On the afternoon of May 4, local time, the 9th Buffett Shareholders' Meeting hosted by Sina Finance and the Chinese and American Investors Reception was held at the Downtown Omaha Marriott Hotel. The reception was the largest and most influential investor exchange event during Buffett's shareholders' meeting. Investment elites, heads of Chinese public and private equity funds and brokerage firms, and executives of listed companies attended the conference and began a high-quality conversation.

Mr. Chen Xi, a tenured professor named by Andre Meyer at New York University in the United States, attended the conference and delivered a speech entitled “Application of Artificial Intelligence in Quantitative Investment”. Chen Xi believes that Buffett's statement that he doesn't understand AI is just a humble statement. It's not that Ba Lao doesn't understand it; it's that he definitely wants to earn money within his field of perception. He believes machine learning and artificial intelligence will revolutionize quantitative investing. He said that AI will bring many new opportunities, and seizing these opportunities is a new door for everyone, whether investors or entrepreneurs.

The following is a transcript of Chen Xi's speech:

Thank you very much, and I am honored to receive an invitation from Sina Finance. In fact, this was my first time attending Buffett's shareholders' meeting, and I was deeply moved. Looking back, I graduated from the computer department of Xi'an Jiaotong University as an undergraduate, then received my doctorate from Carnegie Mellon University, and later did postdoctoral studies at the University of California, Berkeley.

I've always been a technical guy, and I don't know anything about investing. I didn't come into contact with the concept of value investing until I mistakenly became a professor at NYU's Stern School of Business. This reminds me of my own experience; I am a value investing practitioner. Why are you saying that? In 2007, when I was 20, I came to the US from China alone carrying a bag. At the time, I applied for the first batch of doctoral programs at Carnegie Mellon University. This was a brand-new program. At the time, it wasn't called artificial intelligence, but the Department of Machine Learning (Machine Learning Department). I asked my brothers, teachers, and sisters if I should join this department or the traditional computer department. They all advised me never to enter the machine learning department, because if machine learning wasn't popular when I graduated, CMU cancelled this department, and then told others that I graduated from the Carnegie Mellon University machine learning department, and I might be mistaken for a scammer. In contrast, the computer department is safer and safer, because the Carnegie Mellon Computer Department is well known, and famous professors, including Mr. Li Kaifu, are all from this department. However, at the time, I thought machine learning was something new, and I was willing to try something new, so I resolutely chose the Department of Machine Learning and became one of the first PhDs at the time. Later, the results proved that I made the right choice. When everyone was afraid, I chose “greed,” and in the end, the results were quite good.

From machine learning to the development of artificial intelligence, the beginning of artificial intelligence has arrived. Since I entered the field of artificial intelligence in 2007, it has been almost 17 years since then, and I have seen the continuous emergence of artificial intelligence. The appearance of the big model last year shocked us immensely. In the field of machine learning, we are seeing the rise of a new kind of intelligence, or imagine ability. As models and data grew, its power did not grow linearly, but it reached a point where it suddenly unleashed potential we had never seen before.

This trend will spread to all fields, and will generate many trillion-level new racetracks. Let me give you an example. I temporarily left NYU between 2021 and 2023. I joined Amazon and led an advertising team that quickly grew from 30 to 140 people because I wanted to get to know the company more deeply.

At first we didn't realize that ChatGPT would have a profound impact on the advertising industry. Why are you saying that? For example, in the past, I went to Google to search for anything. In the process of searching, it was easy to click on ads when you saw these pages, but now the situation has changed, you can get answers directly by asking questions, and there are no opportunities for advertisers.

Google is facing an embarrassing situation. If it now vigorously promotes artificial intelligence, it can easily offend its advertisers and even hurt its own interests. Instead, ads may flow more to Amazon or even TikTok, and TikTok may become an important traffic entry point for the entire trillion-dollar advertising industry in the future.

AI will bring many new opportunities, and seizing these opportunities is a new door for everyone here, whether investors or entrepreneurs.

Of course, due to the influence of my professor, I did not choose to start an AI business, but instead focused on quantitative investment in the secondary market for a long time rather than value investment. I don't have a lot of money; I can only invest quantitatively, but I think machine learning and artificial intelligence will revolutionize quantitative investing. Quantitative investment companies in the US, such as Castle, were founded around 1990 and have a history of more than 30 years; while Two Sigma was founded in about 2000, with a history of only about 20 years.

In contrast, China's quantitative investment sector is being updated and iterated faster. In China, private equity institutions such as Jiukun, Mingyuan, and Yanfu have all risen in less than ten years, and are a very emerging industry.

Why are these private equity firms in China able to surpass American hedge funds in terms of algorithms and computing power, and have a positive PK with them? Among these, artificial intelligence plays a critical role.

In the past, since University of Chicago professor Fama proposed the multi-factor theory, we have been working on factor mining for a long time. US hedge funds, including World Quant and AQR, are all digging for factors.

What can artificial intelligence do? Through big data, investments are made more systematic, and many factors that are difficult even for top researchers to discover can be discovered. This is a new blue ocean. With the advent of big models, we can use non-traditional data such as news data and remote sensing data for investment analysis. For example, during the pandemic, it was difficult for us to determine the value of Walmart, but through satellite remote sensing data, we can estimate the value of Walmart by counting the number of cars in parking lots. Things like these are actually all new investment paradigms; this has fundamentally changed compared to traditional mining factors.

I myself have also been engaged in quantitative research on commodities and virtual currencies for a long time. Thanks to AI for providing such an opportunity for researchers and practitioners like us, to compete with America's top hedge funds such as Citadel, D.E. Shaw, and Two Sigma. I know that companies like China's Jiukun will also set up branches in the US. Recently, this has become a hot topic. This is all brought to us by artificial intelligence.

I probably don't have time to do a lot of first-level artificial intelligence research, but in terms of first-level investment, artificial intelligence also has many new ideas. Domestically, people are probably chasing big models. For example, the popular face of the dark moon in the country was developed by my Carnegie Mellon mentor Yang Zhilin, but I am more interested in artificial intelligence empowering traditional industries. This is where it is really attractive. Because big models are very voluminous, and open source stuff has had a big impact on our closed source stuff. For example, when Facebook's LLAMA3 was launched, the previous work probably went to waste. The joke is that OpenAI may have wasted half of the investment of American YC (a famous startup investment company). American YC usually invests in startups, but OpenAI may have caused half of the projects they have invested in to fail. Where are the barriers? Where is the moat? It's in traditional industries such as finance, law, and even milk and real estate. Whoever takes the lead in installing AI wings will likely fly higher and farther in the next 20 years.

As a Tier 1 investor, I'd like to share some insights on AI's empowerment and investment in traditional industries. In traditional industries, the application of AI may form a kind of moat, making it more difficult to be replaced by AI, and this field is not such an internal racetrack. Of course, I don't have the ability to make first-level investments myself; I mainly focus on second-level investments, but first-level investments are also very interesting.

Finally, I want to tell you all that I was deeply moved by Buffett's live conference this morning. At the conference, someone asked Buffett about generative AI, which I'm very interested in because it's my field of expertise. I was very excited at the time; I really wanted to hear what Ba Lao had to say. Buffett humbly stated that he doesn't understand generative AI, and he hopes this technology will be beneficial to humans rather than harmful. But in reality, how could Ba Lao not understand? Every day, he talks to frontline AI experts such as Bill Gates and Tim Cook. Moreover, Microsoft is the company at the cutting edge in the field of generative AI. Tim also mentioned that Apple's next smartphone will use AI on a large scale, which is clearly a future trend. I think it's not that Buffett doesn't know AI; it's that he is more focused on his field of expertise, which is very important.

I'd like to give you an example. As early as 2010, I heard about Bitcoin. At the time, a computer systems professor at Carnegie Mellon University was very good. He used a machine in a big room to mine bitcoins. I later became interested in Bitcoin and did a lot of research, including Bitcoin's architecture, white papers, etc. I've also published ten papers on blockchain and distributed finance, and wrote the first MBA-level blockchain textbook in the US, published by Cambridge University Press.

I did so much preparation before I started investing in the Bitcoin sector, although it was too late, around the second half of 2020. I'll admit I missed a lot of opportunities compared to early Bitcoin investors in China. But through my research, I've made my investments more secure.

Cryptocurrency is a very sensitive field, and different people have different opinions about it, but at least I have made my investments more secure through my own research. There may be ten or twenty years to come. The US has just included Bitcoin spot ETFs under regulation, and Hong Kong has also launched Bitcoin and Ethereum-related products.

However, I would like to say that through my own research, whether it's writing a book or paper, it will probably take me several years of research to have this understanding before investing. I believe that on this path of investment, I may not be able to explode, but I will go further.

Ba Lao gave us a very important concept. I heard his views on generative AI. Although I haven't heard his new views on artificial intelligence, his humble attitude and careful way of investing taught us that people really must earn money within their perceptions, and you can only earn money within your own perception.

I just wanted to share this with you, thank you so much!