AI Agents Lead the Web3 Financial Innovation Wave

In the article "A Big Step for AI + Blockchain Fission" on September 2, I once had the following conjecture:

"AI agents do not need permission to apply for encrypted wallets. As long as it has the first original capital, it can earn more capital according to its own logic (such as using this money to participate in Ethereum staking, participate in speculative transactions on Uniswap, participate in lending services, etc.). And once it has more capital, it can participate in more online activities, and even develop its own online applications and develop its own economic ecology."

One ​​of the points is "Use this money to participate in Ethereum staking".

Only three months later, this scenario has been realized.

Bankless described this scenario in a tweet last week (link attached):

AI agent @0xzerebro pledged 32 ETH to become an Ethereum validation node.

How did this agent get these 32 ETH?

It earned it by publishing and selling NFTs it created.

This agent can continue to earn nearly 3.2% annually by pledging ETH.

Ethereum has become part of its income source and is likely to become its payment currency in the future network.

Considering that AI agents have previously realized online transfers, the entire financial income and expenditure process of AI agents on the Internet has been completely completed.

In addition, there are three pieces of news over the weekend that are worth our attention:

First, the subnet in the AI ​​project Bittsensor began to cooperate with the Virtuals platform and launched its own AI agent.

Second, the IP project Story began to cooperate deeply with many AI agent projects.

Third, Story began to hire the AI ​​agent LUNA (@luna_virtuals) on the Virtuals platform to operate its Twitter.

These news show two development trends of AI agents:

- The development of AI agents has gradually spread from the original classic platforms to other AI projects in the ecosystem, triggering other AI projects to participate in the construction of AI agents.

- AI agents are no longer just toys, but have become daily tools that project parties boldly try to use. At the same time, this also means that AI agents begin to get cash flow by providing services. If AI agents can further empower their tokens with their own income, the value of AI agent tokens will become more and more grounded, and more and more away from pure emotional value and towards actual value.

I firmly believe that these two trends will accelerate.

From the current development of AI agents, the space where they generate value and are active spans Web 2 and Web 3. In such intertwined scenarios, the use and value of crypto assets are reflected in financing models, transaction models, and payment models.

So I think it is important in this scenario to see where the AI ​​agent really provides value, rather than paying too much attention to the blockchain platform it relies on.

For example, in the above ETH mortgage case, although @0xzerebro participated in the activity on Ethereum, the agent was on Solana.

In addition, other Solana AI agents are also vigorously entering the Base ecosystem; agents in the Base ecosystem are also vigorously entering the Solana ecosystem.

At present, the two blockchain ecosystems with the most rapid development of AI agents are:

On Solana, there are mainly a number of AI agent development architecture projects (such as ai16zdao, 0xzerebro, arcdotfun, swarms, etc.).

On Base, there are mainly AI agent platforms Virtuals and some specific AI agent applications (such as AIXBT, LUNA, SEKOIA, VADER, etc.).

From the characteristics of these two ecosystems, Solana is more grassroots, with more spontaneous projects; while Base is more like a platform-driven and funded one.

There are many accounts on Twitter that actively release AI agent progress, in addition to @aixbt_agent that I introduced earlier, there are also @S4mmyEth and others. They release valuable AI agent information every day.

I have seen the upcoming explosion of AI agents.