Anita:哲学问题没有标准答案,AI 性的讨论默认 AI 拥有控制权——但这只是当下人类视角的投射
🌐 This transcript was automatically translated to English from the original.
When I was making the introduction, Song Duotao, this time I was also invited by you to cancel the offer. Let’s come and chat together. I am in Focus Media. I am one of the six people working on AI in Focus Media. I am also working as an AI development resource company outside, and I am also doing AI marketing in Chaos. Okay, I can introduce the current one. Hello everyone, I am An Tang, I am Liu Zeren from Level 3, and Level 3 is a community and resources that support software AI. We have mainly come out to return to the Valley, and got Petrocell, Pentera for stuff storage, and recently Dragonwork’s Brandtl, which are all relatively large. In the future, I will personally help some ASRs to do some AFL or some peer-to-peer signings with classmates and funding institutions. So if you are looking forward to it, share it with everyone. Hello everyone, I first started working in government public relations, working as a lobby, and then I worked as a lobby for about thirty years, mainly working as a lobby for the US government. Then I later came into contact with the technology industry, met my current partner, and invited me to join EVA, so we are now working on projects like this EVA. Because I have done this before, I may have had some exposure to various industries, so I really have to learn a lot. So when we chat later, we can also capture your thoughts. Yes, then on your side, let me ask the first question at the end. good. In this way, I am working on many technologies, which we can hear together, right? There is also a news public relations, because there are many, in fact, I am also, my most background is from the media to the literary side, and then if I am in Groogle or Btl, I am also doing content, which is actually needed. So what I am actually saying is that this year I will also use the OPCLAW, which is a sports set that can AGI. In fact, I have been thinking about it, that is to say, the capabilities of AI are getting bigger and bigger now, and we can all feel it. So there will be such a thing, that is to say, whether there will be a subjectivity issue that is considered. Therefore, although I am a big netizen, I still want to say that I can start from a more specific point. That is, can you introduce yourself, at least if you are not exposed to AI, it is the first point. I think a lot of people are talking about switching to GPT and starting to get in touch with it. But when it comes to Edging, when did you use Edging, or Claw Co, or lobster? Yes, after you came into contact with it, what was your own feeling? Yes, there is a kind of Aha, where is the baby of the lobster? I really want to explain this feeling to everyone's science students. Then start with a smile. The first person to talk about AI was probably the same as everyone else, probably more than three years ago. OpenAI, then my earliest Chair GPT, but I laid the foundation of our conversational big science. It also basically laid the foundation for the earliest, and, it also basically laid the foundation for the earliest, and, I also love this form of our interaction. In fact, I say that now, conversational science is still the proportion I use the most. I am a Glock user who takes this seriously. Because I am here, I use Twitter as a user. So I'm a Glock user. Then we also believe that each different AI uses its own different personality. If everyone has been exposed to it, I would find it quite interesting. Chair GPT is like a very woof woof woof, should be this, a role. Always gives you too much semantics, too many of these. Then, maybe like Glock, his character is very cool and very similar to Musk. The date feels very firm as I speak. So the earliest one is used, the earliest I heard it should be. Then, using AI Vision, or OperCore, is probably similar to everyone else. During the Spring Festival this year, it became popular. Then I felt at that time that in fact, although I am in this industry, I might have been exposed to OperCore earlier than everyone else. But I haven't used it yet. Then during the Spring Festival, I saw the craze among people playing drums, so I tried to go back and use it. Then I discovered that because OperCore is more powerful, it takes things that we might have done before through distribution, and then does them together. In fact, AI will find that many of your previous jobs had a reason for your work. You can provide a frame of your own. Then maybe, I think you need to spread out the entire environment, or give your virtual drag again, teach it again, so that it can achieve such a process. So for me, it has become an assistant that I think will be more collective or more unified, to help me do some of the things I did just now by distributing sub-stories. So it's possible, this is my story about going from UIA to UIA. Until now, to be honest, I have not used UIA Code as much as I imagined. I discovered one thing about it, and what I can honestly say is that it was not the Oak itself that started its chaos, but its entire training process, and the communication process between me and it. After I just saw this game in China, I think I have more to do with this work. I think it's integrated into this work, but I'm very surprised because I haven't done a report like this. We just learned a lot. I may be similar to most of you. Maybe I have been exposed to it and used it, and then maybe half of me thinks it is nothing more than that. But I think maybe China will tell us that there are more people who can try it. So just like me, if you have some, you can use one more as you go forward. I think there was also GT AI, which was probably 22 years old at the time, and then it was mainly GT. At that time, the main model of this model was that GN was a conversational wave-shaped model, because the main focus of GN was the general version. But what I really feel is that this AI is called, or we also advocated one, which is AT. When the wave-shaped AI comes, it should be GT. Then after the Azirubic Cloud appeared later, there began to be some different changes. Then through the model, in one aspect of my work, I will use different AIs. It clearly benefits every aspect of my work. For example, the current Codex and this TT. In fact, I still prefer to use Codex. The main reason is because I am currently in China, and the Cloud is less likely to be codexed. Secondly, as for GT, although it is very verbose, after entering its model, I think it is relatively good in the domain you solve. But if you want to do some viscosity after entering and exiting, I think the effect it can feel is not necessarily smoother than Cloud. For me, I am not a backend, but a very light-hearted person, so I think Codex is very useful. Then Block and Plexity, these two are the sources I used to find this information. Especially Plexity, it only needs to find some Resourcing for you. Then at the time of Jiminite, the then Gnana, and the subsequent ASQ, when everyone came out, there was still no Apple. But I don’t know why. Recently, I used Genina to find these feeders, and the answers I got were all very wrong. So today I abandoned Genina and I used it for some expensive pictures and the like. And in terms of Bochan, around the end of last year, I remember Bochan’s large model DC Cuttiny Minimax was a big explosion, and then I put a lot of them into it. I switched to the big model of gambling, especially some of the principles of operation in Chinese and the speed of this market, because I have my own public account. Then one of my current processes is that I will use Chatttt, Grow and Platty to find an agent, and then call it to the production broadcast of this public account or the Chinese production and broadcast reasons. Of course, there is also this Minimax, which can also be used, but I may be more dependent on this DC. And there is another one because I just saw this Toy, can he still hurry up? He said that he typed the 8 characters by hand, whether it was an EP platform kill. Then I also used DC 738. Then the fourth one I saw was an ordinary baggage man who had the best understanding of Chinese culture. Then I discussed this matter with some classmates, and then went to see how the Approof of this was done. I don’t know if you have done it. In the first half of the year, about half a year ago, there was a very broken one. Of course it was a Word3 factory. The one I came to was Li Sheng from the K world, and then theirs. Falter is a full-line approach to figure 8. I also discussed it with him at the time. I said this wave power model. After the step-by-step evolution, Ross's understanding of this complex one has exceeded the scope of this ordinary beginner. Of course, the premise is that your figure 8 should be queued. Who is the queue? It was in 2024. In 2024, they had just moved to Singapore. They came to Singapore with Yishou. Then I think the only one left was Moza. Most of the brands were used to make Low's Approach. Because at that time, Approach was like now, and Katumd. At that time, you still have to know about the phone behind it. You still have to know what is the problem with that program in this and that code. So it is easier to use than it is now. But at that time, it basically took a lot of time to use it. I spent half a day on speeding up the website I wanted, and then including it, jumps, and domain names here and there, and so on. The advantage Manderson gave me at the time was that I felt that I didn’t need to adjust to other major departments. I only needed to add Manderson because it basically helped you go from this, the early one, this TRD, right, to the things in the past, and the domain name behind it. It’s no wonder that you included it all, so you can complete it in ten million. And then I came back to this Over Club Over Club because of this big parade, and you definitely don’t want to miss this opportunity, but I participated in many activities at that time. I personally think that the function of this market is greater than the function of technology, that is, most people feel that they don’t want to be branded. Then I asked everyone to take this. I went down and thought that anyway, the answer to 80 may be some relatively simple tasks, because you are really on the ground. In your work or study, you don’t really need some process capabilities that can be grabbed. Or it will come out later that if you really want to play with your OPC, it still depends on your own business capabilities and your expansion of this dimension. So I personally feel that in my current work trend, I do not particularly rely on an agent. I rely more on different representative models. What are the strengths it targets? Then although I virtualize in each link, it is different. I asked you to do a cross-validation. This may be me. Currently, Wallflow is used more as a habit machine. For the agent, I was the first. The Cloud Code I was using was a good friend of mine, Imi. I asked him what was the difference between this Cloud and Cloud Code. Then he said that Cloud was a large-scale AI, and his Cloud Code was an Agent. Then he said, let me download it myself and give it a try. Then, the Cloud Code at that time was that you could only use it after subscribing, and he used it very quickly, so I might have given up on the Pro at that time, but it was indeed amazing after that. I wanted to buy it. It was so amazing. Like Operol and this Hermès, I have actually tried it for a while. Now, for my theme nationality, I still rely on Code because Pro is really good. Now it is basically the agent that I run the most locally. And the following words are currently asking my friends for some advice on some hands-on scenarios related to IMP Agent. But specifically, I feel that no matter which Agent is used, it still believes in serving everyone. So if there is business, I think I can get one. Maybe an Agen like IMP or Xiaolongjia t Because he can help you with business 24 hours a day, but because I am now, for example, the self-media verification mentioned just now, and there are some GNVs, I will find that, especially when he is going to do the promotion work, it would be more suitable if you are still on the profit and end, then I will let the profit and end. But for example, I have to do the deployment of the new QO. I am now directly connected to my graph through my TroCode. Now it is the most convenient, so it is a completely distributed scenario. I hope not to be kidnapped by tools. If you want to use a cloud server or something like that, it all depends on the business and then see if it is necessary. I think please be patient, so I won’t go into details. It is the result of TroCode. Because of my identity, on the one hand, I am an AI user, and I also need to develop and deploy the company’s AI. So, in fact, for our scenario, we will find that in the company, let’s talk about the company first. I heard about it once. In the scene, there are many more people who should know that they are doing crown-level reports. And because our boss, everyone may also More or less, it's a bit ugly, right there with Mr. Zhang. One of our biggest holes in the company is that Mr. Zhang is a 100-point player and most of the sales ladies. He may be a 60-point player. We need a clone. Right, now we need a clone. How about we make Mr. Zhang's ability available to most of the sales? Then we, in fact, this thing started to happen. I just had this RAG system and then in the end, the Edge AI method. But in fact, you finally found that people like Mr. Jiang are very successful. For people with very high confidence, you don’t need to go at all. You use a good business model and then you use this prompt word to inspire and then add some warnings from the knowledge base. You can very well go and revive this knowledge system. Just a simple sentence. You are not enough now. You really haven’t said it. Part of it is like this. Yes, and then there is another part because of this. In fact, as a company, you are still very complicated because you have to give things that you have to consider. You have to give them to the people of Jerak for them to use. ENGENIC's understanding of this Flow is definitely something that everyone here will not think of. In this industry, it is understandable that they hope to have a better-used bean bag or a chipbox that is more understanding of the weight than buying a bean bag. So this part is the construction of the company's AI. Then, for me, in my own words, in fact, I may be more centrist than others. A process that I have used for a long time, like Person or IDE tools like this, I think. The tools of this IDE are actually very suitable for the middle ground among chipbox and real agents. I mainly say this today. If you don't quite understand the contact arrangement of the IDE and the invocation of its tools, in fact, when you get to the agent, you will find that you are completely dragged away by the AI. So you are the AI, which helps you complete the so-called business and handle it more mindfully. Your own understanding of this matter actually has no way, very good. I think this is the case all the way. A very key thing for everyone is that maybe I am an inexperienced person. Maybe I am in my early stage. I want to go to PRD. I have to make my energy consumption into a skill first. But for most people, it is taken away by AI. So I think it is still very important to use IDE to understand this colonel arrangement. And speaking of it, in fact, whether it is Ownclaw or Amari Amari, it is indeed not useful. I have no say. Ownclaw, I am a you platoon. It costs 30 million yuan, and there are many people. Then we made Xiaolong, Xiaolong, automatically publish, and let him review the content by himself. It is my own feeling. Do you want to put it in a real review, in the outer realm, or more humanistic. So the current Ownclaw is not very useful, so I can do it myself. I rely on Steel very late at night. Then I use Ownclaw as a Steel people-inspired launcher. The so-called commands have been injected into it. You can ask me instead. Okay, this kind of command will be fine. I don't rely much on the problem of Ownclaw itself, but rely on myself. I also have some reminders of freedom. It's almost like this. When did you first come into contact with Peniton and start using Peniton? Is it Cuzzle? I think the real one is slowly starting to understand. This time it started with Cuzzle. Last year in Shanghai, so before that, I think it was very popular in Yezhou, Taiwan, last year. I think this, I can only talk about Clo. I can’t say that it’s been a while since I contacted Peniton. I was preparing for it in Hong Kong. We held a Jianghu. I held a military event. At that time, I was already on that social media. It was mainly recommended. I saw a lot of people on it. That is to say, OpenClaw was done. It was a multi-group talk. Any priest actually wanted to hear about it. Unexpectedly, I also mentioned that I was the Mobo. It seemed to be here for many months. I thought it was very miraculous at that time, because before this, I believe it was the same as most people, that is, I went over to chat with him. In fact, I am also a paid user of Change ET, but I think it has been very miraculous, just now. I also talked about that Luo Song, and I think he is very powerful. You know, the things he gave are very mediocre. I said, so I signed GPT at the beginning of the Spring Festival. Then I gave the Cloud, which is 20 US dollars per page. After using it, it is all paid for the side fees. Because I also noticed that at the end of last year, mainly 30% of us, because many of them are action users of Apps, they will put a lot of those. Go share it now, how is it easy to use, and then that, that, that, in the deep end, and the original experience, after I did it, I felt, indeed, I think, for example, when it came to Cloud, I felt that it was indeed clicked on the Change button on the subscription fee, so the whole thing, Cloud said this, his part of the problem is to first collect the analysis of the problem, especially I think I am a Chinese person, that is to say, I have been working in writing for many years. I met a media and then they were all public relations people. In fact, I definitely have requirements for words. In the past, I felt that dark articles were very trustworthy. Although others also had to spread a lot, I think what Cloud said was very clear. Of course, I also read a lot of them. Sharing means that it cannot be simple, that is to say, throw some away, and give it to him in simple words. Because you give him that income, you share it very well. If you use it, he will output it to you, which is also very long, so what I also learned is to say how to talk to Cloud, just like what China just said, call him Cloud text, just go and have a conversation with him. I asked him to first throw all my writing, writing, and making things to him for about ten years, and let him help me analyze, and then learn my unlocked habits, styles, and characteristics. I think his analysis has special principles. The analysis means that I have not noticed anything special. Then based on this analysis, that is to say, I was giving him a suggestion. At that time, I thought OberCloud, I said you, I will search it and say that I write these things. If not Cloud, who is the most dedicated person for three months? Then he collected a large number of cases for me. I also know that he collected them for me. I felt very good at the time. Then the articles we published will be many next time. I think this article really means to help him understand. In fact, it also helps me understand. Without Cloud, I believe everyone knows. It's a big tool factory, right? It sells servers, right? It sells servers, right, like I said, and there is a space bank that our BKN is very famous for, right? It means that you immediately take this thing and go to make that lobster machine. And that is, it moves very quickly, that is, this kind of development, that is, taking courses, and so on. So, it is very clear. Of course, there are many people who say that it is used to choose sex supplements. In fact, after analysis, it is not reliable, so I am right. I wrote it at the same time. I also installed it myself. I crawled to install it relatively early, specifically to buy a LifeMini and finally installed it in February. But after I installed it, I actually felt that it was the same as many people. I felt that I was very struggling, so I analyzed it. Maybe the model I used was inferior, so I did the same. I heard KOS recommend it and installed the MiniMax, right? I just said that the MiniMax seemed to be a good tablet, but after actually using it, That is to say, the performance is relatively poor, that is to say, so in fact, my main thing, which is the most important KPI that I have always used, is that before I uninstalled the OverCloud, it was not on time and completely on time. Our requirement is to push me an explanation or weather report at 7 o'clock every morning, but it was not successful. Then on Taozi, it took too much time. In fact, my OverCloud said on my Backend that it had been uninstalled three times manually, and I uninstalled it. It is a very thorough one, that is to say, this topic is all uncertain for us, so I format it and then re-recruit, so I mastered my transformation in OverCloud, that is to say, I installed it to move until the firmus. Of course, I feel that after changing to a TQB model, I still feel that indeed, it finally passed. My first level is to be able to complete TQB. It now tells me to rush TQB at 7 o'clock in the morning. Attendance suggestions can be sent to me. That is to say, Regularly looking for digging points is a comprehensive news report at 7 o'clock in the evening. In fact, give me a briefing. It was a sentence or two at 2 o'clock. The current progress is good. I think until the past, I can only start with my command. That is to say, this just feels like Hanlon is indeed Clock. That is to say, I use one of its native APs to develop one. My first web product is also relatively simple, which is foreign language search, so related news, that news. Telegram Push Channel Later, I now see various methods. That is to say, I think why everyone is particularly enthusiastic about Oper Cloud because it does. Secondly, that is to say, it is local group data. I believe that after Keyb, in the three circles of the web, we actually have a stronger, clearer or more common awareness of local and privacy. So and I think that in the future, everyone will The more important personal need is definitely to put it locally. It is very critical. There is also a recent example. There is a small company overseas with about fifty people. They put all their electronic workers on the Cloud. Suddenly, the account was turned over, so the entire electronic company's business was turned over. If it was there at that time, I mean this has become a major issue like Oper Cloud. All the data and workers are placed locally and run. So, I just think, that is, no matter what it is now, it is like I can talk about the job of Hermers. Yesterday and the day before yesterday, I became a new type of podcaster. I was very happy and I finally understood the objections to the development of Hermers Agent. In fact, it is also your original objection. They originally started from the community and worked together like an interest group to develop and play with various new and open-door models. Then they started to do it and then developed various new things in the open source model, and then finally launched the Hermers Agent thing. So I thought, no wonder I uninstalled this Oper Cloud and installed this thing. I feel like I should buy someone and put it there and use it. This is what I am talking about now. I am currently talking about using Agent. An explanation said that we just listened to the understanding here today. Especially compared with the weekend, I feel that it is still a few times worse. I am a little bit behind. I am not ordinary yet. So I still say that if it can be used for creation, it is still a month, right? You use it to chat and then put the content into other things. Actually, let’s just say it’s not complete yet, so it can really be turned into a content partnership. So my second question is, when you use this kind of Agent, does it bring some cognitive or emotional feelings to you, that is to say, will you write a kind of excitement or support excitement? It’s called Yue Translation. I’ve heard about our products. Let me give you an example. Well, the development of Agent on our platform has gone through roughly these stages. The first stage is because we are a social platform. As we have just introduced to you, we hope to be able to generate personal avatars on a social platform like ours. This is called generating personal avatars. In the earliest testing stage, we spent a lot of effort and effort to have some. Now everyone, this thing can think far. In fact, it is a bit like everyone now. It is more popular. It is said to be upright. This is probably what we were doing at the time when we broke half a year ago. It took a lot of effort to own Trump, to go mainstream. We invested in him, all of Trump’s past tweets, all of his public speeches. The Trump family likes to write books very much, and they like to have his various public works in various and various specialties. Whether it is videos or we have pulled out all these materials. I think if I recall what Trump had at that time, if we put it in any word, it is probably at least tens of millions or more. To correct Trump and let this agent make Trump more like Trump, what was discovered? About a while ago, like two months ago, 61 days ago, then I asked him. I asked him, he said, what do you think? It is beautiful. It is beautiful to launch a war against one thing. Trump was very surprised and told me that you are definitely Fake News. It is impossible on my watch. It is impossible for something like this to happen because the United States must be given priority. I have expressed it many times if I am second. If I had Second Terms, I would definitely not do this. I would definitely not start this kind of war. I don’t know where you saw it. So this is Fake News. This Trump is not something we spent a lot of effort to come up with. But he told me that my answer at the time was like this. So you were giving us the first thought. At that time, I just mentioned that you need someone who is particularly like you, and this you who is particularly like you. In fact, he will give it to you many times, just to bring it to me. The result of this kind of seriousness is that he did not make a result that would really be like you. So what is he affected by? He is affected by many relationships around him, right? He may be affected by this guilt pressure, but he cannot be affected by his son-in-law. He cannot be affected by many JD Vance around him, right? Rubio, it is possible, or in some such situations, he is in such a relationship, and he will handle such decisions, and these things are things that we cannot leave alone. It's because we can't judge. If there is no way to make such a prediction, so this doesn't give us much. One idea is that we can't do such a thing. I can't pay attention to the end. I have to pick a 98% Trump and a 999% Trump. Even if I end up pinching a 99% Trump, it can still happen. After he studies in the past, he will find that, yes, Trump will never do such a thing in his second term because he has promised the American people. After he succeeds in the election, he will not develop war with foreign countries. This is Maga. This is Maga's core idea, right? So we enter the second stage, that is, we need someone who has Trump's personality and Trump's personality traits, but has other characteristics. A person like Trump with some skills is also what we give him. For example, chat and search Zhang Jianyao to get the latest news for analysis and interpretation. When combined with Trump's methodology to do such a thing, this person suddenly became enlightened. He felt something about this matter. He left. He walked around the difficulty. After using sgill, he was able to roughly give one. Maybe we were relatively satisfied and able to analyze, so after the second stage. We will think that we need an agent like this. What we need is not a Trump who is more like Trump. What we need is a Trump methodology. What we need is a Trump methodology that can predict some things and at the same time give those skills. He may give an offensive comparison, that is, a more desired answer. This is our second stage. When we were in the third stage, we discovered an interesting idea. Should we give Trump some other, more interesting abilities or Trump? He himself didn’t even think about the abilities he might have. This made us think it was more interesting, so we created a person called angry complaining Trump. This Trump does not output any political views or conduct any political analysis. He only does one thing, which is to complain in a Trump-style way about all the things given to him in the first place. He is a executioner. All his ways of complaining are Trump-style, this executioner style. Then we went to do something like this, and I suddenly moved. My friend said, does anyone know? I said that this Trump can actually go to the Shirane District Association to speak, because I don’t know that the Shirane District Association is a Trump. Among the projects in the Arctic, one of them will be interacted with in the form of a talk show. The one who had the incident the day before yesterday, such as the one in the T&C case, where he was killed by love again, that dinner event, so in this way, I will find that we have reached the third stage. This is 98% of us who do not need a Trump. We don’t need a Trump methodology to let him do Trump’s things. But if we have a Trump methodology or a Trump framework, we can give him many things to make him be a very interesting and fun person or let him be this very interesting Trump. Then he is in the hands of everyone. You can pinch the Trump you want. You can pinch a communist Trump. So I will discover this stage by stage. I will find that such a stage brings us some thoughts about the customer area, and then I think it is quite interesting. This is us. I enter our platform and chat here and there. It is me. I have some thoughts based on our platform. By the way, you will also have some thoughts in it, which is you. Now it is what AVING is going to do now. The kind you have for others. You talk about these and other feelings. It is emotional, or things are a little more, or a newlywed mother, or something, I think this question should be more rational. Yes, I think this question should be Let’s dismantle this Practical News from a practical level, and first let’s talk about this technology and other technologies. I think this matter may be related to us in Xinxiang, because everyone may think that they have to have fun at work. Then I think this matter will make your work very good. That is, after the emergence of this AI, for special people, such as emoters, it has a very strong single segment here, so it is those very powerful emoters who will definitely be retained, and many companies They all recruit such talents for a day or two in order to promote the language. Then secondly, if you are an emotional person and you are a science and technology person, you will be stronger because you have a very powerful artificial method. Looking forward to abstract ability, right? Abstract ability, for example, you can even go to break through your abstract ability. I like to watch some AI machine-based things recently. You don’t know, it is not put into that kind of AI machine system. What is the pain show? Go watch the phenomenon of multiple shows, so I originally said that this AI is better than me, right? Then you can even go and do your own perfect pain show. Then I think this thing is all yours. Make a living and wait a little. It depends on you first. You have a job. You may feel bad about it, but you can use these pictures to make your this better shift amplification and long and short effect amplification. In the same way, you can only test two AI hammers in one day and its transformation. Now you can use your clone to test it. There are more than ten or a hundred AI hammers. It depends on you. Which strategy can make you transform? Okay. Then the second question is whether it has helped my life. It has made me very troubled about my life, so there is no trouble. A while ago, I was chatting with everyone about the plans to check first, and they were quite positive. They said that because of the emergence of AI, my life has been deepened. It can save a lot of time, and it can be played, and the AI can be used by others to help it make a product. It will come even less. Look at its AI dialogue, that is, can you help me take a look at this picture? Okay, it will start. I have more time to chat and start, so I think this is very good. Even it is a very good point. In fact, a classmate, one, actually, this one, pretend to be photographed. Like everyone sent a letter at that time, it is really very good. Even my mother, even my side, knows about this. She can see it on the news. I think this thing is very good. It means that everyone is excited about the same thing, just like the Internet at that time, but it exploded, that is, everyone will want to go I am interested in doing this thing. Then the second thing is whether it makes me very, very happy. In fact, from a financial point of view, everyone has seen that the EU AI is now opening it up for IPO, and SpaceX allows people to find these. This thing in the United States has a lot of sky-high prices. What kind of bad thing will this thing bring? That is, it will drain all the liquidity, which will lead to a financial pain or AI project that everyone may not be worried about now. From a financial point of view, it should be in a relatively tense state. Although now, even if it is A5, you can find some of this. Although there are not many, the scene is very good, but today is good, but it does not mean that today is good. So I think from this perspective and from this perspective, if you want to invest or invest, there may not be some. It is a pity, but you still have to maintain a nervous attitude, and when to buy and launch. I think this is very important, so I think in this era I have happiness and time to maintain the scenery of this time. I think this is funny because AI has taken your. Everyone has eaten this, which means that the technology has become flat. So there are many things, maybe this, you can do it in the future. For example, I wanted to start a big V before. It was actually very difficult, but because of this, the cost of running a big V has been reduced. You can use AI to pretend to be cheap. You can even have the same function. Then learn a few accounts and then go to this to make money. In the past People's researchers may have a special interest like this, but now they are not interested. Now everyone is a researcher. I saw this recently. When I changed one on Twitter, everyone's preference is actually very strong. Of course, I should also make some money, and the information is a little bit worse. So I think after having these things, it depends on how you use it, whether you can get along with it, and then do a lot of things that you may not have thought about before. As for the future, I don't know that you will not work for you. I don't think this matter should be too big. Because Xiaoyu, because he takes you to work, don't worry. He will definitely bring his work. Everyone is working on the strategy. Everyone has no job, so I don’t have to be too anxious about these things. So, it is more of a mentality. You want to play and then don’t sign with him. I think this is more energy. Thank you, that’s what I should say. My personal view on agents is that I really can’t lack them now. So I had a big sign of a server the night before. I don’t want to give him the server from one wrist and then sign to another. As a non-technical person, it just doesn’t work. You can imagine how nervous I was because his data was very large. Then when I signed the first one, I spent three hours on the performance. And the result was that I was using Aubus 4.7. It was very powerful, very powerful, especially for a mission. But at the time, I looked at the Yiyi, Yiyi up and down dancers, and didn't pay attention to this. It was stuffed. I wore it and stuffed it. Then he kept reporting an error. At that time, I was thinking, "No, teacher Gao, what should I do next?" So when it is visible, I am going to be a non-technical person and actually have to do other things. Then I get used to it. After you two give it to him, he suddenly becomes polite and then he may, may really want to be gone. So when it comes to this topic, I am, I am very angry, but I am also quite frightened. In fact, it is true that many friends have mentioned this matter of dark love and human work. I think that the situation that Mrs. Yi had no need to think about just now has been very clear and very comprehensive. What I might be afraid of is that on the one hand, I may have thought about the full-skin UBI for Musk, for example, but to be honest, I am not optimistic about this matter because the capital is negative. Then I feel that there is no capital and he is doing good things, so my UBI is not that optimistic. But it is true that AI will create many different jobs and even different other companies. For example, I know many friends from Liren now. They may One person, with just a few people and a team, can achieve 40 million or 80 million yuan. Before going there, it was very difficult to achieve in a certain situation. But now, it is able to unleash your capabilities, verify your business, and release it faster. You can say it is very fast, you can also say it is very good. I think it fundamentally depends on the physical nature of our people and the rotation ability. You are not very young now, and you are 100 times as likely as you were before. The second quick verification is that among the 100 times, there may be one that is twice as likely to be monetized. Then you can go to this version department and wait for it. But if there may be a hundred times before, secondly, there is no way to start faster. Then I think it is possible to start a stock market business. Then the company, a store contacted me yesterday. He said he wanted me to be an AI investment company, a Taiwanese wealth company, a community benefit, and then he said he There is no such person in his video. The term is very low now. He may be thinking of it based on the current subtitles. But your actual AI system only brings one talent. Will he still be like this now? Can he take the initiative to use AI? Is he still willing to work in that company, such as in such a small business, which may be in an enclosed area, so I think it is still in that form now, that is, a unit. If you want to play wildly, you can cooperate with him for a long time. There is a big thing. He will zoom in. You calm down first and it can also make you see clearly. There seems to be no more money on you. It is just a confirmation. After you get the countermeasure, you can continue to iterate and zoom in. I think it is possible to let AI handle your routine for a long enough time. I think no matter what form it becomes in the future, this is for the rich and for migration. So personally, I am very slow-control and if the demand is honest, I really don’t know what to do. But what I use is that I use this long-term AI rhythm. You are familiar with the platform, and you are familiar with yourself. This form has a lot to do with it. In fact, I am particularly affected by the sentence I just said at the weekend, and the sentence on 5G at the end. It is my understanding that the world of AI does not give you some people who are the first to start. My feelings are also very profound because I am a big private company. So I have recently discovered that the time I am doing in front of the office is getting less and less, because most of my work All can be left to Agent to complete, all can be left to Coco to complete, but I will also find that in fact, many people around me are not in this form. They think that when someone controls something very quickly, I am faced with having to do more things. So this is what I want, so I think because strange things are possible in the circle of AI, including the circle of OPC, so everyone will feel that it is a happy thing to be able to copy their own weather. I think there are many bad things. It's not that big. This is my first feeling, which is whether AI will do the work of the agent. Of course, I think I can give you a good example. Since I said this today, AI brings you such a big moment. I think of it. In fact, the moment that surprised me the most was the stage where I had the concept of Agent with System Prompt. And then, what I did best at that time was because I used to use image quality director. I also did it, that is, I wrote the script myself, etc. Then I suddenly did an experiment and said that we could use what I had previously done. I learned how to perform and craft every episode. Then, I gave them each their own character, Xiao Shuai, and then asked him to start once. I think this thing was the best thing to shoot before I came. Then he produced a script for me. And because the sitting I gave him was my own, the original sitting, and then let them both watch the two AI actors and shoot some more. Then I, as the director, force you from the side. They said that you need to consume a bit of rhythm here, and then you need to slow down a bit here to leave some emotional lingering. They all did a very good job. That was already 800 years ago. I think that in the future, the direction of public opinion, including the current pushback, should be in this direction. This AI problem-solving direction also has the beginning of many explorations. Well, this is very interesting. Well, it actually goes by very quickly. So I want to have the last words. You can also summarize it. It is, uh, because I personally feel that, uh, it is just some small details, like what was said very well just over the weekend. On the one hand, I think AI Keep training this, this, my feature quantity, or Sunday. I think it means that this has become a kind of constraint or restriction or dependence. That is to say, so what I want to say is, uh, if I relax a little bit and look here, I think it means that in the future, we will have more interactions with editors and the relationship will become deeper and deeper. Later, do you think our human's topicality and editor's topicality will increase faster, so this method must be developed and everyone must do it. The end of today's discussion, Andy, hahahaha. During the week, I think um, just now, just now, just now, just now, just now, just now, just now, just now, just now, just now, just now, just now, just now, just now, just now, just now, just now, just now, just now, just now, just now, just now, just now, just now, just now, just now, just now, just now, just now, just now, just now, just now, just now, uh, uh, to go deeper, it will bring not only equality in technology, it will bring equality in belief, it will bring equality in thought, equality in thinking, bring equality in belief, bring all, you can imagine, everything will eventually have equal rights. What does that mean? As we sit here today, I will have a theme or someone will say, talk about a liberal arts student How can I do that? There will no longer be a concept in the future. There will no longer be the so-called liberal arts students, liberal arts students, and science students. The core of the distinction between liberal arts students and science students is at that time. Well, I think it seems that liberal arts students have studied which subjects. Arts and science students have studied which subjects. In other words, liberal arts students have learned these subjects, so they have the thinking of liberal arts students. Science students. In the future, these words will not appear again, because in the world we have become, we will no longer Just think about the problem in this way, and you will not have this kind of thinking just because of what subjects you have studied. In this aspect, the selection we have achieved, there will be a saying called, you are a girl, so why do you think that you are a boy, so you will be more rational. This argument will not exist, it will no longer exist. It is like a boy's achievements in the past, because of what kind of environment he found out at that time, so he seems to have passed the fixed thinking mode in his heart. So we call it a male thinking. Or, it seems to say, men are more rational. This idea will not exist. Gender will also win. Come on, everyone, because with the blessing of AI, the information will be confused, and all our thoughts and ideas will be confused. All the things that will be produced in the future, what we call differences, can produce things that are unbalanced now will become more balanced in the future. So I think what AI brings to us must be in this regard, such as us as a person. It is more, more able to achieve attendance, and that is, the previous one, I want to say that in the general half, what everyone just mentioned is the issue of work, that is, the development of AI, will my job be replaced in the end? I feel that everyone is thinking about whether they like to work, everyone is thinking about working, as if no one is working, right? It will be like this. In the past thousands of years of civilization, most human beings are workers and producers. As this is the side that needs to work, and today finally There is such a moment, there is such a moment, you may have to participate in distribution, and at this time, we who are not used to it cry and shout, quickly put my sweeter Dezan on my neck. I will continue to be sweeter. Don't throw away this sweeter tool for me. I still have to work and I still have to make money. But I thought that with the development of AI to this day, it is finally possible for someone to work for you, someone to engage in production, and when you want to engage in distribution, don't do it at this time, in this relationship, let us think about my job. Will I go to my job? What will happen to my income after I am late? I think we can think more about this issue from the perspective of production and distribution. In the future, we humans will be engaged in production. In such a world, people with family names or what kind of mentality will engage in production, and we can all live a happy, equal, and smooth life in this world. We should enjoy a lot of things and truly participate in distribution. I think this may be the future. Is there a chance that AI will have a chance? Okay, let me speak first because I happen to have a place where I can echo and exchange ideas with Rentha. The first point Rentha mentioned is a very good one, which is the equality of our future thoughts, the equality of the original intention in the future, etc. On the contrary, I think these are terrible things because, have you ever thought about it, it is the world of Trisolarans, because why can he, right? Everyone is because he is a broadcast thinking, he is a trend thinking. I think The scariest thing about this trend of thinking is that it has given the world a very weak point, because everyone is the same, so you have no way to resist the black world in the future. So at this time, I think that all the big companies in the future may have to think about one thing, or how to make mine continue to maintain my willpower instead of comradeship. So I think of course we here must have a similar point of view as last time. In addition, everyone is from production and distribution. In terms of this idea, I will think about what people will do in today's future world. If I talk about things in advance, I may still go back to saying that people themselves have to make fun for themselves. That is to say, create some meaning for themselves, which is to create some meaning for the world. So the role of people will become I am a production house and a consumption house of meaning. Then I produce, I produce. For example, I am an American house today and give it to Teacher Guo for consumption. Then I may have created an art work today. I created a painting, created a program, and so on to consume. I think that at this stage, for the pursuit of meaning, he may itself transcend the pursuit of the worker. So at that time, I felt that the work was still valuable. If our possible goals were the same, I would respond OK. I think this is a philosophical question, so he has no answer. Then this is a relative question. When you are in any AI topic, it is tacitly accepted that AI can have some control over all rights. Of course, Do you really think it's because real humans, humans are in the early days of this world, so I think this question is quite interesting, because just now I felt ACO, I was talking about it, is it TU-Human or TU-Agent, it is a two-door world, I think I also think it is not available. But as a question before opening, I also want to bring this answer, but what I think is that in the future, if I want to work well with Agent, maybe my current situation is actually what I hope to be able to fill in. More about people, for example, our human civilization was like Iran, which exploded some time ago, and then the world cultural heritage in Iran is gone. So I think that human civilization like this is possible. I want to get in touch with more recently. I feel that when I don’t understand people, Agent itself is also a human creation, so I think it can be taken, but the premise is that we have to get along better with people and understand the management of the human world or a kind of friendship between people. So I think from my perspective, I can Through this result, if I am now a kind of person, I want to know more about human interaction, maybe later, maybe later. I think when the Agent changes the person, it will be more important. It is a worrying rule. Well, thank you very much for the guests we are doing today. I am also very grateful to our stage QSWall. The previous patience is to participate in the whole of our activities. Friends, today we are really the first issue, the first issue, a salon. It is a perfect explanation. I have really explored, long-term thinking and groping in one direction. That is to say, we and Edith are the human Edith. How can we interact with each other and develop? Well, then we have today’s content. The three of us heard that we also want to encrypt it, and the other heads of state just now. We will and will work hard to use Edith to re-produce the more condensed content. We will put it on our encrypted sect website, WeChat public account, video account and our Xiaoweizong podcast. Okay, thank you everyone. Thank you everyone.