Teenage-AGI provides an AI agent with persistent memory and a "thinking" capability, inspired by projects like BabyAGI and the "Generative Agents" paper. It's designed for users interested in exploring AI memory and conversational continuity, enabling agents to retain information and context across sessions.

How It Works

The agent leverages OpenAI for language processing and Pinecone for vector storage. Upon receiving a query, it vectorizes and stores it. It then retrieves relevant past memories and queries from Pinecone, "thinks" about the next action, stores this thought, and generates an output based on the thought and retrieved memories. Both the query and its answer are then stored in its persistent Pinecone memory.

Quick Start & Requirements

• Install dependencies: pip install -r requirements.txt
• Configure API keys: Create a .env file from .env.template and add OpenAI and Pinecone API information.
• Run: python main.py
• Docker: docker-compose run teenage-agi
• Requires OpenAI API access and a Pinecone account.

Highlighted Details

• Persistent memory via Pinecone, allowing agents to retain information across sessions.
• "Read" and "Think" commands enable direct memory insertion and agent contemplation.
• Demonstrated ability to remember characteristics and maintain conversation without a traditional context window.

Maintenance & Community

The project is maintained by seanpixel, a student founder. Contact is available via Twitter.

Licensing & Compatibility

The repository does not explicitly state a license in the provided README. Users should verify licensing for commercial or closed-source use.

Limitations & Caveats

The project is experimental, with the author planning to add a context window. The lack of an explicit license may pose compatibility issues for certain use cases.