The absence of any information about Wei Dai can be explained by his work and the ideas that he promoted. Dai is a talented cryptographer who has created and still manages the Crypto ++ library of cryptographic algorithms based on the C ++ programming language. He is also an active participant in discussions about artificial intelligence, ethics, epistemology, and other philosophical issues on the Less Wrong forum. Eliezer Yudkovsky, a famous artificial intelligence researcher, co-founder of the Machine Intelligence Research Institute (MIRI), formerly known as the Institute of Singularity, praised his beliefs.
Dye’s interest in philosophy and politics is quite understandable. Back in the 1990s, a young bachelor of computer science from the University of Washington was fascinated by the writings of Timothy May, one of the founding fathers of the cypherpunk movement. Dai was inspired by the crypto-anarchist ideas that May preached; This new ideology was based on the belief that cryptography and software code can provide security for economic and political freedoms much better than any government system.
“I’m just fascinated by Timothy May’s cryptoanarchy,” Dai wrote in 1998. “In contrast to the goals of traditional anarchist communities, in a cryptoanarchy, the government is not destroyed temporarily but is banned forever, as it becomes absolutely unnecessary. This is a community where the threat of violence is negated because violence is simply impossible — participants cannot be tied to real names or geographic location. ”
By the mid-1990s, Dai had already discussed many topics in the framework of the electronic distribution of encryption systems, including digital reputation systems, game theory, privacy and anonymity of electronic money systems. Moreover, Dai introduced several of his own proposals: trusted temporary seals, encrypted TCP tunneling, a secure file-sharing system, and others. Shifropanki recognized that Dai makes a significant contribution to the common cause, but then they did not even know if he was a man or a woman, because they knew nothing about his personality.
But the real fame I give brought the idea, which he said in November 1998, immediately after graduation.
“Effective cooperation needs a medium of exchange [money] and a way to enforce contracts,” Dai explained then — the protocol I proposed would allow anonymous subjects to cooperate more effectively without the threat of tracking by providing that very medium of exchange and the way to enforce contracts. I hope that this is an important step in putting the ideas of cryptoanarchy into practice.
Traditional digital monetary systems use a centralized registry to track existing balances. Central banks, commercial banks, VISA, and other payment service providers control the database, which contains property rights.
According to Dai and cryptanarchists, the problem in this context is state control over financial flows through regulation, and participants in the system almost always have to disclose personal information.
“The goal of creating b-money was to introduce online economies that operate solely on voluntary terms … which cannot be taxed or regulated through the threat of the use of force,” he explained.
It is for this reason that Dai developed an alternative solution, or rather, two whole solutions.
The mechanism of the first decision assumed that all participants should keep copies of the register of transactions instead of a single centralized organization. As soon as the next transaction was made, the participants had to update their registries. Moreover, instead of real names, Dai suggested using public keys. Using such a decentralized approach eliminated the likelihood that any single organization will be able to block transactions. At the same time, users received a sufficient level of anonymity.
Dai himself considered the first version of his proposal impractical since it implied “simultaneous and synchronous use of the message channel.”
In other words, he did not solve the problem of re-expenditure. Alice could simultaneously send 2 units of b-money to Bob (“B”) and Victoria (“B”), sending a message about the transaction to different parts of the network. Both Bob and Victoria will sell the product to Alice, but only after they find out that only half of the participants verified their balance.
In this version, a copy of the registry was not kept by every member of the network. Instead, new concepts were introduced: regular users and servers. However, only servers that are nodes of the network kept copies of the registry. To verify whether the transaction was successful, Bob and Victoria would have to refer to a random sequence of servers. In the event of a conflict, the sellers would simply refuse to transfer the product to Alice.
The concept of b-money has never been implemented. It could not be implemented because “the b-money mechanism was not very practical,” Dai admitted on the LessWrong forum several years ago. Moreover, he did not expect universal acceptance of the concept even in the case of implementation.
“I think that b-money can act as a mechanism for providing remittances or fulfilling contracts for those who cannot or do not want to use the alternatives offered by the state,” he wrote to the participants of the distribution of encryption.
It is noteworthy that some of the problems of b-money are still not resolved or are not precisely indicated. Probably the most important problem was the proposed consensus algorithm. Now we know that the concept of Proof-Of-Stake has some weak points that Dai could not have foreseen: for example, it is not entirely clear how exactly to establish the fact of unfair behavior. And this is not taking into account the smaller nuances – lack of confidentiality due to the traceability of funds and the potential centralization of emissions/mining. Some of these problems are not solved in the Bitcoin network to this day.
Dai, who worked at TerraSciences and Microsoft after offering b-money, did not work on solving these problems.