a16z CTO: from Memecoin to network token, how to understand token design
Photo by KayMapDao
Original link: https://mp.weixin.qq.com/s/yCKTTZIYKVD66PvofwYvw
Statement: For the purpose of reproduction, readers can obtain more information by linking to the original language. If the author has any objection to the reproduction, please contact us and we will proceed with the modifications requested by the author. Reproduction for information-sharing purposes only does not constitute any investment proposal and does not represent the views and positions of Wu。
This article, translated by Cikey, proofreading by Punkcan, is based on the first speech by Eddy Lazzarin (a16z crypto's CTO) at Crystal Startup Accerator. Eddy Lazzarin will take us from Memecoins to tokens, in-depth analysis of the differences and similarities between the various tokens, and the path of gradual decentrization。
I have long been looking forward to this sharing, but before doing so, I would like to ask the question: who here is planning to issue a token
I have omitted or simplified much of this lecture because the subject of tokens is very complex. So don't hesitate and ask questions directly。
I have been discussing tokens with our entire investment group almost every day, and I have been studying them for several years, but I still feel that I have not said enough. So you have to talk with questions。
Normally, me and our team discuss tokens in dialogue. They share views on agreements and products, which can help me make more targeted recommendations。
Today we are discussing tokens, especially their practical application. I am not a lawyer, so this is not a legal advice, but some of my views as an engineer and an practitioner。
I've heard it so many times that tokens are essentially just an account balance in the encryption system. They can be achieved in many ways, some of which are not even invented. I hope you can invent some new ways. However, today we will focus on some of them。
First of all, let us start with the “Memecoins” because they have recently returned to the tip of the tide. Although they look novel, they actually existed even before the birth of the encrypted currency, even before the Etherwood. Initially, people used it in a way that actually used the original bitcoin coding fork, made minor changes, then named it and released it。
If you look at the 2013 Wayback Machine and Archive.org data on CoinMarketCap, you'll find many of these in the top 100. They are not new innovations. I'm not personally interested in these coins. I also hope that no one here will do it, and we will discuss the reasons later。
In short, Memecoins is a token with no clear purpose, which is their key feature. If they have a clear purpose, they are no longer purely Memecoins, but may involve the design and legal aspects of the agreement. It is precisely because of this “unpurposefulness” that they become purely speculative instruments, which essentially bring them closer to gambling, which is allowed under many legal frameworks. This characteristic therefore requires consistency。
Because of the speculative nature of Memecoins, their prices are extremely volatile. But it is not a flaw, but a characteristic of them. It should be noted, however, that Memecoins may be decentralised, but there may also be information asymmetries, even frauds. Some may plan to sell, control information in advance, or hide loopholes in codes. While these Memecoins seem ordinary, they may actually hide behind doors or other problems。
Next is the Stablecoins, which, although we will not discuss in depth, have important applications in the agreements, especially as collateral on the chain. Although the encryption circle has long been discussing the potential of stable currency in payment, its main use remains collateral. This situation may change over time。
The third category of tokens is Arcade Tokens, which are sometimes referred to as “soft coins”, usually the second in a two-currency model. Such tokens can be transferred along the chain, but prices are usually suppressed due to issuance and foreclosure mechanisms. Their supply is usually unlimited and can be compared to the small coins used in the game hall, which are useful in specific scenarios but do not have the potential to increase prices。
Next is the Network Tokens, which represent a go-to-centre agreement, integrated with the depth of the agreement. These tokens may be more volatile, but they are part of the economic model, such as the Inn, Maker, Uniswap and Compund. Such tokens are designed not only for the token itself, but for the operation of the agreement as a whole. Their distribution, redemption, creation and destruction are closely linked to larger systems。
In designing these tokens, we would like to pay particular attention to the balance between the issuance and destruction mechanisms. The issuance of tokens may expand the value provided by the agreement, but if liquidity is not controlled, it may lead to inflation. The design of a sustainable economic model is therefore essential and a balance needs to be found between distribution and destruction。
Before we proceed, any questions may be raised at any time:
Q: Would you consider something like Fly Tokens
"Fly Tokens" is not currently a widely accepted trade term or a specific token project, but may be a hypothetical or conceptual currency referred to in a given context. - Explain from gpt
At the moment, Fly Tokens is not a real token for the time being, because it's beneath the chain and cannot be moved on the block chain. However, if in the future it is transferable and part of a larger system, it can well be considered a token。
I had some interaction with the Blackbird team, and they were thinking about it. Fly Tokens may eventually be classified as a game token. For example, you can imagine paying for Fly Tokens at the restaurant, while the restaurant might take some measures to ensure that Fly's values are redeemed for relative stability。
When considering tokens, we should not easily assume their value or function, but should carefully analyse their practical use and mechanisms. Even now that Fly Tokens is non-transferable, if there is a chain process or macroeconomic mechanism in the future to deal with its issuance and foreclosure, it may be classified as a game token。
Indeed, more and more teams are exploring similar practices. I will discuss this in detail later。
When discussing tokens, I like to use a simple psychological model to help you understand. Imagine "water tap" for your ability to create tokens. It's not just a programming technique, it's the key to determining the conditions for issuing tokens. This is a very powerful tool. For example, in the Ether Workshop, the system can ensure that the certifying officer is remunerated by issuing a new token, rather than extracting it from a fixed pool. This does not mean that you are free to issue tokens without restriction, because that would cause serious inflation。
We cannot underestimate the power of the tap. When you increase the flow of tokens, you should ensure that they are used to pay for something of great value, thereby improving the quality and quantity of goods or services provided by the agreement. At the same time, however, you need to consider the “water drain” — that is, the mechanism for the eventual destruction, locking or removal of these tokens from circulation. This balance is important because it helps to maintain the price of tokens and makes it valuable in future payments。
The ultimate goal is to create a sustainable system in which the issuance and destruction of tokens can be balanced. While this balance is not easy to perfect, when designing the system, you must always remember that the ability to issue tokens is your basic tool, and that “drainholes” are key to ensuring that they are linked to the actual value created by the system。
Another point is that, in understanding agreements, it can be seen as a market with both supply and demand. Here, “water tap” and “water drain” represent supply and demand, respectively. The supply side refers to those who create demand for commodities, while the demand side refers to those who pay to consume them。
THE MODEL OF THE TAIFENG IS A GOOD EXAMPLE. IN THE ETHER WORKSHOP, THE NETWORK DEALS WITH INFLATIONARY INCENTIVES BY PAYING THE CERTIFYING OFFICER A SECURITY BUDGET, A PROCESS THAT REPRESENTS THE SUPPLY SIDE. THIS MEANS THAT THE SYSTEM ENSURES THE SAFETY OF THE NETWORK BY ISSUING NEW ETHS TO CERTIFYERS. THE DRAINAGE VENTS, ON THE OTHER HAND, ARE INCURRED WHEN PEOPLE PAY FOR THE CALCULATION, SOME OF WHICH IS PAID TO THE CERTIFYING OFFICER AND THE OTHER IS DESTROYED. FOR EXAMPLE, THROUGH THE EIP-1559 MECHANISM, A PORTION OF THE TRANSACTION COSTS WAS DIRECTLY DESTROYED, THUS REDUCING THE SUPPLY OF ETH IN CIRCULATION。
The demand side is reflected in the need for computing, storage and verifiable computing, for example, when users deploy smart contracts or trade in the overn, they have to pay for these services. The supply side includes network capacity and block space, and the availability of these resources determines how much demand the system can handle。
Overall, the design of the agreement requires a balance between supply and demand to ensure that the mechanism for issuing and destroying tokens supports the sustainable operation of the system as a whole。
Q: DO YOU HAVE A SIMILAR PSYCHOLOGICAL MODEL FOR INFLATION AND DEFLATION ECONOMICS
With regard to the economics of inflation and deflation, it was very difficult to design a system that would achieve a perfect balance, so I do not recommend pursuing that balance from the outset. Indeed, I do not believe that there is a perfect balance in nature that has always been maintained. I tend to think that moderate inflation is better than strict fixed supply, because inflation means that you can guarantee payment for certain things. Another important point is that the time when costs are charged in the system should be when valuable items are consumed。
For example, in Uniswap they pay transaction costs through consumer mobility; in Maker they pay for leverage and system-generated risks; and in the Etherport they pay for block space and costing。In short, it is important to ensure that people pay for valuable things。
Q: We used to have something like Bonding Curves, dealing with some problems. Do you have any advice on that? For example, should Bonding Curves or similar mechanisms be used? What specific issues do you think these mechanisms apply
With regard to the use of mechanisms such as Bonding Curves, Bonding Curves is more effective in creating markets with very limited liquidity, and they may set a fixed price curve, which may not fully reflect market demand。
Q: WHAT HAPPENS WHEN YOU MEASURE THE SYNCHRONOUS EFFECT AND SYNCHRONOUS DESIGN
in measuring the impact of drainage vents and drainage designs, there are visible official drainage vents, such as direct burning of costs, and unofficial drainage vents, such as contracts sent to “dead address”. these drains are difficult to measure in ecosystems. i divided them into visible and hidden:
- Visible drainage vents: For example, costs are accounted for and paid for, burning or other direct treatment。
- Invisible drainage: for example, trust and possession of tokens or their use in the market。
in the case of collateral in some other uses, there are many hidden destruction mechanisms (“syns”). i am cautious about the design of agreements that rely on these implicit destruction mechanisms, which are more difficult to quantify. while these hidden destruction mechanisms may be very powerful, they should be considered in the context of measuring them as much as possible, but the system should operate independently of these hidden destruction mechanisms. the hidden destruction mechanism may be very dangerous, and some of what appears to be the destruction mechanism is not。
It's often said, "Oh, people use it to pay, so it's a destruction mechanism, because they have a little bit of it in their wallet." But how do we measure this? Should your system rely on this? Moreover, this is to some extent a burden, as such a destruction mechanism could quickly disappear, which is not a physical double bookkeeping。
Then I set out some good and bad reasons, which are a brief summary of comparison, and I can say a lot about expansion。
First of all, on the issue of incentives, it's bothering me a little. Sometimes someone comes to me and says, "Eddy, I want to be a token, and I really want to be a token." And I said, "Well, what does it do?" And they said, "Oh, it'll pay for that." It's bad. It's not really a protocol design. It's just an infinity printing and spending。
In fact, it's easy to get caught in the wrong spot: you can create a token to pay for something, like having an unlimited advertising budget that you don't pay for. It's not a real protocol design, or even worse, it's going to be a huge problem with your reasoning about the system。
As an entrepreneur, your goal is to find product market convergence, you're making a product, and you want to make sure that this is something that people really care about, are willing to use and pay, because you create more value than you put in. If you spend your tokens on the wrong incentives, you may distort your understanding of product market convergence. Like if someone says, "I have five million in income, and I feel good," but you check and find that their advertising budget is 20 million, then you know it's not true because they're paying for it。
In traditional companies, it'll be on the balance sheet, as you can see. In the area of encryption, however, it is difficult to measure a token if it exists elsewhere. You would be surprised how little has been done to quantify the cost of spending on some kind of token incentive for the clients of the plan。
So what I'm saying is that you shouldn't use tokens to pay for things, you should use tokens, and good agreements will use tokens to pay for the right things. But if you simply use tokens as a way to squander money, you will not learn much, nor will you create long-term values that go beyond single-manufacturing cycles. There is a lot to say, but what you really have to do is pay the network effect and promote the network effect of the system. It is difficult to make sure that what you pay for is valuable, and you want to guide the supply of development to be real, more needed and effective. But there's a way to do that, for example, Compund is the first income farm I know, and they'll pay for it by giving the user a token, when they deposit several collaterals into the agreement。
It's a relatively simple way of authenticating, because you know what collateral is, and you choose which collateral to use, so you can know it's useful and valuable, and you can pay properly. That's all the system needs. But if you consider your agreement, what is the key to developing the network effect? What is the supply side? What is the demand side? How can we upgrade quality supplies
When it comes to the need for subsidies, extra caution is required, as it may make the system more complex. You need to focus on those factors that contribute to the growth of the network effect. Another issue requiring special attention was security, which applied not only to the underlying agreements such as the ETA, but also to many others. You have to pay for security, and “security” here is a very broad concept。
Trying to find a different word to replace this, but it is difficult。
By way of example, it pays security through the issuance of tokens and compensates the certifiers to ensure that the supply of block space is of high quality. This is what happens — they ensure the quality of block space, that computing resources are paid for and replicated and validated in the system. This is at the infrastructure level. But you can also consider this at the application level. That's why Compund and Maker regularly pay for safety。
For Compund, the concept of security is that all assets are concentrated in one pool and there is a risk that dangerous or low-quality collateral assets are added to the pool. This poses a high risk, as if a poor collateral were added, someone might deposit a large quantity of poor assets and take away all valuables. Thus, the protocol was designed specifically to protect the pool deposits, which it has been doing. The currency holders of Compund will carefully vote on the admission of new encumbered assets and on the debt ceiling for the encumbered assets. Their goal is to ensure that suppliers do not lose their collateral or deposits, and borrowers pay their fees and debts. The agreement thus ensures the security of the system, and the existence of tokens is intended to ensure that such security is achieved。
The same is true of Maker. I mentioned a little earlier, Story Protocol, and if you have a network where there are a variety of intellectual property rights that can be commercialized or derived, you can imagine a very critical concept of security that is that I commercialize intellectual property rights legally. It is difficult to measure in the real world, and you need to do a lot of legal work and due diligence to ensure that this is a true and legitimate intellectual property right and to declare that it is indeed owned by those who own it. But if you take that for granted, if the agreement ensures the legitimacy of all assets, then the product of this network will be very high quality. I can say this to any application in the field of encryption, and the product needs to ensure authentication and credible neutrality. If you can pay for it in your tokens, it could be a long-term agreement。
Q: CAN YOU IMAGINE THE IDEAL AGREEMENT AND ALL ITS REQUIREMENTS AND THEN BUDGET FOR THEM, SO THAT IS THE BUDGET OF THE AGREEMENT
Yeah, think about what your deal offers. And when users get to it, they ask, "How's the product? Is it reliable? Is it of high quality?" You need to make sure that it remains of high quality. While “quality” sounds like a simple term, its core is “certification” — although the word “certification” sounds less attractive, its importance cannot be ignored。
Q: DO YOU THINK YOUR DEFINITION OF SECURITY IS AN INTRINSIC PART OF THE VALUE OF NETWORK TOKENS
More details。
Q: THESE NETWORK TOKENS, AS YOU DEFINE THEM, SEEM TO BE A KEY FACTOR: YOU CAN'T HAVE A NETWORK TOKEN WITHOUT PAYING FOR SECURITY。
My next example does not include this, but the key is that network tokens should pay for safety. The value of network tokens comes from the fact that people pay for safe things, that they consume, and that tokens should capture a part of those values in order to maintain their own prices, thereby maintaining the functioning of the system. All of this lies in making the system sustainable. There are also systems that pose risks, such as Compund and Maker, which have systemic risks. These risks, if not diffused among protocol users, can be internalized to the token holders. The advantage of token holders absorbing these risks and pushing them in is that you can pay people to manage them. People should be paid to take risks, and when they take risks, they can work to mitigate risks and ensure that the system continues to function properly. Some systems do not require ongoing payments for safety。
For example, non-changeable systems can have very low security costs. In the case of Uniswap, for example, Uniswap is completely immutable, each version is independent and does not require a sustained security budget. The reason for this is that the design attributes of Uniswap externalize security issues. Each Uniswap pool is independent, and traders accept the risk of the pool when exchanging assets in the pool. As a result, there is no need to pay security costs on a continuous basis and there are no systemic risks that need to be carefully monitored. But not all agreements can be designed to be immutable。
For example, you cannot achieve immutability in Compunding because it needs to determine which collateral assets are safe because liquidity is shared among assets。
Therefore, if you can make the system immutable, you can save significant costs and possibly minimize or eliminate governance. Some people create tokens for bad reasons, for example for voting. What's the point of this governance token? The vote did not mean that you should not vote in the system, but voting was the last resort. Politically, the same is true: voting is a last resort, that is, we cannot organize society at the right organizational level, so we need to do the most painful, expensive and difficult things。
The vote was very cumbersome。
We ask everyone to weigh a complex, multidimensional issue. You should use the ballot only if it is necessary. So far, the low level of participation in token governance and the high opportunity costs make it difficult to motivate experts to conduct research. It is not completely ineffective. It is a cumbersome and slow process that should be used only as a last resort, if necessary。
This should be done for all things that can be automated and built into systems or entrusted to market mechanisms。
If your claim to the value of a token is just to vote, you are deceiving yourself. No one really wanted to vote, and voting was the responsibility of the currency holders. What you really want is for all the stakeholders in the network to be inspired to work towards the same goal. This is a very powerful attribute of tokens, which I think is very difficult to achieve compared to equity。
The basic idea is that you have something close to you that is important to the functioning of the agreement。
You have suppliers, users, technical partners, etc. You can make deals with them, but it's better to have a token than a simple small cash transaction, and they want the agreement to work, and you can establish links between these different adjacent areas, which could have been competing. This combination is much more effective than a stand-alone one。
In this setting, you can deploy tools such as belonging, milestones, locking。
I'll say more later. You can imagine some automated methods, some manually. For example, early manual approaches include commercial development transactions, partnerships, exchange of conditionalities, etc. Automation methods, such as depositing collateral, locking them in, and committing themselves to do verifiable work, will be remunerated when it is done and will be unlocked over time。
We can deal with this very creatively, and I think that the effects are really underestimated, because the most visible effects are often subtle. For example, mutual consent, lack of competition, sharing of open source codes, promotion of each other ' s products in social media are important components of the development of encryption。
I mention this only because it often happens: payment. Some say the token is paid for. I'm talking about Internet tokens, not game tokens, not stability coins. I'm talking about protocol coins. The evidence to date shows that people are reluctant to pay in the agreed tokens, and that in practice there is friction. Depending on the mode of operation of the existing network, one must first obtain the balance of the token and consider the risk of holding potentially volatile assets. People do not like to pay with volatile assets, even if they consider it reasonable。
People may be reluctant to sell tokens if their value is considered high; if they do not think it is worth it, they may not want to hold large amounts of tokens for payment. This is not to say that it will never work, but we need to make many interesting evolutions of these tokens to make them effective. This part is somewhat vague. Finally, if the payment is not a direct synchronization, the withdrawal of the costs from the payment can be a synchronized method, but it is not a direct synchronization that makes payment itself subject to the use of tokens。
Q: WHAT IS THE POTENTIAL CONTRIBUTION OF GOVERNANCE TO SECURITY IN PRACTICE
Governance and security, in the example you have mentioned, like Compund, is indeed an aspect of helping security. Governance is not necessarily a bad thing, and it can be achieved if you know how decisions are made to protect networks that cannot be automated, but one person must be involved. For example, determining which tokens are sufficiently secure for network security to be included in a mortgage pool requires human intervention rather than fully automated decision-making. Governance can achieve this, and there is evidence that it can operate effectively, as has already been done in the DeFi agreements such as Maker, Compund and others, and some security committees deal specifically with system gaps. Governance does help security, but it also introduces a potential failure and political risk. You have to think seriously about how to design it. Compund and all other lending agreements are devoting considerable attention to how experts can review their agreements and pay for them. They need to ensure that at least some experts publicly weigh these changes, and their token holders are assessing them. I'm not saying it's impossible, just it's a pain. Can you do it without relying on it? Can you minimize it? To the extent that you can minimize it, you make the system stronger. Of course, I do not think it can be completely eliminated in all cases, but that is the framework。
Q: TRANSLATING SUBJECTIVE INCENTIVES INTO OBJECTIVE ASSESSMENTS
MARKETING AND ASSESSMENT, FOR EXAMPLE, IN TERMS OF MARKETING, IF YOU SPEND $20 MILLION AND YOUR INCOME IS ONLY $5 MILLION, YOU NEED TO FOCUS ON THE ACQUISITION COST (CAC) AND LIFE-CYCLE VALUE (LTV) RATIO. IF THIS RATIO IS NOT GOOD, IT MEANS THAT YOU PAY MORE THAN YOU GET, WHICH MEANS YOU NEED TO OPTIMIZE THE COST TO THE CLIENT。
Protocol design. In protocol design, you need to consider what you pay for. For example, the early encryption project Steam set is a social network in which users obtain tokens through comments. Such incentives have led to a large number of garbage reviews, as people are concerned only with obtaining tokens rather than the quality of the reviews. This suggests that it is essential to verify the quality of supply in incentive mechanisms。
How can you objectively measure the content of your incentives? It would be preferable to motivate those elements that can, to a certain extent, automatically validate their quality, which directly produce high-quality network effects rather than vanity indicators. For example, there should be incentives not for meaningless random transactions but for high-quality supplies. You know, quality dialogue is difficult to measure. Capital deposits are relatively easier to measure。
Q: BASICALLY COMBINE SIMPLE MARKETING INDICATORS WITH THE ACTUAL OBJECTIVES OF THE PRODUCT。
Yes, but not just marketing, but also security, mobility and any value created by our systems, which may or may not be marketing。
Q: DO YOU THINK A PROTOCOL SHOULD BE DESIGNED AROUND INCENTIVES OR WHAT YOU WANT TO PAY FOR? HOW DO YOU SEE INCENTIVES CHANGING OVER TIME
Well, that depends on the application of the scene and the way it is used, and whether the goal you want to stimulate changes over time. What you want to promote may change over time, and how can this be considered
That's why I'm particularly fond of manual incentives. While automation is crucial in achieving true decentrization and sustainability, the ultimate goal is to decentrize and ensure that even if you leave the system, it will function properly, which is a good test. It is clear, therefore, that automated payments are required to maintain the system, but at an early stage you should be active in distributing tokens manually, bringing in the right partners, contributors, etc. And in this case, the content of your incentives should change over time, because you will constantly get signals from the market about what works and discover new bottlenecks and another problem after solving one problem. You should adjust incentives based on verifiable information。
Q: PEOPLE USE TOKENS AS “PRINTERS” AND THEN VOTE TO RATIONALIZE THE PROCESS, BASICALLY DOING EVERYTHING THAT IS NOT RIGHT. THE VOTE IS PART OF THE DAO. WHAT WILL YOU VOTE FOR? WHAT CAN IT DO? NONE OF THIS IS CLEAR. CAN DAO BE PROPERLY IMPLEMENTED? WHAT DO YOU THINK
Yes, I think so。
In fact, there are things that are important, such as how they are correctly implemented in law. If you haven't studied or learned, I suggest you take a look at the Aragon framework, and we suggest you use it to manage DAO. It's a legal entity in the United States, but you don't need to use it in the United States. This framework can be an alternative to a company and suitable for DAO members. It allows them to vote, to express legally binding opinions and to protect their responsibilities in such voting, while paying taxes and complying with the law, without the need for a system of hiring and dismissal systems such as those of regulatory regulators or traditional companies. So, first of all, it helps a lot in implementing the DAO. Many DAOs have failed to properly address legal risks when trying to achieve their goals, resulting in many unnecessary risks。
On the other hand, it is necessary to clarify the purpose of DAO. I think it's ridiculous just to pay and vote, which is not the function of DAO at all. It seems to me that DAO should be a group of different stakeholders, representing the human beings involved in the agreement. Uniswap, for example, is a simpler example, because it does not have to pay for safety, but it does charge. These are users of tokens, users of Uniswap, and users of Uniswap who build applications on Uniswap. They share common interests around expanding ecosystems and hope that agreements and foundations can represent and implement these interests。
THIS IS A REASONABLE DAO FOR ME, AND NOW THEY'RE DOING IT, AND IT'S REALLY HARD TO DRAW CONCLUSIONS, BUT DAO SHOULD BE ABLE TO ALLOCATE FUNDS FROM THEIR VAULTS, SUPPORT SPECIFIC PROJECTS, UPGRADE AGREEMENTS, ETC., RIGHT
That's understandable. The question of how to avoid the right to vote being tied only to the time of accession is indeed interesting and subtle. I briefly refer here to the fact that simple token governance, like that, is not good, is that you have the right to vote as much as you have. In fact, there are many more interesting ways in which we can have different voting mechanisms, automatic committees, anti-witch attacks, even random voters. There are many creative and interesting possibilities. I can talk all day about more rational distribution of voting rights. Indeed, there is a certain advantage in simply looking at who has more coins, and it is true that those with more interests should have more say. But the key is how to obtain high-quality signals from the whole ecosystem, a complex and delicate task。
Q: We also discussed how the value of tokens guarantees the security of networks and agreements. In the case of MakerDAO, they have two tokens, one for governance and the other for stability. The two tokens are clearly separate, but we can also ask whether they need to be separated. Can currency holders pledge directly in their original currency rather than in two different currencies
Actually, two of the Marker's tokens are because one is the product itself, and the other is the network token. I will not call it a “util currency”, it is actually a stable currency. If you're a user, then you use a stabilizer, but if you get it by leverage, it's like a debt certificate waiting for you to buy back and redeem your collateral. So it's kind of like a "dual-use" token, with some nuances. One of the most important design decisions is that the Marker team has simplified the responsibility of DAI holders. People with DAI need only care about one problem -- DAI can stay stable. This makes clear the social contract of the product, and all risks and problems are borne by MKR holders, while ordinary users simply have to believe in DAI stability。
Then we talk about the path to centralization. We have a chart, and the vertical axis represents the degree of decentrization, the more the downwards represent the centralization; the horizontal axis represents the protocol function, the stronger the right representation function. Our goal is to reach the bottom right -- highly decentralised and powerful agreements. Some might ask, "Why are we pursuing centralization?" It's actually a part of the product. While many users are less interested in the technical details of decentrization, they are concerned that their money will remain valuable, that it will not be eroded by inflation and that their banks will not freeze accounts at will. So people really care about these decentrization-related issues, but their needs are reflected in market choices。
When you create a decentralised network, it attracts many developers because it is the most solid infrastructure. Decentralised and high-quality protocol functions are our ultimate goal. We recommend a gradual decentrization — starting with a highly centralized but powerful state, and then gradually moving to the lower right corner。
We haven't issued a token yet, so I'd say, here's the thing:
Our goal is to begin with a good start. These points on the left are not an exhaustive list, but are examples of things that, if they happen, mean that we may become more central. We have diverse stakeholders, more people in economic activity, different clients, systems that are not licensed, and we have reduced single-point failures. These features begin to depict a decentrized system. As we become less central, we become closer to having a transferable token。
AT THE OUTSET, IF YOU PUBLISH A NEGOTIABLE TOKEN IN A CENTRALIZED CONTEXT, YOU WILL FACE GREAT LEGAL RISK. HOWEVER, AS THE PROCESS OF DECENTRIZATION PROGRESSES, THESE RISKS ARE GRADUALLY REDUCED. I'LL USE A LITTLE CHART TO ILLUSTRATE THIS LATER. YOU CAN USE THE CREDIT SYSTEM, AND THE BENEFIT OF THE CREDIT IS THAT YOU CAN INSPIRE PEOPLE WITHOUT FEAR OF THEIR LOSS. THE SO-CALLED LOSS IS THAT A LARGE AMOUNT OF COINS WERE PURCHASED, BUT BECAUSE YOU ISSUED IMPROPER TWEETS THAT CAUSED THE PRICE TO FALL, THEY WOULD SUE YOU AND THE SEC (THE SECURITIES AND EXCHANGE COMMISSION) WOULD INTERVENE. THIS IS CLEARLY THE SITUATION THAT WE WANT TO AVOID。
In order to avoid this situation, we can make tokens non-transferable. In this way, people can earn something, but it has no market value and they cannot know its value exactly, but they know that they are accumulating something. Indeed, there is ample empirical evidence that these incentives do respond. As decentrization deepens, perhaps one day we can make tokens transferable. We may need to lock teams, foundations and investors, which is good for a long time. Then we can allow people who have received coins through airdrops or contributions to transfer them and to form market prices。
With regard to testing networks, one strategy is to integrate all functions and token transfers, claiming that this is only a testing network. And then, when you feel that everything is ready, you can say, "Now we're going to do the same thing on the Web."And move the balance of the test network to the main networkI don't know。
But if you keep the same balance when you move, this so-called “test net” may already have the characteristics of the main network from the beginning. So, in this case, you need to be extra careful. The greater the blurring of the boundaries between the testing network and the main network, the greater the potential risk. If it's actually the main networkThe user may still bear the lossAs a result, the market also fluctuates。
It is difficult to describe exactly, and I am not a lawyer, but the question is how it looks. I'm not kidding. I mean, it's really about whether you're contributing to the formation of a high-risk market, whether you're committed to money and people are giving you money and eventually losing it. If that were the case, they would strike you hard, no matter how cleverly you use legal terminology。
So, if you are sincerely moving in the right direction and making this clear, while remaining honest, without a mad commitment, and when it is shown that it is difficult to create a well-functioning business, to create a de-centralised and well-functioning encryption agreement, you may be in a relatively safe position, even if you are in some respects in a grey zone, provided that your intentions are extremely clear and good, and that you do your utmost to avoid harming people。
Many networks are now launched with a test network, all of which can be operated and transferred and then accessed by air drops that are highly relevant to the activities of the network. If people have to earn future airdrops by actually working, you tell them, “This is your score, you do some real jobs”, then it's not a money investment, but they are legally helping to build networks. If you let people transfer money to you and promise them a lot of valuable points in the future, it sounds like a sales act。
A simple test of decentrization: if you disappear, if your team disappears, if the issuer disappears, what happens to the price of the token? If the token price is zero, then you're not really decentralised. If the currency price would be zero as a result of a single failure, the system would probably not be decentrized。
If people see tokens as an investment, it is a grey zone. But it's a simple test that allows you to think about the degree of centralization。
One might ask, "Why don't we just start at the bottom?" Why not issue a token at the outset and then gradually increase its functionality? The reason is complex, but very clear. First of all, this is the legal risk you face at all stages: the risk of issuing tokens in a state of centralization is extremely high; the risk is reduced as decentrization deepens. And when you have a rich protocol function and are more centralized, the risk is further reduced。
In cases of complete decentrization, it would be very difficult to find market convergence for products. Because you need to convince 95% of the people who use your system to try an experiment that might or might not work, which is almost impossible. In the case of centralization, it is easier to find product market convergence, because you can control the situation and try constantly until market feedback tells you it works. So if you overlap these paths, you can see a clearer path: Market convergence of products is found in the case of centralization, which is then gradually decentralized while maintaining functionality。
There's one interesting thing about decentrization and Memecoins. If you publish a & nbsp; Memecoins, and it's legal, and you're doing everything in accordance with the regulations, then you might think about trying to find a product match after a while. But once you begin to profit from this memecoins, for example, by selling them in large quantities, or by promising to use them as the main token of the agreement, you may actually be able to reduce the degree of decentrization and bring yourself back from the safe zone to the dangerous zone. This is not a one-way pass, and you can't guarantee that you will never become a security. In fact, you might become securities again. So, how to convert Memecoins to a network token is more complex, and that's why we recommend this path。
Q: MUCH OF THE DISCUSSION REVOLVES AROUND THE DESIGN AND METHODS OF TOKENS, BUT WHEN IS THE BEST TIME TO START THINKING ABOUT ISSUING THEM WHEN BUILDING A PRODUCT? IS THERE A POINT THAT WHEN YOU SEE SOME COMPANIES, YOU THINK THIS IS A GOOD TIME TO START THINKING ABOUT THE DESIGN AND DISTRIBUTION OF TOKENS
The sooner the better. In my opinion, you should first design an agreement and then consider the role of a token. In fact, I think there are no exceptions, and when you complete the idea of the protocol design and the system works clearly, you find the location of the tokens obvious。
In practice, agreements may not function properly without tokens. There is a model in which you can build an agreement before having a centralized application and operation. There is no reason to combine them. Chris Dixon, for example, has a point to consider: hypothetically Coinbase made bitcoin. They could have done it, though they didn't. As you can imagine, the first exchange might say, "I want a digital asset that you can trade." They eventually created Bitcoin, and it is now completely decentralised, open-sourced, all the characteristics we like have, perhaps early miners, with a considerable amount of bitcoin, but it really belongs to everyone, and it is a completely decentrized agreement。
At the same time, they could have a traditional business as clients of the agreement. So your business is a subsidiary business. Consider how they complement each other and do better together. It's a very convincing and reasonable way to play: you can have a company you like, but you need an agreement, or you like an agreement, but you can think of a company that is willing to be a customer of the agreement, and then develop, promote and make it self-sustaining, like the Uniswap Foundation, which is really independent of Uniswap Labs. The two may go in different directions and even compete with each other, which is how the world is encrypted。
Q: Besides Uniswap and possible DYDX, which companies did well
Apparently, Bod did very well, they all came from a16z. I would like to know how other companies have successfully combined centralized operations with tokens and how value is distributed between tokens and equity. If someone asks you, "Is the value going to equity or currency?" How do you answer that
There are some early examples that we can discuss below. With regard to the second question, because of the limited time available, I have to answer briefly. When we invest, we usually like to have both tokens and shares. This is not because we want everything, but because we want to keep up with entrepreneurs. Entrepreneurs should take this decision at a later stage, which should not be a constraint for entrepreneurs. In many cases, they would have thought that the project would go in a certain direction, for example, we thought that it would be a single agreement and a token, that equity would be a tool, but then the company might disappear and then they decided to do a neighbouring business, or, conversely, they decided that equity was not so important and that they wished to become one of the many companies built on that agreement. Therefore, at an early stage, I suggested that it would be difficult enough to maintain an open mind and to find effective market convergence, rather than limit itself in the early stages and allow it to be explored freely. You will learn gradually in the development process。
Next, I'm going to go through the last part quickly。
This is a hypothetical schedule for the issuance of tokens, which contains much information. According to this schedule, it will take at least six months to issue a token. This period includes not only protocol design, network selection and implementation of protocol codes, but also preparation for the generation and release of tokens. In the six months during which tokens are generated and issued, you need to negotiate with the exchange, conduct security audits, deal with the institutional custodians, process the custody of assets of staff and teams, establish foundations and governance structures, file files, and communicate with market manufacturers, analyse air drops and other currency distribution activities. The amount of work involved in issuing tokens is a hypothetical schedule, and in fact each project is different. In short, the issuance of tokens is not just done by deploying them to the main web site, but also by dealing with a complex range of matters. Fortunately, these things are getting clearer and I can help answer the relevant questions。
Finally, I did not have enough time for in-depth discussion, but only four ways to distribute tokens。
The first is the sale of tokensIn other words, the sale of tokens to investors is not problematic and the law is clear; but sales to the public are a major problem and do not do so. This is one of the most dangerous single risk factors。
The second approach is positioningI DON'T KNOW. WHEN TOKENS ARE ISSUED, ESPECIALLY CONTEMPORARY ONES, AS PART OF THE AGREEMENT, PEOPLE NEED TO VISIT THEM, WORKERS NEED TO SELL THEM TO COVER OPERATING COSTS AND FUTURE CONTRIBUTORS NEED TO BUY THEM. YOU WANT MARKETS TO MATURE, BUT IF THEY ARE WEAK AND UNSTABLE, THEY SEND BAD SIGNALS. A MATURE COMPANY HAS UNDERGONE MANY TESTS DURING THE IPO PROCESS AND REQUIRES SUFFICIENT LIFE TIME TO CARRY OUT THE IPO. WHEN TOKENS ARE RELEASED EARLY, THEY ARE VISIBLE TO ALL AND MARKET PARTICIPANTS ARE INVOLVED, MAKING TOKEN PRICES A PUBLIC VIEW. BY CONTRAST, NORMAL COMPANIES USUALLY EXPERIENCE SIGNIFICANT EVENTS WHEN STOCK PRICES FALL SHARPLY. HOWEVER, CURRENCY PRICE FLUCTUATIONS ARE SOMETIMES CONSIDERED NORMAL. I HOPE YOU'RE ABLE TO STABILIZE THE PRICE OF THE COIN AND MAKE MARKET BEHAVIOR LESS CRAZY, BECAUSE YOU DON'T WANT PEOPLE TO BE WRONG ABOUT THE EARLY AGREEMENT。
The third way is automatedI don't know. As I mentioned earlier, these procedures include security budgets, payment for verifiable quality work, supply side, etc. Essentially, these procedures are the automatic issuance of tokens through mechanisms to reward the type you wish。
The last way is manualLike airdrops. While airdrops have significant advantages and risks, there are also some advantages. Like the Grant program, awards and partnerships promoted by Optimism, these are all possible approaches at an early stage. Typically, you can consider how to distribute tokens to long-term stakeholders who will develop, raise funds, establish their own business or bring in users on your agreement. Such an approach would best contribute to the long-term success of the agreement。
