differential allocation of compute and hashpower of telic mining pools will automatically inform knowledge graph link prediction where manual compute configuration is only needed on the boundary of the pools simulated understanding