> To take it one step further I'm convinced at this point that the social conundrum presented by Twitter and other social media platforms has extended far beyon

> To take it one step further I'm convinced at this point that the social conundrum presented by Twitter and other social media platforms has extended far beyond the algorithm and the platforms themselves, and into how people interact in general. A cryptographer once mentioned attacking the PGP WoT by uploading a shadow WoT that has different keys with the same emails/names and signing-graph structure. Similar to what you note-- the real problems come when some bona fide WoT user fucks up and accidentally signs a key from the shadow graph. Even if it's just a single accidental signature before the shadow WoT is spotted and cleaned up, the admin would probably want to quickly go out of band to actually fix the original WoT at that point-- otherwise perhaps the shadow admin got to the user first. :) (And of course if the shadow graph were uploaded incrementally over a wider period of time, the problems probably become a bigger pain in the ass.) The problem you mention is in the same category. But rather than being hypothetical, it is real and probably as bad an instance as someone could design given current technology. To make the WoT attack above comparable, I'm imagining a Debian user accidentally signing a shadow key, then receiving messages from that key that convince them Debian keyring is in fact the shadow key ring and they should use their permissions to destroy Debian. Edit: and I guess in keeping with the analogy, the purpose of the propaganda emails is to generate as much outrage in the recipient as possible, so that they may spend more time reading the entire email on the off chance they also spend one or two seconds extra looking at the ad in the email signature about a special on a nice pair of designer pants.