Recap of PoW vs PoS discussion on Fork It Podcast — Part 2 by Nervos Network Nervos Network Sep, 2022

Terry: Someone launched a survey poll in the Ethereum community. If validators comply and censor at the protocol level as requested by the regulator, will you consider the censorship an attack on Ethereum and burn their stake via social consensus or tolerate it? Vitalik chose the former one. What do you think of Vitalik’s choice?

Cipher: I guess it’s best not to comment. “Consider the censorship an attack on Ethereum and burn their stake via social consensus” if Vitalik had not chosen this, but someone else did, he would have been under attack because the logic is very muddled. Burning validators’ stake via social consensus is impossible and not feasible.

First, there is no answer to what social consensus means. How do you organize the vote if it’s a community vote or a decision made by the majority? How do you guarantee that the vote is fair and accurate?

The cripto community has outstanding checks and balances. For instance, there are several independent checks and balances in the Bitcoin community, including miners, developers, wallets, and exchanges. The community acts as a check on one another. If the developers submit a feature that miners want to reject, miners can do a hard fork.

There used to be checks and balances in the Ethereum community. The mines were weakened by EIP-1559 and will be ruined by the Merge. In other words, the Ethereum community lacks the potential to build social consensus, as there is no such check and balance power. Ethereum developers lack control since validators can reject them and won’t vote against them. Therefore, in the end, only Vitalik can stand out and inform the community of which validators should be prohibited since they are enforcing censorship. Calling this social consensus is ridiculous, as it is not the consensus but the will of Vitalik. Why would Vitalik choose this option on the poll? Because the second option, tolerating censorship, is unacceptable for Vitalik. So burning validators’ stakes via social consensus is not possible and feasible.

Terry: I was baffled when I saw the word social consensus. It seems that you are also confused.

Cipher: I guess Vitalik doesn’t know what social consensus means either.

Terry: The CEO of Coinbase said that they would shut down the staking service if the validators were to apply censorship. Although it hasn’t happened yet, I think he’s making a very noble choice. If protocol-level censorship occurs, and all validators must make a noble choice, isn’t that a problem?

Cipher: Yes, it is. It’s a moral choice, not an economic one. The most significant appeal of blockchain is that it takes advantage of the selfishness of all people to achieve a public good. You can’t ask validators to be selfless and noble and not let them earn money because of censorship or unreasonable demands.

I’m sure Coinbase is sincere about this, but it doesn’t help nor solve the problem of avoiding Ethereum being censored at the protocol level.

Terry: Would things be different if it happened with PoW?

Cipher: I think it would be different. Firstly, shifting from one PoS mining pool to another is difficult, as the settlement cycle is very long. However, shutting down one PoW mining pool and creating a new one is very easy. So if the PoW mining pool doesn’t want to obey the regulation of censoring, it can shut down and create a new one in other regions.

Secondly, the connection between miners and the mining pool is weaker than that of PoS. Miners don’t deal with transactions, they are working hard to solve the mining puzzles, and therefore they are not capable of applying sanctions. Moreover, miners can switch their hash rate from one mining pool to another at any time.

To conclude, it’s more difficult for regulators to ask PoW mining pools to censor transactions at the protocol level.

One more thing, Ethereum employs the account model, not the UTXO model. The UTXO model used by bitcoin makes censorship more difficult. Because one UTXO blockchain address can derive a list of new addresses, making it more challenging to track as the sanctions are often lagging. The OFAC published a list of banned bitcoin addresses, but it had little effect.

MEV and Flashbots can potentially weaken the decentralization of Ethereum. MEV, which refers to Maximum Extractable Value, is built to extract value from Ethereum users through block production. Above MEV, there are Flashbots, which can arbitrarily include, exclude, or re-order transactions from the blocks. Flashbots is centralized, although it is open source now. Forcing it to censor transactions is much easier as many Ethereum mining pools use it to maximize profits.

Terry: My friends think it’s possible to decentralize the Flashbots since many PoW mining pools used to provide tools similar to Flashbots, which allows you to send transactions anonymously. After the open-source of Flashbots, they might upgrade the tools. We skip this topic as we are not experts on this. I’m wondering if miners’ revenue will also incentivize them to be more proactive in switching PoW mining pools. If one mining pool has to ban several types of transactions, the fees paid by these transactions will not be gained by this mining pool. So, in this case, will miners switch to a pool that doesn’t need to censor transactions to maximize profit?

Cipher: I think they will. The gas fee paid by these banned addresses will be much higher to attract mining pools (without censorship) to pack and confirm their transactions. Rational miners will likely switch their hash rate to these pools for higher profits.