A little reverse thinking about Web3.0
Original Author: zephyr
Original source: Mirror
The best critique of web3 I've read recently is this one:
I haven’t seen anyone talking about this article on Chinese social media (it’s already very popular on Twitter), so let’s sort it out briefly.
The author Moxie told a very interesting story: he made a picture himself like everyone who is trying NFT for the first time, and mint turned it into NFT and sold it on opensea. But he noticed that the NFT itself did not have any verification of the picture. In essence, it just stored a link to the picture address, so he specially set the server of the picture to show different appearances for different IPs. What you see on opensea and what you see after you buy it will be two different pictures. What is displayed on opensea is a cool digital art picture, and what you see after you buy it is a pile of shit.
(His starting point is not lying, he is just verifying that this situation can be done. Yes, services such as IPFS can avoid this, but no one is forcing you to use IPFS to do NFT.
Then something interesting happened. As a centralized platform, opensea quickly delisted his NFT. But it doesn't matter, since NFT is based on an immutable blockchain, at least he still owns this NFT, right? But something magical happened: the NFT in his own metamask little fox wallet also disappeared.
why? Because the little fox wallet does not directly scan the blockchain, but only scans the opensea API. So after opensea is taken off the shelves, although technically the NFT is still there, it cannot be seen in the wallet. Little Fox does this because it is obviously the most convenient, centralized services (such as opensea) are always more efficient, and it is expensive to verify the "truth" stored on the blockchain.
The author pointed out very profoundly: The key here is not that Opensea "does evil" (as a centralized platform that wants to go public, Opensea has the right to choose what works can be put on the auction shelves) nor is it the laziness of metamask, but the whole here reflects an unstoppable Avoid the trend of sliding from decentralization to centralization: the blockchain is indeed immutable, but no one really works directly on the underlying blockchain (too troublesome), and everyone will naturally rely on various Ready-made tools, and these tools will naturally tend to be centralized in order to compete for efficiency. As ordinary users, we can obviously run an Ethereum node on our computer, but few people really do this, we just use Little Fox directly. For the same reason, the little fox directly calls the API of Opensea.
In other words, blockchain does not solve the centralization problem of the final interface to the average user. In theory, web3 is ultimately intended to be used by ordinary people like your parents. If your parents find that something is missing in your wallet, you go and explain to them: Ah, your things on the chain are still there, but you These commonly used wallets just refuse to show it, but it doesn't matter. Will your parents accept it?
The author has two paragraphs that I think are very beautiful:
Platforms always evolve faster than protocols. One of the complaints people have about web2 is that platforms are always bullshitting, but the basic idea of web3 - building a decentralized protocol - doesn't really compensate for that. The final winner of the competition is almost always web2 under the guise of web3.
Users are lazy, and users don't want to run a server by themselves, just like users in the email era, everyone can set up their own email server but still prefer to put a lot of privacy directly in gmail. What web3 needs to do is to ensure the verifiability of information even when the infrastructure is likely to be centralized (you can use opensea, but you must have a way to easily know whether opensea is lying to you).
My own understanding is that motto "Trust But Verify" in the blockchain world is a concept that is difficult to apply to ordinary people. Sounds okay, but not practical. If web3 can't find a way to break through this layer of barrier between nerds and ordinary people, it is likely to become a self-enclosed high in the end. Just like the author said in the article: You can still say today that this is still an early stage, and it is normal to have problems. But if you're actually going in the opposite direction all the time, you can't expect the early problem to go away eventually, because it's probably written directly in the genes.
V god just wrote a good response to this article, I will add the gist here. See his tweet for the original text of V.
The problem Moxie points out consists of two arguments: Centralized web3 services are easy to use but not trustworthy (in particular, there is hardly any cryptography-based verification, and Moxie sarcastically said that who would have thought that the vast majority of services in the cryptocurrency field are simply not encrypted), and the non-centralized bottom layer is too far away from users. V said: This is the status quo, yes, but this is not what web3 should be like. The real web3 world should have a continuous transition spectrum. There are a large number of transition states between the easiest-to-use centralized platform and the hardest-to-use self-hosted server to adapt to different application scenarios, but the middle part is missing today. .
This lack is a historical legacy. The blockchain world is too young, and people start out based on wanting to make something that can be used. Of course, the fastest path is to establish the most centralized service, and talents are also readily available. (Here V makes a statement that is almost destined to attract criticism: until four years ago, the entire industry had no money. There must be many onlookers saying: yuck.
V's belief is that this missing transition must be built, and, as Moxie's critique suggests, relies heavily on cryptography.
(But my understanding is that this belief of V is essentially the same as his belief about PoS. So it must be very controversial in itself.
