DeepSeek recently completed its first round of external financing since its establishment. The total amount raised in this round exceeded 50 billion RMB (approximately $7.4 billion), with a pre-investment valuation of about 367.5 billion RMB (approximately $54.3 billion). In the investor lineup, DeepSeek's founder, Liang Wenfeng, personally invested 20 billion RMB, Tencent invested 10 billion RMB, CATL invested 5 billion RMB, while NetEase, JD.com, and IDG Capital each invested 3 billion RMB, and the National Artificial Intelligence Industry Investment Fund invested 1 billion RMB.

Prior to this, Liang Wenfeng proposed a principle of "no financing, no IPO, no commercialization." This large-scale financing marks DeepSeek's formal entry into the capital market and has drawn widespread industry attention to its commercialization path and technological vision.

At a recent investor conference, Liang Wenfeng elaborated on DeepSeek's organizational culture, open-source logic, technical roadmap, and views on the industry competitive landscape.

Below is a compilation of Liang Wenfeng's speech transcript obtained by Tencent Tech from a nearly 4-hour exchange meeting, classified by theme, totaling 118 points. The text has been edited for clarity while preserving the original meaning as much as possible.

Vision and Restraint

1. When we started this company, we didn't think about how much money we would eventually make, going to the capital market, or getting listed. The initial dozens of people never thought about these things; if they had, they wouldn't have come.

2. We started this company with a very great benevolence towards the world. We believe this is something useful to humanity, something beyond money. Our original intention, our vision, and the vision we have maintained to this day were not pursued in a way that maximizes commercial interests.

3. Managing a large company relies not on your rules and regulations but on vision. Vision is not a slogan on the wall; it's about how you do it, not just what you say—it's about how you actually operate.

4. We are not organized; we are vision-driven and organized by a vision. We don’t operate by saying, "I want to achieve this KPI, no evaluations"—it's only about the vision.

5. This vision is not even documented; it's not written down. We haven't written anything down. This vision is in our way of doing things, our attitude towards the world.

6. We don’t have many other advantages; we don't have any special talents, we are not wealthier than others, and we don't have better employees than other companies—actually, we don’t. When we started this company two years ago, we didn't have much money, cards, visibility, or appeal; we were just a group of very ordinary people.

7. The more we restrain ourselves, the more likely we are to succeed, or at least that has been evidenced so far, and at least so far, it has been logically explainable. Otherwise, there is no way to explain how we were able to achieve it: we had no weapons, a very low starting point, very few resources, and our people are actually just a random group of ordinary individuals.

8. AI is a huge matter, with huge benefits. We are very restrained, knowing that as long as we can achieve it, the benefits will be substantial. Just allocating a small portion will yield significant benefits, so there is no need to think about which part of these benefits to take or how to take them. I believe there is no need to consider this because the benefits are significant enough.

9. Last year during Chinese New Year, we suddenly had a lot of users, but we did not try to retain these users, or use these users for monetization, or seek to seize these commercial interests and cash in on the users. We did not try to acquire users, did not try to make money, but we worked very hard to find ways to serve the users well.

10. We do not have the idea of becoming the next super app, then competing with someone, or becoming the next ByteDance, or becoming the next Tencent. We completely lack such ideas. I think the future opportunities with AGI should be very significant, and the opportunities with AGI in the future will always be very significant.

11. Restraint is a strategy. It's about sometimes being able to give up some things in exchange for more others. Not open-sourcing is also the same, it can be considered our pressure, or it can be considered our concession.

12. I understand this kind of restraint, in the long run, can increase our probability of achieving AGI. When considering something, I have no doubt that AGI will have significant commercial value. So, on this basis, what I prioritize is not how I can increase my share, or how I can take more share, what I prioritize is how I can increase the probability of what I can achieve.

13. We have always been very restrained, unwilling to become competitors with any internet giant or small company. I hope I can empower them, or hope I can help everyone to do this, hoping to help everyone do this.

14. I think that by adhering to this attitude, we have not missed out on anything, we have not gained less because of open-sourcing, and we have not lost anything because of our goodwill or because I provided help to others. Instead, there may have been additional gains. This may seem counterintuitive, but it is indeed the case.

15. Our goal is AGI, but we have always been commercializing, that's why we have C-end users and B-end revenue. From historical experience, this strategy has been successful.

AGI Roadmap

16. If you can describe a problem clearly, with full context and instructions, it surpasses human capability. But there's a caveat: you must provide full context and complete instructions.

17. AI cannot replace your employees. However, if AI has the ability of continuous learning and undergoes a two-month learning process like your employees, then it can replace anyone in the world. Therefore, we are just one step away from the next stage, which is continuous learning.

18. The development of AI can be seen as a staircase. The step we took last year was the Chain of Thought. We discovered that through the Chain of Thought, we can elevate intelligence to a higher level.

19. This year's step is the Agent because we found that using the Agent approach, even more can be accomplished. Its scope of ability will be greater, and its intelligence ceiling will be higher. The Agent needs to use the CoT, and the CoT also needs to use the previous step, which is the language model, so no step is taken in vain.

20. After the Agent, we believe the next problem to address is continuous learning. How can we make the model learn continuously, not requiring intense training but being able to learn for an extended period like a human?

21. Following continuous learning, we might reach a singularity. This singularity occurs when a model can learn continuously and can perform all tasks that humans can do. It can develop its versions, conduct research on its own, and create more advanced artificial intelligence models.

22. This singularity is not an abrupt event; it is a gradual process. This process may also be a lengthy transition, not a sudden change. However, habitually, we tend to view it as a singularity.

23. This is our speculation, and this is the timeline we believe should be followed: first, solve the learning-to-learn challenge, then reach the intelligent singularity, a self-iterating singularity, and finally achieve Embodied Intelligence. With Embodied Intelligence, it will enter the real world, capable of household chores and eldercare.

24. If we first address continuous learning, then the self-iterating singularity, and finally Embodied Intelligence, the journey will be smooth. Because later on, you can use the earlier technology to aid in developing subsequent technologies.

25. We are solely focusing on the AGI mainline. The field of AI is vast, and there are many aspects we believe do not align with this mainline, such as 3D, video generation, which I think may not have a significant connection to the core of intelligence, and we will not pursue them.

26. When video generation first emerged, it seemed to be a must-have, as if you weren't a proper AI company if you didn't do it. So I was very curious because if you think about it carefully, it has nothing to do with the roadmap to intelligence.

27. From a business perspective, it's a good business. But this has nothing to do with intelligence. We wouldn't do it just because it's good for business; we would only do it because it's something on the intelligence roadmap.

28. From our perspective, the world model and intelligence are not the most important things at this stage. The most important thing is AI training and how to facilitate continuous learning after AI training. This is our company's view, and of course, every company has a different assessment.

29. We are more inclined to believe in a narrative where AI can accelerate AI research. In other words, it's not linear because you can use AI to speed up your own research, so it may become nonlinear in the end.

30. I think embodiment is definitely necessary, ultimately embodiment. Because for a normal person, their need is not a computer, right? Because for a normal person, they eat, drink, have fun, clothing, food, housing, and transportation; they don't need a computer. What they need is, so they still need embodied intelligence to address specific human needs.

31. What do we hope AGI can do? It can help me iterate on the next version of models, just like it can help me iterate on the next version of models. If, after embodiment, we hope that what it does is also to iterate on the next version of embodiment, to create the next version of robots.

32. The core ability of the next-generation model must have the ability of continuous learning for it to be called the next-generation model. Before that, what we can do is reduce costs, improve performance, and increase speed. But to make a breakthrough, it should have continuous learning.

33. The current Agent's capability is limited because it cannot learn continuously, cannot learn continuously effectively. If we can first achieve continuous learning, AI's ability is very strong, and it can greatly enhance the efficiency of our research.

34. Once continuous learning is achieved, achieving general intelligence may become much easier. It would be effortless to use it. That's why I say this is a result we would like to see, a more efficient approach, making it easier for us. Otherwise, currently, if you want to manually achieve general intelligence, it is a tiring and laborious task, data-intensive, labor-intensive, and not cost-effective.

Team and Talent

35. The insight from our previous experience is that the AGI vision is very powerful. This talent advantage is not about my people being smarter than others, but about how I organize these talents, how I motivate them, and then how we collaborate.

36. Gathering smart people together does not mean they will naturally cooperate, naturally be very passionate about pursuing a goal, and naturally be able to achieve it, so you need a vision.

37. Our primary core interest is to maintain team stability. This is our biggest core interest, and can even be considered the sole core interest. As long as I can maintain team stability, I can definitely achieve AGI, it's that simple.

38. Money is definitely not the issue, resources are not the issue, and other factors are easily obtainable. For us, there is only one core interest, only one that cannot be compromised: we must maintain team stability.

39. This is also a very big challenge we face, or I believe, the biggest risk. Of course, this risk has been greatly alleviated with our recent financing round. Because everyone received quite a lot of options, the amounts were quite substantial.

40. In terms of team stability, as long as the most important and oldest employees can stay stable, then others are less likely to leave. Even if others have fewer options, less income, they will not leave. Because they are not solely chasing money, everyone hopes to do this in an environment where AGI can be achieved.

41. Everything else is just a matter of time, at most causing us to be delayed by half a year, a year, but it will not prevent us from succeeding. We definitely do not lack money, we definitely do not lack resources, in fact, these are not lacking at all.

42. The main gap between us and the United States is primarily in resources, and the gap in talent is not that big. Regarding talent, there is almost no gap, because it is the same group of people, possibly Chinese. When Chinese people go abroad, some stay in the country, some stay abroad, and some go abroad, it's not like only smart people go abroad, that's not the case.

43. Talent is not the bottleneck, resources are the biggest bottleneck. Resources first impact talent development, because of the lack of computing power, we have fewer opportunities for experiments, so overall, our talent is somewhat behind compared to the United States. The gap in talent is fundamentally also due to the gap in computing power.

44. The shortage of AI talent is also temporary, and we have already seen a significant alleviation. Because AI talent is really not lacking, every company will quickly cultivate talent, and talent cultivation is very fast.

45. Currently, there are quite a few domestic companies in China engaged in modeling, perhaps too many. In the United States, there may only be three such companies, while in China, there are too many companies involved in foundational modeling. In the end, it is certain that there will not be a need for so many people to engage in foundational modeling, and it will definitely converge.

46. In our company, management actually follows two approaches: one is a top-down approach, and the other is a bottom-up approach. With the bottom-up approach, each person can choose what they want to do and proceed with it without supervision or KPIs.

47. Generally, we hope that employees have half of their time unallocated, allowing them to do whatever they want. This provides a scope for exploration, enabling individuals to explore based on what they consider important, without any predefined requirements.

48. We usually do not work overtime. There are two reasons for this. First, research requires a relatively relaxed environment. If you are under too much pressure, you cannot conduct research. Since it is based on your own interest and you need to think about these issues in your spare time, exploration is only possible in a relaxed environment.

49. The second reason is our strong focus. With a strong focus, it means we have very few tasks to accomplish. If I don't have many tasks, I don't need to work overtime. This aligns with the previous principle of restraint.

50. Our company is fundamentally built on consensus. I don't make all decisions by myself; instead, I seek consensus. My authority and influence within the company are based on consensus.

51. This decision-making mechanism is actually a consensus-seeking mechanism. This does not mean that I can push through anything as long as there is consensus; rather, I seek consensus first before pushing through with a decision.

52. As we expand our team, we will make adjustments. Actually, we should make these adjustments immediately because I am already in the process of making them. If we don't make these adjustments, many things will not be able to progress. Indeed, many departments should have an organizational structure.

Computing Power and Resources

53. How much hash power do we need? Certainly, more is better at the moment. Within our capacity, more hash power is undoubtedly better, so our current strategy is to buy as much hash power as possible at a reasonable price.

54. In reality, it is very challenging to spend so much money and buy a large amount of hash power. It's very difficult to buy that much hash power, especially when the prices are high. We can't just spend a very high price to buy hash power; we need to ensure that the price is reasonable. If we can spend 20 billion this year, then it means our procurement department has done an outstanding job.

55. The biggest gap between us and the United States is in terms of resources. On one hand, domestic mining resources are simply not available here, and on the other hand, our capital investment is less than that of the United States. We lack a significant amount of capital investment, and within that, the proportion of talent-based salaries is very low. You see, they may pay out a salary of one billion US dollars, but when you calculate it, talent salaries still account for a very small proportion, with the majority still going to computing power.

56. All the differences we see, including differences in talent, model capabilities, and applications, can be attributed to differences in computing power resources.

57. The gap between us and the United States may be that we are lagging behind the U.S. by 12 months, maybe 12 to 18 months behind, or let's say 6 to 12 months. In simple terms, we are two years behind the United States, and yet we only use one-twentieth of the computing power of the United States to accomplish this.

58. The narrative here is being two years behind but using only one-twentieth of its computing power. So in the future, what we need to do is rewrite this narrative to use a fraction of its computing power but compress the timeline further, down to 6 months, 3 months. I believe this is a goal to strive for.

59. Scaling, we believe in Scaling, and it is certain that the larger the scale, the better the effect, unlocking more functionality. What actually hinders our Scaling is the computing power, not that we don't want to Scale, but we do not have enough computing power to do so.

60. The reason we train such large models is not because I think such large models are enough, but because I happen to have this much resources. I calculate based on my resources, what size of model I can accept and train, it is derived in this way, not that this model is sufficient.

61. When Silicon Valley talks about reaching the limit of Scaling, that is for Silicon Valley; for Chinese people, we are far from that, we are nowhere near Scaling to that extent. This Scaling includes scaling of data, model scale, and training costs.

Domestic Chips and Ecosystem

62. NVIDIA's CUDA moat is rapidly being eroded. On one hand, now that we have AI, establishing this ecosystem is much easier than before because AI can code.

63. The market for computing cards is now larger than that of gaming cards, so there is no reason these two should still be coupled. The current trend is that they will no longer be coupled in the future. Thus, specialized chips, whether from Huawei or NVIDIA itself, will all be specialized chips in the future, not like the previous ones.

64. The replacement of domestic AI chips now has a historic opportunity. We believe that within the next year, we will see something validated: there is absolutely no issue with the domestic chip ecosystem. Previously, there were concerns that it was problematic, too expensive to use, or not user-friendly, but I think in the future, within a year, we can reverse this perception, or reality will reverse these views.

65. There are no issues with the hardware and ecosystem of domestic AI chips; the only problem is insufficient production capacity. There is no barrier for domestic card adaptation, and Nvidia cannot stop this. If in a normal commercial environment I can buy Nvidia cards, then it is relatively difficult for domestic alternatives to compete. However, in a scenario where Nvidia cards are unavailable, everyone is forced to turn to domestic chips.

66. During V3 training, it still uses Nvidia cards, but it no longer relies on Nvidia's ecosystem. V3 uses Nvidia cards but does not use Nvidia's ecosystem; instead, we first developed an advanced compiler called TileLang and then completed all other tasks based on the TileLang ecosystem, almost completely independent of Nvidia's ecosystem.

67. I am quite optimistic about domestic computing power. I think in this regard, Nvidia is digging its own grave. Huawei's super node, Huawei's 950 super node, can be a direct replacement for Nvidia's GB200 and GB300 in terms of performance and price.

68. Four Huawei cards outperform one Nvidia card.

69. The gap between us and the United States in chips, I believe, will not widen further in terms of the ecosystem, but in terms of chips, it will quadruple and take two more years.

70. We are mainly cooperating with Huawei now. Huawei adapts on its own, but we will participate in this ecosystem and be deeply involved in Huawei's ecosystem. Huawei's main issue is still insufficient production capacity.

71. I don't believe that five years from now, we will still be stuck on production capacity issues. It is certainly a bottleneck now, and I think maybe this year, next year, and the year after, we may still be constrained by production capacity, but after five years, I think it may not necessarily be the case. I am still quite optimistic.

Competitive Landscape and Industry Analysis

72. The ultimate gap in the effectiveness of various models should be a comprehensive one. When comparing model performance, it must be done at the same cost to be meaningful. Just as when you compare two cars, you also compare cars of the same price range.

73. Will Anthropic surpass OpenAI now, is this a long-term trend? I don't think this is a long-term trend; it is definitely a phase. In the future, OpenAI and Google will most likely continue to alternate in their ascents.

74. When it comes to leading the division of labor in global AI, Chinese companies are likely to play a role based on their potentially largest output. Logically, our production capacity is the largest, including chips, where our production capacity may be the largest, and we have the most electricity.

75. Chinese people will make this product the cheapest, and then focus on the performance. After all, many foreign products, at present, do not have much difference between a product made in China and one made in the United States. In the future, AI may be the same, but AI produced in China may be cheaper. This affordability may be a systemic low cost, just like how Chinese services provided in other industries may be cheaper.

76. The ultimate difference should be in three aspects: cost, time, and user experience. Apart from these, there may not be much difference.

77. Cost is definitely a differentiating factor. I think cost is probably the first difference. The second is time, when you can achieve it. If you are early by a few months or late by a few months, it will be different.

78. OpenAI initially believed it could truly dominate the world, but in reality, it will encounter many challengers. When faced with challenges, it will not be so easy. The United States will face challenges, and in the future, it may also face challenges from China, because Chinese people are willing to accept less in order to provide you with this service.

79. Those who ask for more will be defeated by those who ask for less. Even if you don't really ask for more, if your vision is to ask for more, you will be defeated by those whose vision is to ask for less. In fact, no one has received any money, it's just a vision. If your vision is to ask for more, you have already lost, and you will face greater difficulties.

80. For us, we are not looking to take the most profit, or seek to maximize returns, but only to earn a reasonable return. This is an explanation. I believe in this, I am not trying to find reasons for this, because there is no need to find reasons.

81. I think in many aspects of experience, we may be able to do better than the United States. In terms of product, our product ability may not necessarily be worse than that of the United States. Costs should also be lower than in the United States, so China will still be competitive.

82. Cost is easy to understand because they do not need to do it, so they do not develop this capability. They certainly do not value this as we do. We consider it a very important matter, but for them, it is not important.

83. Possibly, with large models, we may already have enough with not just two large companies or two small companies. The difference is only in two things: time and cost. Therefore, it is unlikely that any one company will make huge profits. Those who control costs well will earn a little more, and those who do not control costs well will earn a little less, and that's all.

Model Research and Technology

84. About half of our company may think that OpenAI is better on a day-to-day basis. In fact, Anthropic has a first-mover advantage, but this advantage should diminish quickly, as it is not something it can sustain in the long run. All three of these companies are very powerful, with Anthropic being the most efficient in terms of cost, expenses, and money spent.

85. We have been working on multimodal layouts. It is crucial for the product and for end-user products. However, for the ceiling of intelligence, it is a component, not the main focus.

86. We are likely to introduce relevant models, meaning our V4 and subsequent versions will support native multimodality. But for multimodality and for intelligence, it is a component; we do not consider it as intelligence itself.

87. Regarding the scaling of language models, I currently do not see a limit. We have not yet reached the intelligence level we have now, or that the United States has, so we have not seen a limit.

88. The idea among many of us internally is this: we need to be useful to ourselves first and foremost. That is the quickest way to achieve AGI. When our own use is effective, it may mean that others can also benefit, but first, we must ensure that we find it useful.

89. The models we create are not primarily intended for everyone to use effectively but for us to use effectively. It must be useful to us first. After it is useful to us, I will be able to develop the next version of the model more quickly.

90. We call this "Lucky Draw." The barrier to entry is very low, and anyone can try, but what each person can come up with, I may not know if it depends on talent or something else. Therefore, there is no need for us to allocate resources here. The difference between us and other companies is that we will spend time discussing this issue, thinking about it, and considering it an important matter.

Commercialization and Pricing

91. Our API pricing represents a reasonable profit, roughly equivalent to buying a batch of equipment back from the market and recouping the cost in ten months. I think this is a reasonable profit.

92. If the goal is to maximize profit, the price should be set higher. Because in this price range, user demand is inelastic. If I double the price or raise it even further, the difference in token consumption is not significant.

93. For one of our models, at the beginning, we were concerned about too much demand, so we initially set the price relatively high, which the team was not happy with. Later, I reduced the price to a quarter of that, and everyone was very pleased.

94. The upper limit of To-Business (To B) should still be demand. In the current era of AGI and AI technology, the demand for To-Business should be limited. It will grow rapidly, but it is not infinite, ultimately constrained by demand, not computing power.

95. I now think that if something is achievable, then we should aim for it. If, for example, this year I can generate billions of dollars in B2B revenue, combined with our user base on the consumer side, then we already have a certain business foundation. By next year, if our B2B revenue continues to increase, and if this demand can further expand, the company is not far from reaching profitability, and may even be profitable.

96. In the worst-case scenario, selling APIs alone could potentially support a publicly listed company. Even if there is no further technological advancement behind our current technology, and our progress stagnates, then focusing on selling APIs to the fullest and delivering these services well should be sufficient.

97. Looking at our current situation, I believe the most reasonable approach should be to fully focus on building a general-purpose agent, with other agent priorities lower in the hierarchy, including financial and medical agents. We should prioritize developing a Coding Agent because a Coding Agent can achieve many capabilities, along with many vertical agents. At this stage, we believe the most critical aspect is still the Coding Agent.

98. I think that achieving low cost is primarily a consequence. Our model has indeed always been moving towards a lower-cost model architecture, which is related to our vision. We also have many methods in algorithms where costs can be further reduced.

99. Another reason for cost reduction is that the lower the cost, the more significant models I can train, the more significant models I can afford. With the same amount of computing power, in a situation where computing power is limited, if my computational efficiency is higher, I can bear more substantial models.

Open Source Strategy

100. I think we will go open source, and our most robust models will likely also be open source. Because I don't see any benefits in closed sourcing, I don't see any inherent benefits. ByteDance's model is closed source, what benefits does it have? I don't see any benefits.

101. Even if the model is open source, and you tell everything to others, the barrier is still very high. It's very challenging for others to use it. Secondly, for them to use it, the cost must be very low, which is also very challenging, not so easy.

102. Open-sourcing will not affect revenue. I believe that open-sourcing will have no impact on our business model.

103. I am not worried at all about others deploying our model and then competing with us. We actually hope they can deploy it. We aim to assist the open-source community as much as possible in deploying our model.

104. When engaging with the outside world, our attitude is: we only focus on AGI. When dealing with the outside world, we are very willing to assist and help anyone, even our competitors, including Alibaba, Zhipei, and the Dark Side of the Moon, to do better. Because we don't lose anything, we were open-source from the beginning.

105. Is the open-source model we provide the same as the one we deploy ourselves? It is the same. We will not open-source a subpar model and then use a better one for our deployment; they are the same.

Data and Post-Training

106. Data should almost equal half of the model. There is also an issue with labeled data. Regarding data annotation, it is related to our capital investment. With the structure of our capital investment, we cannot afford the cost of a large amount of high-quality data annotation because the cost is very high.

107. The cost of data annotation in the U.S. is not much different from the cost in China. China does not have a cost advantage in data annotation, especially in annotating high-end data, which makes it difficult for us to invest in data annotation like in the U.S. This path is challenging in China because data annotation is too expensive, whether outsourced or done internally, it is very difficult.

108. We are basically pursuing a dual-track approach now. It's not that we cannot annotate at all, but because data annotation has some low-cost and some high-cost aspects, we annotate the low-cost ones first.

109. You can also think that now half of our company's employees are annotating data. Half of the core researchers, the most important people, are annotating data. We are focused on data annotation. Solving the AI problem at this stage relies on data annotation.

110. The bottleneck of high-quality data annotation, I think, is time, it just takes time. Because for OpenAI, for foreign countries, for Anthropic, they started earlier, have more capital, and face more restrictions.

111. The illusion problem of large models significantly affects user experience. The illusion problem can be solved through a method, but this is a long-term issue. The illusion problem can be seen as one that can be addressed and improved through better post-training.

Organizational Positioning and Philosophy

112. First of all, we do not have a model to imitate. Every step is taken based on the actual situation, seeking truth from facts, making decisions based on reality, and figuring out what we should do. Therefore, it is a product of the times, or a response to the real situation; it is not a result of imitation.

113. We are clearly aiming for commercialization. In the end, we have to survive; after all, we are a company, and the government will not give us a single penny.

114. Essentially, we are still a company. It's just that we consider what money to make, when to make money, how much money to make, and how to make money. We have trade-offs. Many companies have done great things because they have a pursuit beyond profit. This pursuit not only does not affect its commercialization in the end but also allows it to commercialize even better.

115. As for partners, our financing was carefully selected. First of all, I think the interests are relatively consistent. It is with those whose interests are most aligned with ours, who have no hostility towards us, or who most hope we can succeed. Not everyone wants us to succeed because we have harmed the interests of many others.

116. AI currently does not lack taste and intuition; what it lacks is the ability to learn continuously. AI's taste and intuition are not an issue. If you ask it to write an article, its taste and intuition, I think, are not a problem.

117. We hope to focus on just one thing. I think AI is a big thing, and I don't need to... I focus on just one aspect. If we focus, and I believe the business interest here is large enough, as many trillion-dollar companies will emerge in the AI era, I think we are one of them.

118. We hope to support more people, but we don't have that much energy. We have the intention, and there is no conflict of interest, but whether we have done it is another matter. At least there is no conflict of interest here, and we hope for a win-win cooperation.

Welcome to join the official BlockBeats community:

Telegram Subscription Group: https://t.me/theblockbeats

Telegram Discussion Group: https://t.me/BlockBeats_App

Official Twitter Account: https://twitter.com/BlockBeatsAsia