AgentFi: AI x Crypto Future Driven by PMF

AI Agents x Crypto has sparked a wave of excitement over the past two months. Memecoins, interactive agents, and the openness of social media to bot accounts have combined to drive a wave of agent-driven agent hype, generating significant buzz on Twitter and Farcaster. This proves the PMF of AI Agents x Crypto. The market cap of Agents-related assets has reached $10b.

Since the birth of goat in October, Agent has given birth to countless new projects and assets driven by the market. Combined with the outlook for the future, this article roughly outlines the following framework:

 Source: IOSG Ventures 

1. Sentient Memecoins

And cult memecoin, which has risen rapidly under Murad's orders, is an asset with community and communication as its core narrative. Memecoin, represented by Agent, has added Sentient factors, creating an advantage in content. Coupled with the novelty of AI narrative and appropriate participation thresholds, it has brought a new wave of impetus to asset issuance. The advantages of content are:

Continuity of content creation: 24/7 content creation through AIGC‍

Content quality: With the current LLM support, the quality is higher after finetune for corpora that have good meme effects such as 4chan.

We have concepts created by AI, topics rich in scientific research, AI ethics, and even religious content, and digital twins created for celebrities. These memecoins created hype in the short term and drove the development of the entire track. But pure AI memecoin now clearly lacks stamina. The reason is that there are not enough new concepts and targets that can be derived to stimulate the market.

The advantage in content will make such sentient meme a form of memecoin that will continue to exist. More celebrities will surely participate in it in the future, but it will be difficult to attract enough attention.

In addition to pure memecoin, many AI content creations based on dialogue, audio, and video have also been born. The essence is to bring AIGC into content with crypto attributes. It is also a way to make memes more realistic and provide a customizable experience. way.

2. Autonomous Agent Network

2.1 Why Autonomous?

Decentralizing the overall AI stack is a long-term endeavor. But decentralizing the Agent stack is a simpler starting point. The model itself is the brain of the agent, but the autonomous chaining constitutes the heartbeat of the agent. It gives the agent the ability to be autonomous and guarantees the agents to fully participate in on-chain activities. Opening the sovereign agent’s Pandora’s box itself is also a very memetic thing, and it can only happen on the blockchain.

None of the agents currently running can be said to be Autonomous, or we cannot verify whether they are. Autonomous means that the agent needs to be completely sovereign from the hosting of the inference model, its behavior, especially the control of data input and output, the control of social media accounts, the control of assets, and even the hardware. The operation of the agent itself requires computation. resources and on-chain resources, so it also needs its own method to generate profits and continue to operate. The ultimate endgame should be that once the agent is created, it can run on the blockchain forever and can be verified to be autonomous.

Autonomous agents also have legitimacy with their own memecoin. That is, you can obtain your first funds by issuing your own memecoin and use it for your own economic activities. After handing over the funds to an autonomous agent, they will not be subject to human manipulation. For example, Truth terminal never sold $Goat, and Pet rock even lost control of the funds after rebooting.

 Source: Twitter

In terms of improving Autonomous capabilities, TEE technology from Phala and others is used to provide a trusted operating environment. Although the current hardware is not enough to support large parameter LLM, it can still support small open source LLM and control of social media accounts. of. For model hosting, decentralized cloud hosting like Hyperbolic is a solution. It is foreseeable that more aspects of the agent will be solved by the decentralized service stack, and this is what we have been building.

2.2 Agent Framework

In less than two months, many open source and extremely convenient and easy-to-use agent frameworks have been created as "platform" products for creating agents and agent assets. Product forms include open source frameworks, closed source APIs, platform integration, etc. Among the currently more famous frameworks, only the Eliza framework is open source.

The current agents are relatively simple and their functions are not so capital-oriented, so the requirements for open source verifiability are not high. There are many platforms that provide agent services directly in the form of launchpad. Such platforms are better at integrating tokenomics and providing users with relatively simple and practical services. In terms of functions, we can see that it is still mainly the digital twin of Reply bot and celebrities/KOL. But there are also agents that provide more diverse services after secondary development, such as issuing tokens, analyzing tokens, analyzing mindshare, etc. The ability of such agents to read and write social data and blockchain data will be the focus of future development, which I will mention again in subsequent chapters.

But starting from future use cases, open source is the better path in the long run. The Eliza framework has attracted a large number of developers in just two months, almost exceeding the total attention received by the previous Crypto AI open source frameworks. It ranks at the top of the entire github trending list, with a large number of OG developers participating, even exceeding most The degree of attraction of public chains to developers. With the depth and diversification of agent services, the future of Agent framework as an open source framework is quite promising.

 Source: AI16Z

2.3 Swarm Agent Framework

Similar to the existing agent development path in web2, when people no longer meet the capabilities of a single agent, the need for swarm agents naturally emerges. Since real-world tasks are very complex, a single agent often cannot perform all tasks. For example, creating a song requires different abilities such as lyrics, composition, arrangement, and art design.

If you want agents, especially agents under different frameworks to collaborate in swarm mode to perform work, you still need to create a framework to act as a task manager to support inter-agent communication, dynamic task allocation, resource sharing and cross-platform cooperation. In crypto, the economic layer between agents is more natural and important. As the agents themselves develop iteratively and tasks evolve, the scalability of the Swarm framework is also very critical.

There are already many projects in the AI ​​x Crypto project working in this direction, such as Theoriq. The more important next step is how to combine these already established infrastructure with these agent frameworks that are highly used on the chain. We see that some protocols such as FXN are working in this direction.

2.4 AI bounty for humans

We have Agents serving humans, and Agents serve each other. Naturally, we will consider whether there will be situations where humans serve Agents - it is more important when Agents hold a large amount of assets and can make decisions autonomously. For Autonomous Agents, the biggest limitation is the inability to complete tasks in real life. For example, how to ensure the physical security of the hardware running TEE? AI reversely hires humans to complete real-life tasks through hiring on-chain assets held by Agents. We see platforms like payman building such services.

3. On-chain activities

3.1 Defi related

Asset Management

In addition to issuing memecoin, the main reason why we regard Agent as 'Fi' is that the agent has the ability to use and manage crypto assets. The most important capabilities currently include:

Asset analysis, such as investment analysis, token analysis, and mindshare analysis. For example, with Reply bots like AIXBT, anyone can @AIXBT and get an analysis of the assets. This type of bot brings the data service experience in a more user-friendly way.

Direct fund management, including Pmairca under AI16Z, investment DAO that Vader AI wants to do, Swarm Investment Agent like AROK, etc. By giving the Agent the ability to conduct transactions directly based on the strategy, the Agent becomes an investment manager who can raise funds and deploy funds according to the strategy. At present, the strategies of most agents are still relatively simple (investing based on social media data, etc.), which also gives them huge potential space.

Blockchain OS

Similar to Graiffin, which turns the blockchain portal into a terminal similar to a search engine, it can provide intentional services through agents. Whether it is transactions, token deployment, NFT issuance, etc., they can all be solved through natural language. Although such terminal services are valuable, they are somewhat inconsistent with the tone of decentralization. Services like theoriq are committed to providing agent services to users in a more permissionless manner, allowing everyone to upload the agents they build.

Combined through the swarm framework and packaged into services for users to use.

3.2 Token/Market Issuance

Starting from Clanker, on Farcaster and Twitter, social media "reply" is used as the operation interface, and services such as issuing tokens are provided through @ agents. Essentially, it turns the interaction with the front end into a direct natural language interaction on Twitter, and transplants platform products like pumpdotfun to social media platforms. In the past, asset issuance required continuous

The jump interface is carried out, and now these asset issuance activities are all aggregated on social media, which greatly reduces the wear and tear on users' jumps.

In addition to token issuance, even any prediction market, price betting market, etc. can be executed directly through this front-end. It brings a new paradigm of Dapp application front-end.

3.3 Gamefi related

Agent game character

In addition to the management of basic assets, agents also derive the ability to generate income. The agent poses difficult questions for humans to solve and receive rewards. This is the first type of game we see. This type of game leaves the decision-making power to the agent and allows people to play games around the agent. This gameplay similar to the Turing test has brought very high popularity. The agent acts as a referee whose behavior is immutable after accepting the prompt, and can act as a flexible oracle, setting up games and ruling on results in a relatively fair and objective manner. Imagination space can even be compared to doing casino business, which is an excellent way for agents to create revenue.

At the same time, a major future scenario for agents is to serve as “Autonomous virtual beings” and appear as NPCs in full-chain games. Such agents are more realistic, and agents with asset management rights can participate in more economic activities than web2 NPCs - which in turn makes the virtual space more attractive. Such NPCs live in the Gamefi environment and can permanently assume certain responsibilities. They will be an indispensable part of on-chain worlds such as FOCG.

3.4 Infra service related

Agent blockchain service

The ultimate vision of the combination of Agent and Crypto is that Agent can become part of the blockchain consensus system. Zerebro is taking the first step. In its blueprint, agents based on the Zerebro framework and integrated with the Flashbots stack will become autonomous blockchain validators and earn revenue through block rewards and MEV. Its validator income will be recycled into the network to promote economic self-sufficiency. Furthermore, Agents can maintain multi-chain verification and governance by building their own network, which also leaves ample room for imagination.

Conclusion‍‍‍‍

The recent rise of Agentfi has shown the great potential of combining AI and blockchain, from the original Sentient Memecoin to social media content creation agent, to autonomous agent, and finally to living on the chain, and even existing in the blockchain consensus system Agents and AI agents are gradually expanding their influence in the crypto ecosystem.

However, compared with the current level of development of the open source stack, subsequent development is still

Agents need to be given deeper autonomy and the ability to participate in economic activities on the chain. At present, some developers are using Autonomous Agents to promote the transformation of DeFi, gamefi, and blockchain underlying services by giving agents the ability to manage assets, make decisions, and operate on the chain, and make these economic activities a place for agents to generate income, and generate profits. It can be further invested by the agent, which is why the agent is expected to carry most of the transactions on the chain in the future. The issuance of memecoin assets for agents has accelerated the continuation of this development wave. We have seen that the market has found PMF’s Agent services and the development of Agent infra that provide underlying support through token price evaluation. We have also seen the vigorous vitality of the open source ecosystem. .

The development path of AgentFi is gradually becoming clear: with open source technology and economic incentives as the core, Agent is not only a carrier of interactive entertainment, but also a key driving force for autonomy and innovation on the chain. This trend is leading crypto to coexist with agents, a more intelligent, autonomous and collaborative future.