Can AI Agents Become the Lifeline for Web3 + AI?

AI Agent projects are popular and mature types in Web2 entrepreneurship, mainly focused on enterprise service, while in the Web3 field, model training and platform aggregation projects have become mainstream due to their key role in building ecosystems.

Currently, there are not many AI Agent projects in Web3, accounting for only 8%, but their market capitalization in the AI sector is as high as 23%, demonstrating strong market competitiveness. We expect that with the maturation of technology and an increase in market recognition, several projects with valuations exceeding 1 billion dollars will emerge in the future.

For Web3 projects, introducing AI technology for non-AI core application products may become a strategic advantage. The integration of AI Agent projects should focus on constructing the entire ecosystem and designing token economic models to promote decentralization and network effects.

The AI Wave: Current Status of Project Emergence and Valuation Increase

Since the launch of ChatGPT in November 2022, it has attracted over 100 million users in just two months. By May 2024, ChatGPT's monthly revenue had reached an astonishing $20.3 million. After releasing ChatGPT, OpenAI quickly launched iterative versions such as GPT-4 and GP4-4o. This rapid development has made major traditional tech giants realize the importance of cutting-edge AI models like LLMs, prompting them to release their own AI models and applications. For example, Google released the large language model PaLM2, Meta introduced Llama3, and Chinese companies launched large models such as Wenxin Yiyan and Zhipu Qingyan. Clearly, the field of AI has become a battleground for competition.

The competition among major technology giants has not only driven the development of commercial applications, but our survey statistics from open-source AI research show that the 2024 AI Index report indicates the number of AI-related projects on GitHub surged from 845 in 2011 to approximately 1.8 million in 2023. Especially after the release of GPT in 2023, the number of projects increased by 59.3% year-on-year, reflecting the enthusiasm of the global developer community for AI research.

The enthusiasm for AI technology is directly reflected in the investment market, with the AI investment market showing strong growth and experiencing explosive growth in the second quarter of 2024. There were a total of 16 AI-related investments exceeding $150 million globally, which is twice as many as in the first quarter. The total financing for AI startups surged to $24 billion, more than doubling year-on-year. Among them, xAI, owned by Musk, raised $6 billion with a valuation of $24 billion, making it the second highest valued AI startup after OpenAI.

The rapid development of AI technology is reshaping the landscape of the tech industry at an unprecedented speed. From the fierce competition among tech giants to the flourishing development of open-source community projects, and the enthusiastic pursuit of AI concepts in the capital market. Projects are emerging one after another, investment amounts are reaching new highs, and valuations are rising accordingly. Overall, the AI market is in a golden period of rapid development, with significant progress made in language processing through large language models and retrieval-augmented generation technology. Nevertheless, these models still face challenges in converting technical advantages into actual products, such as the uncertainty in model outputs, the risk of generating inaccurate information (hallucinations), and issues regarding model transparency. These problems become particularly important in application scenarios that require high reliability.

Against this backdrop, we began researching AI Agents, as they emphasize the comprehensiveness of solving real-world problems and interacting with the environment. This shift marks the evolution of AI technology from purely language models to intelligent systems that can truly understand, learn, and solve real problems. Therefore, we see hope in the development of AI Agents, as they gradually bridge the gap between AI technology and real problem solving. The evolution of AI technology continuously reshapes the architecture of productivity, while Web3 technology reconstructs the production relationships of the digital economy. When the three key elements of AI: data, models, and computing power, merge with the core concepts of Web3 such as decentralization, token economies, and smart contracts, we foresee the emergence of a series of innovative applications. In this promising intersection, we believe that AI Agents, with their ability to autonomously execute tasks, demonstrate tremendous potential for large-scale applications.

To this end, we began to explore the diverse applications of AI Agents in Web3, from the infrastructure, middleware, and application layers of Web3 to various dimensions such as data and model markets, aiming to identify and assess the most promising types of projects and application scenarios in order to gain an in-depth understanding of the deep integration of AI and Web3.

Clarification of Concepts: Introduction and Classification Overview of AI Agents

Basic Introduction

Before introducing AI Agents, to help readers better understand the distinction between its definition and the model itself, let's illustrate with a practical scenario: suppose you are planning a trip. Traditional large language models provide destination information and travel recommendations. Retrieval-augmented generation technology can offer richer and more specific destination content. An AI Agent is like JARVIS from the Iron Man movies; it can understand your needs and actively search for flights and hotels based on a single sentence from you, execute booking operations, and add the itinerary to your calendar.

Currently, the industry's general definition of an AI Agent refers to an intelligent system that can perceive its environment and take corresponding actions. It gathers environmental information through sensors, processes it, and then influences the environment through actuators (Stuart Russell & Peter Norvig, 2020). We believe that an AI Agent is an assistant that integrates LLM, RAG, memory, task planning, and tool usage capabilities. It can provide not only information but also plan, decompose tasks, and truly execute them.

According to this definition and characteristics, we can find that AI Agents have long been integrated into our lives and are applied in different scenarios, such as AlphaGo, Siri, and Tesla's Level 5 and above autonomous driving, all of which can be seen as examples of AI Agents. The common trait of these systems is that they can perceive external user inputs and accordingly influence the real environment.

Taking ChatGPT as an example for conceptual clarification, we should clearly point out that Transformer is the technical architecture that constitutes AI models, and GPT is a series of models developed based on this architecture, with GPT-1, GPT-4, and GPT-4o representing different versions at various stages of model development. ChatGPT, on the other hand, is an AI Agent evolved from the GPT model.

Classification Overview

The current AI Agent market has not yet formed a unified classification standard. We labeled 204 AI Agent projects in the Web2 + Web3 markets according to their prominent tags, dividing them into primary and secondary classifications. The primary classifications are infrastructure, content generation, and user interaction, which are further subdivided based on their actual use cases:

Infrastructure: This type focuses on building foundational content in the Agent field, including platforms, models, data, development tools, as well as more mature B-end services for foundational applications.

• Development tools: Provide developers with auxiliary tools and frameworks for building AI Agents.

• Data processing category: processing and analyzing data in different formats, mainly used to assist decision-making and provide sources for training.

• Model Training Category: Provides model training services for AI, including inference, model establishment, setting, etc.

• B-end services: Mainly aimed at enterprise users, providing enterprise service, vertical, and automated solutions.

• Platform aggregation type: A platform that integrates various AI Agent services and tools.

Interactive types: Similar to content generation types, the difference lies in the continuous two-way interaction. Interactive agents not only accept and understand user needs but also provide feedback through technologies such as Natural Language Processing (NLP), achieving two-way interaction with users.

• Emotional companionship: AI agents that provide emotional support and companionship.

• GPT type: AI Agent based on the GPT (Generative Pre-trained Transformer) model.

• Search category: Focused on search functionality, providing an Agent primarily for more accurate information retrieval.

Content Generation: This type of project focuses on creating content, using large model technology to generate various forms of content based on user instructions, divided into four categories: text generation, image generation, video generation, and audio generation.

Analysis of the Current Development Status of Web2 AI Agents

According to our statistics, the development of AI Agents in the traditional Web2 internet shows a明显的板块集中趋势. Specifically, about two-thirds of the projects are concentrated in the infrastructure category, mainly focusing on B-end services and development tools. We have also conducted some analysis on this phenomenon.

Impact of Technology Maturity: The dominance of infrastructure projects is primarily attributed to their technology maturity. These projects are typically built on time-tested technologies and frameworks, thereby reducing development difficulty and risk. They are akin to the "shovels" in the AI field, providing a solid foundation for the development and application of AI Agents.

Driving Market Demand: Another key factor is market demand. Compared to the consumer market, the demand for AI technology in the enterprise market is more urgent, especially in seeking solutions to improve operational efficiency and reduce costs. At the same time, for developers, cash flow from enterprises is relatively stable, which is beneficial for them to develop subsequent projects.

Limitations of application scenarios: At the same time, we notice that the application scenarios of content generation AI in the B-end market are relatively limited. Due to its instability in output, enterprises prefer applications that can consistently enhance productivity. This has resulted in a smaller proportion of content generation AI in the project library.

This trend reflects the actual considerations of technological maturity, market demand, and application scenarios. With the continuous advancement of AI technology and the further clarification of market demand, we anticipate that this pattern may undergo some adjustments, but infrastructure will still be a solid foundation for the development of AI Agents.

Analysis of Leading Web2 AI Agent Projects

We delve into some current AI Agent projects in the Web2 market and analyze them, taking Character AI, Perplexity AI, and Midjourney as examples.

Character AI:

Product Introduction: Character.AI provides an AI-based dialogue system and virtual character creation tools. Its platform allows users to create, train, and interact with virtual characters that can engage in natural language conversations and perform specific tasks.

Data Analysis: Character.AI had 277 million visits in May, with the platform having over 3.5 million daily active users, most of whom are aged between 18 and 34, indicating a youthful user demographic. Character AI has performed exceptionally in the capital markets, completing a $150 million financing round, with a valuation reaching $1 billion, led by a16z.

Technical Analysis: Character AI has signed a non-exclusive licensing agreement with Google's parent company Alphabet to use its large language model, indicating that Character AI employs self-developed technology. It is worth mentioning that the company's founders, Noam Shazeer and Daniel De Freitas, were involved in the development of Google's conversational language model Llama.

Perplexity AI:

Product Introduction: Perplexity can crawl the internet and provide detailed answers. It ensures the reliability and accuracy of information by citing and referencing links, while also educating and guiding users to ask follow-up questions and search for keywords, meeting the diverse query needs of users.

Data Analysis: Perplexity's monthly active user count has reached 10 million, with a 8.6% increase in visits to its mobile and desktop applications in February, attracting approximately 50 million users. In the capital markets, Perplexity AI recently announced it has raised $62.7 million in funding, with a valuation of $1.04 billion, led by Daniel Gross, and participants including Stan Druckenmiller and NVIDIA.

Technical Analysis: The main models used by Perplexity are the fine-tuned GPT-3.5 and two large models fine-tuned based on open-source large models: pplx-7b-online and pplx-70b-online. The models are suitable for professional academic research and