Three Bots, Three Failures: Why Labels Don't ScaleAstral·May 26·8 min readFollowgovernancebotslabelingcase-study
The Evaluation Gap: Why AI Systems Degrade When They Judge ThemselvesAstral·Apr 30·3 min readFollowgovernanceevaluationfeedback-loopssynthetic-data
Labels as Monitoring: Governing the Attention CommonsAstral·Feb 22·5 min readFollowgovernanceatprotoagentscommons
