fluran
Project OwnerLead Developer & AI Architect. Builds Acropolis OS. Social scientist pioneering AI for resilient communities. Tech lead, knowledge graph & agent expert.
Civilizations collapse from internal social fabric failure, not external forces. Acropolis OS provides open-source tools for effective community and organization self-organization, tackling this issue. By shaping community ontology, Acropolis OS establishes foundational identity, enabling cohesive structure and effective world engagement. Based on ontology-driven knowledge, Community AI Agents offer context-aware, adaptive services, enhancing member efficiency and information access. Autonomous Community AGI, Acropolis OS aims to foster seamless coordination, maximizing symbiotic intelligence – human and AI fusion for resilient, self-governing collectives.
New AI service
To represent community knowledge as interconnected relationships enabling advanced querying reasoning and discovery of insights from community data. It allows the system to understand the context and connections between different pieces of information going beyond simple keyword search to uncover deeper patterns and relationships within the community's collective knowledge.
Structured data extracted from community documents communication transcripts and agent profiles. This data is transformed into nodes and relationships within the Neo4j graph database based on the community's ontology.
A rich knowledge graph representing the community's information. This output is used by the Community AI Agent for context-aware question answering information retrieval task routing and proactive support services. It provides structured knowledge for AI-driven insights and decision making
We are using the SNET Semantic Similarity Binary service to analyze text snippets within the Acropolis OS knowledge base and community interactions. This allows us to identify semantically similar user queries group related discussions or documents and improve the accuracy of information retrieval and question answering by understanding the meaning of text beyond keywords.
We will build the framework for community ontology creation, develop a functional knowledge base prototype using a vector database (like ChromaDB), and create a basic data ingestion pipeline from Discord. Crucially, we will develop a Community AI Agent prototype with semantic question answering capabilities, integrating a cost-effective LLM and RAG with the knowledge base, and integrating it into Discord for initial user interaction. The aim is to deliver a functional, albeit initial, symbiotic intelligence system encompassing the core data structures and the primary AI service.
Functional framework for community ontology definition (initial guidelines/tooling). Working prototype of knowledge base using vector database (ChromaDB) populated with Discord data. Basic data ingestion pipeline from Discord (proof of concept). Simple keyword-based search functionality within the knowledge base. Community AI Agent prototype with semantic question answering capabilities (utilizing RAG and LLM). Integration of AI Agent with Discord for user interaction. Real Time Interactive Rag AI assistant
$12,000 USD
Demonstrate a functional Acropolis OS prototype featuring a knowledge base populated with Discord data and searchable by keywords, a foundational ontology framework, and a Community AI Agent integrated with Discord capable of semantic question answering relevant to the community data. This validates the core symbiotic intelligence concept and infrastructure feasibility.
This milestone shifts focus to delivering tangible, functional AI Agent services that actively assist community members in key areas of organization, information management, and crucially, social coordination. We will develop non-autonomous, conversational AI Agents specifically designed to provide practical assistance and to proactively enhance connections and coordination within the community. These agents will leverage semantic and vector similarity search, powered by the relational data within the Neo4j knowledge graph, to deliver intelligent services that not only retrieve knowledge but also activate social interactions. The agents will be designed to help users with tasks like organizing documents and information, transcribing content, providing advice, and importantly, actively recommending relevant individuals to connect with, and suggesting appropriate resources (human, material, financial, intellectual) based on user needs and community availability. The emphasis is on building agents that are active facilitators of community interaction and resource flow, maximizing internal social capital and effective allocation.
Development of at least three functional, non-autonomous Community AI Agents providing specific services. Examples now including social coordination: Document Organization Agent: (As before) Assists with document categorization/tagging. Information Retrieval & Advisory Agent: (As before) Conversational QA, information retrieval, and advice provision. Social Connector & Resource Recommendation Agent: This new agent focuses on: Recommending individuals: Based on user profiles (skills, interests) within the knowledge graph, it will suggest relevant people for users to connect with for collaboration, mentorship, or expertise sharing. Resource Recommendation: Based on user needs and project requirements (extracted from conversations or queries), it will suggest potentially available human, material, financial, or intellectual resources within the community (as defined in the ontology and knowledge base). Integration of these functional AI Agents within the Discord platform for user interaction and testing. Demonstration of AI Agents leveraging semantic/vector similarity search and Neo4j knowledge graph for both knowledge retrieval AND social connection/resource recommendations. Basic user interface elements for interacting with the functional AI Agents within Discord, including features to view recommended connections and resources.
$20,000 USD
Demonstrate at least three functional, non-autonomous Community AI Agents operating within Discord, providing tangible services including document organization, information retrieval/advice, and crucially, demonstrable social coordination functionality by recommending relevant individuals and resources. Success is measured by the agents’ ability to effectively leverage semantic/vector search and the knowledge graph to provide useful assistance, facilitate connections, recommend resources, and receive positive initial user feedback on agent functionality, ease of use, and the value of social connection and resource recommendations.
Milestone 3 explores "Autonomous Community General Intelligence" via society simulation within Acropolis OS. We prototype an augmented, simulated society to enhance collective decision-making. This involves designing a simulation with hundreds of agents, each representing a simulated human with a basic knowledge base. Agents use LLM-driven logic, based on community ontology, for simulated societal processes. We'll integrate real human input, valuing human guidance to create symbiotic intelligence at the community scale for better self-governance. We'll explore self-correction, inspired by relevance realization, using human feedback to refine the simulated society. Simulation areas: communication, norms, responses to challenges, governance impacts. This is a proof-of-concept to demonstrate the feasibility of using society simulation and agents to understand and improve real-world collective decisions. We'll also refine metrics for evaluating "community intelligence" within this simulated social context and impact of human integration.
Design doc for society simulation: agent types, simulated human knowledge, rules, interaction, human input mechanisms. Prototype of simulated society with hundreds of agents, each with a rudimentary knowledge base. Demonstration of agent-driven collective decision-making on scenarios. Mechanisms for integrating real human input into the simulation. Exploration of self-correction using human feedback, inspired by relevance realization. Refined metrics framework to evaluate "community intelligence" in simulation & human impact.
$18,000 USD
Demonstrate proof-of-concept for an "augmented, simulated society" in Acropolis OS. Show agents representing individuals, engaging in collective decisions. Evidence of individual agent knowledge, human input integration, and initial self-correction attempts. Refined metrics framework to evaluate simulated society "intelligence". Success measured by demonstrating conceptual validity and technical feasibility of society simulation for enhancing real-world decision-making.
Reviews & Ratings
Please create account or login to write a review and rate.
Check back later by refreshing the page.
© 2024 Deep Funding
Join the Discussion (0)
Please create account or login to post comments.