Several years ago, I had an idea to use Stanley Milgram's ideas on social distance (see his Small World experiment for an example) to help fight spam.

The idea is a simple one: Spam is a relative term. One person's spam is another person's valid email. Similar groups of people will have similar opinions on spam. And just as you trust your friends' opinions, and base that trust on their similarity to your own and a basis in fact, you can do the same with spam.

I spent about a year tooling around with the concept on and off in spare moments, and then spent a fair amount of effort getting a proof of concept working, getting the community to notice it, and building momentum behind the idea. Sadly, I took a job that required me to stop development on the project, so stop I did.

This isn't the place to explore the entire concept in depth. Instead, head on over to Sufficiently Advanced Technologies, Bitshift's sister-site of sorts, and read all about it, as well as download working code.