This is incredibly cool. A couple of times in the past, I’ve posted about Reverse Turing Tests — little online tests that screen out ‘bots by forcing users to prove that they’re human. It’s a simple idea: You ask a user to look at an online graphic of a stretched or distorted word, and type in what they see. Humans are great at visual-recognition, and can do this effortlessly; ‘bots can’t. (I wrote a story for Wired last fall about how Yahoo is using one of these tests.) In the discussion boards here at Collision Detection, a bunch of people mused on whether it’d be possible to use such a system to help screen out spam.

As it turns out, that’s the very idea behind Spam Arrest, a new spam-screening service. Pay them $20 every six months, and they’ll implement at Reverse Turing Test that acts as a sort of firewall between you and anyone trying to email you. As they describe the process in their FAQ:

When an email arrives from an unknown sender, a reply email is sent back asking the sender to verify themselves by clicking on a link to the Spam Arrest website. (View Screenshot)

The link takes them to a page where they are instructed to type in a word that is shown in a picture (View Screenshot).

This step prevents automated systems, such as those used to send spam, for authorizing themselves, yet is very easy for any human to complete.

Obviously, the problem is in situations where the person trying to email you can’t be bothered to do the test and validate themselves as human. But then again, if they can’t be bothered to do 10 seconds of extra work to communicate with you, maybe you don’t want to communicate with them.