Community Radar.

“Donate your coding sessions to an open CC-BY-4.0 dataset to help train open-weight and open source models”

Anthropic and Open AI are getting so much data from the Claude Code and Codex usage, and I'm quite scared this will create an oligopoly because only their models will be trained on it, leaving the open-weight and open source models behind. So I'm trying to launch a little initiative called Trace Commons and encouraging people around to donate their coding agent traces into an open dataset [https://trace-commons-web.hf.space/](https://trace-commons-web.hf.space/) so that other model labs can also train on them Let me know if you have any feedback and hopefully we can have a nice open dataset soon !