Deconstructing the AI Framework: From Intelligent Agents to Decentralization Exploration

Introduction

Recently, the narrative of the combination of AI and cryptocurrency has developed rapidly. Market attention has shifted to technology-driven "framework" projects, with multiple projects emerging in this niche track, each reaching a market capitalization of over a hundred million or even a billion within just a few weeks. These projects have given rise to a new asset issuance model: issuing tokens based on GitHub code repositories, and Agents developed on the framework can also issue tokens. With the framework as the foundation and Agents on top, a unique infrastructure model for the AI era has formed. This article will start with an introduction to the framework and explore the significance of AI frameworks in the cryptocurrency field.

1. What is a framework?

AI frameworks are a type of underlying development tool or platform that integrates pre-built modules, libraries, and tools to simplify the complex process of building AI models. The framework can be understood as the operating system of the AI era, similar to Windows, Linux, or iOS, Android. Each framework has its strengths and weaknesses, and developers can choose based on their needs.

Although the "AI framework" is an emerging concept in the cryptocurrency field, it has developed for nearly 14 years since the birth of Theano in 2010. There are mature frameworks available in the traditional AI circle, such as Google's TensorFlow and Meta's Pytorch. The framework projects emerging in cryptocurrency are built based on the high demand for agents driven by the AI boom, and they have evolved into different fields, forming AI frameworks in various domains. Here are introductions to several mainstream frameworks:

1.1 Eliza

Eliza is a multi-Agent simulation framework designed for creating, deploying, and managing autonomous AI Agents. Developed in TypeScript, it offers good compatibility and is easy to integrate with APIs. It mainly targets social media scenarios and supports multi-platform integration, including Discord, Twitter, Telegram, and more. It supports processing PDF documents, link content, audio transcription, video content, image analysis, and more.

Use cases supported by Eliza include: AI assistant applications, social media personas, knowledge workers, and interactive roles. Supported models include local inference of open source models, OpenAI API cloud inference, and more.

1.2 G.A.M.E

G.A.M.E is a multimodal AI framework for automatic generation and management launched by Virtual, mainly designed for intelligent NPCs in games. Its feature is that it can be used by low-code or even no-code users.

The core design of G.A.M.E is a modular design that works through the collaboration of multiple subsystems, including the Agent prompt interface, perception subsystem, strategic planning engine, world context, dialogue processing module, and several other components.

1.3 Rig

Rig is an open-source tool written in Rust, designed to simplify the development of large language models (LLM) applications. It provides a unified interface for accessing multiple LLM service providers and vector databases.

Core features include a unified interface, modular architecture, type safety, and high performance. The workflow involves provider abstraction layers, smart agent invocation tools or query vector storage, retrieval-augmented generation, and other mechanisms.

1.4 ZerePy

ZerePy is an open-source framework based on Python that simplifies the process of deploying and managing AI Agents on the X platform. It provides a command-line interface, supports modular design, and allows for flexible integration of different functional modules.

ZerePy supports OpenAI and Anthropic's LLM, integrates X platform API, allowing Agents to perform various social operations. Future plans include integrating a memory system, enabling Agents to remember previous interactions and contextual information.

2. The Replication of the BTC Ecosystem

The development path of AI Agents has similarities with the recent BTC ecosystem. The BTC ecosystem has gone through the BRC20, multi-protocol competition, BTC L2, and BTCFi stages. AI Agents, on the other hand, are developing faster based on a mature traditional AI technology stack, which can be summarized as: GOAT/ACT - Social type Agents/Analytical AI - Competition among Agent frameworks.

However, the AI Agent track is unlikely to be homogenized and bubble-like like the BTC ecosystem. AI framework projects provide new infrastructure development ideas, resembling future public chains, while Agents are similar to future Dapps. Future debates may shift from the EVM versus heterogeneous chains to framework disputes, focusing on how to achieve Decentralization or chainification, as well as the significance of implementation on the blockchain.

3. What is the significance of going on-chain?

The combination of blockchain and AI needs to address the question of its significance. Referring to the successful path of DeFi, the reasons supporting the chainization of agents may include:

1. Reduce usage costs, improve accessibility and choice, allowing ordinary users to participate in AI "rental rights".

2. Provide blockchain-based security solutions to meet the needs of Agent interacting with real or virtual wallets.

3. Implement unique blockchain financial strategies, such as agent-related computing power, data tagging investment, etc.

4. Achieve interoperability that is more attractive than traditional internet giants through transparent and traceable reasoning.

4. Creative Economy

Framework projects may provide entrepreneurial opportunities similar to the GPT Store in the future. A framework that simplifies the agent construction process and offers a combination of complex functionalities could have an advantage, creating a more interesting Web3 creative economy than the GPT Store.

Web3 may be more equitable than Web2 in terms of demand and economic systems, introducing community economics to enhance Agents. The creative economy of Agents will provide ordinary people with opportunities to participate, and future AI Memes may be smarter and more interesting than the existing ones.