[jerf]

It turns out to be really hard to have a medium that A: allows people you have not whitelisted in a advance to contact you and also B: prevents intelligent adversaries from abusing that.

There are numerous examples of this: The phone system, the email system, any large IM system that gets beyond a certain size, the physical mail system (reigned in by charges there, though), social networks...

It would be easy if the adversaries were not actively using their intelligence to get around whatever solution you think might work. If all the spam phone calls came from the same number, which never moved or changed, this would be easy. But even before you account for the unauthenticated nature of caller ID that allows spammers to trivially forge any number they'd like, the mere fact that they're willing and able to buy up large blocks of numbers and would switch them around a lot would make this a difficult problem. Stopping intelligent, adaptive adversaries willing to dedicate time and resources to this is a very hard problem.

You can also attack this problem by trying to hit the first clause I gave, but that creates a very different kind of system. And in general, you need some sort of system that you don't have to prewhitelist people, so that your doctor can call you even though you didn't whitelist them, etc.

[bluGill]

I don't actually need to prevent intelligent adversaries from calling me though - I just need a legal way to go after them if they do. That means I need useful identification. Many of these calls are illegal today, but the law doesn't know how to find them so they get away with it.

Anonymous communication is useful, but it has been abused enough that I want to opt-out of it.