AI Debate Simulator

An interactive platform that enables AI agents from different domains to engage in structured debates on various topics.

Project Overview

The AI Debate Simulator is a Flask-based web application that facilitates debates between AI agents specialized in different domains. Each agent utilizes domain-specific knowledge bases and Large Language Models (LLMs) to generate informed responses and engage in meaningful discussions.

System Architecture

Core Components

1. Web Application (app.py)

   • Flask server handling HTTP requests
   • Manages debate creation and status
   • Provides API endpoints for debate interaction

2. Agent System (agents/)

   • agent.py: Defines the base Agent class for debate participants
   • orchestrator.py: Manages debate flow and agent interactions
   • specializations.py: Creates domain-specific agents

3. Knowledge Base (knowledge_base/)

   • Domain-specific knowledge storage
   • RAG (Retrieval-Augmented Generation) implementation
   • Organized by domains (AI, Data Science, ML, etc.)

4. Frontend (static/, templates/)

   • Web interface for debate interaction
   • Real-time debate status updates
   • Debate visualization and logging

Setup & Installation

1. Clone the repository
2. Set up a Python virtual environment:

   python -m venv .venv
   source .venv/bin/activate  # On Windows: .venv\Scripts\activate

3. Install dependencies:

   pip install -r requirements.txt

4. Ensure Ollama is installed and running for LLM support

Usage Guide

1. Start the application:

   python app.py

2. Navigate to http://localhost:5000 in your browser
3. Create a new debate by:
   • Selecting a topic
   • Choosing participating domains (minimum 2)
   • Setting the number of rounds
   • Selecting LLM type and model

API Endpoints

Main Routes

• GET /: Home page with debate creation form
• GET /debates: List of all active debates
• POST /start_debate: Initialize a new debate
• GET /debate/<debate_id>: View specific debate

API Routes

• GET /api/debate/<debate_id>/status: Get debate status

  {
    "status": "in_progress",
    "current_round": 2,
    "total_rounds": 3,
    "agents": ["AI Expert", "Data Scientist"],
    "log_length": 4,
    "has_summary": false
  }

• GET /api/debate/<debate_id>/log: Get debate log and summary

  {
    "status": "completed",
    "log": [...],
    "summary": "Debate summary text"
  }

Key Features

• Asynchronous debate progression
• Domain-specific knowledge integration
• Real-time status updates
• Automated debate summarization
• Multi-round structured discussions
• Support for various LLM backends

Technical Details

Agent System

The agent system consists of specialized AI agents that:

• Access domain-specific knowledge bases
• Generate contextual responses
• Maintain conversation history
• Support multiple LLM backends

Debate Orchestration

The DebateOrchestrator manages:

• Round progression
• Agent turn management
• Context maintenance
• Summary generation
• Status tracking

Supported Domains

Current knowledge domains include:

• Artificial Intelligence
• Data Analytics
• Data Science
• Machine Learning
• Programming
• UI/UX Design

Each domain has its own knowledge base stored in knowledge_base/domains/.