Author: PermaDAO
The word “narrative” is popular in the crypto industry. It originates from a sociological concept "metanarrative" proposed by the French philosopher Jean-François Lyotard, which is used to describe stories that serve as universal ideals and have basic legitimacy in political propaganda.
However, the encryption industry is increasingly misled by the concept that it seems that the "narrative" determines the value of a project, rather than the increase in production efficiency or improvement in production relations that the project itself can provide.
Since Hyperparallel Computing Network AO was released on February 28, it has mostly been interpreted as a narrative upgrade for Arweave. It allows Arweave to evolve from pure decentralized storage to "storage + computing" and become a member of Ethereum's killer lineup.
It would be a deep pity if the understanding of AO only remains at the "narrative" level. When you try to get closer to understanding AO, you will find that it has huge potential and has the opportunity to bring far-reaching changes to the encryption industry.
Unlike the blockchain network we have in mind, AO seems to have no boundaries, no limits, and everything is possible. We have noticed that in less than a month, a series of exciting applications based on AO have emerged, including instant messaging software, social platforms, games, etc. Perhaps in a few years, the entire crypto industry will realize that our current position is not just a crossroads of narrative upgrades, but the beginning of a real paradigm shift!
In order to allow developers, researchers and even investors in the encryption industry to have a deeper understanding of AO, PermaDAO invited two AO founders to conduct an in-depth conversation about AO on its origin, evolution and future vision. They are Sam Williams, founder of Arweave and CEO of Forward Research, and Outprog, founder of EverVision, proposer of the SCP paradigm, and early AO concept designer.
The development history of AO
First Sam recounted to us the development of AO in his memory. During the early COVID-19 lockdowns in 2020, the Arweave team discussed and developed a concept via Zoom meetings: creating a neutral decentralized computational logging system capable of inferring the state of any program. Based on this concept, they quickly developed a basic smart contract system SmartWeave and released it as a proof of concept. SmartWeave quickly became a success, attracting many teams to join the Arweave ecosystem in the summer of 2020 to build SmartWeave-based projects, which became the core of the Arweave ecosystem. Subsequently, Outprog proposed the Storage Computing Paradigm (SCP), extended the concept of computing logs to all types of data logs, and wrote a series of articles to lay the foundation for understanding the core theory of AO.
The idea of AO further proposes the implementation of SmartWeave inside bundled interactions. These data items can have their own labels and are not limited by scalability, but have different data availability properties, indicating the possibility of an arbitrarily scalable computing system. The core design of AO is message passing, which abandons the interactive lock in traditional SmartWeave and uses a message log. These messages can come from users or other processes. This design began development in the summer of 2021 by the Forward Research team. By the end of 2022, AO’s architecture was proven to be feasible and capable of horizontal expansion. The internal testnet was launched in January 2023 and was quickly joined by 30 active developers. Subsequently, the testnet was publicly released at the end of February. In just three and a half weeks, approximately 3,000 developers were attracted to join and build the project, demonstrating the community's broad acceptance and interest in AO.
Outprog also described the birth history of AO based on his experience with Arweave. Since July 2020, he has joined the Arweave ecosystem and started researching it, attracted by its potential as a consensus data store, which he believes opens up endless possibilities for computing. His goal is to drive mass adoption of blockchain technology, not just for cryptocurrencies and NFTs, but to become part of everyday life. He developed everPay, the first multi-chain payment network on Arweave and a bridge between the Arweave ecosystem and other blockchains, and received investment from Arweave in May 2023. everPay is completely 0 gas and requires no waiting. All data will be uploaded to Arweave to ensure verifiability. His AO idea originated from Arweave Day in Asia in Singapore in 2022, and then he discussed parallel computing issues with other developers at the Arweave ecological gathering in Berlin in the summer of 2023. In order to solve the problem of trusted communication between SCP applications, he designed the Message Protocol, which can be regarded as the predecessor of AO.
AO design and improvement process
Sam mentioned that in the early stages of the project, the team envisioned building AO as an extension layer on top of Warp, but they quickly realized that this approach would require fundamental changes to the SmartWeave architecture, which was not practical in practice. Realistic. As the team continued to focus on the development of AO late last year, at the end of the year, they ushered in a breakthrough "almost discovery" moment, confirming that the designed architecture was not only reasonable and feasible, but also capable of horizontal scalability. When using AOS for the first time, the team experienced a completely new way of operating: they were able to log in and launch any number of processes within the system, all running in an unbounded flat space. Users can easily log into processes, issue commands, and communicate seamlessly with other users' processes in a way that is not only highly composable but also highly scalable.
Why is AO called a global parallel computer?
Sam said that AO provides a single system image that allows users to experience the underlying structure of many different computers in a unified and single computing environment. This system supports running an unlimited number of processes in parallel in a decentralized environment, achieving true parallel computing, as if creating a powerful decentralized supercomputer. Its design is inspired by the idea of distributed operating systems, which aim to increase the capabilities of a home, office, or workplace by adding more computers, while making the user feel like they are logged in to a single computer. AO's test network runs on more than 220,250 different nodes. Anyone can connect their own computing unit to this system, thereby improving the overall computing power and making AO a democratized supercomputer. Users can device, such as a laptop or mobile phone, to log in and perform computing tasks. This design not only promotes broad user participation, but also improves the accessibility of decentralized computing, whether by running nodes or using services to perform calculations.
Outprog recalled that the original idea of AO was an all-powerful state machine, similar to an all-powerful and trustworthy computing platform. The client, operating system and all running programs in the user's hands originated from Arweave. AO itself does not solve the problem of verifiability. With the SCP paradigm, verifiability is provided by the consensus data on Arweave.
The role of AO is actually to replace the traditional HTTP communication protocol with the Message Protocol, providing a decentralized messaging method for SCP applications, thereby promoting communication between decentralized applications. According to the vision in July 2020, if both user requests and server responses are processed through AO, it will be possible to rebuild a truly decentralized Internet and migrate the entire Internet to Arweave. The value of AO goes beyond storage. It can also record current stories and distribute future value in a decentralized manner, turning Arweave's "Library of Alexandria" into a platform for recording and distributing value.
The difference between AO and traditional blockchains such as ** Ethereum, ** Solana, and ICP, which is also a global computer
Himself :
Comparison with Ethereum
Ethereum has been described as a "world computer," but it's actually more of a "world calculator." While it enables many innovative applications, it doesn't live up to the original vision of a global computer. AO adopts a different architecture. It does not optimize the global shared address or memory space, but allows each contract to have local private memory and interact through message responses. This is fundamentally different from the Ethereum model.
Comparison with other blockchains such as Solana
Solana and some other blockchains (e.g. Sui, Aptos) employ widely different architectures, but the key differentiator of AO is its messaging approach. AO believes that in order to achieve scale, optimization of the global shared memory space should not be adopted, which is based on practical experience outside the world of distributed systems and traditional networks.
Comparison with ICP
ICP handles the calculation itself and the achievement of consensus through consensus, while AO adopts a lazy evaluation calculation method. AO has a dedicated subnet, the computational unit, for performing these calculations. This approach differs from the way ICP handles computation and consensus reaching. The ICP community seems to have challenges in forming consensus on the fly, while AO, through its subnet of computing units and staking mechanism, is able to provide more reliable and economically guaranteed computational outputs.
What makes AO unique?
The design of AO was inspired by historical lessons from distributed operating systems, specifically the limitations of shared memory approaches in distributed operating systems learned in the 1980s and 1990s. AO takes a messaging approach, which is the foundation upon which the Internet is built, delivering messages and computation across a broad network via TCP/IP. AO brings this concept to the blockchain, achieving equally good scalability on-chain and off-chain. The AO does not directly form a consensus on the results of the interaction (i.e., the computational output), but rather on the inputs to the interaction and the order of those inputs. This means that the AO is concerned with the order and availability of data rather than directly agreeing on the calculation results. Although AO's consensus does not directly target the calculation results, the output of the calculation is still trustworthy and verifiable. This is because AO has a dedicated subnet of computing units, and even on the testnet, there are a large number of computing nodes waiting to perform calculations and provide staked, verifiable responses. Since AO uses a deterministic virtual machine and there is consensus on the order of interactions, there is also consensus on the calculation results. This ensures that calculation results are consistent and predictable, if not instantaneous.
Outprog :
AO is a new interaction layer
Traditional blockchains such as Bitcoin and Ethereum mainly solve the Byzantine Generals Problem, which is to reach consensus on a single event in a distributed network. AO and SCP (Storage Consensus Paradigm) no longer focus on the issues of blockchain and objective truth, as these issues are already well solved in blockchain systems through Proof of Work (PoW) or Proof of Stake (PoS) Solution, and AO is to build a new interaction layer on the original blockchain infrastructure.
Separation of consensus and calculation
SCP emphasizes that consensus occurs at the storage layer, using Arweave for immutable storage to ensure security and verifiability. On Arweave, there are Merkle trees and consensus on the order of the data, rather than the state of the data computation. Essentially, AO does not solve the problem of verifiability, and the functions of AO and AR are completely separate.
AO calculation processing
AO is responsible for computing and organizing data sequentially on Arweave to generate state. Since AO cannot change the order of data on Arweave, it cannot change the consensus. Ethereum, Bitcoin, Solana and ICP all use on-chain computing, which is significantly different from the design concept of SCP. SCP allows AO to perform computation on top of Arweave's storage layer by separating computation and storage, while the storage layer is responsible for maintaining the immutability and security of the data.
What use cases do developers have that were not possible in traditional blockchains before but can be implemented in AO?
Sam provides some reference directions for developers
Decentralized computing infrastructure
AO provides a new form of cyberspace that has the potential to decentralize the world's computing infrastructure, making the cost of running computations comparable to running on cloud services such as Amazon EC2, while also providing verifiability of smart contracts sex.
Large-scale computing as smart contracts
AO allows developers to build large-scale computational smart contracts, which is not possible in other existing smart contract systems. These smart contracts can perform financial interactions, such as automated algorithmic funds, which automatically adjust buy and sell orders based on market data and news.
Trust-minimized financial services
Developers can create trust-minimized financial services, such as decentralized algorithmic funds, where users can deposit tokens without trusting the fund manager and all operations are performed automatically on the chain.
Decentralized data and information processing
By using Oracle services like 0 rbit, developers can process and verify large amounts of data from the Internet on AO without having to trust a centralized Oracle provider. This service enables trust-minimized access and processing of HTTPS-encrypted data across the entire Internet.
Autonomous AI agent
On AO, in the long term, one can envision a cyberspace populated by autonomous AI agents that live and interact in their own virtual universe, unconstrained by the laws of real-world physics.
A new financial system
AO enables developers to build a new financial system that contains autonomous financial agents that can make decisions autonomously based on market information and allow users to deposit or withdraw funds at any time.
What is the next development and roadmap of AO? When will the mainnet be launched?
Sam explains AO’s future roadmap
current stage of development
The mainnet V1.0 version was built and is currently running as a testnet. The core data protocol has been completed and no major changes are expected. The data protocol itself does not provide security, but already forms the basis of AO.
Second stage goals
A staking mechanism will operate using a set of nodes and three different subnets, each responsible for performing specific tasks, and staking is required to ensure that these nodes and subnets operate as expected and enhance system security. The staking mechanism not only guarantees the normal operation of nodes, but also provides security guarantees for computing, message delivery, and process scheduling. Although related smart contract designs should be kept simple, they must be adequately tested to verify their reliability. Nodes will also assume the role of "gatekeepers", responsible for monitoring processes, child nodes, and conducting network voting when violations of security policies are detected. AO's core protocol infrastructure is composed of an overlay of underlying data protocols, off-chain implementations of nodes, and smart contract layers that allow for staking and penalties, thus building a decentralized and trustworthy computing environment.
The third stage goal
The long-term development goal of AO is to gradually realize strict security protocol execution of computing units, messaging units, and scheduling units by combining a hybrid authority proof system and a pledge proof mechanism. This process will involve creating software to enforce security rules and smoothly upgrade the system after a period of operation, allowing scheduling units to be staked and users to run schedulers autonomously. AO aims to gradually increase the degree of decentralization and eventually become a permissionless decentralized proof-of-stake system. The transition to mainnet previews and release candidates is expected through a series of testnet versions, which may take several years until the data protocol and smart contract layers reach 99.5% stability, at which point the system will officially become mainnet.
How will AO develop in the future?
The AO team hopes to take an approach similar to Solana’s “mainnet beta” early on until version 2.6 is reached. Over time, the team can get better at this. It is now almost certain that a solid foundation is in place to provide the world with a protocol-based guarantee that the core architecture of the system will not undergo large-scale changes. Although it has taken about six and a half years to develop AO, it is hoped that future development processes can be shorter and more mature. The team ensured an architecture that was nearly optimal from a scalability perspective before starting to build AO, recognizing that it could do better with naming and versioning early on. I believe the team will adopt a more mature approach to AO to optimize the development process and timeline.
What’s next for EverVision? What contributions has EverVision made to the AO ecosystem and how will it participate in AO construction in the future?
Outprog :
everVision is a laboratory whose goal is to explore large-scale applications of blockchain. AO may be a best practice for the current large-scale application of blockchain. The team will reconstruct the work results of the past few years according to AO's specifications, which means that EverID can become a universal identity standard on AO, making it easier for ordinary users to use these applications. At the same time, everPay’s cross-chain capabilities will also be transformed into native cross-chain capabilities on AO. AO's "sub-ledger" can greatly solve the problem of payment expansion. These are the key points that we need to integrate technically.
Permaswap has launched the AO test token CRED. Now you can build liquidity for CRED/AR. You can also easily trade CRED on Permaswap, or provide liquidity to earn handling fees. Permaswap is a very mature DEX product and can be regarded as the Uniswap of the AO ecosystem. Next, Permaswap will also highly integrate AO. The tokens on AO will be able to easily create trading pairs on Permaswap and support multiple market making curves.
everPay will also deeply integrate AO and become a bridge connecting the AO ecosystem with other ecosystems. EverID is committed to becoming the universal identity layer of the AO ecosystem and is committed to solving the challenges that the high-threshold public and private key account system brings to users entering Web3.
Sam said the losses from protocol changes that occurred on other blockchains will not be repeated in AO
Changes on Ethereum have caused damage to projects such as Aragon, and the risks that protocol changes may bring to users. People have concerns about building applications on AO, especially when the system is marked as a testnet. They are concerned about the stability and integrity of the system. Although this could pose communication challenges, the team chose to clearly communicate the current status of the system and possible future changes. When building the protocol, the team valued the importance of providing users with immutable and guaranteed rights that should not be controlled by any one person or organization. To better manage user expectations, the team uses terms like “mainnet preview” and “mainnet release candidate” to describe the current state of the system and anticipate possible future changes. When the system reaches an almost completely certain state, it will be called the mainnet and will have appropriate permissionless upgrade and fork capabilities, ensuring the long-term stability of the system and the protection of user rights.
How to understand the concept and role of sub-ledger in AO?
Himself :
Sub-ledgers provide a mechanism that allows the system to add additional computing threads by creating new processes (i.e. sub-ledgers) when the single-threaded computing power of a single process reaches saturation. This approach is similar to creating new tables or shards in a traditional database to spread the load and increase overall processing power. Subledgers allow computing power to be naturally dispersed across a decentralized network, meaning that computing power can be dynamically adjusted and scaled based on the computing resources available in the network.
The interaction and data transfer between sub-ledgers is designed to be flexible enough that users and application developers do not need to be aware of the complexity behind it. Users can seamlessly conduct transactions and interact between different sub-ledgers as if they were operating on a unified ledger. Tokens between sub-ledgers can be interoperable, that is, tokens on one sub-ledger can be seamlessly transferred to another sub-ledger, and the two can be merged into a unified balance. This design increases the liquidity of the token and simplifies the complexity of managing multiple accounts for users.
The concept of sub-ledgers makes it more feasible and efficient to build a payment network on AO. In theory, this could support the creation of a global payments system where users can easily pay in any currency, anywhere, via their mobile phone. Through the design of sub-ledgers, AO achieves unlimited scalability, which means that the system can expand infinitely as demand grows without encountering scalability bottlenecks common in traditional blockchain systems.
Summarize
Through this in-depth conversation about AO, we learned that SCP theory is one of the theoretical foundations of AO. SCP applications have unlimited scalability, but lack interoperability, while AO originated from solving the problem of interaction between different SCP applications. Operational effort. Later, AO was designed as a single-image parallel computing network, achieving almost unlimited scalability through multi-threaded parallel computing. At the same time, AO uses a single system image to hide the complexity behind it, making users and developers feel What you get is still interacting with a single network. In the future, AO will work steadily through iterations of different mainnet versions to find a balance between iteration speed and system stability.