Integrating AI and Web3: Value Creation in the Data Age

Author(s): Harriet Gaywood

Originally published on Towards AI.

Integrating AI and Web3: Value Creation in the Data Age

Perspectives from Meitu's former Chairman, Mike Cai

What is Web3 and how does it differ from AI?

Internet pioneer, Web3 visionary and former Chairman of Meitu Mike Cai explained the differing functions of AI and Web3 and how they will integrate as he shared his vision for the data age during a keynote at BEYOND Expo 2025 in Macao.

Mike Cai explained AI can be simply understood as a technology that improves productivity making it easily accepted in any country or company. As the third generation of the internet, Web3 differs by improving relationships of productivity — changing organizational structures and the way in which information flows.

From Web1 to Web3

25 years ago the first generation of the Internet, Web1 was illustrated by websites such as Yahoo and Sina. Newspapers, magazines, and videos were uploaded to the Internet for people to browse but they were unable to interact with the publishers. The Web2 era saw the arrival of interaction between users on platforms such as Facebook, Twitter, Weibo, WeChat, and Douyin. Content could be posted to be liked or shared by friends.

Web3 solves inherent problems such as preventing a platform from becoming too powerful. For example currently when you register for WeChat, if you want to use the name “Cai Wensheng”, but it has been registered by someone else, you can’t use it. Or if you make some inappropriate remarks, WeChat may block your remarks and may even face government accountability. So if users feel that WeChat is not easy to use and decide to leave, all their previous content on WeChat, including the text, videos, address books, etc., will belong to WeChat. This is because when registering for a Web2 platform, users usually agree to all the terms, resulting in all data belonging to the platform.

Therefore, the emergence of Web3 brings two key features: (i) the ability to use a platform’s services without opt-in permissions (ii) decentralized management.

An example of using Web3 is for cryptocurrency wallets where you can freely choose your username. After registration, all the information and assets generated in the wallet, such as Bitcoin and Ethereum, belong to you. If you don’t like the service of this wallet, you can leave at any time and download another wallet. By entering a 12-word mnemonic, your information and assets can be safely transferred so you can have full control over your data and assets. This is a core feature of Web3.

However, when using a Web3 wallet, if you lose your details, you can’t get them back. Therefore, you need to keep the 12-word mnemonic by yourself, because once it is leaked, others can easily transfer your assets. This is one of the shortcomings of Web3 so in the future, social products on Web3 should have an authentication mechanism.

How AI Evolved

As early as 1940 during World War II in England, Alan Turing, successfully built a machine to crack German Morse code. This is regarded as the world’s first machine with artificial intelligence characteristics. In the 1950s, Claude Shannon who introduced the binary coding system combining 0 and 1 that we use today, first proposed the English term “artificial intelligence”.

Geoffrey Hinton, who won last year’s Nobel Prize in Physics first proposed the concept of large models, integrating multiple machines for deep learning. His student Demis Hassabis is famous for AlphaGo. AlphaGo’s breakthrough performance in the field of Go inspired Sam Altman to develop ChatGPT. In the 11 years before AlphaGo, the development of the field of artificial intelligence was relatively stagnant, and many investments failed to achieve significant results. It was not until the emergence of ChatGPT that people once again paid great attention to artificial intelligence. The breakthrough of AlphaGo tells us that the Go knowledge accumulated by humans for thousands of years may not be completely correct.

With the development of artificial intelligence, we may gradually realize that much of our existing knowledge may be wrong. For example, the success of DeepSeek is precisely because its founder Liang Wenfeng did not rely on traditional capital and manpower investment, but proposed a brand new method — DeepSeek Zero. This method no longer relies on the original artificial intelligence model and data, but adopts a brand new technical route and has achieved great success. Today, global technology giants including NVIDIA, Google, and Apple are learning DeepSeek’s innovative model, and its greatest success lies in breaking through the constraints of traditional theory.

AI — Breaking Traditional Thinking

The key to understanding AI is to realize that it can help us break through the limitations of traditional thinking and re-examine and subvert the existing knowledge system. The biggest significance of AI is that you can now use tools such as DeepSeek and ChatGPT, which can directly give you an answer, saving you the trouble of sifting and distinguishing information. Of course, the accuracy of AI has not yet reached 100%, and sometimes hallucinations or errors will occur, but its progress is huge.

AI has leveled the gap in knowledge acquisition. In the future, whether it is a primary school student with certain abilities or a college student, AI can provide them with similar answers. The difference in the future will not be the answer itself, but how you ask the question. The same answer, but different ways of asking questions may bring better information and affect subsequent results. Therefore, we need to learn how to use AI and how to ask questions in the future. This is the most important thing. As the saying goes, the people who know how to ask questions, not the people who answer them, lead the world.

Looking back at history, AI is the development of productivity, while Web3 is the development of production relations. In the past, we invented carriages and cars, to make us walk faster; we invented various machines to make our hands and feet more dexterous, but never involved the brain. Now AI is changing this.

Similarly, Web3 is also changing our organizational form. From tribes to countries, from religions to companies, all of these are improving productivity. Bitcoin is neither a company, nor a country or a religion, it is a completely new form of existence. Bitcoin is a moral blockade, and the future form of companies may shift in this decentralized direction.

Data is the new Currency

In an agricultural society, the most important means of production is land, while in industrial society, efficiency was improved through various tools. In the information society, the most important means of production is information. Since the 1970s, the richest people in the United States are not engaged in industry or agriculture, but investors like Buffett or information industry practitioners. Before 1978, China was still an agricultural society. In the past 45 years of development, China has experienced the coexistence of agricultural society, industrial society and information society. This is why China’s richest people include internet entrepreneurs.

Conclusion — The Opportunities of the Data Age

Looking ahead to the next 20 years, the richest people are likely to be those who have information and data. We have entered the data age, and the most valuable thing in the future is data, and it is circulated across borders. I believe that we will enter the era of data trade, that is, the era of data navigation, which may transcend national concepts and promote the further development of global trade.

Whether you choose to start a business or not, if you can skillfully use AI or Web3 thinking at work, you will get twice the result with half the effort. This is the biggest opportunity for everyone.

Internet pioneer, Web3 visionary and former Chairman of Meitu Mike Cai was speaking at BEYOND Expo 2025, Asia’s largest tech expo and ecosystem held in Macao.

Join thousands of data leaders on the AI newsletter. Join over 80,000 subscribers and keep up to date with the latest developments in AI. From research to projects and ideas. If you are building an AI startup, an AI-related product, or a service, we invite you to consider becoming a sponsor.

Published via Towards AI