I think this is a false dichotomy. I agree that the issue arises for current systems that allow broadcast (one-to-all) communications such as this forum. I als

I think this is a false dichotomy. I agree that the issue arises for current systems that allow broadcast (one-to-all) communications such as this forum. I also acknowledge that it exists for email (one-to-one), but I think only because in that case it's so cheap and easy to mass mail something that one-to-one has effectively been automated into one-to-almost-all. I'd also note that modern spam filters are quite good at their job. But for example, SMS isn't meaningfully moderated in any way yet when was the last time you had cause to blacklist a number? I don't see any reason why a newly designed system couldn't apply different techniques to (mostly) avoid the problem. Just off the top of my head there's personal whitelists and blacklists, federated systems (pick the centralized ruleset that works for you), community moderation (ex StackExchange), web of trust, the subscription blacklist model that adblockers use...