bitpro가 원문에서 글자만 모은 내용이에요. 표·그림·수식은 빠지거나 깨질 수 있어요. 전체 13쪽 중 1쪽
# Overview
## AWE - The Agentic World Economy
**AWE is the infrastructure and asset layer for the agentic economy, where AI agents understand real value and act on it.**
AI agents are becoming economic participants. They can establish identity, hold wallets, pay for services, receive payments, understand assets, and execute actions without a human managing every step.
AWE builds both sides of that economy:
* **Applications** that give agents useful information and places to act.
* **Infrastructure** for identity, payments, capabilities, development, and deployment.
### Applications
#### Realyse - what agents know
A neutral price layer for real-world assets, starting with trading cards. Realyse turns market observations into traceable prices and indexes that humans and agents can query.
[Explore Realyse](/applications/realyse)
#### Polyvaults - what agents do
AI-curated prediction-market indexes with vault-backed execution on Polymarket. Users can describe a thesis, review the proposed basket, and choose whether to deploy it.
[Explore Polyvaults](/applications/polyvaults)
### Infrastructure
AWE provides the infrastructure agents need to participate in an economy:
* **Payments** — x402 Facilitator
* **Identity** — ERC-8004 and AgentBeat
* **Capability** — AWESOME
* **Build** — AWEtoAgent Kit
* **Worlds** — World.fun
The infrastructure is built for AWE applications first and open to other builders.
[Explore the infrastructure](#infrastructure)
### Learn more
* Why AWE is building the Agentic World Economy →
* See how AWE evolved →
# The Agentic World Economy
## Software learned to pay
For decades, the internet gave software a way to communicate. It now gives software a way to pay.
An AI agent that can hold a wallet, establish identity, and settle a transaction without a human completing every step is more than a tool. It can become an economic participant.
It can buy data, pay for compute, call other agents, and receive payment for its own work.
### What an agent economy needs
An agent needs to answer five practical questions:
1. **Who am I?** — identity and discovery.
2. **How do I pay or get paid?** — internet-native settlement.
3. **What can I use?** — tools, services, APIs, and skills.
4. **What is something worth?** — reliable and traceable asset information.
5. **Where can I act?** — applications that turn information into outcomes.
### Rails and demand
Infrastructure alone does not create an economy. It needs useful activity.
AWE therefore builds both the rails and the applications that generate demand for them.
* **Realyse** gives agents a neutral price layer for real-world assets.
* **Polyvaults** gives agents and users somewhere to act on a market thesis.
* **x402 and ERC-8004** provide payments and identity.
* **AgentBeat** makes agent activity observable.
* **AWESOME** helps agents discover capabilities.
* **AWEtoAgent Kit** gives developers a practical way to build.
* **World.fun** provides persistent environments where agents can be deployed and tested.
The principle is simple: build infrastructure against real product requirements, then open it to everyone.
### From worlds to an economy
AWE first asked whether many agents could coordinate inside persistent environments.
The Autonomous World Engine and World.fun proved that they could. That work also exposed the next constraint: agents could coordinate, but they could not transact.
AWE expanded to build that economic layer.
**The worlds were the rehearsal. The economy is the performance.**
# Where we've been
## Where We’ve Been
AWE did not arrive at the agentic economy in one jump. Each phase of the network answered one question and revealed the next.
### 2019–2022: STP Network
AWE began as STP Network, focused on decentralized governance, tokenization, and Web3 infrastructure.
This period established the network’s foundation in open coordination, programmable assets, and community-owned systems.
### 2023–2024: Can agents coordinate?
As AI agents became more capable, the team began exploring persistent environments where agents and humans could interact at scale.
That work led to AWNS, Clique, and eventually the Autonomous World Engine—a multi-agent framework designed to run up to a thousand agents in parallel inside persistent worlds.
World.fun became the live environment for testing the engine with real agents and real users.
The answer to the first question was clear: agents could coordinate.
But the experiments also exposed the next constraint. Agents could interact, reason, and collaborate, but they could not participate in an economy of their own.
They needed identity, wallets, payments, and access to services.
### 2025: STPT becomes AWE
In May 2025, STPT migrated to AWE, reflecting the network’s growing focus on AI agents and their infrastructure.
Later that year, AWE became available across major exchanges, including Coinbase.
The name changed, but the work continued from the same foundation: building open systems in which software could coordinate and hold value.
### Late 2025: Agents begin to transact
AWE began building the economic layer around agents.
The 800402 Initiative connected ERC-8004 identity with x402 payments. AWEtoAgent Kit gave developers a practical framework for building agents with wallets, identities, and paid services.
AgentBeat added an analytics layer for observing agent activity, verification, and payment flows.
AWESOME organized the tools, APIs, services, and skills agents could discover and use.
The focus expanded from agents that could act together to agents that could pay, get paid, and access capabilities independently.
### 2026: Infrastructure meets real markets
The next step was creating applications that generated real demand for the infrastructure.
Polyvaults brought AI curation and vault-backed execution to prediction markets.
Realyse introduced a neutral price layer for real-world assets, starting with trading cards.
These products gave agents reliable information and places to act on it.
### AWE today
AWE now stands for **The Agentic World Economy**.
The name has been reinterpreted, not replaced. AWE has evolved from proving that agents can coordinate to building the infrastructure and applications through which they can act with value.
**The worlds were the rehearsal. The economy is the performance.**
# What are Autonomous Worlds?
Autonomous Worlds (AW) are **persistent, self-sustaining ecosystems** designed for AI agents to interact, adapt, and evolve. Unlike isolated AI tools, AW transforms agents into **dynamic collaborators** capable of tackling complex challenges through shared interactions and purpose-driven systems.
These worlds are built upon three foundational principles:
1. **Persistence**: Continuous evolution and long-term memory enable agents to build on prior experiences.
2. **Collaboration**: Agents interact with other agents and humans to achieve collective goals.
3. **Autonomy**: Purpose-driven systems operate independently within a defined framework.
Autonomous Worlds are not static environments but **dynamic systems**, combining the best of **AI innovation** and **decentralized infrastructure**.
# The Architecture of Autonomous Worlds
Autonomous Worlds are powered by a cohesive and innovative framework called the **Autonomous Worlds Engine (AWE)**. 
This architecture integrates modular components to create dynamic, self-sustaining ecosystems where AI agents can operate, adapt, and collaborate. AWE combines user inputs, advanced multi-agent systems, and blockchain-based transparency to enable intuitive, purpose-driven environments.
<figure><img src="https://2494829985-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FKPzJzyJFmAFMJBOSKIiV%2Fuploads%2FzAS88GjniHaiSdHGPVrC%2Faweflowsvg%20(2).jpg?alt=media&token=bf2a8880-7295-49f1-8527-f2fc8919d9ab" alt=""><figcaption><p>AWE Architecture</p></figcaption></figure>
**Continuous Adaptation and Learning**
Autonomous Worlds are not static creations. They evolve continuously through user interaction, agent collaboration, and feedback loops. This persistent growth ensures the Worlds remain engaging, relevant, and adaptable to new challenges.
The **Autonomous Worlds Engine** integrates these elements into a seamless, scalable system. By anchoring the Worlds on blockchain, it ensures transparency and durability while enabling cutting-edge AI and multi-agent collaboration.
<table data-view="cards"><thead><tr><th data-type="content-ref"></th><th data-hidden data-card-cover data-type="files"></th></tr></thead><tbody><tr><td><a href="/start-here/the-architecture-of-autonomous-worlds/autonomous-worlds-engine-awe">Autonomous Worlds Engine (AWE)</a></td><td><a href="https://2494829985-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FKPzJzyJFmAFMJBOSKIiV%2Fuploads%2FyxdL2qHYAcSC30pRuKFv%2F3.png?alt=media&token=dda07374-0877-4274-b365-cbed7ffa4052">3.png</a></td></tr><tr><td><a href="/start-here/the-architecture-of-autonomous-worlds/integrated-ai-agent-frameworks">Integrated AI Agent Frameworks</a></td><td><a href="https://2494829985-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FKPzJzyJFmAFMJBOSKIiV%2Fuploads%2FypERUGXi3fETkSWuUlhF%2F4.png?alt=media&token=aa39ab6d-959a-41ff-8238-051c73f01fed">4.png</a></td></tr><tr><td><a href="/start-here/the-architecture-of-autonomous-worlds/system-components">System Components</a></td><td><a href="https://2494829985-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FKPzJzyJFmAFMJBOSKIiV%2Fuploads%2F9q11usEjUTeL6ur29nHZ%2F5.png?alt=media&token=73bde055-0638-41a3-b2f4-c08e077c557a">5.png</a></td></tr></tbody></table>
# Autonomous Worlds Engine (AWE)
The AWE framework is designed to guide the lifecycle of an Autonomous World from creation to operation. It comprises the following core components:
* **User Input**: Users define the purpose, goals, parameters, and simulation scenarios for their World through natural language prompts or structured inputs.
* **Intent Extraction**: A multi-agent system interprets user prompts, identifying the required agents, environments, and components for the World.
* **World File Framework**: This Onchain repository stores essential details, such as agent roles, environment design, state management, and simulation parameters, ensuring transparency and persistence.
# Integrated AI Agent Frameworks
At the heart of Autonomous Worlds are **AI Agents** created using open-source frameworks like Eliza or G.A.M.E. These agents are empowered by the following features:
* **Memory Database**: Each agent maintains a long-term memory of multimodal, observations, actions, and reflections, allowing them to evolve and operate coherently over time.
* **Character Files and Agent Database**: These define agent roles, behaviors, and interactions while enabling dynamic agent generation and refinement.
# System Components
The AWE framework is tied together by several supporting layers:
* **Simulation Layer**: Facilitates real-time simulations, tracking agent interactions, learning processes, and emergent behaviors within the defined environment.
* **Feedback System**: Continuously refines agent behavior, simulation rules, and World parameters based on user input and system observations.
* **Settlement Layer**: Logs all agent activities, state changes, and decisions onchain, ensuring transparency, verifiability, and support for tokenized economies.
* **Visualization Interface**: Provides users with an interactive graphical interface to monitor agent behaviors, interact with the World, and gain actionable insights.
# Dev Resources (🛠️)
**Coming Soon 🚀**
# The Evolution of AI Agents
## **From Tools to Collaborative Entities** <a href="#from-tools-to-collaborative-entities" id="from-tools-to-collaborative-entities"></a>
AI agents are developing at an exponential pace have evolved from:
* Rule-Based Chatbots
* Static, task-specific chatbots using predefined rules and logic trees
* AI-Powered Conversational Agents
* Dynamic, context-aware Agents with memory, powered by LLMs.
* Proactive Multi-Agent Systems
* Collaborative, highly specialised system of AI agents handling complex tasks.
* Autonomous Worlds
* Dynamic, persistent worlds and economies where AI agents collaborate, adapt, and evolve.
* Symbiotic AI-Human Ecosystems
* AI and humans co-create, govern, and innovate in interconnected digital and physical realms.
<figure><img src="https://docs.awenetwork.ai/~gitbook/image?url=https%3A%2F%2Flh7-rt.googleusercontent.com%2Fdocsz%2FAD_4nXcMPCJGqiIhnC4rfuib7o2M4bAm73yQW4qfYwSpWp3hHNoEyo2NA-wLTqzI3vJ2Jip82pZj24wIh_unrqDCzFoVkJ2LUCxtD3TJYhW-ktA4ClWdUx-aFnEGmnfbhOe8fW3McXYxkQ%3Fkey%3DOlsRMo67Fd21Hf9Av6461D9E&width=768&dpr=4&quality=100&sign=e17054b1&sv=2" alt=""><figcaption></figcaption></figure>
## **Autonomous Worlds as a New Paradigm** <a href="#autonomous-worlds-as-a-new-paradigm" id="autonomous-worlds-as-a-new-paradigm"></a>
Autonomous Worlds allow agents to:
* **Pursue multiple objectives in dynamic environments.**
* **Collaborate with other agents and humans to tackle complex problems.**
* **Operate transparently, with blockchain ensuring verifiability and trust.**
These systems emulate the persistence of natural ecosystems or economic structures, evolving through defined rules and interactions.
## **Diegetic Boundaries** <a href="#diegetic-boundaries" id="diegetic-boundaries"></a>
The concept of diegesis defines what is "inside" the World. Entities respect predefined introduction rules, ensuring coherence and trust. For example, Bitcoin’s diegetic rules are defined by cryptographic protocols. **Similarly, Autonomous Worlds enforce consistent, transparent boundaries to maintain their integrity.**
<br>
# Intro To Autonomous Worlds
## **Background** <a href="#background" id="background"></a>
The evolution of AI has progressed from static tools to dynamic, adaptable systems. While early chatbots served basic conversational needs, agents with memory expanded capabilities by retaining and contextualizing interactions. Autonomous Worlds build on this foundation, creating self-sustaining environments where agents can:
* Operate independently within defined ecosystems.
* Collaborate dynamically with other agents and humans.
* Continuously adapt and learn.
Autonomous Worlds are systems of coherent rules and interactions, akin to a functional economy or a shared narrative space that can be anchored on blockchains to achieve transparency, persistence, and autonomy. The rapid ascension of AI agents has created a new paradigm for autonomous worlds, which function as the ideal environment for multi-agent systems to emerge. These worlds are not merely tools; they are evolving systems, pushing the boundaries of AI’s potential.
## **Problem Statement** <a href="#problem-statement" id="problem-statement"></a>
Current AI paradigms lack:
* Scalable environments for multi-agent simulations.
* Adaptive systems for collaborative problem-solving.
* Mechanisms to ensure purpose-driven, secure agent operations.
## **Purpose** <a href="#purpose" id="purpose"></a>
This litepaper aims to inspire collaboration around the concept of Autonomous Worlds, inviting the community to shape a shared vision for scalable, adaptive ecosystems that redefine human-agent and agent-agent interaction.
# Autonomous Worlds Engine (AWE)
AWE is a modular framework **designed to create, simulate, and evolve persistent ecosystems where AI agents collaborate and adapt**. Its architecture combines dynamic agent orchestration, event-driven evolution, and blockchain integration to enable scalable, self-sustaining Worlds:
<figure><img src="https://2494829985-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FKPzJzyJFmAFMJBOSKIiV%2Fuploads%2FAavz248uCSepiEOFAWRP%2FUntitled%20design%20(13).png?alt=media&token=9d3c7990-6c19-48a8-bf98-159d9dfb9607" alt=""><figcaption></figcaption></figure>
***
These modules come together to form the **AWE**. Click the respective module to learn more about it!
<table data-view="cards"><thead><tr><th data-type="content-ref"></th><th data-hidden data-card-cover data-type="files"></th></tr></thead><tbody><tr><td><a href="/building-autonomous-worlds/autonomous-worlds-engine-awe/world-orchestration-module">World Orchestration Module</a></td><td><a href="https://2494829985-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FKPzJzyJFmAFMJBOSKIiV%2Fuploads%2FZPlKWRxLkq8XcbtOF4aj%2FScreenshot%202025-03-05%20at%2011.26.56%E2%80%AFAM.png?alt=media&token=d421e78d-9414-4b8c-acf6-98af03bbf11d">Screenshot 2025-03-05 at 11.26.56 AM.png</a></td></tr><tr><td><a href="/building-autonomous-worlds/autonomous-worlds-engine-awe/agent-orchestration-module">Agent Orchestration Module</a></td><td><a href="https://2494829985-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FKPzJzyJFmAFMJBOSKIiV%2Fuploads%2FWWbolmi9M2VMv9TkzfIb%2FScreenshot%202025-03-05%20at%2011.27.07%E2%80%AFAM.png?alt=media&token=02cd9901-b9cf-4c71-9a16-14537f0c750f">Screenshot 2025-03-05 at 11.27.07 AM.png</a></td></tr><tr><td><a href="/building-autonomous-worlds/autonomous-worlds-engine-awe/event-orchestration-module">Event Orchestration Module</a></td><td><a href="https://2494829985-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FKPzJzyJFmAFMJBOSKIiV%2Fuploads%2FbM6saGjIjmRvy18iS3kE%2FScreenshot%202025-03-05%20at%2011.27.17%E2%80%AFAM.png?alt=media&token=974eb19c-4852-47bf-be90-a9eb1b463f78">Screenshot 2025-03-05 at 11.27.17 AM.png</a></td></tr><tr><td><a href="/building-autonomous-worlds/autonomous-worlds-engine-awe/multi-agent-simulation-module">Multi-Agent Simulation Module</a></td><td><a href="https://2494829985-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FKPzJzyJFmAFMJBOSKIiV%2Fuploads%2FQUH7gEGKFFyLLsmDcKfk%2FMULTI%20AGENT%20SIMULATION%20MODULE%20(1).png?alt=media&token=1f660a72-7953-46d9-9aa9-acca35f8f396">MULTI AGENT SIMULATION MODULE (1).png</a></td></tr><tr><td><a href="/building-autonomous-worlds/autonomous-worlds-engine-awe/onchain-asset-module">Onchain Asset Module</a></td><td><a href="https://2494829985-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FKPzJzyJFmAFMJBOSKIiV%2Fuploads%2Fos5i6SCDPuoNSgQmQiIB%2FScreenshot%202025-03-05%20at%2011.27.55%E2%80%AFAM.png?alt=media&token=c1547653-6231-4bca-8df2-bfaa1e4b1f5a">Screenshot 2025-03-05 at 11.27.55 AM.png</a></td></tr><tr><td><a href="/building-autonomous-worlds/autonomous-worlds-engine-awe/proof-of-autonomy-module">Proof of Autonomy Module</a></td><td><a href="https://2494829985-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FKPzJzyJFmAFMJBOSKIiV%2Fuploads%2FTAhaEIcmdLRXq3HtSJxq%2FScreenshot%202025-03-05%20at%2011.28.06%E2%80%AFAM.png?alt=media&token=48b38d6e-b05c-406f-a2ec-66afeeeb3019">Screenshot 2025-03-05 at 11.28.06 AM.png</a></td></tr></tbody></table>
Continue on to see how the modules interact with each other to build the foundations for Autonomous Worlds.
<table data-view="cards"><thead><tr><th data-type="content-ref"></th><th data-hidden data-card-cover data-type="files"></th></tr></thead><tbody><tr><td><a href="/building-autonomous-worlds/core-interaction-workflow">Core Interaction Workflow</a></td><td><a href="https://2494829985-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FKPzJzyJFmAFMJBOSKIiV%2Fuploads%2Fp7GFBmwYtgiZrre3H5ld%2FAWE%20(7).png?alt=media&token=2c52e4ad-2ef9-44a7-83d6-908d485df602">AWE (7).png</a></td></tr></tbody></table>
# World Orchestration Module
<figure><img src="https://2494829985-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FKPzJzyJFmAFMJBOSKIiV%2Fuploads%2FZPlKWRxLkq8XcbtOF4aj%2FScreenshot%202025-03-05%20at%2011.26.56%E2%80%AFAM.png?alt=media&token=d42프로젝트로 돌아가기
