IMO the strongest applications are blockchain-related, but here are some other ideas which have been suggested - - One can selectively disclose certain aspects

IMO the strongest applications are blockchain-related, but here are some other ideas which have been suggested - - One can selectively disclose certain aspects of their identity. E.g. given a commitment to my DNA, I can prove that it doesn't match that of a suspect. (Though the police need some way to trust the commitment, perhaps a signature from the firm that sequenced it.) - We can build trustless games, like an implementation of battleship where cheating is cryptographically hard. - We could have trustless computation services, like ec2, or like SETI@home, etc. Though it's difficult to make this practical given the computational overhead of ZKPs. - ZKPs can turn any delay function (like an iterated hash) into a verifiable delay function (VDF). VDFs could enable trustless lotteries/sortition. - Given a general-purpose ZKP, it's trivial to build a signature scheme, e.g. by proving knowledge of a hash pre-image. This can sometimes lead to very practical signature schemes, such as Picnic, one of the leading hash-based signature proposals. Ring signatures are another generally-useful application. - I think ZKPs have also been used as a building block in some FHE and MPC systems, though I don't know much about those areas. - One can prove the existence of an exploit before divulging the details. This might be useful for e.g. convincing everyone to disable a certain TLS primitive which was just broken.