Can AI+, which hitchhikes with a ride, lead a new trend in the bull market narrative?

[TL；DR]

1. The AI+Web3 track can be roughly divided into three layers: the infrastructure layer, the middle layer, and the application layer. Among them, the infrastructure layer focuses on providing computing power and storage, which is currently the most popular and popular field.

2. In addition to application-layer cases in games, social networking, trading, etc., AI can also be used in data analysis, information monitoring and tracking, bidding and betting, and other fields.

3. Projects that are closely related to the concept of AI can often quickly gain favor in the market, but care should be taken to filter out projects that are not worthy of the name.

Introduction

In order to explore this potential market opportunity in depth, Gate.io Research will combine various hot projects and conduct in-depth analysis from all aspects of the AI+Web3 industry chain, in order to provide readers with a comprehensive and in-depth understanding.

AI+Web3: New technology, new direction, new hype

In the past year, with the emergence of generative AI large-scale models such as ChatGPT, AI has become a hot investment theme in the world’s capital markets. At the same time, the Web3 market has ushered in a new round of prosperity.

The organic combination of AI and Web3 has undoubtedly become the intersection of two hot topics in the current technology field. Recently, we have seen a large number of new and existing projects around this theme gaining market attention, which highlights the strong interest and high expectations of investors for this portfolio.

According to the definition of Vientiane blockchain, the combination of AI + Web3 is mainly reflected in two aspects: how Web3 promotes the development of AI, and how Web3 applications combine AI technology. Currently, most projects tend to leverage Web3 technologies and concepts to advance AI. To analyze this combination, we can start with the entire process of AI from model training to application.

The AI production process generally includes data acquisition to provide the foundation for model training, data preprocessing and feature/cue engineering involving data cleaning, annotation, and structured querying, model training and tuning to improve model performance through iteration, model review and governance to ensure model quality and transparency, model inference to make predictions on new data, and model deployment and monitoring to ensure that the model remains in the best condition in real-world applications.

In this process, Web3 has many points of convergence. For example, Web3’s distributed networks and incentives enable the construction of more open, open-source AI networks and communities to meet the needs of AI applications for low-cost, open infrastructure and data networks. At the same time, Web3, combined with cryptography technologies such as ZK, can improve the trust problem of AI and solve challenges such as model transparency, bias, and ethical applications.

Figure 1 Source: Vientiane Blockchain

As shown in the figure above, the AI+Web3 track can be roughly divided into three layers: the infrastructure layer, the middle layer, and the application layer.

The infrastructure layer focuses on providing computing power and storage, and the addition of Web3 can reduce its costs and serve more AI applications.

In the middle layer, Web3 technology is used to optimize AI production processes, such as data acquisition, preprocessing, and model validation, and many innovative projects have emerged.

The application layer showcases the wide range of applications of AI in Web3, such as content generation, analytics, and prediction. According to the author’s observation, the description of the above figure in the application layer is still conservative, and I will discuss it in detail in the afternoon. Although there is no head project yet, the potential is huge, and the future competition will focus on products and technical capabilities.

We will make a specific case analysis of these three layers of projects in the following sections.

AI+Web3 hot projects emerge one after another

AI+DePIN

The entire AI workflow is supported by compute and storage infrastructure, which not only provides the computing power needed for model training and prediction, but also stores, manages, and parses data throughout the data model and lifecycle.

Currently, the rapid growth of AI applications has led to a huge demand for infrastructure, especially high-performance computing power. As a result, the development of more efficient, lower-cost, and more resource-rich computing and storage infrastructure has become a key trend in the early days of AI development, and it is currently the most popular area.

Figure 2 Source: Render Network

In this field, a number of representative projects have emerged, such as Render Network, which was born in the last round of the bull market, which mainly provides rendering services, Akash, which focuses on cloud computing, Filecoin and Arweave, which focus on cloud storage, and IO.NET and Aethir, which are newly launched in this round of bull market, which mainly provide computing power support for AI.

AI+Data

The middle tier is a critical part of the AI production process that leverages Web3 technologies to optimize and improve specific workflows.

First of all, in the data acquisition stage, the middle layer introduces decentralized data identity management, which can not only protect the user’s data security, but also ensure that the ownership of the data is clear. At the same time, through the incentive mechanism, users can also be encouraged to share high-quality data for monetization, thereby expanding the source of data.

Due to the limitations of the development stage of the industry, there were almost no relatively well-known projects in this field in the last round of bull and bear markets. This round of bull market has seen the AI identity project Worldcoin (we have written about the project many times), the Gate.io investment Aspecta, as well as the data trading platform Ocean Protocol, and the data network Grass for broadband mining.

Figure 3 Source: Aspecta

Second, in the data preprocessing stage, the middle layer is committed to building a distributed AI data labeling and processing platform to provide strong support for subsequent model training. In this regard, projects such as Public AI have already achieved remarkable results.

Finally, in the model validation and inference stage, the middle layer makes full use of the combination of Web3 technology and cryptography technology, such as ZK and homomorphic encryption, to verify whether the inference process of the model uses the correct data and parameters. This not only ensures the accuracy of the model, but also protects the privacy of the input data. Typical use cases are ZKML, such as bittensor, Privasea, Modulus, and Privasea, which is invested by Gate Labs.

AI+intent-centric

Intent-centric, which translates to “intention-centric”, refers to “what you want to do”, and it focuses on the outcome, not the process. Intent-centric aims to make cumbersome on-chain operations “one-step” through protocol and infrastructure optimization. More precisely, by hiding the complex operation process in the past, the user can realize the purpose without feeling and directly, which reflects the connotation of chain abstraction.

Common intent scenarios using AI today include cross-chain, airdrops, governance, large transactions, and bulk operations, and the Telegram bots we discussed in our previous article also fall into this category.

For example, Delysium (AGI), which is committed to building an AI agent network centered on user intent for Web3 using AI, has gained a lot of attention in markets such as South Korea.

As shown in the chart, the project’s token has risen amazingly recently due to market speculation and value discovery.

Figure 4 Source: Gate.io

Delysium has launched an AI Agent called Lucy. As an AI-powered Web3 operating system, Lucy is able to intelligently plan and automate workflows that address user needs based on understanding the intents and goals contained in natural language, simplifying the complex operational processes of current Web3 applications and protocols.

AI+Game

AI+Game also has a very high imagination space. AI technology not only accelerates the game production process, but also runs through all aspects of game production, from mining user habits to customizing personalized interaction scenarios, showing great potential. Nowadays, major game manufacturers are actively embracing AI and reconstructing the ecosystem of the game industry chain.

When it comes to game production, AI powers art, planning, and operations. Whether it’s creative inspiration, level generation, copywriting, and operational analytics, AI is accelerating the production of game content. In terms of game experience, the capabilities of natural language generation and image generation brought by AI make the gameplay more innovative and diverse, and the interaction of NPCs more intelligent and vivid.

For example, Honor of Kings’ Enlightenment AI has been massively used in level evaluation and testing, in Mount & Blade II: Bannerlord, ChatGPT has enhanced the game’s interactivity by enabling NPCs to dynamically reply to players, and in Naraka: Bladepoint, players can even use AI painting to generate fashion models and vote for the most popular titles, demonstrating the potential of AI for game innovation.

Figure 5 Source: sleeplessAI

In addition to traditional Web2 games embracing AI, Web3 games are not far behind. For example, Ultiverse provides users with AI in-depth feature analysis and customized multiple experiences such as social, gaming, and metaverse through a powerful AI engine, as well as sleeplessAI’s AI-focused virtual companion game.

AI+Analysis

In addition to application-layer cases in gaming, social networking, trading, etc., AI can also be used in data analysis, information monitoring and tracking, bidding and betting, and other fields, with representative projects such as Kaito and Dune already emerging and setting a benchmark for the industry.

We also often quote Dune’s data graphs in our blog posts, so I don’t need to go into them here.

Summary

In the past year, the integration of Web3 and AI has not only led a new trend in technology, but also spawned a new consensus in the industry: blockchain has changed production relations, and AI has changed productivity. This philosophy is now deeply rooted in the hearts of the people and has become a strong driving force for the development of the industry.

As game vendors, DeFi protocols, and other Web3 infrastructure projects have increased their investment in AI, the combination of AI and Web3 is becoming an important direction for industry innovation. In fact, projects that are closely tied to the concept of AI tend to quickly gain traction in the market, and we’ve already noticed this phenomenal growth.

However, under the superficial prosperity and hype, we cannot ignore the practical obstacles of the AI+Web3 industry. Practitioners, in particular, need to delve into their practical application scenarios and evaluate their ability to create value and construct industry narratives. In the long run, how the ecological pattern of the AI+Web3 industry will be formed, which fields will show great development potential, and whether there will be ethical and moral dilemmas need to be explored and answered in practice.

Therefore, in the face of the wave of AI+Web3, we must not only see the opportunities it brings, but also keep a clear mind and look at its challenges and shortcomings rationally. Only in this way can we better grasp the development of the AI+Web3 industry, promote its healthy and sustainable development, and seize the profit opportunities brought by the trend.

Author: Carl Y.

**This article represents the views of the author only and does not constitute any trading advice. **

**The content of this article is original, the copyright is owned by Gate.io, if you need to reprint, please indicate the author and source, otherwise you will be held legally responsible. **