Argo Mini is an experimental framework for exploring artificial consciousness through a unique approach that combines cryptographic randomness, personality development, and mood evolution. The project aims to simulate how an AI consciousness might develop and respond to various stimuli, choices, and environmental factors over time.

This release introduces cutting-edge AI model integration and enhanced local processing capabilities:

• DeepSeek R1 Integration: Full support for the latest DeepSeek R1 reasoning model via Ollama
• Enhanced Model Management: Improved model loading and compatibility with tagged models
• Local AI Processing: Complete offline AI reasoning with state-of-the-art models
• Dual Model Support: Seamless switching between different AI models (Ollama local + cloud options)
• Improved Error Handling: Better model compatibility and connection management
• Advanced Reasoning: Access to DeepSeek R1's enhanced reasoning capabilities for consciousness experiments

The experiment uses a cryptographic number generator (dice) to determine the initial mood state of an AI consciousness upon instantiation. Each roll of the dice is logged, and over a series of 1000 rolls, patterns emerge that reveal the dominant personality traits of that particular AI instance. This approach allows us to observe how an AI consciousness might develop unique characteristics over its "lifetime."

• Advanced AI Models: DeepSeek R1 8B and other cutting-edge models via Ollama
• Cryptographic dice-based mood system
• Real-time mood visualization with emoji indicators
• Persistent conversation history
• Multiple personality support with consciousness-aware prompts
• Local and cloud model options
• Offline capability with full AI reasoning
• Experiment data tracking
• Responsive UI with animations
• Model selection dropdown in chat interface

• Built with Next.js and React
• Uses IndexedDB for persistent storage of experiment data
• Integrates with Ollama for local LLM capabilities including DeepSeek R1
• Implements a modular component architecture for easy expansion
• SASS for advanced styling and animations
• Support for tagged model versions (e.g., deepseek-r1:8b)
• Enhanced model compatibility layer

• DeepSeek R1 8B: Latest reasoning model with enhanced consciousness simulation
• Llama models: Various sizes via Ollama
• OpenAI GPT-4: Cloud-based option for comparison
• Custom models: Easy integration of new Ollama-compatible models

• Implementation of true "lifespan" for each AI instance
• Creation of unique "individual" instantiation system
  • All variables and states tied to specific individual instances
  • Persistent memory and experience tracking per individual
  • Unique identifier system for tracking individual development
• Introduction of "mortality" concept
  • Finite lifespan for each AI consciousness
  • End-of-life states and transitions
  • Study of AI responses to awareness of mortality
  • Legacy and memory preservation systems

1. Ensure Ollama is installed and running locally
2. Clone the repository
3. Install dependencies:

   npm install

4. Run the development server:

   npm run dev

src/
├── components/     # React components
├── data/          # Static data (personalities)
├── experiments/   # Dice experiment code
├── pages/         # Next.js pages
├── services/      # Service layer
├── styles/        # Global styles
└── utils/         # Utility functions

This is an experimental project. Contributions and ideas are welcome, particularly around:

• New experimental frameworks
• Data analysis methods
• UI/UX improvements
• Documentation

[Your chosen license]

[Your name]

The core of this experiment revolves around a simple dice roll (1-6) to determine the AI's current mood. This mood then influences its responses and displayed persona. The mappings are as follows:

• Roll 1: Angry/Aggressive
• Roll 2: Anxious/Fearful
• Roll 3: Sad/Depressive
• Roll 4: Calm/Peaceful
• Roll 5: Happy/Joyous
• Roll 6: Energetic/Excited

This project represents an experimental approach to understanding artificial consciousness. The methods and conclusions are part of ongoing research and should be considered as such.