1) I really hadn't thought about recommender systems at much length before. What really interested me was how several things I use every day can be reccommender systems and the realization that I used them as such. Nothing in this article was particularly disinteresting.

2) The mesage of this article wasn't as direct as others. It was a bit of an overview of reccommender systems. It listed various capabilites and approaches as well as some potential. The challenge of linking people with similar preferences had particular emphasis.

3) Much like anyone else, I probe people who have watched a movie to see if they liked it or not. I don't participate in any formal groups along these lines, but from time to time I ask the advice of people I know and trust. The key is that I know what they like and can judge their responses in light of that.

4) The biggest one I use is google. I use it because it works very well. I also spend time on slashdot but go there more for insight into various happenings and community rather than looking for reccomendations. At Amazon, I've also paid attention to their thoughts, but really pay more attention to the written comments of reviewers and purchasers than "you might also like..." features. Reading reviews can give you some insight into the person writing them and what their biases may or may not be.