The World Computer in a Decentralized Network: The Vision and Challenges of AO

Executing arbitrary code without trust and being able to share it with the whole world is the core concept of a world computer in a decentralized network. After Ethereum, many infrastructure projects have attempted in this direction, among which the upcoming AO network is a typical representative.

For a "world computer," we can roughly divide its functions into three main parts: data computation, access, and storage. For a long time, a certain storage project has played the role of the "world hard drive," while the newly launched AO network (Actor Oriented) introduces general computing capabilities and provides smart contract functionality.

AO: General Computing Network Based on Actor Model

Current mainstream decentralized computing platforms can be divided into two main categories: smart contract platforms and general computing platforms. Smart contract platforms, represented by a well-known smart contract platform, share global state memory and reach consensus on the processes that change state. Since consensus requires a lot of repeated computations, they are primarily used to handle high-value transactions at a high cost. In contrast, general computing networks do not reach consensus on the computation processes themselves but validate the results based on business requirements and handle request sequences, without a shared state memory. This approach reduces costs and allows the network to scale into more fields of computing applications, represented by certain computing power networks.

There are also some projects based on the assumption of virtual machine security that integrate general computing with smart contracts. These types of networks only reach consensus on the order of transactions and verify the results of computations, while multiple state changes are processed in parallel among network nodes. The virtual machine of the computing environment ensures the determinacy of the results, so as long as the order of transactions is consistent, the final state will also remain consistent.

Such networks have low expansion costs because they do not share state memory, allowing multiple tasks to be computed in parallel without interfering with each other. These projects are usually based on the Actor programming model, and AO also belongs to this category. In the Actor model, each computing unit is treated as an independent agent processing transactions, and the computing units interact through communication. AO standardizes the message passing of Actors, achieving a Decentralization computing network.

Unlike traditional passively triggered smart contracts, AO has general computing capabilities and can actively run smart contracts through a "cron" method triggered by fixed time cycles, such as trading programs that continuously monitor arbitrage opportunities.

The rapid scalability of the AO network, its massive data storage capability, the Actor programming model, and the ability to actively trigger transactions make it very suitable for hosting AI Agents. At the same time, AO also supports bringing large AI models into smart contracts on the blockchain for execution.

Characteristics of AO Network

The AO network adopts a modular design, consisting of three basic units: Scheduling Unit (SU), Computing Unit (CU), and Messenger Unit (MU). When a transaction is initiated, the MU receives the transaction and verifies the signature, then forwards it to the SU. The SU serves as the connection point between AO and the underlying storage chain, helping the network to sort the transaction sequence and upload it to the storage chain to achieve consensus. The current consensus method is POA (Proof of Authority). Once consensus is reached, tasks are assigned to the CU for specific computations, and the results are returned to the user via the MU.

The CU collection can be seen as a Decentralization computing power network. Under a complete economic planning, CU nodes need to stake certain assets and compete based on computing performance, price, and other factors to provide computing power to earn profits. If a computation error occurs, the node will be penalized by confiscating assets, which is a standard economic security mechanism.

Comparison of AO with Other Networks

Compared to traditional smart contract platforms, AO's advantages as a general computing platform are obvious. Although a certain storage network has also launched its own smart contract platform, it is an equivalent state consensus architecture to EVM, which does not match the user experience of traditional smart contract platforms.

Compared to other decentralized computing networks, AO retains smart contract capabilities and maintains a global state on its underlying storage. Architecturally, AO is most similar to a certain asynchronous computing blockchain network, inheriting many of the latter's design concepts, such as ordering transactions only by sequence, trusting the deterministic computation of the virtual machine, and using the Actor model for asynchronous processing.

The main difference is that a certain asynchronous computing network maintains state based on containers, while AO has a shared state layer. This enhances the network's Decentralization capability but may also limit the implementation of certain specific privacy services. At the economic and design level, AO adopts a more open and flexible approach, lowering the barriers to participation and development costs.

However, AO also faces some challenges. The lack of atomicity in cross-contract transactions under the Actor asynchronous model may impact the development of DeFi applications. The new computing model also places higher demands on developers. In addition, the memory limitations of the wasm virtual machine under the AO architecture may prevent certain complex models from being used.

Nevertheless, in the context of the rapid development of AI, AO still has great potential. With the continuous improvement of technology and the gradual establishment of the ecosystem, AO is expected to play an important role in Decentralization computing and the AI field.