Project Synapse-X: The Self-Improving AI Workshop

Project Synapse-X is a demonstration of a new AI paradigm. Instead of using a single, monolithic model, Synapse-X is a dynamic system that learns and forges new specialist skills on demand in response to user needs. It is an AI that grows its own capabilities over time.

The Vision: From a Toolbox to a Workshop

• Standard AI: A toolbox with a fixed set of tools.
• Synapse-X: A complete workshop with a master craftsman (the Meta-Mind). You bring it a new problem, and it forges a brand-new, custom tool just for that job, adding it to its toolbox permanently.

How It Works: The Skill Forging Process

1. Problem Definition: A user describes a new skill and provides a few examples.
2. Synthetic Data Generation: The Meta-Mind (a powerful local LLM like Llama 3) uses this information to generate a large, synthetic training dataset.
3. Specialist Training: A new, lightweight SpecialistOctave is trained on this data.
4. Skill Finalization (HONet Principle): The trained Octave is frozen and saved to the Skill Library, becoming a permanent, forget-free specialist.

---

How to Run the Demo

Step 1: Install and Run Ollama (The Meta-Mind)

1. Download and install Ollama from https://ollama.com/.
2. Pull a powerful model to act as the Meta-Mind:

   ollama pull llama3

3. Keep the Ollama application or server running in the background.

Step 2: Set Up and Run the Synapse-X Project

1. Clone this repository and set up the environment:

   git clone https://github.com/your-username/Project-Synapse-X.git
   cd Project-Synapse-X
   python -m venv venv
   source venv/bin/activate  # On Windows, use `venv\Scripts\activate`

2. Install the dependencies: pip install -r requirements.txt
3. Run the main application:

   python run.py

Step 3: Live Test - Teach Your AI a New Skill!

Follow these steps in the interactive console.

1. Teach it "Pirate Speak" Copy and paste this entire command into the prompt and press Enter: