The Community-Driven Future of AI Training Data

AI progress has been gated by proprietary data. We are changing that.

The Data Problem

Big tech companies have vast training data. Open-source projects do not. This creates an uneven playing field.

Our Approach

We believe in community-driven datasets:

• Anyone can contribute
• Data is open for everyone
• Quality through collective effort

What We Are Building

A dataset of tool-use interactions contributed by:

• AI developers sharing logs
• Researchers contributing benchmarks
• Community annotators ensuring quality

Why It Matters

When training data is open:

• Anyone can build powerful AI
• Innovation accelerates
• No vendor lock-in

Join Us

This is not a top-down project. It is a community effort. Your contributions matter.

Whether you have data to share, time to annotate, or skills to test models - there is a place for you.

The future of AI training data is open. Be part of it.