On September 11 and 12, the 4th Global Blockchain Summit was held in Shanghai, organized by China’s Wanxiang Blockchain Labs. Vitalik Buterin, founder of the Ethereum Foundation and chief scientist at Wanxiang, took to the stage again this year to discuss the possibilities and challenges associated with the blockchain from the perspective of mechanism design in a speech entitled “New Development of Blockchain Technology”. The following is an overview of his talk.
Resource allocation and system design through mathematical research
Mechanism design is a field of economics that considers resource allocation rules and social system design. According to Buterin, mechanisms can be used for “incentivizing the production of things that are very valuable to not one person but to entire communities. They can be used for coordinating behavior in different ways.” Because of advances in mathematical research in recent years, opportunities to apply mechanisms to society more broadly are mounting. “I think that there are a lot of interesting synergies between [mechanisms and blockchains] and how blockchains as a technology can help implement many kinds of mechanisms, be a testing ground for many kinds of mechanisms, and what the limits of blockchain technology are in this area,” he said.
As examples, Buterin gave established mechanisms such as voting, auctions, and markets as well as recent focuses of discussion such as currency exchanges, including decentralized exchanges (DEX), and the Ethereum Name Service (ENS), a replacement for the Internet’s Domain Name System (DNS). ENS uses auctions, incorporating many economic and game theoretic components.
As an example of new ideas emerging outside of the blockchain, he introduced the concept of radical liberalism as found in Eric Glen Weyl’s book Radical Markets. Quadratic voting, also from the book, is an idea to reduce bias in voting, which is useful for optimizing fundraising for non-profit projects. Buterin criticized the huge costs of the current financial market’s high frequency trading (HFT) system in which players compete to buy and sell the quickest, saying that batch auctions offer a clue to fixing this.
He also introduced Harberger taxes, frequent batch auctions, automated market makers, which have begun to be used in blockchains, and combinatorial auctions for exchanging between more than two different kinds of assets, saying, “All of these ideas are ideas that have been thought about by a combination of academics and people looking to apply them in the real world. Even outside of blockchains, I think there is a lot of value from these ideas.”
All mechanisms have the problem of credibility
Solving blockchain problems
Buterin then proposed the following four solutions to blockchain problems.
(1) Frequent batching for miner/validator manipulation
Many mechanisms can be attacked by choosing transactions contained in a single block. The solution is frequent batching. This is, for example, waiting for multiple blocks such as block numbers “1, 2, 3”, “1, 2, 4”, “1, 2, 5”, etc., and finally processing these transactions in a standardized order. Since they are not accepted in the individual block order, even if one of the block proposers is malicious, the mechanism works normally as long as another in the batch is honest.
Mechanisms assume that information submitted into the mechanism is completely private, but blockchains actually do not provide that much privacy. If participants require the concept of privacy, different kinds of of cryptology such as VDFs or threshold decryption could be used. Zero knowledge proofs are another solution for privacy.
(3) Sybil attacks and voting systems
Sybil attacks are caused by one attacker having multiple accounts. Even quadratic voting requires that one participant have only one account. Although this is a difficult problem, several ID management solutions are emerging and actively being discussed.
(4) Collusion and MPC
Theoretically, on the blockchain, if someone were to say, “I’ll pay you ten dollars if you vote in the way that I want,” they could buy voting rights at a very low cost. In order to stop such collusion, it needs to be impossible to prove who voted in what way, so an effective countermeasure is multi-party computation (MPC). Multi-party computation can prevent participants from learning private information from the computation except for the final result. Using highly trusted hardware is another method but recently there have been attacks on hardware, suggesting that MPC is superior.
Reducing barriers to credibility
Although Buterin outlined these blockchain problems and solutions, he added that simplicity is important.
“If you can build something that just uses a blockchain, then this is great and this is the best,” he said. But reflecting on the reality that while many attention-grabbing new mechanisms are more efficient than their predecessors, implementing more complex mechanisms correctly often relies on a central operator, he concluded his speech with the following:
“Blockchains can help solve some of the trust problems in mechanism design, but they can’t solve all problems. And so blockchains should be used correctly and in the right combination with other cryptographic technologies. And between these two things, between blockchains and cryptography and offline solutions, offline oracles of different kinds, you actually can really significantly reduce the trust barriers to actually implementing many of these technologies in practice.”