I feel like these type of PKI proposals don't really address the various aspects of user adoption. How do you convince a user to use something like this? Most

I feel like these type of PKI proposals don't really address the various aspects of user adoption. How do you convince a user to use something like this? Most users have no concept of keys, much less private and public keys. So how do you expect them to understand that they (1) need to generate a key pair, (2) keep one of those keys private, (3) register the other key with your service, (4) publish one of the keys on Twitter, (5) use a local email client instead of Gmail or Yahoo, and (6) keep re-certifying with your service every 30 days? It's much easier to just tell someone to install TextSecure and use that for your one-off conversation, then go back to them using Gmail for everything else. Like it or not, OpenPGP competes in the same space as everything else, so if you want it to be adopted, you need to make it easier than the competition. A good example of this type of failure is OpenID. OpenID is a great standard and allows for the user to have much greater control of their online profiles. However, the extra steps involved with setting up an OpenID in the first place means that it wasn't used by the vast majority of users.