I've been using GoogleReader since it first came out in beta and use it daily to keep up to date with various topics such as Sharepoint, .Net development, Enterprise 2.0 and the Latest Technology in general.
As I discover new feeds I add them and tag them to a category. I usually find these by reading referenced posts on blogs I read and then adding the new feed.
Share your OPML web site allows you to upload your entire OPML file (list of all your feeds) and then cit will suggest people with similar OPML lists. You can then crawl through them to find feeds you have yet to discover.
There is a
top 100 list which shows what the most popular feeds are over those who have registered their OPML files which is great to see what you're missing out on!
A very neat tool with plenty of potential for further growth around recommendations. I'm sure it won't be long before something graphical hits the scene here relating to tag clouds and the like. There's plenty of good content out there and making recommendations is a great way to discover it!
It will be interesting to see whether Google Reader can do something similar with all it's users to recommend feeds based on what feeds you've subscribed to and what similar people have. But their is a huge privacy thing around this., potentially though, if you are clicking to share posts, they could simply just recommend based on what you share...