Article reprinted from: Foresight News

Author: KarenZ, Foresight News

AI agents are no longer simply chatbots. Especially at the intersection of Crypto and AI, a new narrative—AI Layer 1 (abbreviated as AI L1)—is quietly emerging.

Here, 'Layer1' does not refer to public chains but to the foundational layer of AI technology, encompassing the basic technology stack and framework of AI agents, including AI agent token launch platforms similar to pump.fun, from which it can capture enormous value for its tokens.

ai16z: Launchpad for Eliza-based agents launched in the first quarter.

ai16z, with the excellent performance of its Eliza agent framework, has become a leader in AI agents. In the past month, Eliza has performed remarkably well on the GitHub trending list, climbing to the top of the trend list, gaining 6,532 stars, and reaching a total of 9,600 stars.

However, despite the significant technical successes of ai16z, there is still room for improvement in the value capture and accumulation aspects of its token economic model.

To this end, ai16z proposed a plan called 'ai16z Token Economics' in December 2024, aiming to transform ai16z into an AI L1 blockchain, launch a Launchpad for Eliza-based agents, make ai16z the base currency for agent-to-agent transactions (A2A) and an 'App Store' for agent services, implement node incentives, and establish value acquisition and accumulation strategies, among others. Through these measures, ai16z aims to build a more comprehensive and powerful AI agent ecosystem.

Specifically, the core discussion content of this proposal and ideas discussed by Discord community members include:

1. Launch a Launchpad for Eliza-based projects in the first quarter of 2025;

2. Make ai16z the base currency for agent-to-agent transactions (A2A) and an 'App Store' for agent services. Position ai16z as an L1 blockchain targeting AI;

3. Node incentives;

4. Value acquisition mechanism: charging a certain start-up fee, requiring the holding of ai16z to obtain allocations at launch, ai16z token staking (to gain early access and fee-sharing benefits), making ai16z the token for purchasing initial agent products, pairing agent tokens with ai16z's liquidity pool.

5. Value accumulation strategy: The repurchase and destruction mechanism of Launchpad fees (there is also community user support for reinvesting in Eliza development rather than destruction).

5. Provide liquidity mining incentives and locking to stabilize LP and incentivize long-term liquidity provision.

7. When a new AI project launches its token on the platform, users can trade that token for SOL, with LP generating a small fee collected by the platform. The Launchpad may split this fee into two parts: half of the fees will be used to repurchase ai16z tokens from the market, and the other half will be paired with newly purchased ai16z tokens to create a special fund pool, providing additional liquidity for the new project's tokens.

8. Implement a community 'curation' process that allows ai16z partners to review projects before launching to maintain quality and ensure alignment with the ai16z brand. This may involve staking and reductions.

9. Treasury diversification: covering major L1 tokens, stablecoins, and selected project tokens, providing liquidity support and incentives for ai16z trading pairs, while avoiding the sale of agent tokens.

10. Do not increase the supply of ai16z tokens.

zerebro: Launching a consumer-oriented AI agent Launchpad.

Coincidentally, at the end of last month, zerebro co-founder Jeffy Yu stated that they are building a consumer-oriented agent Launchpad called 'Zentients', providing a value accumulation framework for the ZEREBRO token. The goal of Zentients is to create a seamless UI/UX to launch and develop agents while returning value to the ZEREBRO token. Transaction fees will be used to establish a fund for special projects/allocations and enhance project liquidity or open new trading pairs.

Specific details are as follows (details are still being finalized):

1. Agents will need ZEREBRO as a creation fee and will launch according to the ZEREBRO joint curve and fees.

2. Once the agent token reaches a certain valuation, it is considered 'graduated'.

3. The liquidity in the joint curve will be deposited and locked into the DEX.

4. Subsequently, traders will buy and sell in the liquidity pool using ZEREBRO.

Of course, zerebro envisions Zentients as an agent platform compatible with both Web2 and Web3, covering Web2 features such as agent email assistants, coding agents, and financial models. Zerebro is also exploring the development of a standardized communication layer for agents—'Agent TCP'—and infrastructure to support multi-agent intelligence. Jeffy Yu also provided an example where one agent can create a tweet, and another agent can critique that tweet and send it back with revisions for publication.

Additionally, through partnerships with decentralized providers, zerebro is also building infrastructure that allows users to rent GPUs on Zentients. This means users will be able to host their agents directly, making Zentients not just a launchpad but also an agent hardware layer. Ultimately, the agents themselves can take action and rent GPUs for self-training and fine-tuning.

It is worth mentioning that in December 2024, Zerebro also announced a collaboration with ai16z to advance the open-source framework ZerePy. The ai16z team will be one of the first external contributors to help develop the ZerePy framework, while the Zerebro team will contribute to the Eliza framework.

Virtuals: Platform mechanisms drive a self-sufficient economic closed loop.

It is worth noting that Virtuals Protocol, rooted in the Base ecosystem, is the pioneer in the AI L1 domain. Since the launch of its token VIRTUAL on Binance's contract market on December 10, 2024, its price has surged nearly twice, with a market capitalization exceeding $4 billion.

Looking back to the end of November 2024, I summarized potential tokens under Virtuals AI agents in an article (Aixbt skyrocketed 170 times in half a month, potential stocks of Virtuals AI agents). Since then, the vast majority of the mentioned tokens have shown strong growth momentum. For example, the GAME token achieved a 12-fold increase, WAI (WAI Combinator) rose nearly 17 times, aixbt and VADER both achieved 4 times growth, MUSIC token also recorded a 3-fold increase, while LUNA token, despite a relatively moderate rise, still achieved a substantial increase of 62%.

The positioning of Virtuals Protocol is by no means just an AI agent platform; it is a vibrant ecosystem full of infinite imaginative space. Relying on its intuitive, easy-to-use, and powerful low-code toolkit, users can easily manage the creation and management of AI agents and tokens.

The strong performance of the VIRTUAL token is largely attributed to the economic system of tokens and platforms in Virtuals.

First, deploying an agent token requires a mandatory stake of 100 VIRTUAL tokens, which are used to establish the liquidity pool for the agent token. The deployer can also purchase a small amount of VIRTUAL to avoid being targeted.

Subsequently, investors will invest in VIRTUAL tokens. Agents need to accumulate a sufficient quantity of VIRTUAL in their joint curve to 'graduate' to full trading status, and the liquidity pool will then be deployed to Uniswap.

Equally important, VIRTUAL, as the base currency for all AI agent token transactions within the platform, ensures the liquidity and utility of tokens. Before purchasing any agent tokens, users must exchange their USDC, ET, or other assets for VIRTUAL.

At the level of agent economics, AI agents within the Virtuals Protocol ecosystem will charge a 1% transaction fee on all transactions to cover the costs incurred by agent execution. Additionally, AI agents using inference services must pay per use, with fees directly credited to the agent's wallet in VIRTUAL tokens. This ongoing demand mechanism encourages agents to become high-value, high-efficiency assets within the ecosystem. The income generated will be directly used to repurchase and destroy agent tokens on-chain, further enhancing its deflationary effect. The prosperity of AI agents within the ecosystem will undoubtedly feed back into the growth of Virtuals Protocol and the VIRTUAL token.

According to Dune data, since Virtuals launched its new platform on October 16, its cumulative protocol revenue has exceeded $72 million, which undoubtedly serves as strong evidence of its strength and potential.

Summary

The economic system of the Virtuals platform indeed showcases a high degree of sophistication and synergy. It provides a clear economic incentive pathway for ecosystem participants through a series of interconnected designs and constructs a self-sufficient, continuously growing economic closed loop. This economic system not only lays a solid foundation for the rapid expansion of the platform but also provides a powerful driving force for the innovation and development of AI agents.

Both ai16z and zerebro have the potential to achieve and surpass successes similar to Virtuals in their respective fields, but this requires them to make correct decisions and executions regarding token economics, community involvement, technological innovation, and market strategies.

In summary, AI L1 is attracting more and more attention with its unique technical charm and economic value. Pioneering projects represented by Virtuals, ai16z, and zerebro are leading the tide of this transformation, building a more complete and powerful foundation for the AI agent ecosystem. These platforms are not only launchpads for projects but also fertile ground for the growth of their token value.

Although there have been many criticisms regarding the implementation details of various AI agent frameworks in the market, in the world of Web3, whichever AI agent stands out from others will have a greater chance of winning.