Introduction： Vitalik, the founder of Ethereum, delivered something interesting on Twitter at the beginning of 2022. We made some comments on this not for any investment advice but for purely personal opinions.
In 2013, I wrote an article about “How Bitcoin Can Actually Help Iranians and Argentinians”.
Core point: The main advantages of Bitcoin are internationality and censorship resistance, not the limitation on 21 million amounts. I predict that stablecoin will continue to prosper， and I verified my predictions when I went to Argentina last week. At the current stage, the adoption rate of cryptocurrencies stays high, so is the stablecoin as many companies run with USDT. But, this situation may change, if the dollar itself starts to have more problems.
Internationality and censorship resistance are some of the benefits of Bitcoin, but I don’t think this is the most important benefit of Bitcoin, cause gold, silver, and more have the same features.
And, the 21 million upper limit takes the same position with other benefits (but not meaning the benefit of this figure, but the benefit from a fixed aggregate or limited inflation). Otherwise, imagine if each block is issued exponentially, the first 10, the second 20, the third 40… The whole system will collapse quickly and its quantitative characteristics are meaningless. The value of this system changes when we just gave different issuance amounts. So the fixed amount cannot simply be ignored.
The most important benefits of Bitcoin (in order of importance) are as follow: 1. A brand new asset is constructed through a non-cooperative game (algorithms and parameters determine its characteristics), which has a completely different risk-return structure from traditional assets. It boasts the value of investment or Hedge (hedging risk is what Vitalik referred to as Argentina’s value). 2. The asset has no credit risk, which is different from general equity and creditor’s rights that deeply rely on a specific organization; and its equilibrium will be continuously enhanced with technological progress for its game feature, and the assets itself be replaced by advances in physical technology, as the results of more stable risk-return structure.3. It is convenient to save and transfer anonymously without third parties. This is the meaning of anti-censorship, which is often used for some special needs, such as gray trading, tax avoidance, etc.
Published an article on the consequences of “regulated” Bitcoin services in 2013.
Core point: Bitcoin’s censorship resistance is created by its technical innovations and not found in the new legal category. In my opinion, Bitcoin’s decentralization can allow it to survive in a super hostile regulatory environment, but it cannot “thrive”. A successful censorship resistance strategy requires a combination of technical robustness and public legitimacy.
This is what was mentioned in Comment 1 above. Its thrives cannot rely on censorship resistance but rely on the demand from investment or hedging which is the universal pain point. Vitalik has consciousness on it but cannot accurately express it.
I predicted when PoS and sharding will be implemented in 2015, and share a screenshot of my 2015 speech, which seems to be very wrong and ridiculous from now:
What is the core error? In my opinion, I seriously underestimated the complexity of software development. But now, the Ethereum research team pays more attention to simplicity-whether it is the simplicity of the final design or the implementation path, which is more compromised to realize it. Dankrad’s latest sharding design fits this spirit very well.
Many people always make complaints about the slow development of Ethereum and think it is not as agile as Internet projects with the expression of turtle speed. We will clarify it on two deeper levels:
First, high requirements for the system. A small fault cost too much so complex solutions are not available. Concise solutions represent many trade-offs that may not be satisfactory, and disputes will consume decision and execution time. Secondly, the blockchain system is a game equilibrium. It is easy to change the code, but each change will form a new equilibrium, which will affect the interests and behaviors of all parties. Therefore, the real modification cannot ignore the existing history and needs to obtain approval from the community, which is much more difficult than Internet project company decision-making.
“The transaction fee on the Internet should not exceed 5 cents.” This was the goal in 2017, and it is still the goal. I still stick to my point of view 100%, which is why we spent a lot of time researching scalability.
It is useful to research scalability, but obviously in contradiction with the censorship resistance that you admire above when you mentioned “less than 5 cents of transaction fee”. At present, relying on technology cannot achieve censorship resistance, and it can only be completed by designing a certain game mechanism to effectively resist censorship when the system reaches equilibrium.
But, System equilibrium is often costly. As this non-cooperative game brings the increasing marginal cost of performing operations， How can it be controlled below a fixed average cost? And if the blockchain system does not have begged with fiat currency, how does the information of the U.S. dollar reflect on the chain? Even if the U.S. dollar is mapped on the chain according to the current guarantee mechanism, it cannot be effectively stabilized below 5 cents unless it sacrificed the part of “Censorship resistance” (that is centralization)
I should also add that the core concept of Sharding has been continued intact. Blockchain 1.0: Every node needs to download everything. if BitTorrent reaches a consensus, each node only downloads a few things, but there is no consensus with low efficiency. But the core consensus comes from the community, ZK- SNARK, and data availability sampling.
This is likely to be self-conflict, but what we can realize is to improve transmission and download efficiency while reducing costs as much as possible. At present, the blockchain is a non-cooperative game system, and there is no way to reduce the marginal cost of operations on the chain (calculated according to resource consumption). Under a given technological boundary, it is a logical paradox to improve efficiency and the level decentralized.
In 2012, I was a supporter of POW. Fortunately, I was excited about POS in 2013. In 2014, I changed my mind again. This reflects the evolution of my broader thinking: from “X is something I must defend, so everything that benefits X must be correct” to “I like X, but X is flawed and it seems that Y fixes them, I Now supports X+Y”. Soldier mentality → Scout mentality.
Compromises that have to be made when the goal is too ambitious. In fact, the production modeling of the non-cooperative game of blockchain can only be used to do one thing: “alchemy” called casting assets. General goods and services require diminishing marginal costs, which is contrary to this model leading to increasing marginal costs. You have to resort to some kind of cooperative game, as the result of taking some centralization risks.
In 2014, he published an article on smart contracts. Basically, it tries to prove that making the whole society more like a unit system is good. Through comprehensive research, I saw the limitations of this way of thinking: the system with more than two participants is vulnerable.
Precisely, any system based on a given algorithm is incomplete with some loopholes, and there is even no guarantee that there will be no relatively dominant participants!
I liked altcoins very much before they became popular. Browse my article published in September 2013 (2 months before the birth of Ethereum). Three core arguments:
Differential chain optimization for different targets
The lower cost of multichain co-existence
In case the core development team makes a mistake, an external force is needed
Do I still agree? The above parameters are not so important today, because (i) chain is more general, (ii) the application is more complex, so bridging is more dangerous, and (iii) the experiment is feasible in L2. Even so, I still think that there are some things that cannot be done in L2, and there is still room for different L1.
L2 will definitely sacrifice certain decentralization properties, and the demands for new L1 rely on the dependent level of certain indicators of L1 such as block height, time, etc.
I was particularly optimistic about Bitcoin Cash because I agree with the views of the big block rather than a small one in this scaling war. Today, I think BCH is a loser. In my opinion, the communities formed during the rebellion time usually have a long period of hardship, and value courage over ability, paying much attention to resistance instead of moving forward in unison.
It is easy to change the technology, but it is difficult to change the balance.
The posts are basically advocating the construction of Uniswap in 2016–2017. “Do something simple and stupid, even if it is sub-optimal.” Although this is the current concept, it took me a long time to find the basis for POS and Sharding design.
Well, simple and stupid.
The applications envisaged in the Ethereum whitepaper:
ERC20 standard token
Domain name system (such as ENS)
Decentralized file storage and calculation
Wallets with withdrawal restrictions
Although the incentive file storage and calculation have not yet been implemented, many of the above assumptions are correct (basically for DeFi). Of course, I completely ignored NFTs. What I want to mention is that the biggest problem is the collusion problem in DAO governance.
The two overlooked issues are actually related to centralization. Obviously, NFT requires off-chain laws to ensure the uniqueness of information (copying and misappropriation are very simple), not technology. Similarly, DAO’s collusion cannot be avoided in the algorithm. Things that exceed expectations are related to the “abuse” of the blockchain in a sense, and the so-called abandonment is essentially decentralization. Why is there such a contradiction? In fact, the above has been very clear. The current blockchain is essentially a decentralized (non-cooperative game) value production model with increasing marginal costs by On-chain operations (including the impact of auction mechanisms and token price growth). But, general goods and services require diminishing marginal costs, which has great limitations. Solving those problems will inevitably sacrifice a certain degree of “decentralization.”
Published an article on Stablecoin in 2014. A large part of the article tried to refer blockchain data as a price, rather than through oracles. I am more pessimistic about this, especially because POS needs an oracle. If we want to make stable currencies robust against the collapse of the dollar (if this happens, by switching to their own local CPI), we need more active governance.
At that time, my thinking about politics and large-scale human organizations was naive, and focused much on simple and complete formal models…
I did have a good instinct in the early days to avoid the craziest part of Bitcoin dominance. There were a few early mistakes, which I quickly corrected.
Wrong X idea does not mean that any resistance against X will go smoothly, which will cause another difficulty in the form of organization.
In terms of technology, I am more accurate in abstract ideas than in software development. one must learn to understand the latter over time,
I now realize more deeply that what we need is simpler than I thought.
In short, we should not be so dogmatic about decentralization. It is necessary to have active users and not necessary to develop such complicated applications on L1 (because goods and services require diminishing marginal costs!). on-chain assets for gaming are good when users are willing to hold it, and complex applications are deployed on centralized systems such as L2. However, it is easy to change the technology, but difficult to change the balance. It is really not as simple as just talking about it.