We’ve been quiet. We’ve also been busy.
Over the past months, PublicAI has been building something new: Trajector, a network for collecting coding trajectories. If you’ve followed us through our audio collection era, the logic will feel familiar. We built our name on a simple belief, and it hasn’t changed: AI needs humans. Real human work is the best training data and the best guardrail there is.
What’s changed is where that work lives.
Every time a developer sits down with an AI coding agent, something valuable happens. Prompts, tool calls, diffs, tests, failures, fixes, commits. The full story of how a real problem actually gets solved. AI labs need this data badly, and most of what they can buy today is staged: preset tasks performed for bounties. Meanwhile the real thing gets generated every day, by developers like you, and vanishes into log files.
Trajector changes that. Install once, keep coding the way you already do, and share your trajectories on your terms. Everything is scrubbed on your machine before it goes anywhere. You choose what’s shared, you can pause anytime, and you get rewarded for work you were already doing.
Build. Trace. Earn.
This is also the start of something bigger. PublicAI began as a contributor network, millions strong. With Trajector, we’re growing into a developer community platform, a place where the people who write the world’s code share in the value their work creates. Open source gave AI its education for free. We think it’s time the community got paid back.
Humans built AI. Humans should profit from it.