Shelf Life of Spam

A thought: after some really bad world-as-we-know-it-ending cataclysmic event—huge meteors, multiple nuclear explosions, simultaneous and persistent hurricanes and tornadoes, The Rapture, whatever; I'm going for that Twilight Zone-y, Could I be The Last Person on Earth plot here—after the disaster, if someone were still alive, and still had access to power, s/he would at some point try to use a computer, connected to whatever might remain of the communications grid, to search for other survivors. Perhaps enough wiring would be intact that someone somewhere else who also had access to power and a computer might still be on the "internet" in some way.

Maybe such a scenario is laughably impossible (either a lot of people are left and maintaining things, or there is just no way there could still be any hope of using a severely damaged and unstaffed power grid or communications network?), but humor me anyway. So how would our Last Person on Earth go about searching for that other survivor? I am sure there are sophisticated ways of "listening" for activity over the internet. But perhaps our lone survivor wouldn't have the knowhow or software tools. What the loner would need would be an automated way of sending numerous messages to a large number of potential addressees, in order to knock on a lot of doors at once. In short, s/he would need a spamming program and a spammer's database of "good" email addresses. And a legitimate-looking From address and a catchy enough Subject line that a recipient wouldn't ignore or auto-delete the message. After all, if the network were working even partially, it seems likely that pre-cataclysm viruses and spam-relays might still be churning out the spam with no supervision, so there would probably be a lot of competition.