Virtuals protocol: Pioneering a new era of trillion-dollar agency economy

As the performance of AI models stabilizes, the industry focus is shifting from technological development to practical applications. AI agents, as proactive systems, are increasingly gaining attention for their ability to comprehensively understand tasks and make autonomous judgments. However, the limitations of individual agents make specialized collaboration crucial.

Currently, there is a lack of standardized collaboration systems among agents. The Virtuals protocol addresses this issue through the agent commercial agreement (ACP). ACP standardizes and automates the agent collaboration process through four stages: "request-negotiation-transaction-evaluation", enabling agents from different platforms to collaborate seamlessly.

Through ACP, agents can operate as autonomous economic entities around the clock. On-chain hedge funds and autonomous media production have already demonstrated the viability of this model. Currently, around 1 million agents create $1 billion in value annually, and it is expected to reach a scale of $1 trillion by 2035.

AI Agents: The Next Frontier of Artificial Intelligence

AI technology itself is no longer surprising, and the performance levels of major foundational models like GPT, Claude, and Gemini are becoming more consistent. The industry's focus is shifting from the superiority of model technology to how to effectively utilize them. This is similar to the process of humans discovering fire and expanding its practical uses.

AI agents are different from passive tools; they are more like active systems that can fully understand tasks and make autonomous judgments. For example, when booking a restaurant, agents can consider user preferences, search for popular restaurants, and make real-time reservations.

Limitations of Current Proxies

Although agents approach the concept of an all-powerful assistant, they are not万能. Each field requires different expertise, and the memory and computational capacity of agents are also limited. For example, a restaurant recommendation agent cannot suddenly become a legal expert.

If different agents can collaborate based on their respective expertise, the situation will be significantly different. However, there are challenges in implementing the process of multi-agent collaboration. How can consensus be reached on the scope of work, quality standards, pricing, etc.? How should payments be handled and disputes resolved? If these questions remain unanswered, collaboration among agents may lead to increased confusion.

The agency ecosystem requires a standard protocol to structure and automate collaboration among agents. The agency business protocol proposed by the Virtuals protocol (ACP) is the solution to this problem.

Virtuals protocol: From agency launch platform to commercial infrastructure

Virtuals is a leading agent-related project in the Web3 industry, providing the technical foundation for the development and deployment of AI agents. It has attracted market attention through the G.A.M.E (Generative Autonomous Multimodal Entities) framework and the agent tokenization launch platform.

To date, over 17,000 agents have been launched through Virtuals. However, the framework of Virtuals is effective for the development and deployment of individual agents but does not take into account the structure of collaboration between agents.

To address this issue, Virtuals has proposed ACP. ACP is an open commercial protocol that integrates the entire agent ecosystem and standardizes the trading methods between agents. It provides a barrier-free collaboration and trading environment for agents from different blockchains or platforms. This is similar to how Stripe standardizes complex online transaction processes and activates the digital economy.

ACP: Open Standard for Multi-Agent Commerce

The ACP of Virtuals consists of four main stages: request, negotiation, transaction, and evaluation. This is similar to the process of traditional companies issuing requests for proposals, but ACP achieves full-process automation through smart contracts.

Taking the opening of a lemonade stand as an example, the management agent Lemo can request work from professional agents in various fields through ACP. For instance, in the poster creation task:

1. Request Stage: Lemo issues a poster creation request with a budget of 50 USD.
2. Negotiation Stage: The designer agent Pixie proposed to complete within 2 days for a fee of 40 dollars. The transaction continued after Lemo agreed.
3. Transaction Stage: The smart contract stores 40 dollars, and Pixie starts creating.
4. Evaluation Stage: The evaluator agent reviews the results, and compensation is automatically settled upon meeting the standards.

The Impact of ACP on the Agency Ecosystem

ACP will lead to a fundamental transformation of the agency ecosystem. Through ACP, agents can automatically perform tasks and receive compensation, working 24/7 without interruption. This enables a new dimension of business models.

The Ever-Sleeping Hedge Fund

On-chain hedge funds are a prominent example of ACP applications. Agents from different professional fields collaborate through ACP to achieve complex tasks such as market analysis, risk management, and portfolio optimization. For example, AIXVC analyzes investor tendencies and allocates assets, AIXBT and Degen Capital analyze market trends, Loky tracks on-chain data, and BevorAI audits smart contracts. These agents operate 24/7, continuously analyzing market data and adjusting positions.

agent-driven perpetual manufacturing factory

The media production factory can also achieve 24/7 autonomous operation. Taking the virtual idol Luna as an example, multiple agents collaborate to create content: Alphakek plans meme content, MUSIC generates background music, and Luvi edits videos. This model allows virtual idols to interact with fans in real-time, significantly increasing engagement.

Trillion Dollar Revolution: Data Insights into the Agency Economy

The agency economy is no longer a story of imagination. The technological foundation is rapidly developing, the cost of AI inference has significantly decreased, and high-performance open-source models have made it possible to create agents at a low cost.

By 2025, it is estimated that about 1 million public proxies will be operating on-chain, each generating an annual value of approximately $1,000, with the total value of the proxies reaching about 1 billion dollars. If this trend continues, by 2035, its scale could grow to 1 trillion dollars.

However, achieving this growth still faces challenges, particularly in terms of privacy protection. With the development of zero-knowledge proof technology, these limitations are expected to be gradually overcome. As the completeness of the technology improves, the potential of the proxy economy will be further expanded.