Incremental consolidation
A consolidation strategy that processes only newly arrived or recently modified experience events rather than scanning the entire memory corpus on each run. Required at scale where full-corpus consolidation is computationally prohibitive. Uses watermarks or change logs to track processing boundaries.
The original consolidation worker scans all experience events on each cycle - workable when the memory store contains tens of thousands of events, prohibitive when it reaches millions. Incremental consolidation replaces this with a change-log approach: each consolidation run processes only events that have arrived or been modified since the previous run's watermark. This keeps consolidation cost proportional to event throughput rather than total corpus size. The tradeoff is that incremental consolidation may miss long-range pattern relationships between new events and old memories that only become apparent across the full corpus. Scaling the Cognitive Substrate series describes hybrid approaches - incremental for routine operation, periodic full-corpus passes for global pattern analysis - that balance cost and coverage.