Recently, at a wonderful event organized by the WebZero team, Shawn shared a different kind of speech. This time, there were no fancy slides or rehearsed scripts; it was more about an instantaneous clash of ideas and sincere sharing.
Shawn admitted that the content this time was not a speech prepared long in advance but rather a series of reflections and insights he had recently about the entire blockchain industry, Web3 philosophy, and the world computer concept. He led us to revisit people’s initial imaginations of the 'world computer' since the birth of Ethereum, pointing out how the industry has gradually deviated from this vision over the past few years, turning towards 'ledgers,' 'financialization,' and 'composability.' And now, with the emergence of JAM, we may finally be able to return to that starting point—redefining 'what truly is a world computer.'
This is not just a talk about technology; it is more like a call to rethink the goals of Web3, to re-understand the meaning of 'decentralization,' and to redraw the future internet we hope to see.
This article will take you on a journey of thought, from 'What is Web3?' to 'What core elements should a world computer have?' and 'Why do we need a resilient digital infrastructure?' We will peel back the layers of the current blockchain technology to reconnect with the original ideals.
Due to the lengthy content of the speech, PolkaWorld will release it in two parts. This article is the first half, mainly covering:
What is Web3?
What is a world computer?
Components of a world computer
What does a decentralized computer world look like?
The cost of a world computer
Continue reading to see the full content!
Thanks to the WebZero team for organizing this event! Honestly, this talk is not as well-prepared as some of my other speeches; it is more about some thoughts I am currently pondering. So what I want to share is that our thinking about the industry we are building has changed significantly and, to some extent, returned to our initial starting point.
If you remember the era of Ethereum, the initial idea was to create a world computer. However, for some reason, we eventually entered a world of blockchains, DeFi, decentralized ledgers, and distributed ledgers. It wasn't until the innovation through JAM that we returned to the concept of computers. We began to realize that technology could actually help us return to the original vision. Therefore, the goal of this talk is to redefine 'world computer' and make everyone rethink the technology we are using.
I believe that with the development of new technologies and the capabilities and possibilities of what we can actually develop, we can truly rethink everything we are doing. So, I want to reassess the idea of Web3, the world computer, the components of the world computer, the Web3 internet, what it might look like, and the products designed for the world computer.
What is Web3?
Let’s start with Web3. So, what is Web3?
I think this is a very complex issue, but it is essential to approach it from this angle. Web3 is the goal we are collectively striving for. The core value of Web3 is that it helps us distinguish which products are genuinely useful and which are just gimmicks or hype-driven things. I find it challenging to define Web3 in a concise phrase. For example, my often-used phrase 'less trust, more truth' might resonate with some who are not very familiar with our field.
Web3 can be seen as a fairer version of the internet. Recently, Gav described Web3 as a 'resilient public digital infrastructure.' However, I believe that all these brief definitions do not truly address the problems we face. We need to start from the problems themselves to understand what we are actually doing.
Thus, I have a simple hypothesis: centralized power will ultimately be abused. As technology drives globalization, the influence of central powers has become greater than ever, and computers play a crucial role in building public digital infrastructure. If the world’s public digital infrastructure itself is more decentralized and resilient, we can create a more resilient world that can better withstand these powers. What Web3 ultimately describes is this resilient public digital infrastructure. This is also what Gav wants to express; he calls Web3 the 'resilient public digital infrastructure' of the world.
In my recent speech, I presented a slide listing some historical events to illustrate why centralized power tends to corrupt and use that power against us. For example, when the financial crisis occurred, we rebuilt banks and conducted regular information reviews, effectively preventing organizations from spreading true information and thus controlling the narrative. We also limited them through means such as blocking their financial activities. Soon, the exposed facts showed that governments and other institutions were monitoring us, stealing our data globally. I once asked a large language model to list 10 historical events to prove this, and it even provided more evidence. If you ask for 100 such events, you could certainly get 100 of them. Looking back at history, whether a thousand years ago or a thousand years from now, even if you could foresee the future, you would repeatedly see how central powers abuse their power. Even in daily life, similar phenomena can be observed, such as the social media information stream, which is controlled by some algorithm, right? These algorithms determine the news you see about Israel, Palestine, Ukraine, or Russia; who can know what is really happening behind the scenes? These examples show how central power affects and determines what we can see and do every day. The true meaning of Web3 is to recognize these issues and address them through specific tools and attributes.
Web3 consists of a series of decentralized tools and technologies that are verifiable and independently verifiable. They are self-sovereign, interoperable, privacy-preserving, open, public, borderless, neutral, essentially durable, resilient, universal, high-performance, consistent, and accessible. When we discuss what Web3 specifically aims to achieve, it attempts to provide solutions to these issues. Even though this list may not be a complete account of everything we need, it merely aims to describe some tools and characteristics we can use to address the issues of central power.
In simple terms, this is Web3.
What is a world computer?
Next, let’s delve into the origins of the 'world computer.' So, what is a world computer?
The world computer refers to a single, globally accessible, trustless computing platform that is not owned or controlled by any single entity. This computer consists of a network of decentralized participants who reach consensus, and the state of the computer, as well as the applications and programs running on it, will execute according to the written terms. You can prove that the contract or application you wrote is written in a certain way, and it will execute in that manner. It can be proven; it really can be done.
In fact, the world computer is a core component of the Web3 we envision; it is a resilient public digital infrastructure. The concept of the Web3 computer comes from the same principle as Web3 itself. The reason I joined Parity and entered the Polkadot ecosystem is partly because of Dr. Gavin Wood. His work is very inspiring, from Ethereum to Polkadot and now to JAM; his work has consistently revolved around the same goal. Nearly ten years ago, he wrote a blog post about Web3.0; although the video is a bit blurry, nine years ago, Ethereum claimed to be a world computer and showcased this concept to the world at the first DevCon. This was the original vision. However, can today’s Ethereum really be considered the standard of a world computer?
I believe that Ethereum claims to be a world computer because it is Turing complete. I think that at the time, this was fair, but based on our technological understanding today, it is not a good standard to measure a true computer. Turing completeness does not effectively test whether a computer is useful. Many things are Turing complete. Microsoft Excel is Turing complete, and the game I like to play (Magic: The Gathering) is also Turing complete, but (Magic: The Gathering) is not a computer, right?
This is not a good, useful computer. For example, Microsoft Excel, you wouldn't build applications in Excel; these things don’t make much sense. In fact, many blockchains have achieved Turing completeness and added smart contract platforms, but they do not genuinely represent a computer, nor do they behave like one. So what are they? Most of these blockchains are distributed ledger technologies. I believe you have all heard this term.
This term is actually like the business jargon I encountered when I first came across blockchain, 'Oh, these are distributed ledger technologies,' but in reality, they are just the spreadsheet version of Web3. Their task is basically to maintain and track data, perform some small calculations, such as using formulas in Excel to transform data or track changes over time, but ultimately, they are just creating a ledger to record a decentralized storage and data history.
However, you can build applications in Excel; in fact, someone has even tested building games in Excel, but it doesn't provide a good experience. Building applications in Excel feels as painful as building applications for blockchain today. It feels strange and unnatural; you have to build an entirely new set of applications under the constraints of the blockchain environment, and that is precisely because blockchains have not yet realized the functionality of a computer; they implement distributed ledgers.
This is an important 'aha' moment I want to share with everyone. The key point is how we should view the future of blockchain, especially from this point forward. I believe that blockchain technology can be used to do two different things.
Some people are developing distributed ledger technology primarily used for data storage. This is crucial in decentralized finance (DeFi) because ledger technology has always been the foundation of the financial industry—from traditional paper ledgers to electronic ledgers and now to decentralized ledgers. The benefit of this approach is that it meets the market demand for decentralized finance, making distributed ledger technology (DLT) very successful.
But at the same time, there is a new technology called JAM, which is leading a revolution. JAM attempts to build a decentralized computer based on blockchain technology, combining it with other Web3 technologies. Unlike traditional computers that store data, the primary goal of decentralized computers is to perform computations, and its core function is to provide decentralized computing bandwidth, becoming the foundation of the new decentralized internet.
However, this decentralized computer is still a very new and unverified concept, with no guarantee that it will succeed, making it a speculative product. This means it is still full of risks and uncertain whether it will truly be useful.
At present, market demand does not explicitly indicate that people urgently need decentralized computers today, but we are building this technology to meet future needs. This is not something immediately needed in the market, but rather a long-term need for us as humans and as a society. We need a better world, a Web3 world, right? Now we are starting to understand why different blockchains look so different. If we look at the blockchains on the market cap list, we will find that some of them look very different, even bizarre. If you start to think about whether this is distributed ledger technology or a decentralized computer, you can see there are significant differences in the directions people are building.
Components of a world computer
Okay, next, let's talk about the components of the world computer. First, I want to mention a concept that is relatively difficult to understand—we are talking about a virtual computer. It is not a physical computer that exists. While there are indeed computers in reality, these computers combine together to form a virtual world computer. They are not separate physical devices. So when you are building an application, the application itself does not need to know that it is not running on a real computer; it should think it is running on a real computer. And when this application actually runs, the whole process becomes more complex because it is a massive technological system.
What we are actually talking about is transmitting applications to a Web3 cloud, Web3 server, or Web3 computer for processing, and then assigning tasks to multiple nodes globally for execution. Ideally, the application itself does not need to know this. All these operations connect multiple computers through consensus algorithms, just like the 'magic' brought by Bitcoin through solving the Byzantine fault tolerance problem. Through this mechanism, we can achieve consensus among multiple computers, and they present themselves to Web3 applications as a single computer.
If we look at it from the perspective of applications, rather than seeing it as a pile of blockchains and distributed networks, but rather as a computer, we can see its different components. For instance, a regular computer has a hard disk, while a decentralized computer would have a decentralized hard disk. This is achieved through the Merkle-ized blockchain state.
A decentralized computer will also have processors. We know that computers have different types of processors, such as single-core processors, threaded single-core processors, and multi-core processors. Systems like Polkadot and Ethereum are like single-core processors; they have a single state transition pipeline that executes states one after another. Systems like Solana, while also single state machines, allow some degree of parallel processing if it can prove that the data of two transactions is entirely independent.
Polkadot acts like a multi-core processor, where multiple state transitions can occur simultaneously on different cores, achieving a completely different processing method through execution and sharding. This design is more powerful and efficient than traditional single-core processors.
Of course, computers can also have memory, which is a relatively new concept. Many old computers and ledger systems did not have the concept of memory; they only had CPUs and hard disks. But we know that computers need to continuously use memory, which is that non-persistent storage. This concept is introduced by Polkadot through our PVF and data availability.
If you don't know, when we do parallel chain-related work, we need to store data in a 'data availability layer.' But the problem is that this layer is not widely used; it has not yet become mainstream. Recently, projects like Celestia are trying to use data availability as a secondary service, meaning separating it from the main network to create non-persistent storage. But this approach has some problems; imagine if your computer's RAM was connected by cables—how slow that would be. In fact, Celestia's speed is much slower than a local DA layer. Systems like JAM are attempting to introduce local DA layers and local memory, allowing you to access and use them directly as needed.
So, to emphasize again, if you look at the components of the blockchain, you will find that it is not just a ledger system; it is actually a computer. Moreover, there are some novel components within the blockchain that cannot simply be compared to traditional computers. For example, we have a consensus layer used to reach consensus; there is an economic layer that helps the computer influence the real world; and there is on-chain governance that allows the physical world to influence the computer in return. These concepts do not have direct counterparts in traditional computers because traditional computers are typically owned by one person, while these layers in the blockchain are collectively owned and distributed by many people worldwide. These additional layers act like the 'glue' of a decentralized world computer, allowing various parts to work together and ensuring the computer operates correctly.
What does a decentralized computer world look like?
If we continue down this path, we can start to imagine what a decentralized computer world would look like and how it will evolve. Currently, a critical part of computers is the GPU, which is very efficient at handling parallel simple operations. Although no blockchain can fully achieve this at the moment, I believe that if we treat the blockchain as a decentralized computer, or as the technology we are building, we can start to introduce decentralized GPUs to handle a large number of parallel small tasks.
We know that GPUs are very important in computing. At the same time, there are privacy-preserving technologies, which we haven't fully engaged with yet because they might not be very profitable. For example, mix nets and zero-knowledge privacy layers; we hope to introduce these technologies into the blockchain. I believe that although blockchains are inherently transparent, this transparency is somewhat simplistic. It merely verifies whether you and I have done the same thing, but what we really want is to be able to verify this private information while protecting privacy.
If you view this system as a decentralized computer, then the internet connects many such computers. This is the concept of trustless bridging, and in fact, we are still continuously working to improve this part. Ultimately, you do not want just one decentralized computer; rather, you want a larger decentralized computer, similar to the concept of data centers.
This is what we call the 'grid' concept. The grid connects multiple JAM computers to collaboratively provide higher processing capabilities. More creative ideas may emerge. If you discuss deeply with me, I believe we can come up with countless ideas about my computer and how to represent them in a decentralized manner. This perspective is very important. Many people feel that decentralized computers are very abstract and impractical.
So is it really attractive? Who is attracted to it? Do people really need a decentralized computer? When I show the characteristics of Web3, some people resonate, while others do not quite understand. I believe the way to understand a decentralized computer should be to see it as different types of computers.
A decentralized computer, or world computer, is a distributed computer that is inherently fault-tolerant and can provide very high stability and guarantees. This is a very powerful feature. It is a verifiable computer, right? On this computer, people do not need to trust the platform or other users they interact with.
It is a very resilient computer, where applications and services do not have to worry about the laws and restrictions of the country or region they are in and can freely publish uncensored applications. I am not saying you have to be attracted to these characteristics to use this computer, but any one of these characteristics is enough to make you willing to place your application or service on this computer. You can think about it this way: if I were introducing it to a particular business, I wouldn’t say, 'This is a place where anyone can publish crazy memes or lies, and no one can censor it.' That is not its appeal. Perhaps for businesses, its appeal lies in its resilience or its distributed nature, with inherent fault tolerance. When you face businesses or certain target groups, you can explain to them the characteristics of this computer; the world computer embodies all these features.
The cost of a world computer
Now, let's talk about its cost. The world computer will not replace your personal computer, nor will it replace most computers. It is a whole new type of machine that addresses different categories of problems and brings some additional costs. As we know, the world computer is inherently slower than traditional computers, and the cost of using it is higher because you are running the same application on many computers. It is slow and more expensive.
It is a limited public resource. Because it is a world computer, everyone can use it, but this also means that if its bandwidth is limited (which it clearly is), then only a certain number of people can use it at a certain scale, and you may be restricted in how you wish to use it. Ideally, we hope it can scale to allow everyone to have some access, like a public resource, but if you want more access, you naturally have to pay a fee.
This will also bring certain limitations. It is a permanent and immutable design. Using blockchain technology as one of the infrastructures for the decentralized computer means you will have an immutable historical record. All past mistakes or actions will be reflected in it, which may not always be attractive to everyone. Of course, one important aspect of the world computer is that it shifts the burden onto the users. There is no 'forgot password' link, and no support hotline. This computer is unforgiving in some respects, very primitive, but in many ways, that’s precisely why it can be uncensorable, resilient, and not controlled by any one party. You have complete sovereignty, which can be both a good thing and a bad one.
Original video: https://www.youtube.com/watch?v=9tFyfYkorCw