On October 17, the 10th Blockchain Global Summit hosted by Wanxiang Blockchain Lab was successfully concluded. With the theme of "Technology for the Future", the summit attracted hundreds of guests, many important opinion leaders from all over the world, and dozens of surrounding industry activities. At the meeting, Dr. Xiao Feng, Vice Chairman and Executive Director of China Wanxiang Holdings and Chairman and General Manager of Wanxiang Blockchain, delivered a closing speech entitled "Redefining Token in the AGI Era". The following is the full text of the speech, which was compiled from the on-site shorthand, with some deletions that do not affect the original meaning.
Dear friends, everyone has worked hard. After a whole day of intensive communication and sharing, everyone may feel very brain-burning. Here, I would like to bring you some more relaxing topics.
The topic I want to discuss with you today is "Redefining Token in the AGI Era". In fact, the word "Token" existed before blockchain, but its connotation and definition have been constantly evolving and iterating. Therefore, at this point in time, we need to re-examine and define Token to prove that it is not an evil thing, but a legitimate and valuable term.
Three Cognitive Revolutions of Humankind
To deeply understand the core value of Token economics, perhaps we can start from human history tens of thousands of years ago. I recently read a new book called "Above Homo Sapiens". The book mentioned that a scientist divided human history into three cognitive revolutions.
The first cognitive revolution occurred tens of thousands of years ago, when humans invented language. The birth of language enabled human society to form, collaboration to become possible, and ideas to be communicated. Before that, these activities were extremely difficult.
The second cognitive revolution occurred thousands of years ago, when humans invented writing. The emergence of writing not only promoted communication, but also enabled all knowledge and experience to be precipitated and passed on. As a result, human society has developed better.
The third cognitive revolution took place decades ago. With the invention of computer language and code, human society entered a new era. I believe that this revolution did not begin in the era of big models, but quietly began when computer language and code were invented in the last century. Because code, human language and text are all tools for human cognition and expression, they jointly promote the progress of human society.
Today, in the digital society and virtual space, algorithms, computing power and data have become the basis for building a new society. In the AGI era, the term Token is frequently mentioned again to define certain functions of AI or as the basis for pricing.
From the era of code to the era of big models, humans are completing the third cognitive revolution.
Five stages of AGI development
Open AI divides the development of AI into five stages. Two months ago, we just entered the second stage. But with the launch of Open AI's o1, the second stage has been officially achieved. And just a week ago, Elon Musk's unmanned driving technology (Robo Taix) indicated that the third stage is about to be realized.
The first stage is the chatter, and ChatGPT has helped us achieve this goal.
The second stage is the reasoner, and the machine begins to have the ability to think. The "Turing Test" proposed by Turing in 1950 was to test the thinking ability of the machine. With the launch of o1, we can answer Turing's question: machines can indeed think.
The third stage is the intelligent agent, and the machine not only has the ability to think, but also has the ability to act. Musk's unmanned driving technology is a typical representative of this stage.
The fourth stage is the innovator. It is expected that in the near future, AI will be able to innovate from zero to one, no longer relying on human guidance and creation.
The fifth stage is the organizer, which is a more difficult task than innovation. It requires AI to be able to systematically build a parallel universe and operate it effectively.
According to Open AI's division method, we expect that the third stage should be achievable by the end of this year or before June next year.
AGI and ASI
Another division method divides AI into general artificial intelligence (AGI) and super artificial intelligence (ASI). Some people predict that in the era of super artificial intelligence, AI's intelligence will far exceed that of humans, even to the point where "everyone is Einstein."
When will the watershed from AGI to ASI come? When will the singularity occur? Kurzweil predicted in his 2005 book The Singularity Is Near and this year's new book The Singularity Is Closer that 2030 will be the time when the ASI singularity arrives. Recent predictions suggest that between 2026 and 2030, AI will enter the fifth stage of development, which is also the watershed between AGI and ASI.
UBI and UHI
Training AI requires data, but the quality of the data is crucial. Garbage data will only produce garbage results. Using AI-synthesized data to train large models often leads to degraded model performance. Therefore, what we need is random data generated by humans.
As long as we are born as humans, the data we generate is the data needed by AI and the best data in the large model, not the data generated by the large model itself and AI.
I have seen two research groups that have proved that using AI-generated data to train AI large models will cause the large models to crash.
So, as long as you are born as a human being and can generate random data, you should get feedback from AI, because this is important to AI, and you make AI smarter.
Therefore, it makes sense to propose the concept of national basic income (UBI). Because humans provide AI with data that makes it smarter and smarter, AI should give back value to humans.
In addition, with the advent of the era of super artificial intelligence, some people have proposed the concept of national high income (UHI). In this era, AI will create greater economic value, and these values need to be distributed to humans.
Among the representatives of the AI era, Sam Altman of Open AI and Elon Musk of Tesla both have deep roots in Token. Sam Altman not only launched the Worldcoin project, but also commissioned Stanford University to conduct a three-year UBI experiment. Musk may use Token for value exchange in interstellar travel or on Mars.
Value Laws in the Digital Age
In the context of the digital age, the value laws of the digital economy and digital business have undergone earth-shaking changes. The digital economy is characterized by high fixed costs, low marginal costs, or even zero marginal costs. This enables valuable products, goods, and services to achieve exponential growth, following a power-law distribution rather than a normal distribution.
The digital economy is cross-time, cross-space, cross-subject, and cross-organization. Therefore, we need blockchain and tokens to ensure that payment settlement and asset marking in the digital world can pass unimpeded. In the digital world, most of the classical physics may no longer exist, and the production factors have become algorithms, computing power, and data (or "three calculations").
In the agricultural civilization period, wealth was accumulated by land; in the industrial civilization period, it was accumulated by factories and machinery and equipment; in the digital economy era, these are of course still part of wealth, but the newly created wealth is more measured in digital form. Therefore, Token has become a measure of value in the digital economy era.
Three Elements of Traditional Financial Markets and Crypto Financial Markets
With the fundamental change in the way of value creation in the digital era, we are gradually building a crypto financial market that is completely different from the traditional financial market. Regardless of the type of financial market, its basic structure includes three major elements: accounting method, accounting account and accounting unit, and there are significant differences between these elements.
In the traditional financial market, the accounting method generally adopts double-entry bookkeeping, which is the accounting principle followed by all organizations at present. Its accounting account is mainly a bank account, and the accounting unit is legal currency, such as RMB. In contrast, the accounting method of the crypto financial market adopts a distributed accounting method (i.e. blockchain technology). Here, the accounting account does not refer to a specific currency type (although it may involve the US dollar, euro, etc.), but more importantly, its accounting unit and method have taken on a new form, which is completely different from the token or legal currency in the traditional financial market.
The crypto-financial market we are building, as another important field outside the traditional financial market, has undergone fundamental changes in its basic elements. Among them, the accounting method adopts the distributed accounting method, namely blockchain technology. The essence of blockchain lies in its decentralized and distributed ledger characteristics, which provides a global open and transparent accounting platform.
In contrast, the double-entry bookkeeping method relies on private ledgers notarized by a third party. Since private ledgers may be questioned for authenticity, they need to be regulated by regulations such as the Accounting Law and audited by accountants in accordance with regulations. In addition, in order to ensure the authenticity of the ledgers, the guarantee of the judicial system is also required. These measures are all aimed at ensuring the authenticity of private ledgers as much as possible and reducing the possibility of false records.
However, the distributed accounting method does not require these cumbersome safeguards. Due to its inherent immutability, all records are authentic and reliable and cannot be forged or tampered with. Therefore, from a cost perspective, the distributed accounting method is undoubtedly more cost-effective. For the whole society, this accounting method will also be less costly.
Thousand Years of Changes in Human Accounting Methods
Human accounting methods have undergone three iterations so far:
(1) As early as 3500 BC, single-entry bookkeeping was born. This method was discovered in the cradle of human civilization, the Euphrates and Tigris delta, in the Sumerian region of present-day Iraq. Archaeologists unearthed a clay tablet from 3500 BC. After research, it was confirmed that this was the earliest account book in human history, which used the single-entry bookkeeping method and simply recorded income and expenditure.
(2) By 1300 AD, double-entry bookkeeping came into being in the Mediterranean region of Italy. At that time, the city-states in northern Italy, such as Florence and Venice, conducted trade activities through the Mediterranean. These maritime trades were extremely complex, and the single-entry bookkeeping method could no longer meet the needs of accurately and comprehensively recording economic activities. Since maritime trade involved borrowing, chartering, joint ventures, and taxation, in order to clearly record each transaction, the double-entry bookkeeping method gradually emerged.
It is worth noting that the emergence of double-entry bookkeeping in 1300 AD not only marked a major advancement in bookkeeping methods, but also a direct response to the complex maritime trade needs of the Mediterranean region of Italy.
(3) Then in 2009 AD, distributed bookkeeping emerged, thanks to the introduction of Bitcoin blockchain technology. The value of distributed bookkeeping is self-evident. Its magnitude of change is a once-in-a-millennium event, far beyond the annual iteration of technological products. The far-reaching impact and value of this bookkeeping method are still difficult to estimate. I believe that no one has yet been able to tell us exactly how distributed ledgers will reshape human society and what unprecedented benefits it will bring us.
Definition and Classification of Token
Token existed before blockchain. From the perspective of value, the earliest tokens were change and coins. In computer systems, tokens are tokens, which are licenses to use computer systems. In the AGI era, tokens become units of data, used to measure the cost of using large models.
In the digital economy era, tokens are divided into the following categories as value symbols:
Payment Tokens: Countries are legislating and issuing licenses to issue compliant stablecoins.
Functional Tokens: Extended from the tokens of computer systems, referring to the right to use the network.
Securities Tokens: Involving STO, RWA, DeFi and other fields.
Reserve Tokens: Cryptocurrencies such as Bitcoin, as the underlying assets of crypto finance.
Two financial markets run in parallel
The financial market system based on bank accounts and the financial market system based on crypto accounts will develop in parallel. Interoperability between the two is expected to be realized in 2025. Through tools and channels such as stablecoins, ETFs, RWAs, STOs, and licensed institutions, the two markets will be connected and integrated with each other.
That's all I have to share today. Thank you!
Alex
YouQuan
Hui Xin
Joy
Aaron