Grass Network: How to Profit from Sharing Internet Resources?

Positioning and Use Cases of Grass

Grass is a project deployed on Solana on-chain, combining AI, Depin, and Solana technologies, positioned as the Data Layer for AI. It is a Decentralization network crawling platform designed to help companies and non-profit organizations train artificial intelligence (AI) by leveraging unused internet bandwidth. It achieves web crawling through a browser extension application, utilizing individuals’ unused internet bandwidth and rewarding users with Grass Points. Grass aims to redefine the internet incentive structure by allowing users to share unused internet bandwidth resources, enabling users to directly benefit from the network and ensuring that the value of the internet is in their hands. Currently, the network has over 2 million users running Nodes, capturing a large amount of data for AI models.

Technical Architecture

Grass Sovereign Data Rollup is a network specifically built by Grass on Solana, which enables protocol to handle all transactions from data source to processing, verification, and building data sets. The network is built around validators (issuing data collection instructions), routers (managing web request distribution), and Grass Node nodes (used by users to contribute their idle network resources). The specific architecture is as follows:

Validator: Receives, validates, and routes web transactions in batches. Then, generates ZK proofs to verify the on-chain session data. On-chain proofs can be referenced in the dataset to verify the data source and trace its lineage throughout its lifecycle. The validator set will transition from the initial centralized framework of a single validator to a decentralized validator committee.

Router: Connects the Grass Node to the validator. The router ensures the traceability of the Node network and relays bandwidth. Grass is incentivized to operate based on the proportion of total validation bandwidth provided through the Relay. The router is responsible for reporting the following metrics to the validators in the network: size of each incoming and outgoing request (in bytes); latency of each Node and validator; network status of each connected Node.

Grass Node Node: Using the user’s unused bandwidth and Relay traffic so that the network can fetch public web data (not the user’s personal data). Running a Node is free, and the person running the Node (Node operator) is rewarded for the data passed through its Relay.

Zk Processor ZK Processor: batch processing the validity proof of session data for all web requests, and submitting the proof to the L1 blockchain. This operation permanently records every fetching behavior executed on the network. It also lays the foundation for a comprehensive understanding of the source of AI training data.

Grass Data Ledger: This is the link between the captured data and the L1 settlement layer. The ledger is an immutable data structure that hosts the complete dataset and links the data to its corresponding on-chain proof, serving as a data repository to ensure the source of the data.

Edge Embedding Models: This is the process of transforming unstructured web data into structured models. This includes all necessary preprocessing steps to ensure that the collected raw data is cleaned, normalized, and structured in a format that meets the requirements of AI models.

Technical Features

In the above architecture, the Grass network is located between the client and the web server. The client sends web requests, which are sent through validators and ultimately routed through the Grass Node. Regardless of which website the client requests, its server will respond to the web request, allowing its data to be scraped and sent back to the route. It is then cleaned, processed, and prepared for training the next generation of AI models.

This process requires understanding two main additional features: Grass Data Ledger and ZK Processor.

The data ledger is the final storage place for all data, it is the permanent ledger of each data set captured by Grass, embedded with Metadata, recording its original lineage from the moment of origin. The Metadata proof of each data set will be stored on the Settlement layer of Solana, and the Settlement data itself will also be provided through the ledger.

The purpose of the ZK processor is to help record the source of the data set captured on the Grass network. The process is as follows: when a Node (i.e. a user with the Grass extension installed) on the network sends a web request to a given website, it will receive an encryption response, which includes all the data requested by the Node. This is the moment when the data set is born, the moment of origin that needs to be recorded, and also the moment of recording the Metadata. It contains many fields, such as the session Secret Key, the URL of the captured website, the IP Address of the target website, the transaction Timestamp, and of course the data itself. Thanks to these necessary information and the clear data set from the website source, the AI model can be trained correctly and faithfully.

The ZK processor can prevent data that needs to be settled on-chain from being seen by Solana validators. In addition, a large number of web requests that will be executed on Grass in the future will exceed the throughput that L1 can handle. Grass will soon expand to the level of processing tens of millions of web requests per minute, and the metadata of each request needs to be settled on-chain. Without the ZK processor to provide proof and batch processing, it is impossible to submit these transactions to L1. Therefore, Rollup is the only possible way to achieve the planned goals.

In addition to recording the source website of the dataset, Metadata also indicates which Node it is routed through on the network. This means that whenever a Node fetches the network, it can receive rewards for the contributions it has made without revealing any identity information. This allows Grass to reward Nodes proportionally, with more valuable and data-fetching Nodes receiving more incentives. This mechanism will significantly increase rewards in the world’s most popular regions, ultimately encouraging people in these regions to register and increase network capacity. The larger the joined network, the larger the capacity Grass can fetch, and the larger the network data repository stored. More data means that Grass can provide more data to AI laboratories in need of training data, thereby incentivizing the network to continue to rise.

Grass Node operation and security mechanisms

Grass Node operation is free and serves as a gateway to the internet. Node operators (i.e., application users) are rewarded for the traffic through their NodeRelay and receive network traffic based on their reputation score and geographical demand.

Grass Node has two main purposes: to relay traffic (i.e., web requests) initiated by clients and instructed by validators; and to return encrypted web server responses to the designated routers.

The supported systems of Node are shown in the figure above, and the process of running Node is also very simple: create an account, download the Grass desktop application, and connect to the network.

After the connection, the Node will automatically register on the network. The operator is responsible for maintaining the network’s normal operating time so that the Node can forward network requests to public network servers. Every request sent to the Grass Node is an encryption data packet. The data packet only provides direction to the Node at the destination of each data packet. Network requests are authenticated by the Digital Signature of all parties involved. These signatures will verify the legitimacy of the request and determine whether it should be forwarded to the target network server (i.e., a public website). This encryption process prevents data tampering and ensures that validators can accurately measure the reputation of each Node.

The reputation score of the Node mainly includes the following points:

Integrity: Assess whether the data is complete and whether the data set contains all necessary data points required for the expected test cases.

Consistency: Check the consistency of different data sets or data within the same data set that changes over time.

Timeliness: measure whether the data is up-to-date when needed.

Availability: Assess the data availability of each Node.

In terms of security mechanisms, the Grass network does not use user Nodes (i.e. computers) or view any actions performed by users on their computers. It simply routes internet traffic through the user’s IP Address, completely unrelated to the user’s activities. This means that Grass has zero access to users’ personal data, and all captured data is 100% sourced from public network data.

In addition, Grass uses bandwidth encryption to ensure that all users are protected when sharing the Internet connection. Grass also collaborates with leading network security compliance audit company AppEsteem to monitor Grass products 24 hours a day for vulnerabilities, leaks, backdoors, and malware to ensure user safety. AppEsteem certification enjoys a high reputation in the cybersecurity industry, and obtaining this certification means that Grass products are also included in the allowlist of top anti-malware applications, including Avast, Microsoft Defender, McAfee, AVG, etc.

Functionality of Grass token

The holder of Grass token can participate in the Grass network in the following ways:

Transaction and Repurchase: After Decentralization, Grass will be used to support network transaction capture, dataset purchase, and LCR (Real-time Context Retrieval) usage.

Stake and rewards: Stake Grass to the router to facilitate network traffic and receive rewards for contributing to network security.

Governance: Participate in the development of the Grass network, including proposing and voting to support network improvements, coordinating with which organizations to collaborate, and determining incentive mechanisms for all stakeholders.

According to the Dune website, currently, Grass’s stake Annual Percentage Rate is around 45%, with about 33% of grass tokens participating in the stake, and the stake amount exceeds 26 million.

Router stake and returns

The router acts as a decentralized hub, connecting all network Nodes and managing the incoming and outgoing web requests of validators. The Router operates on an incentive basis, with rewards proportional to the amount staked on each router. All traffic through the relay routes of the Router undergoes encryption and measurement to ensure security and performance.

Currently, the stake amount of each Router is as shown in the above figure. Users can stake Grass to the Router to earn profits on behalf, and the commission for each Router is different.

Currently, the Grass stake amount of DBunker is about 1.43 million, with a minimum stake period of 7 days and a commission of 10%. (Data source STAKE can connect Wallet to stake Grass and obtain Router stake income.)

Summary

Grass is committed to building a fair and open Decentralization Data Layer, aiming to address the ethical and data quality issues in current internet data extraction, and oppose the data monopolies controlled by a few large companies. In terms of technical architecture and features, Grass introduces a Metadata mechanism that records the sources of all data sets by building data Rollup. The ZK proofs of these data are stored on the L1 settlement layer, while the Metadata itself will eventually be bound to its underlying data sets, as these data sets are stored on Grass’s data ledger. Therefore, ZK proofs lay the foundation for increasing transparency and providing proportional rewards to Node providers for their work, which is also an important factor in incentivizing the expansion of the Grass network.

Grass focuses on the data at the intersection of Crypto Assets and AI. Unlike traditional participants in closed-source, centralized AI, it is the original source of Decentralization for AI data. As an important participant in the web3 wave, Grass uses Decentralization technology to build a fair and open Data Layer for AI companies and protocols, with promising prospects based on market demand.