Partnered With:

Trusted neutrality: system of mechanism design this | rhythm research about mechanism design series one

Read this article in 25 Minutes

An obvious mechanic is a good mechanic

What is a mechanism? A mechanism is a set of algorithms with reasonable incentives. The algorithm will coordinate and optimize the participants' behaviors to get the optimal solution of the public, and the incentive ensures that the original intention of the participants' behaviors is compatible with the public interests. A reasonable mechanism can make the system sustainable and benign development, and the essence of rationality lies in the credible neutrality of the mechanism. In order to achieve the credible neutrality of the mechanism, the most basic point is to transparently let the participants understand the fairness of the internal design of the system in the simplest form, and to achieve long-term consensus among the participants in the system in an efficient form.

In early 2020, V Wrote this article entitled Credible Neutrality As A Guiding Principle, which dissects the roots of mechanism design and the main design ingredients of Credible Neutrality. Translated by BlockBeats:

Let's consider the following scenarios:

· If the government spends 5% of GDP on specific public projects or industries, people may not be happy about it, but if the government is enforcing property rights policies, even if it leads to a greater redistribution of capital, people will have no problem.

· There is also a backlash if a blockchain project developer issues a large number of tokens directly to a designated person in advance (also known as "pre-mining"), but not if Bitcoin or Ethereum prints billions of dollars worth of tokens to miners in the form of POWs.

· If any social media uses strict censorship measures to restrict the flow of content containing certain political ideas, people will react strongly even if those ideas are not accepted by the general public. However, people do not care about the drivers who lose their jobs due to low ratings on ride-sharing platforms.

One can yell "Gotyou" at these irritating situations and bask in the self-satisfaction that seems to have exposed all sorts of social hypocrisy. Of course, this is sometimes true, but it seems to me a mistake to view carbon taxes as a bad act of state interventionism, while government enforcement of property laws is a strong enforcement of natural law. Not only that, those engaged in block chain security maintenance of miners is equal to thermodynamic industry professionals, think they deserve a certain amount of compensation, while those who block chain development technical staff worked so hard to improve the code, just have a little reward was identified as a "printing money", are also wrong.

Even if people often go astray when they systematize their intuitions, this deep moral knowledge is not worthless. In fact, there is an important principle at work, and it will also play a key role in establishing a free, efficient, fair, equitable and inclusive institutional mechanism to maintain and influence our lives in an all-round way. This principle is: credibly neutral at all times when making decisions about risky returns.

The formation of the correct mechanism requires precise algorithms and incentive measures

The first thing we have to figure out is, what is the mechanism? In essence, a mechanism is a combination of algorithms and incentives that, given input from various parties, use those inputs to judge the value of others in order to generate results and make decisions of all kinds that people care about. A good mechanism can make decisions that are efficient and matched to incentives, where "efficient" means it can pick the best outcome based on people's preferences, and "matched to incentives" means that people can participate "in good faith" under the incentives.

Examples of this mechanism are good, such as the following:

1. In the mechanism of private property exchange, users' redistribution of ownership through donation or trade is the "input", and the "output" is a database that determines the right to use each item, sometimes as a formal norm, sometimes as a hint. The idea is to encourage more production of products that are more efficient, so that they get the most value in the right hands.

2. In the auction system, the "input" is the bid, and the "output" is who can win the bidding and pay the corresponding amount.

3. In a democratic election system, votes are "in" and government seats are "out".

4. In the mechanism of social media, the corresponding "input" is the number of people clicking for, against, like and forward, while the "output" is what kind of content and how to put it. Game theorists would argue that this is just an algorithm, not a mechanism because it lacks built-in incentives, but it is likely to be added in future releases (see Slashdot meta controls for information on past releases).

5. Among the PoW and PoS incentive mechanisms granted by blockchain, the "input" is the block and information edited by participants, and the "output" result is that a chain becomes the longest orthodox chain, and the block-maker gets the due reward, at the same time, it encourages other block-makers to obey the "order".

We are entering a highly networked, mediated and rapidly evolving information age in which the public is slowly losing faith in centralized institutions and exploring alternatives. As a result, these different forms of mechanics become more important and influence the way we interact. Not only that, but these mechanisms are formed through a process of constant selection, the crystallization of wisdom distilled from a thousand varieties of mediocrity.

What is credible neutrality?

Let me now turn to the important concept of credible neutrality. Essentially, if a mechanic has this nature, we can tell by looking at its design whether it is biased or not. Such a mechanism can treat everyone fairly, even in the real world where people have different abilities, needs and differences. In the following statements, "anyone who digs a block gets 2 Ethereum" shows credible neutrality, whereas "Bob wrote a lot of code, we should give him 1000 tokens as a reward" does not. Similarly, trustworthy neutrality is "posts flagged as bad by five people will be taken down", while "posts judged by our moderation team to be biased against blue-eyed people will not be shown". "The government awarded any invention patent period of 20 years" embodies the credible neutral, though invented in determining which conform to the conditions there are serious challenges, but "the government will as the treatment of cancer, and so the nominating committee for those trying to cure cancer in allocating $1 billion" cannot be considered faithful is the neutral.

However, there is no complete sense of neutrality. Blockchain rewards favor those with the resources to control hardware and access cheap energy. Capitalism favours concentrated interests and the rich, discriminates against the poor and over-reliant on public goods, and political discourse is biased against anyone who stands on the scale of social appetites and prejudices. Any mechanism will anticipate the coordination failures it corrects in advance, and those that are underestimated will be ignored. Nonetheless, some mechanics remained neutral and performed far better than others.

The private property system is so effective not because it is an innate human right, but because it serves as a credible neutral mechanism that solves a large, if not all, of society's problems. Similarly, the reason it is possible to filter by popularity rather than political ideology is that it is far easier to convince people that a neutral system can be relatively fair than to persuade different groups to understand the validity of banning certain political views. So it's easy to understand why the revenue of on-chain developers is more questionable than that of on-chain miners. This is because miners are easier to verify than developers, and in reality identifying developers can easily be accused of favoring them.

It is worth noting that the above mentioned neutrality has to be trustworthy, not just neutral. An institution must not only be impartial, but also be trusted, so that its efforts at fairness can be seen by people from different sectors. At the same time, it has to serve those people who are free to collaborate on blockchain, political systems or social media. Mechanism is the foundation of the system, and only when everyone believes in the fairness of a mechanism, it will work well, and people will not have to worry about it being abandoned suddenly one day.

In other words, what we need is something akin to "common sense" in game theory, or, in ordinary terms, a broad consensus on legitimacy. In order for all people, even the uneducated, to understand this neutrality, any mechanism needs to be clearly displayed so that people's trust in it will not be easily shaken by hostile propaganda that makes it seem unfair and unreliable.

There are four main principles for establishing a credible neutral mechanism:

(1) No specific population or outcome should be mixed into the mechanism

(2) Open source and run under public supervision

(3) Keep it simple

(4) Do not modify frequently

The first point is easy to understand, going back to our previous example, "anyone who mines a block gets 2 Ethereum" demonstrates credible neutrality, whereas "Bob gets 1000 token rewards" does not. And, credible neutrality is "posts flagged as bad will be curtailed", while "posts biased against blue-eyed people will not be shown" is not. In these cases, "Bob" would be the designated person, and "bias against blue-eyed people" would be the designated result. At the same time, Bob is probably a top developer who has successfully completed some blockchain project, and these 1,000 tokens are his due, and this bias against blue-eyed people is something that you and I don't want to perpetuate. But in credible neutrality, these particular people and things cannot appear; instead, they are discovered only in people's actions. For example, in a free market, people find out through the price mechanism that Charlie's parts are not as useful as David's. As a result, no one buys Charlie's parts and he goes bankrupt, but David makes a lot of money and can continue to scale up and make more parts to make money. Therefore, the outcome of the mechanism should be determined by people, not rigid codes.

The second is understandable: the rules on how the mechanism works should be open to the public, who can also check that they are being run properly. But sometimes you don't want the public to know what the inputs and outputs are, which means you want to keep your participation secret. There is an article that goes into depth and explains in detail the desirability of high privacy. We are fortunate to be able to combine zero-knowledge proof with blockchain technology to achieve both verifiability and privacy (see here for more details).

However, the third point, seemingly simple, is the hardest to understand. A post on "centralization leads to overfitting" delves deeper into the controversies and deserves to be summarised. The simpler the mechanism, the fewer parameters there are and the fewer opportunities there are to favor or target specific groups, the article notes. If you have a mechanism with 50 parameters, and all these complex parameters interact with each other, you can find the right parameters in it to get the result you want, but if you have just one or two parameters, it's not that easy. It is easier to secure privileges for a broader group, such as firebrand politicians or the rich, but less so for a minority. Not only that, but the longer the timeline is stretched, the harder it is to achieve the desired outcome, because the veil of ignorance that separates you as the creator of the mechanic from the beneficiary of the mechanic grows over time, and in some cases, the benefits are not equal.

To solve this problem, we need to implement the fourth point, which is, don't tweak the mechanic too often. Changing the mechanism will not be easy, but in the process the veil of ignorance will be reset. In this way, the barrier is less, the opportunity for adjustment is more, in a timely understanding of the special positioning of different groups of people, make clear the influence mechanism of different adjustment measures, it is easy to protect friends and enemies.

Neutrality is not the only thing; effectiveness is just as important

As I mentioned at the beginning of this article, many of the most extreme forms of politics have the fallacy of centrism: the belief that if something can't be completely neutral, it shouldn't be done at all. In fact, this is the pursuit of narrow neutrality at the expense of broad neutrality. In this model, all miners are on a horizontal line, mining a block for 12.5 Bitcoins or 2 Ethereum; Similarly, all developers get the same reward, but it's just a thank you, no real reward. As a result, there is much less incentive to develop than to mine. Developers contribute far more to the blockchain than the 20% of miners, but the reality could be quite the opposite, at least under the current payback system.

In a broad sense, social production consists of many things: private goods, public goods, accurate information, good governance decisions, goods that are not valued now but will be valued in the future, the list goes on and on, and some of them are easier to establish credible neutral mechanisms for. However, if we stick to pure narrow neutralism, making no concessions and accepting only the extreme credible neutrality mechanism, only the simple problems of the simple mechanism can be solved. As a result, other user needs will be put on hold, there will be little hope of systematic help and broad neutrality will be difficult to achieve.

Therefore, in order to achieve credible neutrality, we need the validity principle. A good mechanic is one that actually addresses the issues we care about, but that usually means that developers have to listen to others at all times, even in a credible neutral mechanic, because a neutral mechanic is likely to be unpleasant, as the patent system is.

In some cases, this also means using alternative mechanisms temporarily until a credible neutral mechanism can be found. For example, early tokens and developer time-limited rewards in blockchain can be attributed to this mechanism. Similarly, a centralized approach to filtering users to detect special accounts when a decentralized strategy is not mature is also a vivid demonstration of this mechanism. But even so, establishing credible neutrality is vital, and we need to do everything we can to keep moving towards that goal.

Alternative mechanisms tend to be less neutral, which makes some people very worried that dishonesty will become the norm, or even that people will be completely manipulated by politics. This is where "failsafe" measures come in, the schelling Fence (Schelling fence) is a good example. This means that one of the ways to limit the amount of money a developer can raise is to make a fee, rather than a publishing fee, a schelling point, which prevents a developer from getting too much money up front. For example, when the reward system is wearing out and needs to be updated, but there is not enough time, a "freeze period" can be used to alleviate the problem. In "Layer 2" systems, such as ethereum rollup and Ethereum 2.0 implementation environments, fail-safe mechanisms can lock down certain network effects, but can also be disabled through coordination if problems occur. Also, if we foresee a system failure on the Voice platform, we can reduce the risk of failure by improving the free exit mode.

Although it needs to be improved in practice, in theory there is a credible neutral mechanism that can solve a variety of problems, such as:

1. Prediction market: Electionbettingodds.com & NBSP adheres to the principle of "credible neutrality" in prediction market; A website that can be used to predict the results of recent elections.

2. Second-order voting and financing: it can be used to reach agreements on issues of public governance and public goods.

3. Harburg tax: as a more efficient alternative to pure property, this tax could be used to distribute non-homogenous. Chemical and illiquid assets.

4. Peer forecasting: a more formalized "meta-regulation" approach.

5. Reputation system: This contains a transferable trust diagram.

New ideas are emerging before we know whether these ideas are feasible, and to understand the relationship between implementation principles and outcomes, we need to run multiple rounds of experiments to ensure good results under different rules. While encryption has simplified many things by making principles public, execution and results verifiable, and raw input secret, making the rules of any mechanism open to the outside world while still standing up to the test will be a challenge.

In principle, these sound rules are perfectly feasible and have been largely implemented. However, the market for all kinds of software platforms we use continues to emerge, and in this context, ensuring that the system is fair and doesn't favor one group over another is all the more important. We should not let platform operators or any powerful force shake the balance of justice. On the contrary, the most important thing is to establish a system of rules that everyone supports and trusts.