alt Hacker NewsWhy is no one talking about using zero knowledge proofs for solving this? Instead of every platform verifying all its users itself (and storing PII on its own servers), a small number of providers could expose an API which provides proof of verification. I'm not sure if some kind of machine vision algorithm could be used in combination with zero-knowledge technology to prevent even that party from storing original documents, but I don't see why not. The companies implementing these measures really seem to be just phoning it in from a privacy perspective.
Replies
Zero knowledge proofs are only private and anonymous in theory, and require you to blindly trust a third party. In practice the implementation is not anonymous or private.
Technologists engage in an understandable, but ultimately harmful behavior: when they don't want outcome X, they deny that the technology T(X) works. Consider key escrow, DRM, and durable watermarking alongside age verification. They've all been called cryptographically impossible, but they're not. It's just socially obligatory to pretend they can't be done. And what happens when you create an environment in which the best are under a social taboo against working on certain technologies? Do you think that these technologies stop existing?
LOL.
Of course these technologies keep existing, and you end up with the worst, most wretched people implementing them, and we're all worse off. Concretely, few people are working on ZKPs for age verification because the hive mind of "good people" who know what ZKPs are make working on age verification social anathema.
People are talking about it, at least here anyway.
The reason you don’t see it in policy discussion from the officials pushing these laws is because removal of anonymity is the point. It’s nit about protecting kids, it never was. It’s about surveillance and a chilling effect on speech.