Sentient GRID: Paving the Way for Open AGI with Decentralized AI Network

Introduction

In the realm of Artificial Intelligence, the competition between traditional giants and decentralized networks is intensifying. On one side, we have companies like OpenAI and Google building powerful but closed "walled gardens." On the other, there are decentralized networks such as Bittensor raising the banner of openness, yet their applications remain out of reach for ordinary users. Amidst this landscape, the Sentient project emerges, aiming to create a network of specialized intelligent entities, achieving more flexible collaborative intelligence. The Sentient project focuses on creating a decentralized smart network GRID, aiming to explore AI applications in various scenarios through an open and community-built approach.

Sentient GRID: Towards Open AGI

GRID is not a single model but a vast "AI component library" gathering AI models, Agents, data sources, and tools from around the world. These modular and composable "intelligent units" are known as "Artifacts." When users send requests to GRID through Sentient Chat, the system automatically analyzes the tasks and assigns them to the most suitable agents for processing, leading to the integration of multiple results into an optimal and coherent answer.

Key Components of Sentient GRID

1. ROMA Framework: The "AI Project Manager"

The ROMA (Recursive Open-source Meta-Agent) framework is designed to handle complex, multi-step tasks. ROMA breaks down large goals into clear and structured sub-tasks. In the SEAL-0 benchmark test, the ROMA-based system achieved an accuracy of 45.6%, nearly twice that of Google Gemini 2.5 Pro.

2. Open Deep Search (ODS)

ODS is an open-source search framework developed by Sentient, enabling agents to perform complex, multi-step web searches. ODS aims to discover and integrate crucial information hidden deep within the web.

3. Dobby: A Crypto-Native Open Source Expert

Dobby is an open-source large language model fine-tuned to be an expert in cryptocurrencies and the Web3 ecosystem. Dobby is based on Meta's Llama model and fine-tuned using feedback from over 2 million community members.

4. Loyal AI and Fingerprinting

The Loyal AI concept and Fingerprinting technology ensure the preservation of ownership rights in an open-source world. Fingerprinting technology embeds a unique digital signature into AI models during training.

5. Enclaves: The "Hardware Vault" for Confidential AI

The Enclaves framework provides a solution for data and model security. Enclaves utilize Trusted Execution Environment (TEE) technology to create an isolated hardware-based "vault" for confidential AI applications.

Sentient Chat: A Unified Workflow Experience

Sentient Chat is a unified entry point to experience the powerful capabilities of GRID. When users pose complex questions, Sentient Chat intelligently analyzes intentions and activates a dynamic workflow, allocating the most appropriate resources in the GRID network to collaborate. Sentient Chat aims to replace manual workflows that require switching between numerous browser tabs and applications.

The $SENT Token

The $SENT token incentivizes the GRID ecosystem. It uniquely balances the interests of four key roles: Builders earn by contributing AI applications; stakers share in the success of applications they "invest" in; network delegates (community experts) ensure rewards flow to valuable contributions; and active users can earn incentives.

Conclusion

Sentient aims to break the dichotomy between "powerful but closed" and "open but abstract" AI. It invites developers, researchers, and users worldwide to contribute to the development of an open and sustainable AGI ecosystem.