Source: The Intelligent Investor
On March 12, during a forum dialogue held by CNBC in Singapore, $Alibaba (BABA.US)$ Chairman Cai Chongxin discussed changes within the organization, the future of AI, the impact of open-source models, and the direction of China's economy.
This dialogue covered several key topics, among which the most noteworthy is his belief that the outside world underestimates the importance of the recent meeting between the Chinese government and private entrepreneurs, and the signals conveyed by this meeting may become key to driving the recovery of the Chinese economy.
"Many people do not realize that this meeting sent a very clear signal to the entire private sector— the government hopes for private enterprises to return and promote economic growth," Cai Chongxin stated.
Shortly after this meeting, Alibaba announced that it would invest more than 50 billion dollars in computing infrastructure and AI over the next three years. He pointed out that the recovery of business confidence is driving companies to reinvest and hire employees, which will ultimately lead to a restoration of consumer confidence.
Regarding technology, Cai Chongxin expressed unique insights into the future of the AI industry. He emphasized that the value of AI lies not in creating the "smartest" models but in how to truly apply them.
Regarding the trade war and tariff issues between China and the USA that the outside world is concerned about, Cai Chongxin is relatively optimistic, believing that although there are still frictions in the short term, global trade is expected to maintain healthy development in the long term.
On the adjustment of Alibaba and its internal culture.
Host: Today's theme is integration. You once said that media, sports gambling, and Cryptos are gradually merging. Your exact words were, "For the gambling user group, appropriately introducing some Cryptos could be quite interesting." What does that mean?
Cai Chongxin: (laughs) That was probably what I said four years ago, right? Now when talking about Cryptos, I feel a bit "outdated."
Host: Isn't it true that Cryptos are popular again now?
Cai Chongxin: Yes, they have indeed made a comeback. But now everyone is more focused on things that are truly valuable. Almost no one talks about NFTs anymore, right?
Host: Yes, it seems everyone is returning to the basics.
Cai Chongxin: Exactly, people are mainly focusing on.$Bitcoin (BTC.CC)$、$Ethereum (ETH.CC)$Perhaps add a little more.$Solana (SOL.CC)$。
As for the intersection of Sports Gambling and Cryptos, it indeed exists, and we often discuss this Topic.
Now everyone is placing bets, investing, and trading Cryptos on their phones; these boundaries are becoming increasingly blurred.
Another example is Sports Gambling, Cryptos, and the Stock Market. Today's young people, especially young men, enjoy Sports Gambling, trading Cryptos, and investing in Stocks.
Wouldn’t it be perfect if they could do all three things in the same app?
Host, well, this can be considered a potential.Mergers and Acquisitions.Now, let's talk about Alibaba. You resumed your position as Director at the end of May 2023, right?
Cai Chongxin: Yes, in 2023.
Host: When you returned, you mentioned in a letter to Shareholders that Alibaba has characteristics of a large company, but you hope to regain the mindset of a Venture. How important is this mentality?
Cai Chongxin: This is the core reason for Alibaba's existence. We must operate like a rapidly growing Venture, seeing ourselves as a small company, maintaining flexibility and humility.
Large companies act slowly, mainly because the decision-making process is too complex. If a decision needs to go through 19 layers of approval, it won't work. Therefore, we really need to regain agility and act quickly.
Our competitors are all Ventures. New competitors constantly emerge in the e-commerce space, each with different models, such as live-streaming sales, Short Video e-commerce, social e-commerce, and so on.
All of these are Ventures, and if our decision-making is very slow, taking 10 days to make a small decision while they can decide in 10 minutes, how can we compete?
Host: So, how can employees regain the entrepreneurial mindset?
Cai Chongxin: The e-commerce Industry is the most fiercely competitive industry in the world, bar none.
This industry has almost no entry barriers; the infrastructure, including Logistics and payment, has already been built by some companies in the market, and this infrastructure is not monopolized by any single player.
Therefore, any emerging company can directly invest in e-commerce without having to worry about Logistics and payment, which is why the competition is so fierce.
When I returned to Alibaba, I worked with the new CEO Wu Yongming to set the top priority, which was to simplify the company structure. We no longer described Alibaba as six different Business units but streamlined it into two core businesses: e-commerce and Cloud Computing.
This adjustment made everything clearer and also made internal communication more efficient. We must ensure that employees clearly understand the company structure so that they can act more quickly.
Next, we need to empower employees, especially the young management—those who are in direct contact with users and customers every day.
I believe that 'empowerment' is very important. Empowerment means giving them the power to make decisions and allowing them to make mistakes, while helping them learn from those mistakes. You must give them some autonomy and space to experiment.
Host: How is your progress on this goal going?
Cai Chongxin: We have basically completed this transformation in terms of organizational structure, personnel adjustments, and cultural transmission. Now, in our core e-commerce business, we are competing head-to-head with competitors every day, every hour, every minute.
Regarding why Alibaba is developing AI
Host: You recently launched a new AI inference model. So, what do you think about Alibaba's efforts in the field of artificial intelligence? Will it exist as an independent product, or is it mainly used to enhance your e-commerce business?
Cai Chongxin: Before directly answering your question, I would like to take a step back and talk about why Alibaba is so excited about AI and why so many people worldwide are actively promoting the development of AI, competing to create the most advanced AI technology.
If you think carefully about the essence of AI, its core function is to enhance human productivity.
The global GDP is about 100 trillion dollars, of which 60% comes from labor income, meaning that 60 trillion dollars is contributed by human labor.
If we assume that AI can increase productivity by 20% and the cost of AI is 20% lower than human labor, then the total addressable market size reaches 10 trillion dollars.
Which other industry has a market size of 10 trillion dollars?
I checked, and the entire Transportation industry is about 5 trillion dollars, and the Medical Insurance industry is also 5 trillion dollars. As far as I know, the entire Technology industry, including all subfields, has a total size of about 10 trillion dollars.
Therefore, the market size for AI is at least 10 trillion dollars and may even be larger. This is why everyone is so excited about AI.
Host: What is the core goal of your large language model (LLM) and reasoning model?
Cai Chongxin: There are several important reasons for Alibaba to develop AI:
First, we operate a Cloud Computing business, and the popularization of AI greatly increases the demand for cloud computing infrastructure. Whether for training models, running inferences, or post-training, strong computing resources are needed, all of which depend on Cloud Computing.
We sell computing power to the market, and regardless of whether clients use our AI models or others, they require computing resources.
In China, we are the market leader in this field.
Secondly, AI can empower our core business, including e-commerce, map services, food delivery, and various other scenarios.
For example, a portion of our revenue comes from advertising, and the key to advertising is the conversion rate, which is how to turn user clicks into purchases. AI can greatly optimize this process, improving the accuracy of ad placement and conversion rates.
Therefore, we see that AI technology has huge growth potential within our own business.
Third, you may have heard our CEO mention 'AGI (Artificial General Intelligence)' as the ultimate goal.
Now, if you are a technology company or consider yourself a technology pioneer, you cannot avoid pursuing AGI. This goal represents the pinnacle of technological exploration, and I personally think it is a really cool thing.
Host: Does your interest in AGI involve deeper thinking, such as from a philosophical or scientific perspective?
Cai Chongxin: Yes. The research on AGI is not just a scientific issue but also involves many profound philosophical questions.
If we do not know the limits of the human brain, how can we determine whether AI is smarter than the human brain? Therefore, we first need to explore the upper limits of human intelligence. This is also why many of my charitable endeavors are related to research in neuroscience, human cognition, and psychology.
I am very curious, what are the limits of human thinking? How smart can we actually become?
But one thing is certain, the human brain is currently the most energy-efficient "Computer" in the world. In contrast, all AI computations today, whether GPU or large language models, are extremely energy-consuming.
They require megawatt-level electrical utilities, while the human brain can accomplish astonishing computational tasks with very little energy.
This itself is a question worth pondering.
Another philosophical question is, what is the significance of our pursuit of AGI? When someone claims, "I have the most powerful large language model, the most advanced AI," what is the meaning of this technological leadership? Does it truly drive human progress?
These questions are worth deep contemplation.
Long-term thoughts on the value of AI.
Host: Let me interrupt for a moment, because you recently mentioned that 'the AI competition is no longer about whose child is smarter,' and you said, 'the value of creating the smartest AI in a vacuum will ultimately approach zero.'
What does this mean? Why do you believe that the value of creating the smartest AI will ultimately approach zero?
Cai Chongxin: First of all, I must say that many technical experts may not agree with my views.
Host: Yes, the reason I bring this up is that I find your perspective very interesting. There might even be quite a bit of debate within your company about this?
Cai Chongxin: That's right, our scientists within the company are also debating this issue. And I want to clarify, I am not a computer scientist, nor a technical expert; I studied law.
Host: But isn't that quite helpful for you?
Cai Chongxin: I'm not sure about that.
But let me explain from a different perspective; I believe the cultivation and education of AI is very similar to raising children. Many of us have children, so what is our goal in educating them?
Should they become Nobel Prize winners in every subject? Or should they read all the books in the world? These are certainly important, but as parents, we also need to teach our children emotional intelligence, to learn how to make friends, how to interact with others, and how to manage social relationships.
These abilities are not just simple accumulations of knowledge.
The development of AI follows the same principle.
If we want AI to truly think and make decisions like humans, simply making it 'smarter' is not enough; it needs to possess capabilities across more dimensions. For instance, this even raises a question—can we inject hormones into machines?
Perhaps from a technological standpoint, this may be possible in the future, but I have not yet seen a method to achieve it.
Additionally, I have found that 99% of the people discussing AI and AGI are male. And we all know that men and women think differently and have different approaches to problem-solving.
So, when we talk about the 'smartest AI', does this AI stem from a male way of thinking or from a female perspective? This is also a question worth considering.
Host: Returning to the business application level, the strategy you advocate does not seem to be about whose AI model is larger or whose computing power is the strongest, but rather how to truly apply AI in real scenarios, right?
Cai Chongxin: Yes, the value of AI does not lie in whose model is the largest, but in how to apply AI to actual Business, creating real value. For example, in the e-commerce field, the application of AI can greatly optimize the user experience.
For instance: suppose you want to go camping, the only equipment you think of is a tent. But if you consult AI, it will not only tell you that you need a tent but also recommend other items based on your destination, time, and climate data, such as hiking boots, insulated water bottles, mosquito repellent, and even tell you what kind of food to bring.
Such smart recommendations from AI can help you prepare more comprehensively for camping, avoiding omissions.
Furthermore, on our e-commerce platform, you can place orders with one click, and all items will be delivered to your doorstep the next day. Isn't that great? This is the real value of AI.
Host: This is indeed an important application of AI.
Cai Chongxin: I also mentioned the value of open source in my article. The discussion about DeepSeek actually focuses not on whether China's AI is stronger than America's or vice versa, but on the power of open source.
When AI models are open source, anyone can run them on their own Datacenter or even their own laptop, which means that the capabilities of AI become accessible to everyone.
This openness will spur more innovative applications based on these models, driving the development of the entire Industry.
Now it is not just a few giant companies driving the development of AI; small companies, large enterprises, individual developers, and entrepreneurs can all innovate based on these open-source models.
Therefore, competition in AI is no longer a game that only the five wealthiest companies globally can participate in; it is no longer an exclusive domain for companies that can invest 50 billion dollars a year in research and development.
The application of AI will scale up significantly due to open-source, leading to more innovation.
Host: You mentioned the philosophical discussions about AI within the company. So, have you discussed the ultimate impact of AI? After all, we are still in the early stages of AI development, and everyone is exploring what it might mean, especially regarding future work forms, value creation, and even whether it will affect human employment. Do you have concerns about this?
Cai Chongxin: I am not concerned because AI can actually enhance human work capabilities.
We discussed an example, such as AI could completely replace securities analysts. If you need a research report on$NVIDIA (NVDA.US)$or Apple, AI can generate a very high-quality report. It can complete all repetitive, low-value research tasks, and then humans can provide judgment and insights to ultimately make better investment recommendations.
Therefore, it is not believed that AI will completely replace humans, but rather will enhance human work capabilities.
Similarly, my professional background is in law, and the work of a lawyer is actually quite similar to programming. When drafting a contract, one is essentially writing "legal code" where every clause and definition must be logically rigorous and compatible with each other.
AI can assist in drafting contracts and may even complete most of the legal document work.
So, does this mean lawyers will be replaced by AI? Not necessarily.
This is because people seek out the best lawyers not just to draft contracts, but to seek legal advice. Truly excellent lawyers provide judgment, offering valuable suggestions based on specific business contexts, social relationships, and negotiation strategies.
These are aspects that AI cannot replace. Therefore, it is believed that AI will not replace lawyers but will enhance their capabilities, allowing them to devote more energy to high-value decision-making.
Regarding the private enterprise seminar and the Chinese consumer market.
Host: In the time we have left, I would like to discuss the Chinese economy. The consumer confidence in the Chinese market has been a topic of discussion for a long time. There are concerns about insufficient consumption in China, problems in the real estate market, and the savings rate remains very high. What is your view on the current state of Chinese consumers?
Cai Chongxin: Let me explain why consumers are crucial to the Chinese economy, especially at this current stage.
In 2023, China achieved a GDP growth of 5%, but the main driving force behind this growth came from exports. As we enter 2024, due to geopolitical factors and trade barriers, China's export growth may not be as strong as last year. This means that domestic consumption must become the new engine of growth.
We look forward to the government launching more stimulus policies to promote consumption after this year's "Two Sessions." Currently, we have not seen all the details yet, but hope that the policies can eventually be rolled out to further drive consumption growth.
Host: So does the current consumer confidence in China remain relatively weak?
Cai Chongxin: Currently, there has not been a noticeable consumption boom, but we cannot ignore the fact that the financial situation of Chinese consumers is actually very healthy.
Chinese households' balance sheets are very robust, with bank deposits exceeding 20 trillion dollars. This means that many people are still holding back on spending, not because of economic pressure, but rather being cautious.
Host: What can truly make consumers willing to start spending?
Cai Chongxin: The core factors are confidence and market sentiment.
I want to specifically mention the meeting between **** and entrepreneurs a few weeks ago.
I believe many people underestimate the importance of this meeting.
Because this meeting sent a strong signal to the entire private enterprise sector, allowing entrepreneurs to regain confidence in market investments. An increase in corporate confidence will drive more investments, create more job opportunities, and ultimately boost consumer confidence.
We just announced (on February 24) that in the next three years, we plan to invest over 50 billion dollars in computing infrastructure and AI. This has given us confidence, allowing us to reinvest the company's profits into capital expenditure and long-term investments, and continue to hire talent.
When businesses are confident, they will take action.
Additionally, with the current excitement around AI technology, people see companies like DeepSeek, Alibaba, Tencent, and ByteDance in China are all pushing AI forward, leading everyone to think, "We are not inferior."
AI will spawn a large number of new applications, and the entire society is full of expectations for technological innovation.
Therefore, the warming of confidence in private enterprises, combined with the wave of technological development, will eventually translate into consumer confidence.
Host: Why do you think this will ultimately affect consumer confidence?
Cai Chongxin: Because companies are increasing investments, are re-hiring, and are experiencing growth, which will ultimately seep into consumer confidence. When people see more job opportunities and an active economy, they are more willing to consume.
About the impact of China-US trade and tariffs.
Host: Finally, let's talk about economic and trade issues. There are a large number of American goods sold to Chinese consumers on your platform. Currently, tariffs on Chinese goods in the USA have increased to 20%, while the specific countermeasures from China remain unclear. Are you worried about the trade war? Will it affect your ability to provide American goods to Chinese consumers?
Cai Chongxin: Obviously, any business involved in global trade will be concerned about tariff issues. At Alibaba, our cross-border business is two-way—Chinese sellers sell products to the USA, and American sellers sell products to China.
Currently, 50 billion dollars worth of American goods are sold to Chinese consumers annually on the Alibaba platform, and the annual sales of European products also reach 30 billion dollars. This indicates that the demand from Chinese consumers for overseas brand products remains very strong.
Of course, we are also concerned about the potential impact of tariffs, but this is already beyond the control of businesses. It will ultimately depend on negotiations between the leaders of various countries.
From a global economic perspective, tariffs often create too much friction, while a certain level of global trade and business interaction is beneficial.
Host: How do you think the attitude of the USA government will be?
Cai Chongxin: I believe that whether it is the Trump government or other governments, they hope that more American businesses can do business in the Chinese market and make profits.
In the long run, I still hold an optimistic attitude. I believe that tariffs are to some extent a negotiating tool, but ultimately, the situation will improve.
Of course, I also think that certain protected industries will still be subject to tariff restrictions. For example, China's electric vehicle (EV) industry is rapidly expanding in the global market, and I expect there will still be tariff measures targeting China's EV exports in the future.
Host: So, it is unlikely for American consumers to buy Chinese electric vehicles costing 0.02 million dollars in the short term?
Cai Chongxin: These tariffs will increase the cost for Chinese EVs to enter the USA market, so in the short term, Chinese EVs will not enter the USA at extremely low prices.
Editor/rice
Comment(3)
Reason For Report