As the site comes into public beta, we should start to consider how to promote it amongst the relevant communities.
We can host a competition with some kind of prize (not sure what we can offer, whether something small like a library of the top 10 TCS books or a larger cash prize): solve one of the problems with more than 10 votes from this question: Major unsolved problems in theoretical computer science?.
I've used twitter and my blog (geomblog.blogspot.com) with some degree of success. I'll probably also request specific bloggers (the complexity blog, Dick Lipton etc) to advertise it as well.
Update: Here are some specific examples of posts. I've tried to highlight certain posts from the main site that might catch people's eyes: I'll do this semi-regularly from now on.
I've sent out announcements on our internal mailing lists, and I'll pass it along to a few others in the near future.
I'll also give the site a mention in a few of my upcoming talks. The site is a good example of new ways that the web can positively affect work in theory by speeding progress and letting us rapidly close or open new avenues of thinking, which certainly merits mention at the tail of a talk. This is doubly appropriate in cases where the site helps in investigating a problem.