The Myth of Web3 Decentralization

The Web3 movement is all about decentralization - but is it really possible to have a completely decentralized web?

The short answer is no. The current generation of web3 apps relies on several central points of control, whether that be in the form of centralized servers or web browsers or mobile apps or even API endpoints.

That's not to say that web3 can't be more decentralized than it is today. But the idea that it can be completely decentralized is a myth that requires rethinking how the internet itself was created.

Here's why:

1. There will always be central points of control.

No matter how you slice it, the web will always have central points of control. Whether that be in the form of centralized servers or web browsers or your ISP, there will always be some level of centralization. Even if we decentralized the web by moving to a peer-to-peer network, there would still be central points of control in the form of the computers that are connected to the network. In a way, centralization is an inherent property of the web. However, that doesn't mean that we should give up on decentralization entirely. Even if we can never achieve a completely decentralized web, we can still strive to reduce the number of central points of control. By doing so, we can make the web more resilient and open. This is the heart of the Web3 movement.

2. Web3 is built on top of existing infrastructure.

The idea of decentralization is all the rage these days, but it's important to remember that web3 is built on top of existing cloud infrastructure, which is itself centralized. That means that web3 can only be as decentralized as the infrastructure it relies on. Most cloud computing and data storage infrastructures today are both highly centralized, which means that web3 applications are also subject to those central points of control. Looking at your AWS.

Of course, there are ways to work around that such as using IPFS for your assets, but it's important to remember that decentralization is not an all-or-nothing proposition. Most Web3 apps may be built on top of a centralized infrastructure, but that doesn't mean they can't be used to create (mostly) decentralized applications. But it is ridiculously difficult to create completely decentralized applications unless you're building everything entirely from scratch including your infrastructure and application logic, and even then, I'd argue that most UX/UI is still centralized.

3. Most Web3 apps rely on crypto-currencies.

When it comes to decentralization, crypto-currencies are something of a paradox. On the one hand, they're intended to be an alternative to traditional centralized fiat currencies, which are controlled by central banks and other financial institutions. On the other hand, Ethereum and other crypto-currencies are themselves fairly centralized. For example, Bitcoin is controlled by a small group of miners who have a disproportionate amount of power over the network. And proof of state currencies are generally even worse, controlled by an even smaller group of early creators, backers/investors or worst a single benevolent dictator. As a result, many people have begun to explore alternatives to traditional crypto-currencies adding proof of (almost anything) which is almost always an off-chain and centralized data proof.

4. Most web3 applications are built using centralized MVC frameworks

The truth is Most web3 applications today are built using centralized MVC frameworks such as React. The Model-View-Controller (MVC) is an architectural pattern that separates an application into three main logical components: the model, the view, and the controller. The MVC pattern is often used in modern web applications, where the view corresponds to the graphical user interface (GUI), the controller handles user input and interactions, and the model stores data and business logic. This is basically how any web application is built even web3 ones.

The problem with this approach is that it creates a single point of failure. If the server goes down because of scalability or security issues, the entire application goes down with it. Additionally, this approach is not very scalable without a significant amount of parallelism (think more cloud servers). As the number of users grows, the server quickly becomes overloaded and some central process is required to add additional capacity, either manually, by a person or automatically, by an algorithm. Finally, centralized MVC frameworks are not very secure, they're generally made of a bunch of random open source components each with its own set of maintainers.

Also because a lot of the data is stored on a cloud service or interacts directly with an API/RPC or blockchain node, it is easy for hackers to gain access to sensitive information through these central points of interaction.

5. Most Web3 apps Interaction via a centralized RPC

Most blockchain RPCs are centralized. This means that a single entity, such as a company or organization, controls the flow of data that is distributed to the blockchain. However, there are some decentralized RPCs that are controlled by a smaller group of users. These types of RPCs are more secure in that offer more transparency but are also difficult to scale beyond the initial group who create the end points.

The bottom line is that the web will always be centralized to some degree. Web3 can be more decentralized than it is today, but the idea of a completely decentralized web is a myth.