒就你闭着眼睛慢的数 60 下，你会发现这并不是一件容易的事儿。对于现代人来说，这个事情已经变得，我们已经被这种各种短视频，乱七八糟，真的给弄的非常焦躁不安了。那么在这个时候呢，我想能不能够围绕这个老虎，对于我们我们个体，如何去对抗这种所谓的大脑的腐烂，以及就是我们对于 AI 时代，我们下一代的育儿，可能会有一些底层的，很多观念要彻底的发生改变。你可能得转变一些东西，是吧？而且对于我觉得很多时候我就认为家长，你其实就是这种所谓的三观，是吧？就是你对了的话，你其实大事不糊涂就可以了。但如果你三观上就有问题，但但是 AI 时代的教育，孩子的三观到底是啥？我们能不能从这儿来，你来跟我们分享一下？
对，像老虎，我当时说老虎的概念我是分享的。对，我说年度 2024 年的年度词汇老虎，他是有研究员拿这个让模型，让语言模型去看这些对社交网络碎片化内容，看完了之后，他自己也思维混乱了。对，你看你看，哈哈，原模型也思维混乱了，他让思考受阻，他没法形成链式思维了，就是一步推导下去的。对，这是他做的测试，然后叫他 X 上的这个，这个什么，他们有个算法，就是点赞差不多少的。因为社交网络，我做社交网络，我知道大家，我们考核都是按这个互动量考核的。你的东西互动量越好，那用户那你后面跟的广告越好卖，对不对？流量越高。但互动量好的内容不一定是很能让你，就是有的有效收获的内容，他只是会让你满足了一下情绪，价值越上头越好。对吧？所以说最后，社交网络让社会两极化，就是仇恨，对吧？就是让让那个让大家都要泡泡，都是泡泡。是社交网络自然形成的，这个没办法，这是算法造成的，这我们都懂。现在 AI 更是这样子造成。
那么它然后这个东西呢，我觉得特别是短视频，把这个东西加剧了。对，哎，短视频有两个，有两个我觉得有两个不好的地方。一个他太碎片化了，大家短视频你看到的东西都是，你看推荐嘛，基本上是流量好的。流量好的都满足前面的 3 个特征，对吧？哈哈，那个是吧？最快的时候吸引你眼球，最快的时间给你满足，然后最快的时间给你个结论，感觉是个结论，然后实际上什么都没有。对不对？那样说有人看短视频学习，那你能学到什么东西呢？因为学习是一个需要，听我们这门节目，你至少把我 20 个问题都听完，是吧？哈，你才能学到一点东西。对，这是我觉得是短视频要控制，特别对小孩来说。小孩我，我肯定是不让小孩看短视频的，而且他们也挺听话的，他就不看。对，他就不看，就是对他不看，他看 YouTube，我让他看长的可以，对，看动画片都可以，就是不能看短视频。因为他控制不住小孩，控制不住。然后，这是一个绝对要戒掉，对小孩一定要戒掉，哪怕以后建议政府立法把它戒掉，像澳大利亚那样，是吧？哈哈，澳大利亚已经立法了。然后对对，对小孩对大人其实也是一样的。对，这个，这个很多人他因为不像小孩，他也能控制，他就是太累了，他每天是吧？这个这个牛马坐久了，或者说怎么样，工作太累，他就想休息一下放松，刷短视频很快乐的。就是这就叫做及时满足嘛。
所以这个东西呢，你想看，我觉得接下来 AI 让大家的这种牛马工作变少了，是好事。那你的时间就变多了，对吧？变多了之后呢，你就可以专注的做一点自我提升，有这个时间。那你去社群里面去学习，去系统的去做一点什么东西出来，让 AI 帮你。那核心在这个时代，你第二个能力就是你需要有想法，你有好奇心的摸索这个。所以我就怕 AI，但是我觉得可能 90% 的人都没有想法。哈哈，对。有想法的人，那肯定早就把用 AI 用到飞起了。就是你得有想法。哎，这东西从小对小孩教育来说，第二个，让他第一个让他不要接触老虎的东西。第二个，你得让他自己做主，拿主意，自己有想法产生。对，你让他多自己做主，他自己就会有想法了。对对，这东西你不要去太严格的去要求他，让他玩他玩什么都可以，就是但是你不能控制他去上瘾的东西。然后有想法，成年人可能是，我觉得成年人有时候还没有小孩的想法。对，你观察过就知道，成天太累了，真的就不会有想法。哈哈，每天哈哈挣钱都累死了。其实我觉得对 AI 如果真的能够释放大家这个劳动力，那真的是造福人类了。对，当然这个前提是你得，我不干这个工作，我政府也得给我一份钱才行。哈哈，对，那个保底收入我才能够。但是我觉得这个也不也很也会很快的，也会很快的。知道加拿大好像之前有这个想法，自由党哈哈，给每个人每个月发 200 块钱。我操，对对，是。
okay。那其实呢，我觉得到关于不管是小孩子的教育，还是就是对于个人来说，就是人在教育和成长这块，如果让我来想，如果就是他是一个要回答的问题，我就想可能会有三个关键词。其实 Indigo 刚才也说到了，就是，一个关键词呢，其实非常重要，就是还是一个独立思考和做决策的能力。就是让小孩子从小做决策，这是一个特别简单的一件事。比如说举个例子，就是这个中国家长最容易犯的错，什么是哎，小孩子还小，就是啥都给他做主了。其实特别重要，就是小孩子其实特别有意思。我就我就因为我以前地产是跟犹太人合作嘛，然后然后我曾经有一个朋友，也是一个犹太人，是一个这个这个国际教育专家。然后正好他的小孩子，跟我的小孩子跟我们家老二是同岁，我们家小孩子大概很小的时候，我就问他，我说你假如只有一个词儿，你告诉我，他那个，他那个孩子也是个男孩，就是你养育孩子只用一个词，你告诉我是啥？然后他很认真的想了一下，他说了就是就是做决定，他认为是最重要的。所以呢，就是就是小孩子不管怎么样，让他去做决定，哪怕只有两岁，今天晚上你要听啥，你来选。然后呢，小孩子稍微大点，我们去哪玩，你来定，错了的话你得承担。然后甚至于再小一点，吃完饭让他自己端碗，把它放到厨房去，摔了就摔了。哎，这个这个特别重要。
第一是决定。第二呢，我觉得可能特别特别重要的，就是就是好奇心。因为我觉得可能对于假如对于 AI 而言，人类还有什么价值？可能人类的好奇心这块，是一个特别特别重要。因为好奇心它是有两重，一个代表的某种随机探索，然后再一个呢，它代表的这种欲望和目标和理想，和这个它，它是一个定义了一个所谓的效用函数。因为对于 AI 来说，最长的最强的是，你给我定义好这个函数，我来怎么样去来去趋近于它，去找到一个相对最优解。所以，这是一个人和机的一个特别好的结合。
第三个我觉得可能特别重要的，就刚才我们也老在说这个，就是 taste，taste 会非常重要。越来越但这种 taste 要建立这种 taste，其实特别特别难。你需要这种跨学科，你需要它在某种意义上呢，也说可能是无法训练的，是一种涌现出来的结果。okay。
那，那我们再继续，哎呦，时间好快。就是再继续，其实我觉得下面是可能，我再稍微吆喝一下，就是我们的同学加一下我们的那个二维码，然后这样的话，就因为我们的社群的有些资料可以发给大家。我们的那个侯静可以给那个发一下。
那我们现在这一块，我觉得是一个特别特别重要，我想把这个作为一个重点。就今天我们就把这几点把它聊聊，聊的那个就是说白了就是，我想起来，就是去年的时候，阿里就是说那个，就是那个关键词，AI 的询问的关键词，排在第一的呢，是股票，就是想用 AI 选股。这显然不太可能。第二呢，就是算命，就是算我的命。第三，是，怎么发朋友圈？哎，我觉得这个特别有趣，它基本上代表了，这个咱们这个中国人的精神世界。是吧？第一就是搞钱，排在第一的必须是搞钱。是吧？那我们就说说，哈哈，我们这个话题下面，就是就说一下职业和搞钱。
就 Indo 你说了，其实现在就是正在 AI 带来的一个变化，就是它，它会你提了一个这个中空效应。就我们说到了，就是大模型从商业上是消灭中间商的，从职场上也是消灭中间层的。其实这是一个挺可怕的事情一件事情。就我们说这个，就怎么样怎么样面对这种消灭，就是公司层面的消灭和职场上的对于中间的消灭？
对，就这个，这个既然这个事物出现了，它就不可避免。对，因为大家慢的人就做决策，做验收是吧？这是两头嘛。我画过一个图，对，是中间都是 AI 建做的。然后呢，这个是这是很明显的一个趋势。而且能力越强，对，他这个让你验收，或者让做决让让做决策就是越快，然后验收其实就越简单。对吧？这个东西挺难的。而且现在最大的变化，我们做过实验，在组织里面，就是中间层是对中，公司的中层是对 AI 阻力最大的。老板出来看，对下中层去，他中层人员各种 KPI，哈，这个这拒绝，找找麻烦。但呢，下面有很多底层的员工呢，或者说他很积极，有的人因为他自己很自由嘛，我用什么工具，其实 AI 现在是自下而上的。它是由对下面很能力很强的一些，这些在一线的人，他们他要提高工作效率嘛。我要是吧？我要让我自己变轻松，我能够摸鱼，是吧？多的时间摸鱼，那我就应该干了是吧？这个大家都这样想。所以说就有一批人还是很牛逼的，下面。然后呢，在中层就是他用，因为他总觉得自己工作没了要，对吧？就是这有一种这种危险，就是危机感。万一哈，我这个业务线上面有很多东西被替代走了，我怎么办？客服忠诚，这个这个，对吧？他在忠诚，老板说这不需要了，那我就直接从 VP 直接下来，我用 AI 就好了。对吧？对，这个是组织架构，是对 AI 最大的一 个阻力。
所以说，大家现在，我觉得按照现在这个趋势来看呢，我不知道这个社群里面肯定有当老板的，有打工的嘛，是吧？当老板呢，这个比较这个特点就是肯定是要先学好 AI 怎么用 AI 提高自己的这个企业效率和投资效率，是吧？这个这个是必须要掌握的东西。因为而且不能被忽悠了，这个 AI 怎么去精准提升效率？我觉得现在来看呢，在知识工作里面是可以的，一些重复性的工作，信息处理，什么一些客户服务，还有内容生产的，这是可以的，包括程序代码，这是最快的。然后这个要多使用。另外一块呢，可以辅助 AI，可以很好，AI 是一个很好的辅助决策工具，它可以给你提供多种场，我经常用多种场景模拟，然后你来选答案。我不清楚这未来怎么样，让他推演一遍，他基本上他是个平均水平嘛，但是他也不会让你犯重大错误。对，就减少错误率，也是提高成功率的办法，对不对？对，这个这个是可以做决策上的。
所以说大家在这个里面呢，在在 AI 你在这个咱们改进之后呢，如果你有一个生意，或者你有一个业务企业在做，你效率提升了，你也自然就提升了利润率嘛。你可以裁掉一部分人，确实可能裁掉一个重复的工作，或者让你的角色变得更好。对，因为你想看，在现在的 GBT 2.5 这种时代，我我三年前，我之前我去年年底年终的时候，我发了一个发了 AI X，我讲的比较具体，这样形象一点。我对开着自动驾驶 FSD，然后然后我用着什么 GPT 这样聊天，我在疫情期间 2020 年前我想都没想过这种事件。2025 年和 2020 年，五年前，2025 年，2020 年，你根本想不到 2025 年世界会是这样子的。你想看，想不到。我不可能想到我这个车会自己开，然后我会用着比我还聪明的 AI 帮我来做决策，我不可能。这个那个时候想不到。那么现在是 2026 年，你想 2030 年会是什么样子的？想得到吗？我也不敢想。我现在可能我只是幻想一下。对，那么呢，这个里面呢，其实机会挺大。对于这个里面来说，对公司你肯定要用它，你不用它，你不在你不在这个船上面，永远就会落后了，是吧？
对于个人来说，我觉得对于个人，说搞钱这个词不是很好。我觉得搞钱有点急躁。很多钱是怎么说呢？是真的在你自己最擅长的地方，你慢做出来的，你会发现这个事情它有一个向心力。你在做这个事情时候，你会就是说很多人就是打工不一样，就是如果说你真的要自己在做创业，做点什么事情的时候，你并没有，是你在创业完成一份工作，这个就是在做你自己最喜欢的事情，然后顺道做了一个事业，然后把钱赚了。这是最好的一种状态。对很多人创业的时候，做我最喜欢做的事情，对不对？就像以前，以前李佳琦说他就喜欢做直播，他一天播十个小时都可以，这是爱好，哈哈。他最开始一天播十十五个小时，好像播 16 个小时，对，不停的播。就他这爱好，对，所以这个时候对钱就是对自己的一种奖赏。所以说搞钱这个一，这个出发点就不对。这个出发点就不对。不能说你搞钱。
对我我特别赞成这件事，对，对我我非常非常赞成。因为其实钱并不是去捡的钱，因为是一个稀缺的，是在竞争的。你一定要是创造出价值，然后你创造出价值了，你一定要做工，你到底付出了什么？学习并不是做工，你学习是改变你这个做工的机器。所以我特别特别赞成，就是就是，我们会觉得就是现在就冒出各种各样的 AI 的工具，就我们不管它好还是坏，其实最重要的是它能够帮助你的本职工作，能够做好什么？怎么说呢？就是不管是投资还是工作，其实回归到这种特别特别，我就是很本分，我就是我现在到底在干嘛？不管我现在如果是打工，我就好好的把我这份工作做好。我这个这个，虽然这个老板很讨厌，这个同事也很烦，这个工作我可能以后要换，但是对我现在而言，这就是我的工作，我的回报率最高的，就是全力以赴把我这个工作做好。然后呢，我在我的工作岗位上，尽可能的去使用好 AI，尽量的使用好，去学习就我自己的一个感触。
会特别有趣，就是有些相当多的有些人，有一个很奇怪的心理，就说了我不能便宜了老板，所以我就要去混。这是一个很很要命的一件事情，因为其实广义而言，每个人其实不管你是老板，你可以这么去理解，是你在创业的时候，他还在给你发工资，还不要你承担风险。当然这说这是一个。那我们在能不能这个话题，我们也比较那个，就你会觉得在这个时候，就以我们队来说，因为相当多的还是在这个要感觉，要被 AI 影响到的这个中间层。那么对于这个中间层，就大家可以做什么呢？比如说，哎，我可以有哪些能力可以去培养？可以主动的去做些什么？哎，我也会很实在的做好我的工作，我也会尽可能去运用。但是有哪些是我周末，或者我晚上，我可以去自我强化的一些东西？从技不管是技能还是我的原技能，比如技能我可以去学是吧？哎，别人说这个工具我去用，但还有一些原技能，可能我要增加一些某些方面的阅读，我要去干点啥，或者是等等。
对，这个东西呢，我觉得 AI 是可以让你快速的开始学习的一个好方法。如果你想学 AI，可以让你学得很快。而且呢，另外一个呢，AI 可以呢，让你创造内容出来是吧？这个这是可以的。因为现在是原模型嘛，他可以很快速的让你创造内容出来。因为你学习最好的方法就是学完了，要输出嘛。你刚才说了嘛，对你这个闭环得完成嘛。如果你只学你不输出等于没学了，对吧？对你有没有验收你的学习成果？所以挺多人会分享，很多人会分享的。随着你在业余时间，你成为你的这个工种里面的专家，这个不难吧？就你之前可能在你的这个行业里面表现平，突然你每天用 AI 来加强学习之后，你在这个领域里面一脱颖而出，那至少在你的这个社交网络里面，在你的公司的圈子里面，那你就比别人更多的这个能力，曝光机会了。对你做你，你只要是能力变强，你就有多的曝光机会。那你想做更好的职业，或者说用这个能力来带来另外的附加的收入，也是几率也变高了嘛。这个是我，我觉得这是一个比较 AI 强化，你自己输出，然后让你成为你的。我记得我以前聊过一期播客，和那个谁，那个特斯拉 99T，他就是牛马应对方法，一期播客，他聊得非常好，就是让你自己成为你的岗位上最牛逼的人，在公司里面 AI 就干这个事情的。对是吧？哎，我觉得这个特别具体，我觉得很具体。
Indigo 很实在，就是就是很具体，就是说我们别讲那些大道理，是吧？别说我要学这学那，应对这应对那。就就具体一点，就是说比如说举个例子，我如何运用 AI 把我现在手头的事情变得更好，哪怕只有一点提升。然后呢，我在周末的时候试着把它输出，以你喜欢的任何形式，文字的形式，视频的形式，播客的形式，哪个让你舒服，你选一个。哪怕特别特别严，哪怕是朋友圈的形式也可以。而且这年头你根本不用担心不会表达，是吧？你甚至可以通过对话，对话的方哎，我做了这个，这玩意对我挺有用，我对 AI 来讲，你帮我把这个故事写出来，然后我到小红书，我每周分享这么一个。然后在输出的这个过程中，我是强化了我的自我学习，然后我也让这件事有点滚雪球的效应。我觉得这个特别特别具体，我觉得比讲那种各种大道理，或者比起各种恐慌都要具体的多。特别好。
然后我们下一个。okay，那我们就说到了一个，就刚才你也提到了，就是就是这个，就是从管理者的一个价值，就是从叫部署者和验收者，中间就交给了 AI。但现在也越流越来，越流行的就是所谓的一人公司。就来这是一个特别热门的话题，你能不能跟我们分享一下，就是比如艺人公司是一个故事，还是说比如在硅谷，在北美就是做的比较好的，或者已经可行的，已经有哪些可以来跟我们分享一下？
其实现在硅谷很多创业的公司都很小，可能就两三个人，最多 5 个人，一般对，就这么多人。然后呢，在某种意义上也是艺人公司，是吧？可以这么说，也是艺人公司。对对，我记得有一家叫好早以前，其实 AI 流行之前，就有两个很有名的公司，一个负责专门卖数字商品的那个公司，以后我一下忘了名字，那公司就一个人，他很多年就一个人。对，还有大家用的那个，那个叫黑曜石 Opstine 那个软件，那也是一个人。对，还还有澳大利亚的做数据的公司，也是一个人。那个在 AI 之前就有了。这个艺人公司并不是 AI 之后才有的。就之前，就是说这个是用拉巴尔的时候，你把这个软件的，软件的这个网络效应，或者是数字无限复制效应发挥的最大了。你一个人，你是个全栈工程师，你什么都会，你从写代码，做网页，营销全部都会，那就已经干得好了。
但是现在呢，就是说 AI 进来之后呢，你代码不用自己写了，你有想法，那可以写，你只用验收代码就好了。然后你想做这个 marketing，那很多 marketing 自动化 AI 呀，它自动帮你发社交网络，然后自动来做。然后呢，客服是吧？那马上就有，对不对？都有，你要有 agent 就好了，要指挥人。但很多人大家说，我开着一堆 cursor，然后到处是 agent cc 他到时候也剪头，然后帮我工作，我一个人在这指挥就好了。这个时候我觉得 AI 这个时代，让艺人公司的成本变低了，可能性变得更高了。而且很多硅谷有很多创业公司，可能就两三个人，怎么说呢？就是 CEO 是吧？就是基本上简单上什么都做。因为我觉得有时候呢，两三个人公司，大家还是有个有个商量的人，比较好一点。这个人还是需要点有情感联系的，至少是我们要合伙人是吧？但是呢，我自己，比如说我们现在做类似，我我自己做社交媒体，你也是自己做社交媒体，我也可以说我自己是艺人公司，我是这个文化，我是艺人公司。虽然我们有投资公司，但是我自己做的是媒体这块的业务，就是我自己，我就一个人，我就一个人，我自己写，我自己用 AI 技能帮我加速，对吧？这个我觉得都没有太难的地方。对，现在。
而且未来这个东西会，我觉得会出现非常多的这种 KOL Youtuber，国内我们叫做说网红直播的 KOL，美国非常牛逼 KOL 才会有 studio，一般就是一个人，他一个人，他能坑他的粉丝上百万，对，有一点，就他自己形成一个巨大的网络。这个这个东西因为有工具足够强大了吧，他自己来做。对，他自己就就他做视频内容，或者做成短内容，因为平台给他带来了极大的福利，可以帮他做无限传播和分发，他也可以收会员费，这对他一个人来说就足够了。对，他来说，唉，他创造了很多用户内容消费的价值。我觉得这个这个是 AI AI 上让这个人个人能力加强了，加强之后呢，这种艺人公司成本就更低了，可能性更高了。而且我周我周围有认识的一些，比如我，我们中国不是有一些这种圈子嘛，就是做 AI 创作的，做 AIGC，做偶像的，也有天在那分享，分享那个什么 AI 使用技巧的。好多人都是艺人公司，他们都说我光荣离职了，我从字节走了，我从哪离职了，我现在一个人干，我就是个公众号，然后一个 X，然后呢就负责每天做各种教程，发工。因为有个特点，他就是这种，我说我前面那个例子，让他成为这个学习领域最强的人，马上他就脱颖而出，他把他学到的东西马上分享出去，他快速的积累了网络效应。对，就是前面，那我们再说一下，对于他已经积累到可以离开他的工作了。
对，就是我们来说一下，就是其实就是相当于 AI 对个人的能力，对个人的工作，对个人的影响，他像一个放大器一样。然后然后又因为就说就说现在的大模型的，它的这种，他其实是一个很博学的，他是一个跨界的一个通才。对于个体来说呢，其实可以迅速的补齐你的一些短板。对，就它的确会很重要。当然我们另外就特别具体，包括你刚才你也说你会用 agent 的，包括那能不能就是给大家介绍一下，比如说作为我，作为艺人公司，我如何更好地运用好 AI，如何更好地运用 agent？那么对于这个 Vibe coding 这件事情上，我可以做些什么？或者就是，这个给我们介绍一下？
对，我觉得两种方法。有的人先用 Vibe Vibe code 是解决自己的需求。因为我对我，比如说我，我自己现在我在生产内容，比如说我想跟大家做分享，我想组织一套 keynote 出来。你看我上一期的 keynote，简单，因为挺多都是这个 AI 就生成的嘛，就是对吧？因为之前我做这种图，都是我自己一张画呀，我自己画的。对，现在不用了。我现在我告诉你，见面来，我说我要这样做一个什么风格的，我要怎么用，他很快给我出几个，两三个我来选。这把我效率提高太多了，知道吧？而且所以说我现在我想真的去，我来做一些课程，我真的我口述就好了，我说我想这样。比如说我们今天两个人聊的这个内容，我就可以做一套很漂亮的，这个回头，我们可能花只要 30 分钟就做出来了，30 分钟就能做出来。这个是吧？有你可以有做得更好一点，有 AI 技能，可以自动的从从这个我们所有的语音转录，把它变成结构化的文章，然后再生成视频，再配上图，全部都可以 AI 来做。当然这个核心就是说，是我们两个人生成的内容，我再用我们平时讲的内容，反复的变成另外一种数字形式，再做一次传播。这个就是一种传播的复利效应。对于说对于我来说，就因为对于知识博主吧，我们俩也算知识博主，这个是最快的方式的把你的知识沉淀下来，再传播出去的。肯定是你的杠杆效应，复利嘛，对吧？对吧？那那种 AI 就极帮极大的帮我加强了。这是我举的一个例子。
对对，但在这我也可以做一个人的一个小的一个分享。因为 Indigo 它是技术背景的，我不是一个技术背景的。那那其实我自己的一个感悟是这样的，就是你应该直接去用，就不要想太多，就不要想着哎，就是我要怎么样，我要他会不会多复杂，我要用到最好的，他有什么流程，有什么方法。其实并不是，就你拿来就用就好了。然后呢，在这个过程中呢，比如说举个例子，在业务上，就比如说我刚才我自己，就我们在做这个，这个我们的线上人生复利行动营的时候了，他其实并没有那么简单，因为他有很多头绪。你这个时候如果想，比如说我要通过 Vibe code，要把这个东西全部把它给弄好，一下子把它弄好其实不容易的。我们做的一种方式是什么呢？是人加上 AI。就开始时候你一定要有手搓的环节，你得就我觉得也没有必要介意手搓，你真的手搓的哇，这玩意搓不过来了，然后我这时候再想办法，再把它再产品化，再我再考虑去 agent，再考虑在 Vibe code 等等。我觉得这样可能会更现实，将会更好。
这说白了，本质上就和当初的精益创业的思路也是一样的。你先别太早，就是像那几位老兄，就是硅谷那几位老兄经常说的，这就是他其中有种说法叫首先的时候，你应该做不可重复的事情。他其实说的所谓的做不可重复的事情，你只有在这个过程中，才能探索出什么东西是可以重复的，才可以被规模化，才可以加上 AI 的杠杆。我觉得这是一个试错的一个过程。
那在这我想再来 um 问一个，就是请教一个，就说我们说一下这个副业的可能性。这个副业的可能性，当然就是我们刚才首先就讨论起，不管是艺人公司还是个人跟 AI 的结合，我们都提到了是和自己的主业相结合，是吧？得是你的主业，得是你的工作，得是你每天在花的时间，得是你在这方面有一定的专业优势，投入时间的。副业看起来有一点跳。就那么我们到底如何定义这个副业？首先呢，我先说，我先开个头，就说副业可能有两种。一种呢，我是干 A 职业的，然后呢，我去弄一个 B 职业的事情。比如说举个例子，我是我是写自媒体的，然后我的副业是教别人下棋。是吧？这就完全跳开了。没关系。还有一种副业呢，就是我是做自媒体的，然后我的副业是我用 AI 去，我教别人如何用 AI 做自媒体。是吧？这算一个副业。所以他可能我只是大概框架，不然这个问题就不没有被精确定义。
就是 individual 你来说这个副业。对，副业就刚你刚才说的很清楚了。对，一个和你职业有相关联的，一个和你的做的工作没有太大关系的，就是教别人下棋。对，比如说我这做投资，我教别人弹琴，不对不对。对，教别人弹琴，这也是副业。就这个东西，只我觉得在这个，在这个我们温哥华很常见。这一个人说上着上着班，他下午 3 点就下班了，然后晚上有很多时间，他会当老师，当老师最多了，教各种小孩。因为这个课外班很多嘛，很多人都有技能嘛。比如这个，这个很好的一个副业，就是他打两份工。打两份工，这是一种副业。但是这个副业有一种，我建议有一种副业，就是能够产生复利的副业。因为你前面说的副业，你打工，打主工卖卖，你每天 6 个小时劳动力时间，你打辅工，还要卖 4 个小时劳动，都在卖劳动力时间。你天卖劳动力时间，是没有复利的。对吧？哈，这个是。
所以说，我觉得副业做自媒体是比较好的。自媒体是有复利的，因为你的网络效应 build 起来了，你上面跑什么都可以。对吧？这个是比较好的。就是说一般副业呢，就是就是前面从来说了，生产可复制的，就是对，做自媒体，然后你就是你借 build network，是可复的，就是在建 build network 最重要，他是在建立一个运输管道嘛。对，你上面以后想，你想做课，你想教学，你想干什么，你想被别人做点介绍，一点考察性的业务，都可以做。你是一个渠道嘛，是吧？那这个东西，你就可以把它 build 的很快。这个这个是一种比较好的。
所以说，我觉得自媒体是一种很好的副业。而且大家都在 AI 这个时代，怎么说呢？我觉得说一个不好听的话，就是 AI 这个时代，很多 C 端应用可能没有太大价值，因为因为模型及产品，是吧？这是我们前面都说到了。然后呢，当一个技术发展巨快的时候，社会上就有很大焦虑感。焦虑感强的时候，那你就去上，就教课你卖课是最赚钱的。对，哈哈，这个哈哈，这是实话，这是实话。对，这个对，但是因为我，其实你刚才，对对，我觉得你刚才说了一个特别好的一个，就是甚至我们可以玩个谐音梗，就是这个副业，这个副就是那个的副，正负的负，我们把它变成复利的复，是吧？就是对，他是这个复利的业。我觉得特别好。
就是因为就是一个人，我主业是，一个专业我能挣钱，我副业还是另外一个专业，我也能挣钱。其实这种人挺少见的，就说明你两个都没干好。是吧？只有两个都没干好的人才会去，两个都是用时间换钱。我这说的有点那个，但是有些人是爱好，就是两个都没，或者就是比如说以前的那个高盛的很多人是爱好，对，像高盛的 CEO 说晚上去打碟，是吧？那那的确是爱好，是吧？那但是也有很多了，就是真的，我白天做一份工作，就是在这个，其实在加大也挺多的，就是就卖我的时间嘛，然后晚上也是出卖我的时间。但是两件完全不着边，都是消耗我的时间，就很疲惫。
其实呢，就最好了，就是如果从这儿呢，我们其实是可以得出一个小的结论。就说了，最好你的副业是和你的主业相关的。然后 Indigo 给出了一个特别好的模型，是啥呢？你的主业可以用时间去计算，然后你的副业呢，还是这件事，但它是一个有复利的事情。其实 Indigo 给我一个印象特别深的事呢，就是比如说，其实我记得两年前，就是在那个当堂，在文化当堂，听你的介绍，哎，觉得很有趣。因为因为我也看你的那个 Twitter 嘛，那时候 Twitter，其实你会你日常会思考，日常会记录，然后完了之后你会阶段性的把它汇总，然后到了年度又会把它汇总。其实你是在做一件有复利的事情，而且你是在这个这个雪球越滚越大。我觉得这是一个非常非常有趣的事情。
其实我今年呢，比如说我也会在做一个小的尝试。就我过去做的事太多了，太杂了，就是就是当然我会说，反正这是我的爱好，或者这是我的探索。但是我觉得这也是不对的。比如我 2026 年，我就做了一个小的一个探索，是啥呢？我大概是在两周前呢，就在去录了一个混沌的课程，主题是啥呢？就是人生复利。就是因为复利讲的特别多，但复利到底是啥呢？就是就有点像刚才那个 Indigo 里说的，就是谁能够把这个东西讲得最好，弄得最明白。讲复利的特别多，拉瓦尔讲复利，但是他呢，基本上就讲哎，复利挺重要，但复利到底是啥？从数学意义上是啥？从物理意义上是啥？从投资的角度是什么？没有人去讲。然后上网去一搜，你会发现讲复利的最好的一篇文章是谁的了？是我的。所以我就哎，那是不是我应该把这件事把它，哈哈，我有点厚颜无耻，就是那，对，我这有点厚颜无耻的讲出了一个事实。就是那个，然后我就想哎，我干脆把这个东西，把它弄的透一点儿得了。所以我就在混沌做了这样的一个课程。然后呢，同时呢，就我在知识星球的这个社群呢，以前叫 AI 决策岛，那我现在把它叫做人生复利花园。然后呢，中信呢有一本书稿拖了很久，然后最近跟他们的主编，我也说哎，能不能把它变成就叫人生复利。然后包括刚才我也说到，我们做的人生复利性的东西，就我希望呢，我做的所有的事情，它是一个网，就像一个花园一样，我不断地围绕这个花园，它不断地积淀，然后有滚雪球，能够产生这种复利效应，然后它又能够有所生长。就包括 Indigo 里所说的形成的，这样的网络，形成的资源能够聚合在一起。特别好。
那我们再来，我们加快点速度，就是因为我觉得你都能够给出我们特别特别好，特别特别具体的一些那个建议。就是我们说一下，我，我说一下，回到你身上，就我觉得你身上有一种这种，我还是非常觉得特别喜欢，就是这种乐观。就我们说一下你个人化的东西，你个人有没那么乐观的时候吗？还是说比如说你从你原来在深圳到北京，再到武汉，哪段时间就哪段时间没那么乐观？
有确实有。这个刚工作的时候其实有点迷茫嘛，对，刚工作之后，对，我刚我工，我工作及创业的嘛，就是跟别人在小公司做创业，就是有时候哎，就是因为这个时候前途不要我决定嘛，这个我们对这，但是也人少嘛，就是大家也能参与决策。就是那个时候确实不是那么乐观，就是我们有点随波逐流的感觉，就是就是撞了什么是什么，对吧？对，这个，然后呢，后面我把公司卖给新浪了嘛，就是运气好，卖给新浪了，然后面卖进去就马上做微博了。但这刚做微博也不是那么乐观，不知道就是这个东西是个新东西，对这个东西，反正但是新浪就挺惨的，新浪没什么产品，新浪基本上就是命悬一线，对门户也不怎么样，然后播客也不怎么样，然后社交网络也没有。那公司反正就背水一战，你们做呗，反正我们也没什么其他事做。对那个时候，哈哈，那做其实做了半年之后，其实一下子他这个还是有积累的嘛，现在还是大门户嘛，就是他的这个明星效应在，网络效应在，他很快就点燃了，点燃了之后那一下，哎，这半年之内就其实三个月之后，我觉得这个事情就是我们要要干的事情，就是新浪网的未来。哎，那个时候我们有了，我们乐观了。但新浪其他部分人不乐观，他们看不懂，对。然后那个时候公司也分分开了嘛，把我们单独立出去了，然后都说 oh，我们要去新公司创业，然后就就所以说，我们创始团队都去新公司嘛，对，就几个人，四个人，对，然后对这样子就开始嘛。但是那个时候呢，也整体来说，因为我觉得运气很好。
你站在了一个一个事物向上的一个周期上，那移动互联网刚开始。对，你就做一个移动互联网上最代表性的。那你要是还不乐观，那还谁能乐观？是吧？对不对？那必须要乐观，你不乐观，这件事情没法做。
然后呢，这是一个机会。然后来离开的时候，当然离开了之后，我是我是移民到加拿大的嘛。然后我又做投资去了。做投资那个时候，其实还可以，就是我我没有特别像早年那样，不是那么因为早年可能认知不够。后面随着这个认知提升了之后呢，我觉得我对自我的发展呢，就是人生发展，我并没有说定一个很高的目标了，就是我需要创一个什么公司，做很大。对，当时我们有日本做房地产，对，你在国内做，我们在日本做。这但是这个这个时候挺倒霉的，你刚做就碰到民诉法呀，后来碰到疫情。这个这个这个挺倒霉的。
那你怎么看？比如说举个例子，遇到这些倒霉的事情，你会自己怎么会去面对这些事情？你会是哪种？就是比如说举个例子，就我说哪种，我只是举个例子。比如说 Elon Musk 这种人，就是我死磕到底。假如是我，一个特别重要的事儿，假如是巴菲特了，他从投资的角度，这玩意儿不行，我就是叫换船，放弃我就换一个，我就一点不纠结。你你会是哪种风格？就面对这种事情？
当时我们还是想，就是让这个业务能够活着的。对，但是没有说宣布。所以说，我们尽量的让这个业务能够活下来，因为我们知道疫情总是要过去的嘛。对，特别日本很快嘛，很快就过去了。对，过去了之后呢，你一定会带来增长嘛。这个这个还好。这个当时选择是选择问题，就是我们并没有完全，因为很多当时非常公司倒闭了，非常的公司倒闭了。而且我们还从别人手上把资产接过来了。这是一个，这是一个，就是我们当时有有联合创始人嘛，所以说大家意见是一致的。哪怕我们卖资产，我们要公司活下来。对，这个这个是一个决策问题。
但是，所以你你积极的心态还是得有的。我觉得做任何事情乐观，就是我觉得，我觉得之前的一路巴士还出来说的这个，悲观者永远正，不是不是一路巴士说的对，悲观者永远正确。对，乐观者才会成功嘛，是吧？对，是乐观者才会向前。对，悲观者一次当然正确的。你说当然，做投资确实不一样，做投资是手上有棋，可以选择。就是我会放弃这个东西。所以说，所以说我大部分时间在做投资的，我就会我就知道要放弃公司。但是我能够拯救的，能够帮忙，是早期投资，能够帮还是帮。帮不了算了，帮两次第三次就算了，我没必要，也不是我自己做的。如果说能帮两次就够了。对，就仁至义尽了。对，对于股票就更简单了，那你不好了就抛掉就好了。对吧？就千万不要炒股，炒成这个炒，炒成原始股东了。真的炒成再也不走了。对，那没必要。对，所以说这都是选择。
其实我觉得乐观是一个心态，我觉得人做什么事情都要乐观。如果你不乐观，你没有动力，对，你没有动力，而且你心情不好。然后呢，这是这是前提，这是一个心态问题。然后第二个呢，有的这个，原来这个把我笑死了，草人董事长，哈哈。OK，对，这个第二第二个，就是你得有一个选择，就是你的认知。选择很重要。其实我，我觉得我之前过去的很多选择，还是比较正确的。就是基本上关键选择都没有选错，可能有些小错误，但是大的选择都还没有选错。
就是为什么呢？你觉得？你觉得你为什么可以做到大多数选择都是正确的？对，是。就是有时候这个有的动作可能选在了，选在这个事情变革之前，那个时候我就不叫选择那个时，那个时候就好像蒙的一样，就蒙的对。就说那个时候，我们这从移动互联网开始之前，最后我们就要做一个最像移动互联网的东西。对，当时公司，但是这个也有新浪的决定，但是我们的团队是我自己做的嘛，所以说我们感觉我们还是这个选择是对的。但是后面也有很多选择不对。微博，但是微博就把把自己慢变成了社交媒体嘛，很多社交网络的事情做不了。好，对。
然后呢，第二个呢，我当时选择，我说我选择换一个地方生活，我选择在在温哥华。你也也不是做过这个选择吧？就出来就是，我觉得选择这个同时，我也选择做投资，因为我觉得投资会比较灵活一点嘛，比较是我自己的风格。我不是一个特别适合做，我说我从来做不了 CEO。CEO 的人呢，要说他得一根筋一点。对，他得一根筋一点，就是他得很坚持他自己的这个事情。但是我觉得我比较适合，像给他提供策略的，给他提供选择方向的人。我可能有时候会乐观，是乐观对，但是我有可能不会那么坚持在一个业务上面。但是呢，这是一个，这是我的一个性格特点造成的。所以说我就有时候想法可能会比较多，就点子比较多，点子很多。
然后我当时在离开微博之后，我们做创业，我就帮助了好多人，就是做了不同类型的公司和艺术品的，包括像李佳琦他们的，还有还有其他的，还有挺多的，还有一些人。这几个比较比较有名，就是我就跟他们来做 start。对，这样子我就会，我感觉这样，我我比较愿意做这样的事情。哈，这是性格造成的。对，是。
所以其实你刚才说到决策，在我觉得有一个你，你说了一个特别重要的一个细节，我觉得挺挺好的。就你就说，你做不了 CEO。其实很多人不愿意开这个口，尤其是一个人，就是年轻气盛的阶段。这这就是我承认，就是相当于我承认，我在一件重要的事情上，我不擅长。其实很多人不愿意承认的。其实我就干过这种蠢事，老实说，很坦率说，我是干过这种蠢事。其实我是特别有，就是不全面的那种人，就是偏科的那种人。但是呢，从决策的角度来说呢，就我大概前几天，就是公众号文章也写过这个，因为我现在这个人生复利，我有一个方法论的东西，我把它叫做一个所谓的引擎。就所谓的引擎，就第一了，是是 be to be，就是所谓的 to be，就是我要成为什么样的人？就我这辈子，他这种 to be 有点像那个莎士比亚说的，就是关于存在，就是我的这个人生的意义，我到底要成为。因为就是，然后呢，才是选择，然后才是 to choose，你你做出选择。因为假如你不定义清晰，我到底是一个什么样的人，其实你很难做出正确的选择的。
所以我觉得，其实你的这个还挺有趣的，因为我另外一个好朋友，做投资的，他跟我说，他说他很早就发现，他他，他跟他比你还狠。那你是说，你是说你做不了 CEO。他说，他说我很早就发现，我这个人干不了实事，就是我啥啥，你让我具体的事我干，我都烦。哦，这这，其实对一个男性而言，就是也是一个性。然后他就，但是他很坚定地认定了这一点，然后他去做做做投资，他是去做二级市场的投资，然后呢，就是挣到，挣到了就做到了百亿级别。是吧？但是他很坚定的就就知道自己的那种。就就我觉得这点是一个特别好的。所以我觉得从决策的角度来说，很多时候，我们说决策是要做正确的事情。但是如果你没有自知之明，如果你不知道一个正确的我是什么，你就很难做出一个正确的一个决定。我觉得是。
这个也是很。那其实我刚才那个乐观的那个话题，我还是想把它问完。你觉得你的乐观是一种天性，还是说你的一种主动的选择呢？我觉我觉得性格占了很大部分，性格占了很大一部分。对，性格，性格占了很大一部分。然后呢，另外，所以你小时候，所以你小时候是在一个很有安全感的家庭长大的？对，比较安全感对。然后第二个呢，我觉得现在的这个，我们叫科技乐观主义。科技乐观主义，就是我们自己要这样塑造这个，这个硅谷的，硅谷的思路。硅谷现在就是强科技乐观主义嘛，就是他觉得因为现在科技进步很快，进步非常快，因为我们进入到了一个加速时代。如果你在这个时代你还很悲观，那就会错过所有机会。是对对，你要强行让自己乐观起来，然后呢，相信这些东西，他真的就会成为真的。
对，这个，我们我们，我觉得我大家在移动经常说，我们是最好的时代，就现在，但是也是最坏的时代。对，就很乱很乱。但是但是但是确实机会很多。对，你在这个上面，对吧？这个我觉得是，这是一个看待看待世界的态度问题。当然，你像你看像美国，这如果我在北美呆着，你讲美国是非常分裂的。对，这个国家，是吧？你是分裂到快都快打内战的感觉。对吧？但是呢，科技确实厉害，他在硅谷人才密度极高，他就是这么一个二元的一个一个社会。对吧？你看总统有时候发疯是吧？然后然后这个社会也是这样。哈哈，对对，这个但是他也不影响整个这科技的。这个我觉得这挺有趣的。现在这个世界，之前肯定没见到过的。对对。
所以其实其实挺好的。我觉得你其实也给了我这个问题一个特别好的答案。就一个人的乐观呢，长期而言，就具体而言，就生活层面，他的乐观，一定是来自于他的一个成长的环境，他可能在一个有足够安全感的一个环境里面。但从另外一部分呢，其实乐观，对于一个成年人而言，乐观是一个人的一种选择。这种选择可能有很很多种，有的呢，是基于你的人生哲学，有的是基于你的就是生活的一个态度，还有的呢，就像 Indigo 说的，是基于一种对技术的一个信仰。我觉得这种技术的信仰，他有一种再往深层的，就是科学家的那种乐观主义精神，其实是一种特别特别美好的一件事情。我觉得就从这点来说，不选择乐观其实是一个很愚蠢的事情。哈，就是因为因为，咱们就活这么短。哈哈，不是不选择乐观，假如这个，这不好那不好。对对，这是这是对，说什么不 বলল这不好那不好，其实是一个很简单的事，也很容易正确。是吧？很容易找到正确。你看我说的对吧？你看我说的这个，这个不行吧？就是对，这个很容易正确。但是事实上，生命本质意义上来说，就是在荒漠中去寻找一些甘泉，去寻找一些微弱的生，微弱的幼苗。我觉得这是一个特别。
那 OK，那我们最后，我看对，我们就两个小时吧，就别太晚。最后 5 分钟，要么你来分享一下，就是我觉得你前天的那个演讲里面分享的，就是对 2026 年的这个十个预测，我觉得很精彩。要么你就给我们来一个这十个预测吧，然后我觉得这是一个很精彩的十个。我可能记不太清楚了，对，因为那个 OK 技术我不知道，这个，对，没关系，我们可以说一下。我简单说一下，我大概脑子里有印象。对，我觉得 OK，这第一个可能是关于我，我记得我分享的第一个是关于学习，机器学习的一个趋势的。是吧？就是我们叫持续学习嘛，持续学习。这个是 2025 年，是叫做 agent 嘛，就是我们叫做推理嘛。然后呢，所有的模型，在 2026 年，他们都来研究，让模型可以动态的，不停的自我演化。这样一个，这是很重要的，是研究方向。就持续学习，最好的这个团队就在捷米拉，刚才刚才我分享的阿尔伯塔大学，对，就是干这个的。对，对，这个大神在那，是吧？这个 Richard Shadden，对，这个对。
然后第二个呢，就是我做了一个预测，我说这几个模型的公司竞争，我觉得 Grok 会赶上来。只做了一个预测，我说之前我 2025 年，我说 Google 年底会翻盘 OpenAI，还真翻盘了。这么的哈哈。然后呢，但是我没法说这个 Google 能超过 OPPO，超过 Google。但是他至少会在这个里面表现非常亮眼，在 2025 年。对对，然后 2026 年，对，然后二零二零二六年。然后呢，这个这几个模型公司，都有自己的演化方向了。对，就是你看 SOP 那一家，他叫重点干白领嘛，是吧？干代码和白领，其他的他就不供，他也不动了，动保态他也不做了。对，open AI 可能得改做一些收缩策略，这我说的他，不然他贪婪，铺太大了。对，对，这是一个问题。
然后第三个，我就还预测了一下关于哦，这个 agent 电商元年。就是就是，我们现在搜索里面都集成，你看阿里已经做了嘛，你可以在里面叫外卖，点什么东西。对不对？对，然后呢，就是可能会有 50 亿美金左右的这种，这种广告或交易量会来自于你，在这个聊天和和这个 chatbot 里面聊天的这个对话里面呢，产生的购买就是 agent 自动去给你买东西。就是这个，这个东西会成为一个 2026 年，至少再发展个几个月，一个季度左右，就能够成为一种大家一种一种习惯了。对，慢的演化成一种习惯了，在这个里面去购物。所以说这个东西是一个很，因为那个麦肯锡预测，不管怎么样，电商他擅长的交易，规模还是世界上最大的。电商才是世界上最大的生意。对吧？这个 commercial 嘛，是最大的生意，交易嘛，是。所以说对，他服务于最大生意的，这个这个聊天工具搜索引擎这些东西，他肯定是 AG 呢，这个是要吃最吃最大的市场是吧？这个所以说这个是非常重要的。对，这些公司的大家如果做投资，你看 Google，看 OpenAI，它到底能够在 Agent 支付里面和购物里面怎么显示，体验它怎么赚钱，广告怎么出现，2026 年是一个很很有趣的一个年份，我们看得到很多东西。
然后呢，2026 年，有那个说这个我还有预测了一下这个，这个叫做深层次的这个世界模型。对吧？世界模型，这里面会有一些，会有一款深层次的模型游戏出现，2026 年下半年或者年底的时候。那不知道什么游戏，他可能会刷眼球，很酷这东西。哈，不知道哪个公司，但是我觉得 XI 的 game studio 应该会发一些产品计划出来，至少要就看他今年的模型进展怎么样了。这个因为老马也藏，藏了一年多了，这个东西。哈哈，对，所以说我还是很看好，因为我特别看好这个 AI 驱动的互动游戏，它会超过电影的所有的总和。对，因为我们以后会生活在这种世界里面。这个大家对，其实一个小细节，有有有一个小的细节，你说起这倒是有小的细节，那个 lets face on 上面一直有游戏就是一直有，但是他一直藏在藏在角落里，现在很很少，但它一直在做这一块。哎，OK，它在做，它在 Windows 版上有，你可以进去很多游戏同步，可以根据里面。对，然后还有就是很这几个我记得比较清楚，因为写的有点多。
然后那个 a 科学叫 AI 的，这个 for science 才能够攻克一个真正的通过他自己发现，自己研究，自己做调研，完成一个真正的科学发现。2026，对，这是我预测，至少完成一个。对，这个，这个是我，而且我上次整个新年分享里面，我特别看中，我觉得 AI 的这个 c 端市场，你不要跟巨头抢了，不用打了。更多的机会是在更多前沿市场上面。因为那边的这个，大家对 TOKEN 的消耗的经济价值不一样。你说我搞一些博士生研究员，那我花这么多 TOKEN，那我产生经济价值比个人，我给你天写文章高多了吧？对吧？对不对？就你就你那两个问题，你怎么可能有科学家的问题值钱呢？对吧？这是不可能的吧？对对，这个经济价值都在，都在这个更前沿的这个道路上面，企业科研，这种各种世界模拟上面，那个他的经济价值更高，而且能够真的解决人类生存问题，或者解决以后什么战争问题。这个上面会更多。所以说这个投资的视野要打开，这个很重要。当然美股上这种选择很多，各种稀奇古怪的公司都有。这是，这是一个。
然后还有一个预测了一下，可能有一个关于 a 键的智能体的网络攻击。是吧？这块预测了一个可能会震惊世界的。哈哈，对。这个还有还有一个是关于这个，噢，还有一个是关于国家的，就 a 建中，就是 AI 中立国会出现。因为有的像美国中国 AI 都很强大嘛，是吧？有自己的模型，有些国家没有这东西，他不能够把我自己的数据设施，智能设施都放在里边，就核武器一样。对对，必须是对我，我各我各方面都要采用，我要 balance，我要平衡嘛，就跟现在印度去买美国的武器，买俄罗斯的武器一样的。对，这种这种东西是一个 2026 年，应该会至少会成为一个很现实的问题，很多国家会有这种策略，也有很多新闻出现。因为 AI 再发展两个季度之后，它的这个实力，就基本上是到国防级势力了。对，这个这个，我觉得会会是一个。
我说还有什么其他的，我可能，哦，还有一个 2026 年将是 robotaxi 真正落地的一年。这它会大量普及，在美国很多城市都会有。对，然后在这个里面，这个行业，这是这是行业信号，就是自动驾驶成熟。对，然后呢，机器人会在 2026 年会有一些会有一些进展吧，但泡沫不一定。在 2026 年产生这大量的机器人公司会出现，但是泡沫不一定在 2026 年爆的。但机器人这行业里面泡沫是非常多的，但是我全部都是用投资视角分析的，因为我主要是做这个这个行业来观察的。然后其他的什么政治的经济的，我就没有做太多预测了，而且有时候我预测他的政治经济的，我在微信直播里也不好说是吧？哈哈，对对。
是非常非常好。对，特别感谢，就是我们今天差不多两个小时，然后其实有特别的特别大的信息量。然后我们稍晚的时候，也会再做一个整理，然后看能够生成一些文本，能够和大家能够分享。然后我们的那个助理同学，也在这儿分享了 Indigo 的 x 和 YouTube，大家可以关注一下。然后呢，也再请再次请我们的那个，我们的那个那个社群的，我们的富丽花园的小伙伴，加一下我们的助理同事的企微，这样的话，就后续的，我们会把这个和 Indigo 的一个对话，就是会有一些整理能够发给大家。其实大家我觉得，应该能够特别能够感受到，为什么我们 2026 年的第一位嘉宾就请的 Indigo。因为我觉得他一个呢，就是 AI 是现在改变世界的最大的一股力量。第二呢，就我觉得 Indigo 呢，就不仅体现出来的专业，而且就我觉得特别他身上的这种乐观，我觉得其实应该能够给到我们特别深刻的印象。因为觉得这些年，大家都觉得你看就是就好多东西是吧？地产没了是吧？然后呢，各种的裁员，然后下滑，各种都是这种新闻。但是我们可以看到，其实这个世界，如果我们拉长了来看，可能是它处在一个，其实是一个特别美好的一个阶段。很多问题正在被解决，我们的寿命变得越来越长，就我们的就就物质程度，就技术进步程度而言，我们的幸福，处在一个特别加速前进的一个时代。相当多的，我们的感受的不是那么好的一些了，其实来自于就是人类的那种，有一点过于狭隘的，就包括我也有，每个人都有的那种过于狭隘的一些情感或者等之类的。但如果是在一个大的时代背景下，如果我们从一个乐观的角度，我们相信就是技术的一个变革，正在加速的让让这个世界变得越来越好。就我觉得对于个体来说，其实特别特别重要的，就是我们要参与其中。
其实就说起一个特别好玩的事，Indigo，我跟你分享一件事，就是特别好玩。就我原来还在做地产的时候，就一次呢，就我看一本，一个美国人写的一本书，然后他这本书里面呢，就引用了一个人的一句话，然后我当时说，哎，但是这个人呢，毫无名气，就是谁都没听说过他，我周围没有一个人听说过他。但这句话呢，我觉得说的好好。因为那时候就是我的那个地，地产公司，就我们就是有那种内刊，就我就每个月也写一篇，我就用他这个做了那个卷。收益就早到什么程度，大概 0304 年，就是很早。然后特别搞笑的，你猜这个人是谁了？就是 Elon Musk。就在他很早很早就还没有出名的时候，他说了一句话，什么话了？就是在互联网刚开始，他说了，他说互联网是一个正在改变世界的，一个特别巨大的一件事。他说了一个他的态度，他说我的态度特别简单，我不管怎么样，我得跳进去。就是他说我得投身其中，我不想做一个旁观者。我觉得也是一样，我觉得在这个 AI 的时代，其实我们应该快乐一点，像 Indigo 这样，乐观一点，参与其中，想一想他对如何能够让我们自己能够变得更好，我们应该去享受他，享受他给我们带来的美好的那一面。
OK，那我们今天这个就到这，好吗？就是，那我们就谢谢 Indigo。好，再见。再见，拜拜。拜拜。新年快乐。新年快乐。新年快乐。谢你。我怎么哦，结束连麦就好了是吧？对，你有录像，是吧？有吧？应该会有。都会有完整的录。我到时候试一下，看有没有可能把它弄成一个那个播客。对，哎，你你有试着计划弄一个播客或者什么吗？我可以，没问题，我们就是可以做的。这很好，这一期做的播客挺好的。两个小时有点长。对，对，现在我看那个像那个小宇宙什么，他们这都挺长的，一两个小时，两三个小时都特别多。就是你可以，你可以试一下看一看。对对，对，是。OK，那行，那我们今天就到这。谢大家。没问题，如果你方便让我发，我可以发。没问题，对，我可我这有群，可以可以。我可能也会去那个到哪去开一个或者什么之类的。好，谢大家。然后大家就欢迎大家关注 Indigo 的 x 和 YouTube。然后请大家加我们小助理的。哈，那我们就今天就到这。再见，拜拜。拜，拜。