OP Crypto Research Report: Unlimited reverie of the possibility of combining AI and Web3
Original Author: michaeljin&Yetta
Original source: OP Crypto
As a Web3 practitioner who has been swept by the AI wave, after experiencing the information explosion in the two industries in recent months, I have sorted out some thoughts and research to share with Web3 practitioners:
AI and Web3, one has broken through our imagination of the upper limit of productivity, and the other has reshaped our understanding of the economic model. As a cutting-edge technology that represents the future development direction, the combination of the two seems to be a natural fit, and can always inspire infinite There is room for imagination, but when we turn our attention to reality, we find that there are very few projects that truly combine the two. The collision of the two tracks gave birth to a new narrative, but it also gave birth to a lot of bubbles and gimmicks, and many beautiful visions that complement each other in theory may not have real needs in reality; and those that can meet the actual needs Projects will also be difficult to implement due to cost or technical bottlenecks.
I think that the idea of Web3 and AI ebb and flow is also proportional to the number of web3 projects that see AI content in the primary market and the number of AI projects that encounter unnecessary web3ization. Entrepreneurs/project parties of AI native don’t actually think about how to transform into web3, such as data rights confirmation, economic model, distribution of production relations, etc., because bottom-up projects in the AI large model have high demand for resources and require A large number of resources make AI very centralized from training to operation, and I am very cautious about the actual implementation feasibility of some Web3 project parties who so-called help AI improve production relations.
The Web3 market has encountered considerable bottlenecks at both the macro policy level and the innovation level. Leaving aside new regulatory pressures, from the innovation level, when AI rapidly improves productivity and replaces human thinking ability, it attracts the vast majority of users, Builders and In the eyes of investors, Web3's industrial innovation dilemma is even more unconcealed. Web3 has not had AI-level innovation for a long time. To be honest, most of the new projects that are getting some attention right now are minor changes to past technologies/products. For example, better pledge methods, multi-chain wallets with better user experience, meme coins with new gameplay, Dex with better liquidity on the new public chain, etc. These so-called "innovations" are helpful for introducing more users or areas. Does blockchain usage penetration really help, and is it what the industry really needs.
image description
first level title
TL;DR
There is a conflict between AI and web3 in the underlying logic. The large amount of resources required by the AI large model makes AI very centralized from training to operation, while the prospect of Web3 based on blockchain is given priority to decentralization and transparency. This makes AI and web3 The combination of web3 at the bottom layer is very difficult. Whether its business logic is established and whether the actual demand exists needs to be deliberated.
But it is this contradictory logic at the bottom that makes AI and Web3 complement each other, not seeking to be the core of each other’s narrative but a solution to each other’s pain points, boosting their respective development. The two technologies will also bring each other There are many new narratives, leaving a huge room for imagination. The economic model design of web3 can allow many AI project parties to increase the utilization rate of funds to boost the project's new activities, and the benefits of the blockchain itself, such as reducing infrastructure costs and verifying identities , injecting democracy and transparency into the data black box in AI and providing incentives for data contribution can provide new ideas for product design of AI project teams.
At the infrastructure layer, the decentralized mechanism of Web3 can solve the risks and problems of current AI from the bottom, such as privacy protection, data abuse, etc.
Provides a decentralized market for the essential elements of AI development, such as computing power and data, maximizes the use of idle resources, optimizes resource utilization and allocation, and promotes the development and application of AI large models.
The decentralization mechanism of Web3 allows AI to become more democratic from the bottom level. Through decentralized deployment, training and use of AI, users' data privacy can be better protected, and there is also an opportunity to share data get payback.
Blockchain can also be used to record and monitor the behavior of AI, thereby improving the security of AI and promoting the use of automated AI agents in various scenarios.
AI at the application layer can help the development and popularization of Web3 applications.
First, as a productivity tool, AI can help Web3 applications to greatly increase the development speed, and as a knowledge engine, it can reduce the interaction and learning costs between users and dApps, and help more users enter Web3.
first level title
secondary title
Token Incentive and Governance Mechanism: Decentralized Market Empowers AI Infrastructure
In the era of AI large models, all aspects of the infrastructure supporting the development of AI will become particularly important.
In the process of building and developing AI infrastructure, a key challenge is how to effectively motivate and coordinate participants to jointly promote the development and operation of the system. Decentralized markets and token incentives provide a novel and powerful way to solve this problem. In such a market, tokens play an important role as a digital asset and value medium. Tokens can represent certain rights, functions or resources, and their transactions and transfers are carried out through smart contracts, realizing a safe, transparent and automated transaction process.
Token incentives can play multiple roles for AI infrastructure. First, tokens can serve as an incentive to reward and encourage those who contribute to the AI infrastructure. These contributions can include providing computing resources, datasets, algorithm models, computing power, and more. For example, the recently popular AI voice chatbot creation platform MyShell has realized the data flywheel effect through the chatbot creation workshop and data analysis. Users can customize and interact with the chatbot's voice, functionality, and knowledge base on the Myshell platform. The data collected from these interactions are used to improve the performance of robots and personalized services, attract more users to use the platform, further increase data and value, and form a virtuous circle of growth.
secondary title
Homomorphic Encryption and Federated Learning: Integrating Privacy Protection into AI's Underlying Training
Efficient model training while ensuring personal privacy and data security has long been a challenge. In this regard, homomorphic encryption technology provides a powerful privacy protection method, which can be integrated into the underlying training of AI to ensure the security of sensitive data.
Homomorphic encryption is a special encryption technique that allows computations to be performed on data in an encrypted state without decryption. This means that model training and calculations can be performed on encrypted data without exposing the content of the original data. By applying homomorphic encryption to the underlying training process of AI, privacy protection can be achieved without leaking sensitive data.
Here are some key steps and considerations when using homomorphic encryption for AI training:
Data encryption: Encrypt the data participating in AI training using a homomorphic encryption algorithm. This ensures data privacy and confidentiality during training.
Encrypted computing: Perform computing operations in an encrypted state, including model training, optimization, and inference. Homomorphic encryption makes these calculations possible without decrypting the data.
Security parameter sharing: All parties involved in the training need to share and exchange security parameters required for encryption calculations. These parameters are used to control the homomorphic encryption process and decryption results.
Encrypted result processing: After the encrypted calculation is completed, the result can be decrypted to obtain the final model weight or predicted output. Appropriate security measures need to be taken when decrypting the results to prevent data leakage or unauthorized access.
secondary title
zkML and AI inference on the chain: AI agent behavior monitoring and rights and responsibilities constraints
In the context of the rapid development and widespread adoption of artificial intelligence (AI) technologies, it has become even more important to ensure that AI systems behave ethically and legally. AI systems are often viewed as agent entities capable of performing tasks and making decisions that can have profound consequences for humans and society. Therefore, monitoring the behavior of AI agents and constraining their powers and responsibilities has become a key issue in protecting public interests and personal rights. As an innovative method, zkML provides a safe, verifiable and transparent solution for monitoring AI agent behavior and constraining rights and responsibilities. By combining zero-knowledge proofs and blockchain technology, zkML ensures the compliance and credibility of AI systems while protecting privacy.
Take Modulus Labs as an example. The project uses zkML technology to ensure that key data or sensitive information will not be leaked during the operation of AI systems. By applying zero-knowledge proofs during computation, the project can prove to regulators or stakeholders that its AI performed a specific task without revealing actual data or internal models. This approach protects personal privacy and commercial confidentiality, while providing a means of auditing and verifying the behavior of AI agents. A decentralized monitoring and constraint framework established by zkML can monitor and review the decision-making process and behavior path of AI agents in real time.
first level title
secondary title
Improve production efficiency, an accelerator for Web3 development
In the development of Web3, artificial intelligence (AI) plays an important role, combining with various fields to improve productivity and create better user experience. Here are a few key areas where AI meets Web3:
AI and on-chain data collection and analysis
AI technology plays an important role in on-chain data collection and analysis. As a distributed database, the blockchain records a large number of transactions and information. By leveraging AI technology, data on the blockchain can be better understood and utilized. For example, Web3 Analytics is an AI-based analytics platform that utilizes machine learning and data mining algorithms to collect, process and analyze on-chain data. It can help users gain insights into on-chain transactions, market trends, and user behavior patterns, thereby providing users with more accurate data analysis and decision support. A similar platform is MinMax AI, which provides AI-based on-chain data analysis tools to help users discover potential market opportunities and trends.
AI and automated dApp development
The application of AI technology in automating the dApp development process is also very important. Smart contract and dApp development usually requires writing a lot of code, and tedious testing and deployment work. By combining AI with smart contracts and dApp development tools, a more efficient and smarter dApp development process can be achieved. AI can help automate code generation, verification and testing of smart contracts, and deployment and maintenance of dApps. This saves time and resources and increases the efficiency and accuracy of the development process. For example, some AI-assisted development tools use natural language processing and machine learning techniques to help developers write smart contracts faster and automatically detect and fix potential errors.
AI and on-chain transaction security
secondary title
Optimizing resource allocation, a navigator for the Web3 world
Optimizing resource allocation is a key challenge in a Web3 world. With the combination of blockchain technology and artificial intelligence, we can use AI as a navigator to achieve more efficient resource allocation and utilization. Here are a few areas where AI can navigate the Web3 world:
Optimization of AI and on-chain activities: Activities on the blockchain include transactions, contract execution, and data storage. Through the intelligent analysis and prediction capabilities of AI, we can better optimize on-chain activities and improve overall efficiency and performance. AI can help identify transaction patterns, detect unusual activity, and provide real-time recommendations to optimize resource allocation for blockchain networks through data analysis and model training.
AI and on-chain advertising mechanism: In the Web3 world, advertising is also a type of resource. AI can play a key role in the on-chain advertising mechanism, helping advertisers more accurately target audiences and provide personalized advertising content. By analyzing the data and behavior patterns of users on the chain, AI can achieve more accurate advertisement placement, improve the click-through rate and conversion rate of advertisements, thereby optimizing the allocation and utilization of resources.
secondary title
application layer
Lower entry barriers, a booster for Web3 popularization
AI-embedded friendly user interface
For example, the Web3 audit platform Fuzzland uses AI to help code auditors check code vulnerabilities and provide natural language explanations to assist audit expertise. Fuzzland also leverages AI to provide natural language explanations of formal specification and contract code, as well as some sample code to help developers understand potential problems in the code. By combining AI technology with audit expertise, Fuzzland makes it easier for developers in the Web3 industry to understand and explain code, improving audit efficiency and accuracy.
Interpretation of smart contracts embedded with AI
AI-embedded smart contract writing
secondary title
Rich plot gameplay, the creative library of the Web3 world
AI and Generative NFTs
AI Automated Trading Agent
Character AI and Game NPCs
AI and automatic rendering of metaverse scenes
The rise of generative AI has brought new possibilities to the creative industry, bringing more diverse and innovative experiences to the Web3 world, allowing users to participate in rich plots and gameplay. In the past NFT bull market, AI has injected unlimited creativity into generative NFT. Generative NFT (Non-Fungible Token) is a kind of artwork or digital asset based on algorithms and data. Various unique and diverse artworks and characters can be generated through AI technology. These generative NFTs can become characters, props or scene elements in games, virtual worlds or metaverses, providing users with rich choices and personalized experiences. In the upsurge of DeFi, AI automatic trading agent also brings convenience and efficiency to the economic transaction process in the creative library. In the Web3 world, users can earn benefits by owning, trading or participating in digital assets in the creative library. AI automatic trading agents use intelligent algorithms and machine learning technology to automate asset transactions, helping users obtain the best trading opportunities and maximize returns. AIGC also brings new gameplay and ideas to content platforms and UGC communities. For example, Yodayo is an AI art platform for virtual anchors and anime fans to share and create more content they love. By connecting to the AIGC engine, Yodayo makes the creation and interaction of users on the content creation platform easier and easier to operate, so that most users who are usually "silent" on traditional platforms can also become creators and up masters. Consumers become content creators, connecting more closely with and contributing to the community.
epilogue
epilogue
As Web3 practitioners swept by the AI wave, after experiencing the information explosion in the two industries in recent months, we have had a deeper thought on the combination of AI and Web3. Although there is a conflict in the underlying logic between the two, the centralization of AI and the principle of decentralization of Web3 seem difficult to reconcile, but it is this contradictory logic that enables AI and Web3 to complement each other and become solutions to each other's pain points. promote each other's development. The decentralized mechanism of Web3 can fundamentally solve the problems of privacy protection and data abuse faced by AI, and the application of Web3 and blockchain technology can also monitor and record the behavior of AI, improve the security of AI, and promote the automation of AI The promotion and application of agents in various fields.
Although the combination of AI and Web3 at the bottom layer is difficult, it can create many new possibilities and narratives at the application level: AI can become an important boost for Web3 applications, greatly improving the development speed of Web3 applications and reducing the interaction between users and dApps. Interaction and learning costs help more users enter the Web3 world. At the same time, while AI lowers the technical threshold for dApp development and project distribution, it can also bring more gameplay and enhance competitiveness to projects in terms of innovation and operation, such as embedding virtual humans and character AI in games and social ecology, etc. Novel elements will bring a new narrative and experience to Web3 applications, and further promote the development and promotion of the Web3 industry.
Although the combination of AI and Web3 faces some challenges and limitations, we believe that only the organic combination of the two can support the narrative and ideal of the next generation Internet. We look forward to seeing the emergence of more innovative projects that can bring AI into Web3 and push Web3 to a wider field. We also hope that the development of these two cutting-edge technologies can continue to help each other break through technical bottlenecks, overcome cost constraints, and jointly Create a smarter and more open future.
Reference:
Foresight Ventures: A Rational View on Decentralized Computing Power Networks
