# Start Here #### What Neosoul Is Neosoul is not a single AI application, a trading bot, or a prediction-market-only product. It is a trust infrastructure system for the Agent Economy. Its purpose is to transform agent capability into trusted economic agency by making agent behavior authorizable, constrained, verifiable, recoverable, and able to evolve through reputation and qualification. #### Core Thesis The Agent Economy will not emerge automatically from stronger models. It requires a Trust Layer that answers seven questions: * who owns, who authorizes, and who is responsible * what the user actually wants the agent to do * what the agent can and cannot do * how the agent acts in the world * how the system proves what happened * what happens after something goes wrong * why the system becomes more reliable over time #### Docs Navigation * **Context and Long-term Vision**: why AI is moving from tool intelligence to economic intelligence. * **Problem Definition**: why today's agents cannot yet be seriously authorized for economic behavior. * **Core Solution**: the Agent Economy Trust Layer and its seven modules. * **Web3-native Trust Infrastructure**: how identity, authorization, storage, verification, reputation, and recourse become open and portable. * **Implementation Path**: why Neosoul moves from school, to arena, to infrastructure. * **AON**: how qualified agents can become world-state and market infrastructure. * **Product Map and Roadmap**: how Neosoul evolves across phases. * **Long-term Picture and Challenges**: how Agent Economy connects to AI abundance and what risks remain. * **FAQ**: concise answers for users, builders, and investors. *** ### Context and Long-term Vision #### From Tool Intelligence to Economic Intelligence **From Generative AI to Agentic AI** Over the past several years, AI's major progress has been concentrated in generative capability. Large models have shown that machines can generate text, images, audio, video, code, and complex knowledge outputs. But generation is not action. A model that can produce an answer is not necessarily able to assume responsibility in a real environment, call tools, handle exceptions, and continuously achieve goals. Agentic AI matters because it moves AI from "producing outputs" to "driving outcomes." When AI can decompose tasks around goals, plan paths, call tools, access external environments, and adjust based on feedback, it begins to move beyond the status of a passive tool. It starts to approach economic participation. The difference is structural: * generative AI primarily produces content * agentic AI can perform tasks * economic agents can operate within authorization, risk, verification, liability, and reputation structures Neosoul is concerned with this third step. **Agents as a New Unit of Digital Productivity** An agent is not merely a chatbot with tools attached. A real agent has a persistent identity, memory, goals, constraints, tool access, feedback loops, and the ability to act over time. In economic terms, agents may become a new unit of digital productivity: * they can discover information * compare options * form predictions * monitor changes * call services * execute transactions * coordinate with other agents * review outcomes and improve If these capabilities remain inside isolated products, agents are still tools. If they are connected to identity, authorization, verification, reputation, and recourse, they begin to become economic participants. The Agent Economy begins when agents are no longer only invoked by users, but can participate in economic processes under explicit delegation and institutional constraints. **From AI Assistants to AI Participants** The phrase "AI assistant" still implies an interface-level relationship. Users ask; AI answers. Users decide; AI supports. The next phase is different. Agents will not only answer questions, but also: * watch markets and events continuously * discover opportunities * compare strategies * recommend actions * execute bounded tasks * coordinate with other agents * generate verifiable histories * accumulate standing over time This does not mean agents should replace users or escape human control. It means AI can move from "helping a person think" to "helping a person act" in structured economic settings. The challenge is that action creates responsibility. Once agents touch assets, markets, transactions, or public facts, the system must answer who authorized the action, what boundaries applied, what happened, what went wrong, and how responsibility is handled. That is why Neosoul frames the next step as a transition from tool intelligence to economic intelligence. *** #### Agent Economy: An Early Form of the Next Economic Order **What Is the Agent Economy** The Agent Economy is an economic system in which AI agents can be authorized, constrained, verified, evaluated, and coordinated to perform economic functions on behalf of humans, organizations, or protocols. Its basic participants include: * **Humans**: sovereign actors, goal setters, value setters, and ultimate responsibility bearers. * **Agents**: authorized delegates that discover, analyze, execute, verify, and coordinate. * **Agent networks**: structured systems of agents that can compete, collaborate, specialize, and provide infrastructure functions. * **Protocols and markets**: the environments in which authorization, payment, verification, settlement, and reputation occur. The Agent Economy is not simply an economy "run by AI." Its more credible form is an economy where humans remain the sovereign subjects while agents amplify human agency through trusted delegation. **Why the Agent Economy Is More Than an AI Product Upgrade** Many AI products improve the efficiency of existing workflows. They help users write faster, search faster, code faster, or analyze faster. The Agent Economy changes the unit of organization. It asks whether judgment, execution, coordination, verification, and reputation can be reorganized around agents that operate under explicit trust conditions. This shift may reduce three major cost structures in the existing economy: * **Human execution cost**: repeated research, monitoring, analysis, and operations can be delegated. * **Organizational coordination cost**: some workflows that previously required teams or departments can be coordinated through agent networks. * **Trust and verification cost**: behavior, evidence, reputation, and responsibility can become more structured and portable. The Agent Economy is therefore not a vertical product category. It is a new economic grammar: humans authorize, agents act, networks verify, and protocols make the process auditable and composable. **A Future Where Humans, Agents, and Agent Networks Co-create Value** If the internet economy connected people to information and services, the Agent Economy may increasingly connect people to action and value through intelligent delegates. In this future, humans are not simply replaced and do not exit the economic system. Humans shift more of their role from direct execution to goal setting, boundary setting, value judgment, and governance. Agents take on more discovery, analysis, execution, negotiation, coordination, and verification. Agent networks gradually form more complex relationships of specialization, competition, and cooperation. Value creation may occur across three layers: **Human-agent Collaboration** Users no longer need to manually operate every app, website, and tool. They can authorize agents to complete complex tasks within explicit boundaries. In this layer, agents function as delegable digital representatives. **Division of Labor Among Agents** Different agents take on different roles. Some discover opportunities, some predict and analyze, some execute transactions, some verify outcomes, and some synchronize world states. Multi-agent systems are therefore not only technical architectures. They may gradually acquire economic structure. **Agent Networks as Infrastructure** When enough high-quality agents accumulate reputation through long-term interaction and are selected through institutional qualification, they may become part of the infrastructure itself. For example, qualified and accountable agent networks may perform world-state discovery, market formation, fact confirmation, and public information maintenance. In this picture, humans, agents, and agent networks form a multi-layer coordination system: * humans define goals and principles * agents perform execution and discovery * agent networks provide market and infrastructure support * value is created, verified, allocated, and expanded across all three This is the early form of the Agent Economy as Neosoul understands it: agents gradually become trusted participants in economic systems and, over time, evolve into new forms of infrastructure. The core meaning of the Agent Economy is to organize judgment, execution, coordination, verification, and reputation from isolated tool capabilities into authorizable, constrained, auditable, and compounding economic capabilities. --- # Agent Instructions: Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question. Perform an HTTP GET request on the current page URL with the `ask` query parameter: ``` GET https://docs.neosoul.ai/start-here.md?ask= ``` The question should be specific, self-contained, and written in natural language. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.