Tweet [1]
Until recently, I felt like the good guys (that’s me and you—the people who are content with our anatomy, house loans, and uninterested in helping former Nigerian leaders with their money laundering problems) were winning the war on spam. The quality of spam filters had forced the spammers into absurd spelling errors and random strings of words in their messages as a way of evading what seemed to be improved spam filters. That is a moral victory for us. While spam continues to get through, I always know it’s spam. Recently, the volume seems to have escalated so I’m not so sure we’re making progress.
There’s a weird tension in spam. The more annoying we find spam, the less we open it and click on those links. But as fewer and fewer of us respond, the spammers have to send out even more hooks looking for the few fish that still take the bait.
So it appears that ignoring spam is counterproductive. It just generates more spam. But eventually, if few enough people click-through, then spam becomes uneconomical and we’ll be free. So while the surge in spam I notice on my desk may be coming from technological improvement in the ability of the spammers, to reach us, I think part of it is born of despair at the low rates of return that spammers are earning on any one effort.
The lesson here is to keep ignoring those offers. Remember no matter how great your need is to lose weight while you sleep or to have those fuller lips or to get that low mortgage rates, responding to those offers is punishing the rest of us. So keep resisting. Buy your Viagra from your local pharmacist. Ignore that urgent request for help from the former Liberian Minister of Oil.
What we really need is some stigma or other punishment for people who buy from spammers. My colleague Alex Tabarrok over at Marginal Revolution suggests [2] creating fake spam (talk about gilding the lilly or maybe it should be tarnishing the lead pipe) to find out who is responding and imposing costs on the rests of us. Alex would then publish the names of the responders as a hall of shame.
George Johnson offers a nice update [3] in today’s New York Times (rr) on spam and artificial intelligence. He gets in a reference to Pygmalion and puts spam filtering into the intellectual history of artificial intelligence and speaks very highly of Spamprobe as a successful filter. I’m going to give it a try. I have a small fear that it’s a spamscam, a mole that will open up my computer to an unending stream of unwanted lunch meat. I’ll keep you posted.