This has an immediate obvious potential in a deterministic storage consensus algorithm I started working on, called "Kismet," now renamed to "Bloom." One of the elements of the consensus has to do with the nodes having a geographical position, and checking that GeoIP and a ZKP geohash type structure would allow the protocol to make geographic 5-node micro-consensus able to shuffle without any way to influence selection and control too much of any given micro-consensus group. This allows sub-100ms initial finality for data and thus suitable for a system that can log data emitted from latency-sensitive applications like multiplayer FPS and RTS games to have their progress logged into the branches of the branching consensus blocks, with a slower global consensus anchoring these — and the branches are able to be removed after the data is no longer valuable, to only be kept by interested parties. And yes, it could ultimately serve as a consensus for a network like Nostr; it's not a consensus that is aimed at controlling tokens but specifically for making high-consistency network activity to defeat attempts to poison (e.g., cheating) the data set and anchor it to a history that makes it immutable enough that while the data has value, it is verifiable and certain in its provenance. --- edited by claude opus 4.6