Arweave is an "atypical" blockchain project. Most people know nothing about it. Those who know a little about it often regard it as one of the many decentralized storage projects that run with Filecoin. The few students who have the patience to find the white paper/yellow paper for research are inevitably confused after reading it. Because the whole article is about the unpopular concept of "permanent storage of information", there is no expansion, cryptographic innovation, DeFi support, value capture, etc., which can make the students in the currency circle and the chain circle shine. Who needs permanent data storage and pays for it? Life is only a hundred years, why should we care about the permanent preservation of human knowledge and history?

Arweave's founders and core team have their own reasons for being unique. However, as the Chinese translator of the Arweave Yellow Paper, I intend to interpret Arweave from the perspective of a typical coin-chain circle, so that domestic blockchain entrepreneurs and investors will not miss out on this major innovation. First, please allow me to transliterate Arweave as "阿维" (although this Chinese name is still under discussion in the Chinese community and has not been finalized) to facilitate its dissemination in the Chinese crypto community.

Avi and Filecoin/IPFS

IPFS is a pioneer in the field of centralized storage. Since its launch in 2014, it has grown freely like BT and has stored a large amount of data. However, in order to make IPFS a commercially available storage system rather than a random data sharing platform, it must provide service quality assurance. This is the problem that Filecoin wants to solve, namely the economic incentive layer of IPFS. From the proposal of the Filecoin concept to the "imminent" launch of the main network this year, it can be said that it has been delayed for a long time. As Protocol Labs, which has developed hard-core technologies such as IPFS and libp2p, why has it been so slow to get Filecoin?

The Filecoin protocol builds two markets: the data storage market and the data extraction market. Users with storage needs go to the data storage market to declare their needs: I want to store ** size of data, require ** copies, and store for ** days. The storage service providers (storage miners) in the market quote this storage demand. If the user accepts the quotation, he signs a contract with the miner and pays the fee. When the user needs to use the data, he goes to the data extraction market to make a request. The extraction miner then gives a quotation to meet the data access needs. The above process does not seem complicated, but there are several difficulties in implementation:

1. Miners need to provide unforgeable cryptographic proof that they store user data;

2. During the validity period of the contract, the agreement will continuously check whether the miner has saved the data as agreed. If the miner breaches the contract, he will be fined;

3. In order to encourage miners to store data, the capacity of stored data should earn more additional rewards than the idle capacity. At the same time, it is necessary to prevent miners from injecting garbage data to defraud additional rewards.

Filecoin designed Proof of Replication (PoRe) to solve the first problem, and used Proof of Spacetime (PoTS) and pledge mechanism to solve the problem. The third problem is solved by precisely adjusting the economic model1 and introducing the authentication of real users. Although Filecoin has solved the above problems to a certain extent, it has inevitably produced some adverse consequences. The first is the high complexity of the system. In addition to paying the necessary storage costs, miners also have to bear the high cost of proof and the option cost of staking Filecoin losses. It should be noted that computing is relatively more expensive than storage. According to the recommended configuration provided by Filecoin for small-scale mining2, an 8TB SSD hard drive only costs $300, but an AMD 3.5Ghz 16-core high-end CPU costs $700, and at least 128GB of memory costs more than $500 (for comparison, the recommended minimum memory for Avi mining is 8GB). The high cost of mining is bound to lead to high storage service prices in the Filecoin system. In addition, verifying real users is a delicate issue. If the verification is too strict, it will affect the user experience. If it is too wide, it will not prevent miners from pretending to be users, and the verification will lose its meaning. The balance between the two is difficult to grasp.

At the same time, as a crypto asset, Filecoin’s price is highly correlated with the overall crypto market, that is, it is highly volatile. If the price of Filecoin plummets, miners may take losses and leave, resulting in the loss of user data. In addition, large price fluctuations also increase the implicit option cost of miners staking Filecoin. The implicit option cost has been ignored by most PoS economic model studies. I think at least the option cost lost during the unlocking period should be taken into account (some even think that the option cost of the entire lock-up period should be calculated). The unlocking period is the period from submitting an unlocking request to obtaining a tradable token. During this period, the pledger cannot transfer the token, which is equivalent to giving up a current European option (unlike American options, European options can only be exercised at maturity)4. Taking Tezos as an example, assuming that the current price and strike price are both $2.53, the annualized volatility is 185%5, the unlocking period is 14 days (a longer unlocking period means a higher option cost), and the risk-free interest rate is 4% (does not affect the calculation result). Using the BS Option Calculator6, it is found that each European option is worth $0.363 (since the strike price is equal to the current price, the call and put options are equal in value), equivalent to 14.3% of the principal value. It can be seen that due to the high volatility of crypto token prices, the implicit option costs caused by staking should not be ignored.

The Filecoin protocol divides storage and extraction into two markets, which requires the establishment of two incentive mechanisms (and pricing mechanisms), and the user's data access rights are not guaranteed. Suppose you store important data through Filecoin and pay a certain amount of storage fees. Later, you or other users (such as your customers) will have to pay fees based on the market conditions of the extraction market to access the data. If the extraction market price is very high, it is equivalent to the data being "held hostage" by miners, and users are faced with the dilemma of either paying a high price or migrating data.

I read the Filecoin white paper in 2017 and immediately gave up on the project. My programmer’s intuition tells me that complex extrapolation solutions usually don’t work. What is an extrapolation solution? It is a solution that comes naturally to a problem without deep thinking, or a “taken for granted solution”. Filecoin’s extrapolation method is: since miners need to (continuously) prove that they have properly stored user data, the protocol should include a set of cryptographic algorithms to implement these proofs. As for the problem that highly complex proofs inevitably bring high system complexity and high costs, they can only be solved slowly in the future. However, Filecoin’s competitor, centralized cloud storage, does not require proof and verification. The cloud service provider and the customer sign a legal contract, and the law guarantees the customer’s access rights and recourse rights. It can be seen that as long as the cost of proof remains high, decentralized storage will find it difficult to provide competitive prices.

Although Sia, Storj and other protocols are technically different from Filecoin/IPFS, they are all contract-based decentralized storage protocols. That is, users and miners sign a contract through an agreement, users pay the fees specified in the contract, miners assume the obligations specified in the contract, and the agreement (or users) checks (challenges) the miners' performance and punishes breaches of contract. Contract-based decentralized storage protocols all face the basic problems analyzed above. The norm of technological development is that when most people try to solve complex problems with "taken for granted", there are always people who can find another way to solve the problem with a method that others have not expected and is usually much simpler. Sure enough, after three years of observing the field of decentralized storage, I got to know Avi by chance, a decentralized storage game-changer.

Only by understanding the difficulties of Filecoin can we understand Avi's ingenuity. Avi is a complete decentralized storage protocol that is not based on IPFS, or it is equivalent to Filecoin + IPFS. How does Avi solve the problem of miner proof? The answer is no proof is needed. The Avi protocol encourages miners to store as much data as possible through mechanism design, and gives priority to storing scarce data with few copies. As for how much and what each miner has stored, it is the miner's own business, and there is no need for proof or inspection. It is like a school that wants students to study hard. There are two ways to do it. One is that the teacher stares at everyone every day to see if they listen attentively and complete their homework seriously. If they find that they are not serious, they will be criticized and punished. The other way is to take exams. No matter how you study in normal times, the final test results will speak for themselves, and there will be rewards for good test scores. Both methods can improve learning effects, but the latter is obviously much simpler.

Contract-based decentralized storage is similar to "marking people", while the Aave protocol is like "exam", which is called incentive-based decentralized storage. Its advantages can be intuitively understood in this way: Filecoin has to manage thousands of different storage contracts, check the execution of each contract, and provide rewards or execution penalties respectively. The Aave protocol only handles one contract - all data is permanently stored. Therefore, the protocol is very simple, the operating cost is low, and the price and reliability of the service are better than those of the contract-based system.

Avi's PoA access proof is a simple extension of PoW. Each round of PoW puzzles is related to a past block (memory block), and only miners who have stored the memory block are eligible to participate in the PoW competition. Since the memory block is randomly determined and cannot be predicted in advance, the more blocks a miner stores, the greater the chance of participating in the PoW competition and the higher the possibility of obtaining block rewards. If the miner has limited storage space and cannot save the entire block history, he will give priority to saving blocks with fewer copies in the network. Because each block has an equal probability of being selected as a memory block, when a scarce block is selected as a memory block, only a few miners are eligible to participate in the PoW competition, so storing scarce blocks is more beneficial to miners.

Some students may ask, if it happens that all nodes do not store a certain block, then won’t this block be lost forever? Yes, this possibility exists. We can quantify the risk of a single block being lost forever. First, we need to introduce the concept of replication rate, which is the proportion of block history stored by miners on average. For example, if the network has a total of 100 blocks, and each miner stores an average of 60 blocks, then the replication rate is 60%. The replication rate is also the probability that a miner has a randomly selected block. Conversely, if a block and a miner are randomly selected, the probability that the miner does not have this block is 1-replication rate. When there are N miner nodes in the network, the probability that all miners do not have a certain block is (1-replication rate)^N. The probability of a lost block is (1-replication rate)^N * total number of blocks. Assuming that the Ave network has 200 miner nodes, a replication rate of 50%, and a total number of blocks of 200,000, the probability of a lost block is 6.223*10^-61, which is a negligible and extremely low probability event. Currently, the Ave network has about 330 miner nodes, a replication rate of 97%, and more than 510,000 blocks have been produced. The probability of a lost block is much lower than the previous calculation result, and is comparable in order of magnitude to the probability of a private key collision. Moreover, the above calculation assumes that miners randomly store block history. Considering that miners will prioritize storing scarce blocks, the possibility of losing blocks is even lower.

The Aave protocol has only one market, and users only need to pay storage fees, and subsequent access to data is free. This is possible because the Aave protocol adopts a mechanism design similar to BT. All nodes in the network are equal (no distinction between miner nodes and user nodes), and all nodes try to respond to requests from other nodes as quickly as possible. Like BT, the more upstream contributions, the faster the downstream speed. Selfish nodes will be demoted by other nodes and gradually excluded from the network. The best way to fully understand the design of the Aave protocol is to read the Yellow Paper . Although the Yellow Paper is long and has many formulas, don't worry, you can understand it with a basic understanding of middle school mathematics.

Compared with Filecoin, the Aave network has two major advantages. The first is low cost. Although the Filecoin mainnet has not yet been launched, I will make a prediction in advance: one year after the Filecoin mainnet is launched (the economic model enters a stable state), the price of permanently storing a 1MB file in hundreds of copies on the Aave network will be lower than the price of storing 5 copies on the Filecoin/IPFS network for 5 years, and data access on the Aave network is permanently free. Second, the incentive mechanism of the Aave protocol makes data storage and access more reliable. By simply and cleverly solving the biggest problem of decentralized storage, without the need for $200 million in fundraising and three years of development, the Aave mainnet has been online for more than two years. It is not a runner-up to Filecoin/IPFS, but the most promising encryption protocol to make large-scale decentralized data storage a reality.

Aavi and Ethereum

Aave is rarely compared with Ethereum. After all, in the Web3.0 protocol stack, they are at different levels and seem to be complementary. However, if you study the Aave protocol in depth, you will find more possibilities. Ethereum (and other smart contract public chains) are born to support decentralized applications DApp. DApp is an Internet application that is executed fairly and transparently and cannot be controlled by an individual or a few people. From the perspective of software architecture, network applications (including Internet applications and DApp) can be divided into three layers: presentation, business logic, and persistence (data). Let's analyze the development bottlenecks of DApp and the application potential of the Aave protocol from these three layers.

So far, the presentation layer of DApps remains in the same state as centralized web applications, that is, developers deploy them on cloud servers and then download them to user clients for execution. Therefore, developers and cloud service providers still have the right to stop and review DApps, and network outages, server downtime, DNS hijacking and other failures and attacks still threaten the availability and security of DApps. In addition, the IT infrastructure costs of DApps will increase with the growth of the number of users, forcing developers to adopt some kind of monetization method to maintain the operation of DApps. The monetization method is either Web2.0-style, that is, selling traffic; or with the characteristics of encryption protocols, that is, issuing tokens. Once monetization fails, developers may give up running DApps, and users can only look for alternatives. Even if there are alternatives by chance, they still face the same problem. DApps that can maintain operation often encounter the problem of "forced upgrades", that is, the new version is not necessarily more popular than the old version, but users cannot prevent it from upgrading, nor can they continue to use the old version. In summary, the presentation layer of decentralized applications is still centralized and can still be controlled by individuals or a few people.

The application layer of the Aave protocol is called the permaweb, and its main (but not the only) application architecture is serverless. The development of serverless DApps is similar to the front-end development of traditional Web. Developers use HTML, Javascript and CSS to develop the presentation layer of DApps. The difference is that the presentation layer is not uploaded to the cloud server, but packaged and stored on the Aave network. The storage cost is very low, and it is a one-time payment and permanent service. Users still use the original method to access DApps. Aave DNS and TLS are compatible with ordinary browsers, and users do not need to install and learn to use new clients. No matter how the number of DApp users grows, it will no longer incur overhead for developers. Since Aave is a decentralized network, neither developers nor Aave miners can prevent or censor users from using DApps. Developers can develop new versions of DApps, but the new version cannot overwrite the old version. The choice of which version to use is in the hands of the user. It can be seen that Aavi has achieved the decentralization of the DApp presentation layer, so more and more DApps have transplanted their presentation layer to Aavi, including: Synthetix Exchange , Tokenlon , KyberSwap , UniSwap , Oasis App , Curve.fi , etc.

It should be noted that the concept of using decentralized storage to achieve the decentralization of the DApp presentation layer is not the creation of the Aave protocol. As early as 2014, Dr. Gave Wood listed "static content publishing" as one of the four basic components of Web3.0 in his paper describing the Web3.0 network form11. The practical result of this thinking is the Swarm project. Both Swarm and IPFS were once highly expected to solve the decentralization problem of the DApp presentation layer. However, due to various reasons, this wish has not yet been realized. It was not until the emergence of the Aave protocol that the decentralization of the DApp presentation layer had a practical solution.

Ethereum and other smart contract public chains have achieved the decentralization of the DApp business logic layer and data layer, but it is well known that there is a scalability bottleneck. Scalability and price are two sides of the same coin. The scalability limitation comes from the scarcity of computing and storage resources. In a decentralized network, the result of competition for scarce resources is high prices. Since price is easier to quantify, this article chooses to analyze from the perspective of price.

Let’s look at the data layer first. Ethereum consumes 20,000 gas13 to store 256-bit integer data, and 625 million gas is required to store 1MB of data. Based on a gas price of 20gwei (when this article was written, the gas price was often as high as 100gwei or more during the DeFi boom) and an ETH price of $400, it costs up to $5,000 to store 1MB of data on the Ethereum chain, which is obviously an unaffordable price. Most DApps with data storage needs use a hybrid storage solution, that is, high-value data such as encrypted assets and hashes of attachments are stored on the chain, and detailed data, multimedia data, etc. are stored off the chain. If centralized off-chain data storage is used, such as a relational database or a NoSQL database, the DApp is still partially centralized and will still be controlled by an individual or a few people (cloud service providers and developers). Therefore, many DApps prefer decentralized storage, such as IPFS.

In this regard, Aavi provides fully decentralized, low-cost, and highly reliable permanent data storage, thus becoming a powerful assistant to Ethereum. Without sacrificing decentralization, Aavi currently only costs 0.1 cents to store 1MB of data. You read that correctly, it is one-five-millionth of Ethereum. At the current price, the cost of storing 1MB of data in Alibaba Cloud for 100 years is 2.6 cents. Moreover, it only supports redundant replication in the same city, and the network overhead for data synchronization and data access is charged separately. The Aavi network is redundantly replicated with hundreds of nodes on five continents around the world, and data synchronization and access are completely free. You read that correctly, the decentralized Aavi network is already cheaper than centralized cloud storage. No wonder layer1/layer2/DApp protocols such as Solana14 , SKALE15 , and Prometeus16 choose Aavi as the data storage layer. There are also NFT projects such as InfiNFT , Mintbase.io , and Machi X that use Aavi to store NFT media resources, metadata, and code.

Smart contracts are the business logic layer of DApps. Similar to the data layer, the bottleneck of smart contracts is the scalability/computational cost issue. According to Vitalik's estimate, the computing and storage costs of Ethereum are about 1 million times that of Amazon Cloud Services18. The previous estimate of the cost of the DApp data layer can also confirm this estimate. The fundamental reason for the high computing and storage costs of public chains is its fully redundant architecture, that is, all on-chain data is stored by every full node, and all calculations are performed on every full node. There are three ways to achieve public chain expansion: representative system, layering and sharding. For a more in-depth discussion, please refer to my article " Analysis of Polkadot Architecture ".

Ave’s Smartweave smart contract takes a completely different approach. Smartweave smart contracts are programs developed in Javascript and stored on the Ave network, so they are immutable. The contract code is submitted to the network for storage at the same time as the contract code. Unlike smart contracts in Ethereum (and other public chains), Smartweave is not executed by miner nodes, but downloaded to the contract caller’s computer for execution. The execution process starts from the contract’s genesis state, executes all transactions in the contract’s history in a certain order, and finally executes the contract caller’s transaction. After completion, the contract caller submits the input of his or her transaction and the executed contract state to the Ave network for permanent storage. Subsequent contract calls repeat the above process.

That is to say, for a smart contract transaction, the Aave network only needs one node - the caller's own node, to execute (note that the Aave network does not distinguish between full nodes and light clients). Since the caller's node executes (and verifies) all transactions in the contract history, he does not need to trust or rely on any node to obtain a reliable calculation result (i.e., the new state of the smart contract). Therefore, each Smartweave contract can be regarded as Aave's second-layer chain, and executing a smart contract is the full synchronization and verification of the second-layer chain. This design solves the scalability/computational cost problem of the DApp business logic layer. Smart contracts can contain almost any complex calculation without restriction, at a very low marginal cost, because the caller's computing equipment is usually purchased or rented for a long time.

Some students may ask: As the number of transactions increases, won't the execution of smart contracts become slower and slower? This is true, but there is a way to think about it. For example, the caller can name the result state of the contract to form a contract state snapshot. If the caller is trustworthy (for example, the caller is the developer of the smart contract), subsequent callers can specify the state snapshot as the initial state, and only need to execute transactions after the snapshot. State snapshots do not necessarily lead to an expansion of the trust set. After all, the premise of the reliability of smart contracts already includes trust in the initial state.

Of course, Smartweave is still under development, and the current version is V0.3. The above content should be regarded as a discussion of Smartweave's potential. To achieve commercial use, Smartweave still needs to solve many problems, such as composability. From my understanding of the operating mechanism of Smartweave, there is no special technical obstacle to achieving composability. However, I have always believed that the composability of Ethereum smart contracts is "too powerful" to limit the exponential growth of the complexity of the contract system. I look forward to more surprising innovations from the Smartweave team, and to making good use of the double-edged sword of composability.

In summary, the Aave protocol supports DApp to truly achieve full decentralization and solves the scalability/cost issues of computing and storage that have plagued the public chain field for many years. In this sense, Aave should be classified as a full-stack Web3.0 protocol advocated by Blockstack20, rather than just decentralized storage.

Avi and Bitcoin

Bitcoin is the pioneer of encryption protocols and the king of cryptocurrencies. There has always been a topic of endless debate in the industry: Can Bitcoin's kingly status be replaced? Even Bitcoin royalists admit that after 10 years of development, Bitcoin is no longer the most technologically advanced cryptocurrency. But they believe that super-sovereign value storage currency is the biggest use case for cryptocurrency. The Bitcoin protocol has the longest running time, the highest popularity, and the best security. Moreover, the competitive barrier of cryptocurrency is not technology, but liquidity. Liquidity has a network effect, that is, a mechanism by which the utility of a product or service increases as the number of users increases. The Bitcoin protocol has established a liquidity advantage, which will only continue to increase as cryptocurrency becomes more popular. Therefore, Bitcoin's kingly status is unshakable.

Is it possible to break the liquidity network effect advantage? To answer this question, we need to conduct quantitative research on network effects. I believe many people will immediately think of Metcalfe's law, which states that the value of a network is proportional to the square of the number of users. Metcalfe's law is the first quantitative model of network effects, but research in recent years has shown that the value of no network grows according to Metcalfe's law. At least when the number of users is large, the network value growth curve will inevitably become flat.

Studies have shown that the network effect of some Internet businesses is n*log(n), and some are S-curves. The S-curve is an exponential growth of network value as users grow, which is slow at first and then fast. After reaching saturation, the growth rate slows down. An important inference of the S-curve is that the strong get stronger, but it is not a winner-takes-all situation. If the network effects of all Internet platforms conform to Metcalfe's law, then a single oligopoly will be formed in each segment of the Internet industry. But the reality is that in most segments of the Internet industry, whether in the global or Chinese Internet industry, there is more than one platform that exists for a long time.

So what curve (formula) does the liquidity network effect grow along? Assume that for a certain crypto asset, the average daily trading volume of each participant accounts for one ten-thousandth of the total market value of the asset. The daily turnover rate of 10,000 investors is 100%, and the turnover rate of 20,000 investors is 200%. In other words, the turnover rate doubles with the addition of 10,000 investors. If the number of investors increases from 100,000 to 110,000, the turnover rate increases from 1000% to 1100%, which is only one-tenth of the increase. Therefore, the more investors there are, the smaller the contribution of new investors to liquidity, and its network effect is in a log(n) relationship with the number of participants.

The above quantitative models and pictures about the liquidity network effect are all from Multicoin Capital's research 23. The conclusion of this study is very important. For example, exchanges compete on liquidity. After the top exchanges reach a certain scale, the value growth brought by the liquidity network effect will slow down, giving latecomers a chance to catch up. If the liquidity is n*log(n) or even n squared network effect, there will be no Binance, Kucoin, and MXC coming from behind, and there will be no tens of thousands of exchanges. The quantitative relationship of log(n) shows that the greater the liquidity, the stronger it is, but it does not guarantee that the strong will always be strong.

There is another factor that makes Bitcoin's liquidity advantage easier to break, which I call liquidity transmission. That is, new cryptocurrencies can use the established global trading network to share liquidity with existing cryptocurrencies. For example, when Ethereum was born, the industry infrastructure, including exchanges and payment platforms, had been developed for 6 years, and they easily integrated ETH. As long as ETH forms a highly liquid trading pair with Bitcoin, it indirectly has liquidity with major fiat currencies. Therefore, Ethereum no longer needs to go through a long market introduction and infrastructure construction stage to become a highly liquid cryptocurrency.

In a state of free competition, currencies are compared in terms of their currency properties. Currency properties include scarcity, interchangeability, verifiability (difficult to forge, easy to identify), accessibility, divisibility, and the cost of storage, carrying, and transfer. All cryptocurrencies are direct descendants of Bitcoin and have inherited Bitcoin's strong currency properties. Before Ethereum, the theme of cryptocurrency innovation was "better Bitcoin", that is, creating cryptocurrencies with stronger currency properties. For example, Litecoin, Dash, and Stellar have faster transfer speeds and lower transaction fees. ZCash and Monero have better privacy and more guaranteed interchangeability. But none of them threaten Bitcoin's position. Because quantitative improvements are not enough to challenge the advantages of network effects, qualitative innovation is necessary to achieve a "paradigm shift". For example, Microsoft did not invent a better mainframe to defeat IBM, and Apple did not defeat Microsoft with a better PC. Revolutionary innovators all became new kings by implementing a dimensionality reduction attack on the old overlord.

The industry generally agrees that Ethereum is the representative of blockchain 2.0. Because Ethereum is a new level of innovation, by introducing EVM, cryptocurrency has powerful programmability. Generational innovation is not that I can do better than what you do, but that I can do what you can't do. Ethereum smart contracts can realize decentralized asset issuance, fund raising and asset trading. In the last round of ICO wave, ETH was used as the main currency and value storage. The demand for ETH soared, and its market value reached 60% of BTC. Of course, 1CO has serious information asymmetry, which inevitably leads to widespread adverse selection and moral hazard problems, and the bursting of the bubble is an inevitable result. Highly programmable cryptocurrencies have infinite room for innovation. The rise of DeFi will be a new round of challenges for Ethereum to Bitcoin. Unfortunately, ETH's value capture mechanism is not sound. If EIP1559 had been implemented a few years earlier, ETH should have entered a deflationary stage, and the DeFi craze is likely to push its market value beyond BTC.

There are two major investment themes in the crypto asset market: sound money and Web3.0. Sound money is a decentralized, super-sovereign cryptocurrency, represented by Bitcoin. Web3.0 is the application of blockchain technology to reconstruct social production relations, represented by Ethereum. I believe that the two major investment themes of sound money and Web3.0 can be achieved at the same time, that is, a decentralized, highly programmable blockchain platform that can support Web3.0 and whose native crypto assets also have the properties of sound money can have the best of both worlds and become the king of cryptocurrencies in the future. The new king should have the following properties: highly decentralized (implying super-sovereignty), widely used, low consensus protocol externalities, good scarcity, highly programmable, and compliant.

Given the scalability issues of Ethereum 1.0, it will be difficult for it to last even if it ascends to the throne. Which project is the representative of blockchain 3.0? Ethereum 2.0, Polkadot, Cosmos, and Aave Protocol are all strong contenders. AR also has the potential to become the king of cryptocurrency:

● High degree of decentralization, the network will not be controlled by individuals, institutions or governments;

● Widely used, as a full-stack Web3.0 protocol, it is an ideal platform for all kinds of decentralized application innovation;

● PoA consensus does not consume a lot of extra electricity, as discussed in detail in the next chapter;

● The Aave protocol’s native token AR has a low issuance rate and good scarcity. See the next chapter for detailed discussion;

● Highly programmable, smart contracts are Turing-complete. Both DApp and smart contracts use mature Web technologies such as Javascript, which is conducive to the formation of a broad and diverse developer community;

● Aave is very similar to Ethereum, with an ICO before the mainnet launch. After the mainnet launch, functional tokens were distributed. The function of ETH is to pay for the computing and storage costs of Ethereum; the function of AR is to pay for the storage costs of the Aave network. Over time, AR is used by more and more people, and the currency holders are becoming more and more dispersed, which meets the legal definition of bulk (virtual) commodities.

Avi economic model

The economic model of cryptographic protocols is how to coordinate the interests of service providers (miners), service users (users) and coin holders. Miners provide computing, bandwidth and storage resources for the cryptographic protocol network to ensure the security and availability of the protocol. Users have to pay miners for using the protocol. The miners' income is divided into two parts: one is the transaction fee paid directly by the user; the other is the newly minted tokens distributed by the protocol to the miners, that is, the issuance reward. The issuance reward is the seigniorage shared by all coin holders according to the number of coins held. In almost all cryptographic protocol economic models, the main income of miners is the issuance reward (coinage tax). For example, although Bitcoin has undergone three issuance reward halvings, the issuance reward still accounts for 95% of the total income of miners, and transaction fees account for only 5%. This is actually a mechanism for coin holders to subsidize users for using the protocol.

Among all the crypto protocol economic models I have studied, the Aave protocol's economic model is the most friendly to coin holders. In the genesis block, the protocol generated 55 million ARs, and then each block will issue additional ARs. The formula for calculating the additional issuance is as follows:

in:

GAR = Number of ARs in the genesis block: 55,000,000;

Bny = number of blocks in a year: 262,800;

BH = current block height;

Substituting the constants into the equation, the formula is simplified to:

Rinflaton = 29 * POW (2, -BH/262800)

The Aave protocol produces one block every 2 minutes on average. After the genesis block, each block will issue about 29 ARs, and the issuance will be halved every year, with a maximum of 11 million ARs. In other words, half of the remaining ARs will be mined every year after the Aave mainnet goes online in June 2018, that is, 5.5 million ARs will be mined in the first year, 2.75 in the second year, and 1.375 million in the third year... (The Aave mainnet was launched in June 2018, and the block reward was issued more than two months later) At the time of writing this article, the Aave network is facing its second halving (estimated to be around September 10, 2020). The issuance rate one year after the second halving (i.e. the third year) is 137.5/(5500+550+275) = 2.17%. By the fourth year, the issuance rate of AR will be lower than that of Bitcoin in the same period. Another scarcity indicator may be more convincing, the waiting rate = unmined amount/total amount. Currently, there are only about 1.98 million ARs left unmined, with a waiting rate of 3%. In comparison, there are about 2.55 million BTCs left unmined, with a waiting rate of 12%. It can be seen that AR has a small amount of additional issuance and a fast decay rate, which is a typical feature of a highly scarce sound currency.

However, please note that according to the figures provided by the Avi team, the current circulation of AR is about 38 million, which means that there are about 2,600 ARs in non-circulation. I don’t know the ownership structure and unlocking plan of this part of the pass, and can only speculate that it belongs to early investors, teams and foundations. If anyone knows about this, please let the author know, I would be grateful.

The principles of the Aave economic model can be roughly summarized as follows: users pay enough for storage services; miners' income exceeds costs, maintaining a basic roughly fixed profit margin; coin holders receive almost all the benefits of AR token appreciation. Based on AR's current price of $4, Aave miners will receive $5.5 million in revenue from additional issuance this year, which will be shared by hundreds of mining nodes around the world. In comparison, Bitcoin miners receive more than $10 million in additional issuance revenue every day, or more than $3.6 billion per year.

: : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : :

Avi and Crypto Asset Investment

Since I started investing in Bitcoin in early 2013, I have heard many people talk about how to learn about Bitcoin, my first impression of Bitcoin, how to pass by huge wealth, etc. There is a question that has always troubled me: what determines our view on Bitcoin at that time? Most people don’t care. Some people believe that Bitcoin is a capital disk with high-tech cloak. A few people invest in Bitcoin out of various psychology or start mining. A very small number of them persisted and their destiny was changed by Bitcoin.

These very few people are often regarded as geniuses who foresaw the future. But you should know that any novelty has a group of early participants, but among countless novelty things, there are very few talks that have a wide impact on society. Rather than treating those who participated in Bitcoin early and became big bosses as geniuses, they are better to say that they are lucky. But the question is, does such good fortune come to everyone at random with roughly equal probability? With my thoughts on this issue over the years, it may not be entirely true.

: : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : :

: : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : :

People who have achieved great success in the crypto asset market are optimistic and long-termists who pay attention to big problems. The so-called big problems are the basic issues that affect the development of the Internet and even human society. In 2011, 2013, and even 2015, you can list hundreds of reasons why Bitcoin will fail, and these reasons are tenable. But if you focus on the following big problems (or one of them): the Internet needs native value transmission that does not depend on specific institutions; Internet platforms and financial intermediaries have already seized most of the profits of the economic activities of the whole society; the central bank's continuous issuance of currencies to promote economic development is no longer sustainable, etc., you will realize the cross-eracial significance of Bitcoin. Moreover, an optimistic person must believe that although Bitcoin has hundreds of reasons for failure, it may also be successful. As for long-termism, focusing on big problems is two sides. If someone clears Bitcoin's position after getting several times the profit, it is hard to believe that he is really paying attention to big problems.

The permanent preservation of human knowledge and history is of course a big problem, and it is likely that its importance is beyond the right level of humanity. After all, modern people are no different from their ancestors in terms of intelligence and physical fitness. The only reason why we live a completely different life from our ancestors is that we have inherited and utilized the knowledge and experience accumulated by humanity over tens of thousands of years of history.

For the rulers of the Ptolemaic dynasty, the Alexander Library may be just an embellishment of the richness of the country. But for later generations, the Alexander Library is much more important than the Ptolemaic dynasty. Although Caesar was called by historian Monson: the only creative genius of the Roman Empire. However, Caesar's successes could not make up for the fault of burning the Alexander Library. Has the technology today developed to a critical point? The world can no longer rely on individuals, institutions or countries, no matter how powerful they are, to permanently preserve the knowledge and history of all mankind? How lucky we would be to participate in this generation!

So Avi is not a substitute or competitor of Filecoin/IPFS. The goal of Filecoin/IPFS is to subvert the monopoly of centralized cloud service manufacturers in the storage market. This is of course an important issue in the Internet industry, but compared with Avi's goal, it is far from being a "big problem". When I finished reading Avi's yellow book, it seemed like time and space were back to the first time I learned Bitcoin. Will a miracle happen this time?

Citations

1. https://filecoin.io/zh-cn/2020-engineering-filecoins-economy-zh-cn.pdf

2. Labs, P. A Guide to Filecoin Storage Mining. Filecoin Available at: https://filecoin.io/blog/filecoin-guide-to-storage-mining/.

3. https://pcpartpicker.com/user/tperson/saved/H2BskL

4. Venturo, B. The economics of Ethereum's Casper. Medium (2018). Available at: https://medium.com/@brianventuro/the-economics-of-ethereums-casper-6c145f7247a2.

5. https://www.reddit.com/r/CryptoCurrency/comments/982x9l/top_100_cryptocurrencies_ranked_by_annualized/

6. http://app.czce.com.cn/cms/cmsface/option/Calculator/utCal.jsp

7. Project, TA Decentralised storage: Incentives vs Contracts. Medium (2019). Available at: https://blog.goodaudience.com/decentralised-storage-incentives-vs-contracts-b74ee0b7eff1.

8. https://viewblock.io/arweave/stats

9. Bram Cohen. Incentives build robustness in bittorrent. In Workshop on Economics of Peer-to-Peer systems, volume 6, pages 68{72, 2003. [19] Matt Corallo. Compact block relay. bip 152, 2017.

10. Project, TA Arweave News: July. Medium (2020). Available at: https://medium.com/@arweave/arweave-news-july-7905d5e0c84f.

11. ĐApps: What Web 3.0 Looks Like Available at: http://gavwood.com/dappsweb3.html.

12. Swarm Available at: https://swarm.ethereum.org/.

13. G. Wood, Ethereum: A secure decentralised generalised transaction ledger, In: Ethereum Project Yellow Paper 151 (2014).

14. Solana - Arweave Bridge: ArweaveTeam Funded Issue Detail. Gitcoin Available at: https://gitcoin.co/issue/ArweaveTeam/Bounties/30/100023463.

15. SKALE Network - Arweave Bridge: ArweaveTeam Funded Issue Detail. Gitcoin Available at: https://gitcoin.co/issue/ArweaveTeam/Bounties/27/4468.

16. Labs, P. New primary storage for Ignite. Medium (2020). Available at: https://medium.com/prometeus-network/new-primary-storage-for-ignite-94096e2e8506.

17. Project, TA NFT Permanence with Arweave. Medium (2020). Available at: https://medium.com/@arweave/nft-permanence-with-arweave-35b5d64eff23.

18. Wang, B. Ethereum is about 1 million times less efficient for storage, network and computation. Next Big Coins (2018). Available at: https://www.nextbigcoins.io/ethereum-is-about-1-million-times-less-efficient-for-storage-network-and-computation/.

19. Project, TA Introducing SmartWeave: building smart contracts with Arweave. Medium (2020). Available at: https://medium.com/@arweave/introducing-smartweave-building-smart-contracts-with-arweave-1fc85cb3b632.

20. https://www.blockstack.org/

21. Odlyzko, Andrew & Tilly, Benjamin. (2020). A refutation of Metcalfe's Law and a better estimate for the value of networks and network interconnections.

22. The Network Effects Bible. NFX (2020). Available at: https://www.nfx.com/post/network-effects-bible/.

23. Kyle Samani, On the Network Effects of Stores of Value. phoenix Available at: https://multicoin.capital/2018/05/09/on-the-network-effects-of-stores-of-value/.

24. tevador. Randomx. https://github.com/tevador/RandomX, 2019.

25. Shevchenko, A. & Shevchenko, A. Monero Penalizes GPU and ASIC Mining with RandomX Upgrade. Crypto Briefing (2019). Available at: https://cryptobriefing.com/monero-penalizes-gpu-mining-randomx/.

26. [email protected] The Rise & Fall (And Rise & Fall) Of The Top 10 Cryptocurrencies... Merchant Machine (2018). Available at: https://merchantmachine.co.uk/cryptocurrencies/.