There was some good discussion about a P/NP for dummies article a while back, here:

http://news.ycombinator.com/item?id=1605415

And by the same author as the story above, a good list of reasons for believing that P probably != NP:

http://www.scottaaronson.com/blog/?p=122