From AI infrastructure to application scenarios, what Web3 projects deserve attention?

原文标题：《 从 AI 基础设施到应用场景，哪些 Web3 项目值得关注？两者如何结合？ 》

Biscuit & alertcat.eth, ChainCatcher

Openai-owned ChatGPT has reached 100 million monthly active users just two months after its launch, making it the fastest growing app in history. This ability to "fan" quickly spread the popularity of AI to the crypto space. On January 10, Bloomberg reported that Microsoft was considering investing $10 billion in ChatGPT developer OpenAI. All AI-concept cryptocurrencies exploded completely. FET, AGIX, etc., rose more than 200% in a month.

With the help of capital, can these two cutting-edge technologies get together? Artificial intelligence uses computers to solve problems by mimicking the thinking abilities of the human brain. OpenAI provides large amounts of training data to natural language processing (NLP) models to make them more powerful. In the crypto world built by blockchain technology, the daily mass of data on the chain can provide "fuel" for the AI engine, allowing the AIGC to feed back better strategies.

Also, as AI algorithms get smarter, it becomes harder to understand how they reach decisions and conclusions. Blockchain has immutable properties that allow us to access immutable records of the data and processes that AI uses in its decision-making process.

AI concept encryption project (source: Rootdata)

Compared to the massive attention and adoption of artificial intelligence in traditional fields such as Stability AI and ChatGPT, blockchain's greater imagination lies in the economic system that can transform AI models. When the FOMO fades, this article will explore the characteristics of encryption projects that incorporate AI technology. What kind of chemistry could AI combine with blockchain?

AI infrastructure

The common feature of AI infrastructure projects is the distribution and sale of traditional AI architecture (data, models and computing power). They typically use their original tokens as the medium of exchange. They often act as intermediaries between users and service providers, building a decentralized trading market. These are all tasks that need to be completed by traditional AI, such as NLP, AI voice and CV fields, which use DApp as an intermediary platform to carry out transactions. Essentially a decentralized marketplace that uses token pricing and exchange for traditional markets.

Openfabric AI 

Openfabric is a platform for building and connecting AI applications. Through the platform, collaboration between AI innovators, data providers, enterprises and infrastructure providers will facilitate the creation and use of new intelligent algorithms and services. The Openfabric ecosystem consists of four roles: the algorithm creator, the data provider, the infrastructure provider, and the service consumer, with the service consumer paying the other three services.

Algorithm creator: Uses expertise to create AI algorithms capable of solving complex business problems. Data provider: Ensure that large amounts of data are distributed to train AI algorithms. Infrastructure Provider: All the hardware that runs the AI platform. Service consumers: End users who require a particular business product or service.

Oraichain

Oraichain is an AI-powered blockchain prognostic machine and ecosystem. In addition to the data predictor, Oraichain aims to become a complete AI ecosystem in the blockchain space, serving as the base layer for creating smart contracts and Dapps. With AI as the cornerstone, Oraichain has developed many important innovative products and services, These include AI feed pricing, full on-chain VRF, Data Hub, AI Marketplace with more than 100 AI apis, AI-based NFT generation and NFT copyright protection, Royalty Protocol, AI powered revenue aggregator platform and Cosmwasm IDE.

Fetch.ai 

Fetch.ai is a blockchain platform based on artificial intelligence and machine learning that enables anyone to share or trade data. As an autonomous machine-to-machine ecosystem, any network of independent parties can become a network proxy for Fetch.ai, recording any protocols generated between agents on the Fetch.ai blockchain. Fets are the original token of the Fetch AI blockchain, the primary medium of exchange for payment transactions.

Source: Fetch.ai blog

SingularityNET 

SingularityNET is a decentralized AI platform and marketplace. Developers publish their services to the SingularityNET network for use by any user with Internet access. Developers can charge for their services using native AGIX tokens. Services can provide reasoning or modeling training across domains, such as images, video, voice, text, time series, bioartificial intelligence, and network analysis.

SingularityNET ecosystem

The SingularityNET ecosystem will provide AI services to the platform and create mass utilization of AGIX tokens. These SingularityNET spin-offs are being developed in a number of strategically selected vertical markets, including DeFi, robotics, Biotechnology and longevity, gaming and media, arts and entertainment (music), and enterprise AI.

Gensyn

The Gensyn protocol is a Layer1 network for deep learning computing by instantly rewarding supplier participants who invest computing time into the network and perform ML (machine learning) tasks. The agreement does not require administrative oversight or enforcement, but instead facilitates task allocation and payment programmatically through smart contracts. The fundamental challenge for this network is to validate the ML work that has been done. This is at the intersection of complexity theory, game theory, cryptography, and optimization. The Gensyn ecosystem consists of four roles: submitter, solver, verifier, and whistleblower.  

Submitters: Provide the tasks to be evaluated and pay for the completed unit of work. Solvers: Perform model training and generate proofs for the verifier to examine. Verifiers: The key to linking the non-deterministic training process to deterministic linear calculations, replicating a part of solver proof, and comparing the distance to the expected threshold. Whistleblowers: Inspect the verifier's work and challenge it in the hope of accumulating a bonus.

Gensyn's vision is to provide significant infrastructure components for Web3 applications by decentralizing ML computing to reduce Dapps' dependence on the Web2 infrastructure.

Application scenario

In such application scenarios, the project aims to deal with the emerging needs arising from the development of blockchain in recent years in an AI manner.

These requirements could be to allow users to skip the hassle, developers to quickly develop games, to socialize on the blockchain platform, to create virtual people with their own personalities, or to detect fake NFT projects. Different from traditional AI platforms, such projects have strong demand irrefungibility, which gives them a deep moat. At the same time, the difficulty in the development of platforms that take emerging demand as a selling point lies in customer acquisition. How to attract enough customers to prove that the demand of their platforms is sustainable and objective has become a major problem in the development of such platforms.

Chain travel direction

In the mainstream financial system of the crypto game "P2E" model, where the user is faced with constantly changing gameplay and a large number of repetitive basic actions, the AI can provide the player with a stable automated flow and develop a higher probability of winning game strategies. rct AI is an application of AI to provide complete solutions for the game industry. Its core technology, Chaos Box, is an AI engine based on deep reinforcement learning. rct AI has developed the DRL (Deep Reinforcement Learning) model of AI training for Axie Infinity. Due to the number of combinations of all cards in Axie Infinity is about 10^23, and the game features, etc., rct AI's model improves efficiency and win rates in a large amount of simulated battle data.

In addition, AI can provide action prototypes for developers. Mirror World is a game matrix virtual world based on Solana, Has released Mirrama, PVP arena game Brawl of Mirrors with Roguelike gameplay using AI technology. In addition, Mirror World has launched a series of NFT's that can interoperate in games. These NFT prototypes are completed using AI action algorithms.

Related: Conversation rct AI: It's Time to think about how blockchain is changing game Publishing

Social orientation

PLAI Labs, which is focused on using AI and web3 to build the next generation of social platforms for users to play, talk, fight, trade, and take risks together, received $32 million in funding from a16z in January 2023. Currently, PLAI Labs has shown two products to the outside world:

Champions Ascension, a massively multiplayer online role-playing game (MMORPG), gives players the option of owning their avatar in the form of an NFT and the ability to fight in a massive Colosseum arena, do quests, build and compete for and trade digital items in a custom dungeon. AI protocol platform that will help handle everything from user-generated content (UGC) to matching to 2D to 3D asset rendering.

PLAI Labs plans to launch V2 white paper this year, including details of the core economic cycle (leveraging NFT and blockchain to enhance the experience), UGC toolkit (including AI) plans...

Related: Plai Labs: Why We chose Web3

NFT direction

Aletha AI  Proposed the concept of iNFT, a technology that combines artificial intelligence and blockchain. With AI, NFT is interactive, generative, scalable, and unique.  

To put it simply, if NFT is a digital human work, it becomes iNFT with AI, and it has the ability to chat with users. On June 10, 2021, the world's first iNFT Alice was auctioned at Sotheby's for $478,800.

Altered State Machine (ASM)  Is an innovative project combining NFT, artificial intelligence and machine learning to provide training power for AI-powered NFT, with the vision of becoming an ownership and monetization agreement for AI using NFT technology. In ASM ecosystem, AI-based Avatars are called Agents, which are composed of brain and Avatar. The project also issued ASTO tokens to power the ASM ecosystem.

Related: Detailed Understanding of Altered State Machine: An Innovative Quest to Evolve NFT Using AI and Machine Learning

Optic  Is building an AI NFT verification protocol focused on NFT fraud analysis and NFT value discovery within the community, aiming to help the entire NFT market achieve greater authenticity and transparency. The Optic intelligent engine retrieves the NFT collections on the market by learning the real NFT series. Optic will then return a match score indicating how well the inspected NFT matches the true NFT.

Optic completed an $11 million financing round led by Pantera Capital and Kleiner Perkins in July 2022, with participation from Circle Ventures, Polygon Ventures and others. At present, OpenSea has adopted Copymint detection service of Optic.

Optic Analysis: Artificial Intelligence NFT Validation Protocol

Trend analysis

From the current development path of blockchain AI projects, the infrastructure of AI is composed of three parts: data, algorithms and computing power. A normal AI project that wants to achieve AI generation or analysis capabilities needs models and data sets and the software ontology and GUI that invoke the model. The distribution of models and data sets, the training of models (computing power leasing), the formation of intermediaries in the development of software front ends in this field will lead to blockchain AI projects aimed at efficiently meeting customer needs.

For example, Fetch.ai acts as an intermediary that allows a client to use its native token transaction data set. SingularityNET allows customers to purchase computing training services from developers, while Openfabric AI customers need models (algorithms), data sets, infrastructure (software) and other services from providers. Humans.ai is essentially a trained AI model that encapsulates data sets in NFT, purchased by users with native tokens,

Gensyn is essentially a decentralized computing power rental platform. These are all tasks that need to be completed by traditional AI, such as natural language processing, AI speech, image generation fields using DApp as an intermediary platform for trading projects.

Then, decentralized application in blockchain has created new demands.ai projects based on the direction of blockchain, social direction and NFT direction aim to solve the pain points of users in blockchain. For example, rct.ai solves the problem of manual and repetitive operation of users in blockchain, Mirror World solves the development of blockchain. Other projects are focused on blockchain social and NFT.

Right now, in the early stages of Web3 social, AI is being introduced more as a narrative device. In the future, there are some possible directions for AI project research and development:

Enhance data privacy: Web3 can maximize data privacy by using zk technology, and AI can analyze data without compromising privacy.

Smart contract: Web3 technology can integrate AI application into Web3 application through smart contract, so as to realize the controllability of AI model. Such applications can be used in the trading of models and data sets to automate the trading process. And use ZK technology to protect the user's data. However, projects of this type face the impact of open source data sets and open source models. Imagine: if users get the open source data and models on the Hugging face and train them with auto train, why do they trade on the blockchain platform? Faced with the onslaught of Web2 companies, Web3 AI models and data set transactions do not have a sufficient moat.

More Efficient machine learning: Web3 technology can increase the efficiency of machine learning in a decentralized way, thus making AI applications faster and more reliable. This has already been used in traditional AI training, such as KataGo, an improved version of AlphaGo, which uses distributed training techniques to get people around the world to volunteer for computing training if they want to update the AI. Applications in blockchain can be similar to Gitcoin, where donated computing power can obtain POAP, or similar to AMM, which provides incentives for liquidity and becomes a platform for paid computing power rental. However, due to the high volatility of the currency price, such applications do not have advantages compared with traditional GPU computing power rental. Unless the platform itself is in the financial business, enough to subsidize users from the value captured by the agreement, such as Numerai, to use AI technology to profit from the stock market, enough users will be willing to provide the three elements of AI into the platform.

summarize

At present, both the original AI infrastructure of blockchain and the encryption project to realize application scenarios with the help of AI engine are in the nascent stage. The main goal is to build an applicable underlying infrastructure to integrate token economics with artificial intelligence solutions such as hardware providers, data providers and AI algorithms.

However, the integration of the two faces many challenges. First of all, the trend towards complex technologies such as Rollup and ZK on blockchain will create challenges for AI to access data. Second, not enough continuous experimental data has been conducted to support the applicability of AI in the blockchain ecosystem and the ability of the AI engine to adjust to unexpected events. Finally, the frequency of fake projects in the crypto space that piggy-back on the concept of AI makes it easy to lose faith in the field of exploration.

All blockchain AI projects that solve traditional AI problems need to answer the question: Why does the platform need to introduce tokens on the blockchain? This gives an onboarding disadvantage to platforms that use existing metrics in the Web2 market, such as models, data, and computing power.

Token economics, like a flywheel, can change the boom and bust cycle of a project. At present, if you want to forward flywheel, you need to take into account the actual users of the platform, namely the acquisition problem. The irreplaceability of requirements is the moat of a project; without a moat, a project can achieve short-term success, but not have enough users and a thriving developer ecosystem. When requirements are false, economic incentives are unsustainable and project life cycles are shorter. We expect more AI+Web3 projects based on real users and irreplaceable needs. They are designed to fulfill requirements that are not or are not fulfilled well in web2 and thus are inherently required to introduce Web3.

In any case, the integration of AI into Web3 is a future technology trend, and there are already some examples of Web3 applications combined with artificial intelligence. Over time, more relevant Web3 infrastructures and new patterns will emerge.

Original link

Welcome to join the official BlockBeats community:

Telegram Subscription Group: https://t.me/theblockbeats

Telegram Discussion Group: https://t.me/BlockBeats_App

Official Twitter Account: https://twitter.com/BlockBeatsAsia