When Jen-Hsun Huang spoke at WGS in Dubai, he proposed the term “sovereign AI.” So, which sovereign AI can meet the interests and demands of the Crypto community? Maybe it needs to be built in the form of Web3+AI. Vitalik described the synergy between AI and Crypto in the article “The promise and challenges of crypto + AI applications”: Crypto’s decentralization can balance the centralization of AI; AI is opaque, and Crypto brings transparency; AI needs data, and Blockchain facilitates the storage and tracking of data. This kind of synergy runs through the entire industrial landscape of Web3+AI.

Most Web3 + AI projects are using blockchain technology to solve the construction problems of infrastructure projects in the AI ​​industry, and a few projects are using AI to solve certain problems in Web3 applications. The landscape of Web3 + AI industry can be described roughly as follows:

The production and workflow of AI is roughly as follows:

In these links, the combination of Web3 and AI is mainly reflected in four aspects:：

1. Computing Power Layer: Tokenization of Computing Power

In the past two years, the computing power used to train large AI models has increased exponentially, doubling basically every quarter, and growing at a rate that far exceeds Moore’s Law. This situation has led to a long-term imbalance in the supply and demand of AI computing power, and the prices of hardware such as GPUs have risen rapidly, thus raising the cost of computing power. But at the same time, there is also a large amount of idle mid- to low-end computing power hardware on the market. It may be that the single computing power of this part of mid-to-low-end hardware cannot meet high-performance needs.

However, if a distributed computing power network is built through Web3 and a decentralized computing resource network is created through computing power leasing and sharing, it can still meet the needs of many AI applications. Because it uses distributed idle computing power, the cost of AI computing power can be significantly reduced. Computing power layer breakdown includes:

• General decentralized computing power (such as Arkash, Io.net, etc.);
• Decentralized computing power for AI training (such as Gensyn, Flock.io, etc.);
• Decentralized computing power for AI reasoning (such as Fetch.ai, Hyperbolic, etc.);
• Decentralized computing power for 3D rendering (such as The Render Network, etc.).

The core advantage of Web3+AI’s computing power assetization lies in decentralized computing power projects. Combined with token incentives, it is easy to expand the network scale, and its computing resource cost is low and cost-effective, which can satisfy the needs of some mid-to-low-end computing power.

2. Data Layer: Data Capitalization

Data is the oil and blood of AI. Without relying on Web3, only giant corporations typically have access to vast amounts of user data, making it difficult for smaller startups to acquire extensive data. Moreover, the value of user data in the AI industry often does not trickle back to the users themselves. Through Web3 + AI, data collection, annotation, and distributed storage processes can be made more cost-effective, transparent, and beneficial to users. Gathering high-quality data is a prerequisite for training AI models. With Web3, a distributed network can be leveraged along with appropriate token incentive mechanisms to crowdsource data collection at a lower cost while obtaining high-quality and widespread data. Depending on the project’s purpose, data-related projects mainly fall into the following categories:

• Data collection projects (such as Grass, etc.);
• Data trading projects (such as Ocean Protocol, etc.);
• Data annotation projects (such as Taida, Alaya, etc.);
• Blockchain data source projects (such as Spice AI, Space and time, etc.);
• Decentralized storage projects (such as Filecoin, Arweave, etc.).

Data-based Web3+AI projects are more challenging in the process of designing the Token economic model, because data is more difficult to standardize than computing power.

3. Platform Layer: Tokenization of Platform Value Assets

Most platform projects tend to benchmark against Hugging Face, with the integration of various AI industry resources as their core. Establishing a platform that aggregates links between data, computing power, models, AI developers, blockchain, and other resources and roles, with the platform at the center, facilitates the resolution of various needs more conveniently. For example, Giza focuses on building a comprehensive zkML operation platform, aiming to make the inference of machine learning trustworthy and transparent. The opacity of data and models is a widespread issue in AI currently, and it is only a matter of time before the industry calls for verification of model inference through Web3 using cryptographic technologies such as ZK and FHE to ensure correct execution. There are also layers like Focus AI, such as Nuroblocks and Janction, which connect various computing power, data, models, AI developers, and node resources. By packaging universal components and SDKs, they help Web3 + AI applications achieve rapid development. There are also platform types like Agent Network, which can build AI Agents for various application scenarios, such as Olas and ChainML. Platform-based Web3 + AI projects primarily capture platform value through tokens, incentivizing all participants in the platform’s construction. This approach is particularly helpful for startups to grow from 0 to 1, reducing the difficulty of finding partners such as computing power, data, AI developer communities, and nodes.

4. Application Layer: Tokenization of AI Value Assets

The preceding infrastructure projects mostly focus on using blockchain technology to address the construction of infrastructure projects in the AI industry. Application layer projects, on the other hand, primarily utilize AI to solve problems existing in Web3 applications. For instance, Vitalik mentions two directions in the article that I find meaningful. Firstly, AI as a participant in Web3. For example, in Web3 Games, AI can act as a player, quickly understanding game rules and efficiently completing game tasks. In DEX, AI has been involved in arbitrage trading for many years. In prediction markets, AI agents can analyze predictive capabilities by widely accepting vast amounts of data, knowledge bases, and information. Then, they are productized and offered to users. This helps users make predictions about specific events, such as sports matches or presidential elections, through model inference. Secondly, creating scalable decentralized private AI. Many users are concerned about the black box problem and potential biases in AI systems, or fear that certain dApps may exploit AI technology to deceive users for profit. Essentially, this stems from users lacking oversight and governance authority over AI model training and inference processes. However, creating a Web3 AI where the community has distributed governance rights over the AI, similar to Web3 projects, may be more readily accepted. As of now, there haven’t been any standout projects in the Web3 + AI application layer that is hard to transcend.

Summary

Web3 + AI is still in its early stages, and the industry is divided on the development prospects of this field. We will continue to pay attention to this field. We hope that the combination of Web3 and AI can create products that are more valuable than centralized AI, allowing AI to get rid of the labels of “giant control” and “monopoly” and “co-govern AI” in a more community-based way. Perhaps in the process of closer participation and governance, humans will be more “awe” and less “fearful” of AI.

Statement:

1. This article originally titled “Web3 + AI：社区主权的人工智能” is reproduced from [IOBC Capital]. All copyrights belong to the original author [0xCousin]. If you have any objection to the reprint, please contact the Gate Learn team, the team will handle it as soon as possible.

2. Disclaimer: The views and opinions expressed in this article represent only the author’s personal views and do not constitute any investment advice.

3. Translations of the article into other languages are done by the Gate Learn team. Unless mentioned, copying, distributing, or plagiarizing the translated articles is prohibited.