Social theory...

as in, doing theory socially. Three points of interest:

1. Luca Trevisan is soliciting ideas for building a schedule/recommendation engine for FOCS using collaborative filtering. He's on a short time frame, so you need to know what you're doing, but I dare say there's an excellent ALENEX submission waiting for anyone who has the time to work on this. 
2. Anand Kulkarni is proposing a theory overflow site, much like Math Overflow (which many of you inhabit already). I've been relatively happy with MO, and they're quite friendly towards algorithms folks (although sometimes a little confused about the difference between theoryCS and programming). But I do often tire of wading through pages and pages of unrelated questions to get to interesting ones.
   
   I don't know if there's enough global support for theory overflow, but I do know that MO has been a fantastic resource for research-level mathematics, and with enough participation, theory overflow could get there too. If you don't know what I'm talking about, go to mathoverflow.net. If you think that it's a waste of time, I'll mention that among the ACTIVE participants there are Terry Tao, Timothy Gowers, Richard Stanley and Bill Johnson (as in Johnson and Lindenstrauss) 
3. Mark Reid has built a discussion site for ICML 2010 (ICML has been doing this for a few years now). Each paper at the conference gets a page, and anyone can post comments on the page. Authors can opt to get email whenever someone posts a comment, and can in this way interact with discussants. I wonder if something like this might soon become a de facto part of all conferences.