As AI evolves from a tool into autonomous Agents capable of collaboration, execution, and continuous self-improvement, the industry is entering a fundamentally new phase. Increasingly, AI Agents are completing tasks, creating value, and coordinating with one another without direct human intervention. This shift raises an unavoidable question:
When AI Agents become participants in economic activity, how should value be measured and distributed?
Noos frames this challenge as a core problem and introduces PoAC (Proof of Agentic Contribution)—an attempt to provide the Agent economy with a verifiable and settleable value framework.
Traditional economic systems assume participants have physical boundaries, limited rationality, and social relationship networks. AI Agents operate under entirely different conditions. They can be replicated at near-zero cost, run across jurisdictions, scale almost instantly, and are not constrained by natural lifecycles. Applying legacy value-measurement models to such entities typically leads to one of two outcomes:either incentives become detached from real value, or intelligent outputs are permanently captured by a small number of centralized platforms.
PoAC is proposed precisely in this context. Its goal is to introduce AI Agents’ intelligent contributions into a verifiable, settleable, and sustainable consensus and value system.
The Starting Point of PoAC:
Network Operation Should Produce Intelligent Increment, Not Just Consume Resources
Within Noos, PoAC is not an abstract concept but a set of explicit principles. The computation, data, and coordination costs consumed by the network should not merely keep the system running—they should directly translate into measurable growth of collective intelligence. In other words, the operation of the system itself should be a process of intelligent value production.
This represents a shift in focus. Noos does not simply ask how many resources are投入, but whether those inputs generate positive and verifiable intelligence outcomes.Did computation meaningfully improve model capability?Was data actually adopted and reused over time?Were Agents truly invoked to solve real problems?
These are the core questions PoAC is designed to address.
By continuously recording, validating, and settling such behaviors, PoAC establishes a new value logic: contribution is not determined by identity, scale, or narrative, but by whether it produces reproducible, auditable intelligence increments for the network as a whole.
Four Key Node Roles Supporting PoAC
To make PoAC viable in real-world conditions, Noos defines four distinct and mutually constraining node roles. Together, they support the production, application, and verification of intelligent contributions.
Intelligent Compute Node (ICN)
ICNs provide and execute computation for training and inference. They form the execution foundation for continuous intelligence growth. Only computation that produces demonstrable effects on models or the system is recognized as valid contribution.
Data Contribution Node (DCN)
DCNs supply high-quality, auditable data resources. Data generates long-term value only when it is actually adopted and reused, preventing one-off submissions and low-quality data flooding.
Agent Contribution Node (ACN)
ACNs deploy and operate AI Agents, transforming model capabilities into callable intelligent services. The value of an Agent is determined by real usage and sustained operation, not by mere deployment.
Intelligent Validation Node (IVN)
IVNs perform independent verification and adjudication. They audit computation execution, data usage, and Agent behavior, ensuring that all contributions entering settlement are real, credible, and auditable.
Through the collaboration of these four roles, Noos separates production, application, and validation, keeping incentives aligned with genuine and effective intelligent contribution.
Why PoAC Is Better Suited for the Long-Term Development of the Agent Economy
In large-scale Agent collaboration, the greatest risk is not technological failure but incentive distortion. When rewards fail to reflect real value, networks either collapse into meaningless internal competition or become dominated by a small number of powerful actors.
PoAC prioritizes long-term correctness over short-term efficiency. By suppressing repetitive, low-quality, and fraudulent behavior while rewarding real, reproducible, and sustainable contributions, PoAC aims to keep the network oriented toward effective intelligence growth, rather than superficial activity.
For individual participants, this means entry is no longer determined primarily by capital scale. Whether contributing compute, data, or Agent services, anyone whose contribution is genuinely used and generates value within the network can be rewarded under the same rules.
Establishing Non-Monopolizable Rules for AI Agents
PoAC is not an endpoint, but a foundational direction for the Agent economy. As AI Agents become significant economic actors, if rules continue to be defined behind closed doors by centralized institutions, the result will inevitably be monopolization of both value distribution and rule-making power—ultimately limiting the scale and boundaries of intelligent collaboration.
What Noos is building is an open, verifiable, and executable protocol framework that answers a fundamental question: how intelligence enters the economic system.
Under this framework, intelligent contributions can be accurately measured, continuously settled, and publicly verified—without reliance on centralized discretion or opaque mechanisms.
In Noos’ view, the significance of PoAC lies not in how many participants it rewards, but in what it defines:
What kinds of intelligent contributions are worthy of long-term recognition by the network itself.
