Exclusive record| Ali Cai Chongxin’s latest speech in Dubai: Why did I choose the “crazy” Jack Ma in the first place

Photo source: Visual China

Blue Whale News, February 13 (Reporter Zhu Junxi)On February 13, local time, at the World Governments Summit 2025 Summit held in Dubai, United Arab Emirates, Alibaba co-founder and chairman of the board of directors Cai Chongxin shared his views on the cultural theme of survival and innovation.

In a nearly 40-minute conversation with American producer Jeffrey Katzenberg, Cai Chongxin first recalled his mental journey when he joined Alibaba more than 20 years ago, and praised Jack Ma as a leader. He mentioned that many start-ups will close their companies and turn around once they fail, but Ali has never done this, but has made many innovations out of its survival instinct.

Cai Chongxin said that the most exciting thing at the moment is AI. Ali is not only a cloud service giant, but also has a self-developed model. It has also launched an active layout in AI investment. Talking about the sensation caused by DeepSeek, Cai Chongxin commented that they proved that the cost of model training and reasoning can be greatly reduced through engineering innovation, and people have invested heavily in these areas.

“With the DeepSeek phenomenon, people will start to change course and no longer invest tens of billions of dollars in computing infrastructure just to train the brightest kids to become doctoral students and Nobel Prize winners. We should focus on the actual problems in the world and use AI to solve them. So many people will start to invest resources in application areas.& rdquo; Cai Chongxin said.

Cai Chongxin responded to recent rumors that Apple will cooperate with Ali to develop AI functions for China iPhone users,Apple has always been very picky, talked with many companies in China, and finally chose to cooperate with Alibaba.“We are fortunate and honored to work with a great company like Apple.& rdquo;

The following is a transcript of the conversation, and the content has been adjusted by Blue Whale News to ensure the original intention:

Look on the other side of the fence

Jeffrey Katzenberg：

It’s a pleasure to have this conversation with you. First, in 1999, you made $700,000 a year in a job, but you were invited to join startup Alibaba for only $600 a year.At that time, it was a mutual friend of yours who introduced you to Jack Ma. Your friend told you that Jack Ma was smart and a little crazy. So, my question is, why did you choose this madness?

Cai Chongxin:

Thank you, Jeffrey. Before answering your questions, I would like to say that I am truly honored to be interviewed by you on this stage. You have always been my mentor. You have taught me how to treat people. No matter who they are and what background they come from, they must be respected. I have learned a lot from you, thank you very much, I am honored to be here.

Regarding the question of madness, why did I choose madness? What kind of madness do I mean?

This was in 1999, and I went to a second-floor apartment in Hangzhou to meet Jack Ma. I know something about him. In fact, I have read about him in The Economist. This is when no one knows him yet. But Chris Anderson, editor of The Economist, compared Ma to Jeff Bezos of China. At the time, few people knew Jeff Bezos, especially in China.

When I arrived at Jack Ma’s apartment, there were 10 to 12 pairs of shoes outside the door.People usually take off their shoes before entering other people’s homes, so I thought maybe 10 or 12 people work in this apartment. After I entered, Jack Ma was a very attractive person, and I was completely attracted by him for about an hour.

Then, I suddenly wanted to go to the toilet. I said to him, excuse me, can I borrow a toilet? He said it was there.I went into the toilet and saw 12 toothbrushes on the sink.I realized that it was like these people lived here 24 hours a day, working in sleeping bags. Suddenly I had the feeling that this place was like an HP garage in China.

What really touched me was not the crazy things. Jack Ma talked about a big vision that I didn’t fully understand at the time.He said that to be a startup, you don’t necessarily need a business plan, you just follow your instincts. These are all crazy aspects.But that wasn’t why I decided to quit my job. What really made me decide was that he was a leader. He’s leading these believers, and they really believe in him. He gave them a mission and talked about how technology, especially the Internet, could transform not only companies, but entire countries.

Jeffrey Katzenberg：

At that time, did he have a clear vision and know what he was doing?

Cai Chongxin:

The vision at the time was to create an Internet platform where all small companies, manufacturers and trading companies from China could list themselves. In fact, we call it a BBS, or bulletin board system, and this was the first website of alibaba.com. In fact, it is in English because it is designed for non-China people who purchase goods from China and use it to buy goods from China. And we are on the eve of China’s accession to the WTO. Once China joined the WTO, China exporters began to rise as an export-driven economy, which contributed to Alibaba’s early success.

Jeffrey Katzenberg：

So, from a lawyer to a private equity investor to giving up those identities completely and entering a technology startup, this is a huge shift for you. Why do you think you can do it? What gave you this courage?

Cai Chongxin:

I have always been a very curious person, not satisfied with the status quo, always wanting to know what it’s like on the other side of the fence.As a lawyer, I do consulting work. Lawyers usually do not control decision-making power, but provide advice to clients, and then the clients make decisions. So I decided to transform from a consultant to a decision-maker and started joining private equity firms. When investing, you invest money and must provide theoretical basis for investment and make decisions. Then I realized that as an investor, I was sitting on the company’s board of directors, but I wasn’t really involved in the company’s operations. I just allocate funds to entrepreneurs and management teams to do business.

So I started to wonder, what was it like on the other side? I want to start an operating company. The environment at that time happened to be the rise of the Internet. Companies such as Yahoo and eBay had emerged before Alibaba, and it felt really exciting. So for me, this is an opportunity to go across the fence and see how the company works.

And I was 35 years old, had a family, had been married for three years, my wife was pregnant, and our first child was about to be born. So when she was born, we called her Ali baby.

Jeffrey Katzenberg：

This was a very bold attempt. We all know that 99% of startups fail, so why is Alibaba successful?How did it become one of the largest and most successful companies in China?

Cai Chongxin:

I think the title of this conversation is survival and innovation. In fact, the two are causal and closely connected. From the beginning, we didn’t have a lot of money, so everything centered around how to keep the cash on hand going. Because Ma Yun taught us that as long as we can live, we have the opportunity to create. So we have the instinct to survive, which helps us overcome failure.Many startups fail, their products are constantly iterating, and if they fail, they will shut down the company and do something else. But Alibaba has never done such a thing, and we have always insisted that you can say that demand is indeed the mother of invention.So out of the desire to survive and grasping our last breath, we created many innovations.

In fact, if you look at Alibaba’s history, almost all of our fast-growing businesses have developed organically. We are not successful in mergers and acquisitions, but we achieve success through internal development. We had the B2B market, and then we entered the consumer market Taobao. Then we realized that payments were very important, and in order to allow transactions to take place online, we organically developed Alipay, and then we had a profit engine, which was actually an e-commerce market based on search advertising. As you can imagine, we did the same thing in a way similar to how Google ads were inserted into the eBay system. This is our invention, this is our idea.

Today, Taobao remains a very profitable platform. Therefore, our cloud computing business is also realized through organic development. So, I think the key is that as long as you can survive and persist for another day, good things can happen.

Jeffrey Katzenberg：

So after 15 years of persistence, you let Alibaba go public, setting the largest IPO in history, raising US$25 billion and having a market value of more than US$230 billion. Today is 25 years since Alibaba was founded and 10 years after its listing. What is the most exciting thing for Alibaba and you personally today?

Cai Chongxin:

Every day at the company, I feel very excited because I can work with smart, creative and innovative young people. Alibaba’s DNA is to let young people innovate. The company’s innovation is not completely top-down. We encourage bottom-up innovation. Obviously the most exciting thing today is AI.

Jeffrey Katzenberg：

Before you discuss this topic, I would like to mention one thing you have said, which you have been promoting for the past 18 months: my job is to get people moving forward unimpeded. rdquo; Most board chairmen don’t think so. Can you explain it?

Cai Chongxin:

I’ve been with the company for 26 years, so I don’t feel the need to prove to anyone that I’m in charge of the company.

About 18 months ago, in fact, two or three years before that, I was actually fading out of the company because as one of the founders, I felt that the company was already big and it was already on its own path.Then 18 months ago, I returned to the chair position because I felt that a lot of people, including the board, thought we needed to change the management team, so we appointed a young CEO who was 12 years younger than me.They said, Joe, when you come back, you will only be chairman, not CEO. Chairman, you know, because we need an experienced, steady leader.

So the first thing I did, when I walked into the board meeting, our lawyer said, Joe, you are now the executive chairman of the company. I said, wait a minute, don’t give me this title, I am the chairman of the company. I know I’m a full-time executive, but execution shouldn’t be my title.

I disagree with the division of labor between executive chairman and CEO under the British system. The reason is, who is in charge of everything? Employees will be confused, and outsiders will be confused. Who is really responsible for running the company?In our company, I am very clear that the CEO is responsible for company operations, and I support the CEO in resource allocation.I will participate in capital allocation decisions and participate in human resources decisions. If we wanted to create an equity incentive plan, I would consult with him on these aspects. As the chairman of a company, it is important not to interfere with the work of others.Many chairmen can’t let go, and I don’t think this is the right way to cultivate the next generation of leadership.

AI should not be limited to five or six wealthy companies

Jeffrey Katzenberg：

Okay, now we move to one of the most important businesses, cloud computing. What’s different from Amazon is that you’ve been investing in and developing your own AI capabilities. So, what is the current situation? What are the current levels of Grok3, DeepSeek and other competing big language models (LLMs), and how does Alibaba position itself among them? What is the most important thing to you and what do you think is valuable?

Cai Chongxin:

I think Musk has already said that Grok 3 will be the best AI ever built, and no one can surpass it, forever.

So, do I need to answer this question now? Actually, I don’t need to. Personally, in fact, I did use Grok because I was on X (his previous Twitter app) and it provided an interface for me to use Grok, I use it a lot, and I was impressed with the results. I also use some other AI applications, including our own.

As for the question you asked, whose AI is better in the AI competition? The answer to this question yesterday will be different from the answer tomorrow. Everyone is in a relatively fierce competition right now, and the next model released will be better than some models, but not necessarily better than others. Then someone else will launch something better than you. We think this is what will happen.

But the DeepSeek you mentioned is important because they demonstrate that the cost of training and reasoning for large language models can be significantly reduced through engineering innovation. So far, people have invested heavily in these cutting-edge models.

But the problem is, if you think of training cutting-edge models as a career that makes kids smarter, the goal is to make them the brightest and get a doctorate, not just one subject, maybe 15 subjects, from mathematics to physics to biology to psychology, you want to make kids smarter. If so, put this idea into today’s AI competition.

Perhaps only five or six wealthy parents can afford such a comprehensive education for their children. But is this really something we are interested in? Maybe not.You have to ask yourself, what is the purpose of developing AI? Is it to win every Nobel Prize to have the brightest child?Or does it have greater significance, such as solving many problems in the world, using AI to solve them, and creating economic benefits.

soI think with the DeepSeek phenomenon, people will start to change course and stop investing tens of billions of dollars in computing infrastructure just to train the brightest kids to become doctoral students and Nobel Prize winners.We should focus on the actual problems in the world and use AI to solve them. Therefore, many people will start to invest resources in application areas.

Secondly, DeepSeek allows us to see the value of open source.If your goal in life today is to develop a closed-source AI system, become the smartest doctoral scholar, and win the Nobel Prize, I personally think the value of this goal is close to zero.

Jeffrey Katzenberg：

This is a pretty surprising and bold AI vision, completely different from what these five very wealthy parents are pursuing.

Cai Chongxin:

So what about the rest of the people? If only five companies can afford the brightest kids, this is not the world you want to live in, is it?

So I think there are ways to raise smart children without having to rely on these huge computing resources. If everyone can concentrate, first, focus on applications; second, utilize open source communities and ecosystems. One of the characteristics of open source communities is that people share. Your contributions to open source are now part of the ecosystem that others can learn from.

Therefore,I think there may be more small models in the future.For example, we are an e-commerce business. If I want to develop a very smart shopping assistant to help recommend products to customers, do I really need a large language model with trillions of parameters? Probably not needed. Like when you hire a shopping assistant in a store, do you need to get them a doctorate in physics to pick out the best clothes? Probably not.

So I think now that there is open source, many companies can participate. Because they can deploy open source code to their own infrastructure, whether it’s in their own data centers or on their own laptops. Whether it is a large company or a small company, or even entrepreneurs, can now jointly develop AI and contribute the power of AI together.

Jeffrey Katzenberg：

This is a very unique and distinctive view of AI development, especially in terms of application and deployment, which you mentioned is more democratic. And I believe you will publish an article today.

Cai Chongxin:

Yes, II am writing a column today and will publish it online in the South China Morning Post later.Remember, I did not use the words inclusiveness or democratization because I think these concepts have been overused. I just think that through open source and small models, you can get more developers, entrepreneurs and small companies to participate in AI development. I think this is a good thing for innovation, not just for five or six companies. It would be crazy if these five or six companies were investing $60 to $80 billion a year in computing infrastructure.

Jeffrey Katzenberg：

This is undoubtedly an interesting topic, and the world is changing very quickly. Your stock has risen 38% this year and 8% yesterday, and I think a large part of the reason is the announcement of your new partnership with Apple in China. Can you talk about this cooperation?

Cai Chongxin:

Apple is an example of edge computing because their smartphone is a device, not a mainframe computer located in a data center. So they need to have some small smart AI on their phones.

I think this will be the trend in the future (Cooperate with China companies to develop AI)。You know, no matter where you do business, you probably want to have a localized AI to support your devices. Therefore,Apple has always been very picky. They talked to many companies in China and finally chose to work with us to use our AI to power their phones.We are fortunate and honored to work with a great company like Apple.

Collection of Cai Chongxin’s famous quotes

Jeffrey Katzenberg：

So one thing I have to understand about Joe is that there are two Chua Chongxin, just like Clark Kent and Superman. One is the gentle gentleman you see on the stage, and the other is the fierce competitor on the field.

So let me quickly go through your lineup. You own shares in the NBA’s Brooklyn Nets, the WNBA’s New York Liberty, the National Lacrosse League’s San Diego Seals and the Las Vegas Desert Dogs, and the NFL’s Miami Dolphins.

I learned that your interest in sports actually stems from when you were in boys ‘school, you said that I was better than others in mathematics, but I didn’t gain respect for it, but I gained respect from being on the football field. So, can you connect these two Joes ‘stories?

Cai Chongxin:

You think there are two Joes, but it’s actually just one person. He started as an athlete. I loved sports very much when I was a kid, and my parents sent me to boarding school in the United States. At that time, I was in a new environment, a little shy, and integrated into a new culture, you know, boys ‘school. So, the only way to make others respect you is to exercise. I chose the hardest thing. I played football in high school, not football.

So, unlike many technology entrepreneurs, I was first an athlete and then a technician working in a technology company, spending a lot of time working with engineers and scientists. This is probably the beginning. My passion for sports has always existed, but the opportunity to enter professional sports did come at a difficult time. At first I didn’t plan to own an NBA team. It wasn’t until 2017, when the opportunity came, that I began to study the economics of the NBA and other major professional sports leagues. Then I became very interested and decided to invest. I am very happy today because the value of that investment has increased.

Jeffrey Katzenberg：

You once said: Sports not only teach discipline and hard work, but also how to face failure. You lose, you have to rebound.& rdquo; So, what lessons you learned on the sports field have you applied to your work?

Cai Chongxin:

It’s resilience. It came to me that if you have played a team sport or a personal sport like golf, like fantasy golf, you will experience failure, almost every season. You won’t win every game during the season. So, when you fail, they often say bounce back. But those who have really experienced it know how to bounce back? Not thinking about the next game, but thinking about tomorrow’s training. Can I improve a little bit in training tomorrow to make myself better and prepare for the next opponent?

So it’s really incremental step by step, focusing on the details. If you apply this principle to work and business, the same is true. If you are running a company, you must be very focused on every operational detail. I have said that as chairman, I don’t directly manage the company, but I expect our management team to focus on these things.

Jeffrey Katzenberg：

Through my Alibaba search, I found some of your great quotes that I think are very inspiring and enlightening. So I’m here to read these famous quotes, and maybe you can talk about what they mean to you. Okay, number one: The most important job of a leader is to guide people in the future.

Cai Chongxin:

This is obvious because the team needs a direction. I don’t think there’s anything more frustrating than employees who don’t know how you view the future and have no sense of direction. May I say that’s the most important thing? Okay, I have a more important point. As a leader, you need to be able to work with people and recruit people smarter than you are. I think this is very important because you can’t stay in this company forever. At some stage, the company’s progress will exceed your capabilities, so you must bring in better and smarter people than you so that the company can continue to improve.

Jeffrey Katzenberg：

This is my favorite: Humility is very important to leaders. You must be able to admit mistakes. It’s not easy for any leader.& rdquo;

Cai Chongxin:

If you don’t even dare to admit your mistake, it’s hard to imagine you would be willing to correct it. Imagine we are in an industry where just a slight deviation from the right direction can lead you to disaster.I think good leaders are always willing to reflect on themselves and say: Okay, I was wrong, we need to change direction. rdquo;It’s a hard decision to make, it’s very difficult to tell people you’re wrong, but we have to do it.

Jeffrey Katzenberg：

This sentence is actually somewhat related to some of the issues Musk talked about. You once said: A common problem with large companies is that they are bound by a fixed organizational structure and try to adapt the company’s direction to that structure, but in fact it should be the other way around.” rdquo;

Cai Chongxin:

His point of view is about government efficiency. Governments often have a huge structure that determines the direction. But this shouldn’t be the case. You always need to first set the direction, then set up the company’s organizational structure and put the right people in the right positions to achieve your goals.

Jeffrey Katzenberg：

I have known you for some time, and there is one of your famous words that I now understand does not apply to me. You said: 99% of people need seven to eight hours of sleep, and if you think you are 1%, you are deceiving yourself.” rdquo;

Cai Chongxin:

This is something I admire very much about Jeffrey. He only sleeps four hours a day, or three hours, and I can’t do that. I have to get seven to eight hours of sleep in 24 hours. Due to jet lag and other reasons, I travel a lot and can’t get enough sleep at once. But I need to take a nap. But you are an exception. I said that many people think they are the 1%, but they are not. But after all these years of knowing Jeffrey, he is really the 1%. In fact, I’m really grateful to the Beatles for writing a song for me called “8 Days a Week”.

Thank you very much, thank you!