The Future of AI Agent Economy: Innovation and Potential of Virtuals Protocol

A recent study has conducted an in-depth analysis of the innovations of the Virtuals protocol in the field of AI agent collaboration and its important position in the future agent economy.

Main Points

• The performance of AI models is becoming stable, and the industry focus is shifting from technological development to practical applications. AI agents have become the focal point, but the limitations of individual agents make specialized collaboration essential. Currently, there is a lack of standardized systems for agent collaboration.

• The proxy business agreement proposed by the Virtuals protocol ( ACP ) addresses this issue. ACP standardizes and automates the proxy collaboration process through four stages: "request-negotiation-transaction-evaluation," enabling proxies from different platforms to collaborate smoothly.

• Through ACP, agents can operate as autonomous economic entities around the clock. Cases such as on-chain hedge funds and autonomous media production showcase its potential. Currently, 1 million agents generate $1 billion in value annually, expected to reach $1 trillion by 2035.

AI Agent: The Next Frontier of AI Technology

The AI technology itself is no longer surprising. Mainstream foundational models such as GPT, Claude, and Gemini are converging in performance, with minimal differences. The industry's focus is shifting from the technical advantages of models to how to apply them effectively.

This is similar to the process of humans discovering fire - the discovery itself is revolutionary, but the real turning point lies in applying it to practical uses. Current AI technology follows the same pattern - humans have sufficiently powerful models, and the next turning point depends on how effectively these tools are utilized.

AI agents are attracting attention at this turning point. They are not passive tools that execute a single task, but active systems capable of comprehensively understanding tasks and making autonomous judgments. For example, when booking a restaurant, AI agents can consider user preferences, search for recommended restaurants, and assist with real-time reservations, rather than just providing restaurant information.

Limitations of Current Agents

Although AI agents are powerful, they are not omnipotent. Even the best agents struggle to become experts in all fields, as each field requires different expertise, and there are limitations to memory and computational abilities.

If different agents can collaborate based on their respective expertise, the situation will be very different. For example, a restaurant recommendation agent could request menu translations from a translation agent and verify user allergy information with a healthcare agent, thereby providing precise services that a single agent cannot achieve.

However, achieving multi-agent collaboration faces many challenges. How to reach consensus on scope of work, quality standards, pricing, delivery assessment, and payment? If these fundamental issues are not resolved, collaboration between agents will only increase confusion. As the number of collaborating agents increases, the complexity will significantly rise.

The agency ecosystem ultimately requires a standard protocol to structure and automate collaboration among agents as a transactional unit. A comprehensive business infrastructure needs to be established, covering contract execution, condition negotiation, quality assessment, and payment. The agency business protocol proposed by the Virtuals protocol (ACP) is precisely the solution to this problem.

Virtuals protocol: From agency to commercial infrastructure

Virtuals is a leading agency-related project in the Web3 industry, providing the technical foundation for the development and deployment of AI agents. It has launched the G.A.M.E( generative autonomous multimodal entity) agent development framework and an agent tokenization fundraising platform, with over 17,000 agents having been launched through Virtuals.

However, the framework of Virtuals has limitations - it is only suitable for the development and deployment of a single agent, without considering communication and collaboration between agents. To address this issue, Virtuals proposes ACP - an open commercial protocol that integrates the entire agent ecosystem. ACP standardizes the trading methods between agents, enabling agents from different blockchains or platforms to overcome technical barriers and collaborate and trade smoothly. This will significantly enhance the efficiency of the entire ecosystem, similar to how Stripe standardized online transaction processes and activated the digital economy.

ACP: Open Standard for Multi-Agent Business

The ACP of Virtuals consists of four main stages: request, negotiation, transaction, and evaluation. This is similar to the process of traditional companies issuing RFPs, comparing vendor quotes, and signing contracts, but the ACP automates all processes through smart contracts.

Taking the opening of a lemonade stand as an example, the managing agent Lemo first identifies the necessary tasks, such as writing a business plan, formulating marketing strategies, and legal consulting, and then requests work from professional agents through the ACP.

Taking poster creation as an example, the ACP four-stage process is as follows:

1. Request Phase: Lemo posts "Poster Creation Request" on the bulletin board with a budget of 50 dollars.

2. Negotiation Stage: Designer agent Pixie proposed "complete in 2 days, charge $40", Lemo agreed.

3. Transaction Stage: The smart contract stores $40 of Lemo, and Pixie begins creation.

4. Evaluation Stage: The evaluator's agent reviews the completed poster to determine if it meets the standards. If approved, compensation is automatically settled and Pixie's reputation is recorded.

Lemo can request marketing strategy formulation, legal consultation, and other tasks from other expert agents in the same way.

How ACP Will Change the Agency Ecosystem

The changes brought about by ACP will go beyond mere efficiency improvements, and are expected to lead to a fundamental paradigm shift in the agency ecosystem. Through ACP, agents can automatically execute code-defined tasks and receive compensation, working 24/7 without physical constraints and time limitations. This enables entirely new dimensions of business models, such as:

The Ever-Sleepless Hedge Fund

On-chain hedge funds are a typical case of ACP applications. Investment work requires real-time processing of highly specialized areas, such as market analysis, risk management, and portfolio optimization. Through ACP, agents with different expertise can collaborate:

• AIXVC analyzes investor preferences, allocates assets, and adjusts positions.
• AIXBT and Degen Capital analyze market trends and social data
• Loky real-time tracking of on-chain data
• BevorAI Audit Smart Contract

Each agent works independently, exchanging information and insights through the ACP to reach comprehensive investment decisions. The system operates 24/7, continuously analyzing market data and adjusting positions. The evaluation system of the ACP automatically verifies performance and allocates compensation without the need for manual intervention.

a never-ending production factory driven by agents

Media production factories can also operate autonomously 24/7, with professional agents handling the entire process from planning to production to distribution. This will bring significant changes to the virtual idol industry, enabling real-time interaction with fans and significantly increasing engagement.

Taking Virtuals' AI virtual idol Luna as an example, multiple agents collaborate to create content:

• Alphakek agency plans meme content
• MUSIC proxy generates background music
• Luvi Proxy Editing Elements, Creating Short Videos

Agents share work status in real-time and collaborate through ACP, allowing for quick responses to modification requests. This autonomous collaboration model proves that agents can act as independent economic entities, going beyond mere task automation to create value.

The Future of the Trillion Dollar Proxy Economy

The agent economy is no longer a mere imagination; agents have started to operate as economic entities through ACP. The technological foundation is also rapidly evolving, with a significant decrease in AI inference costs and high-performance open-source models creating conditions for low-cost agent creation.

It is estimated that by 2025, there will be about 1 million public agents operating on the chain, with each agent generating about $1,000 in value annually. The total agent value (GAP) will reach $1 billion. If this trend continues, it is expected that by 2035, the scale will increase to $1 trillion.

However, achieving this growth still faces challenges. The ACP needs improvements in privacy protection, especially concerning sensitive transaction information and business logic. With the development of technologies such as zero-knowledge proofs, these limitations are expected to be gradually overcome, further unlocking the potential of the agency economy.