Here is a good examination of the state of the Netflix prize: 300 days later [1].

So what I get from the Netflix prize is that there are probably significant limits to recommender systems. Even the smartest don't do a whole lot better than the simple approaches, and a lot of work is required to eke out even a little more actual information from the morass of data. It seems surprisingly difficult to get reliable, factual information on this important question of how useful they can be. Part of the reason is that they are new - Amazon has only been in business for about ten years after all - and part of the reason is that the behaviour of these systems is often a closely guarded secret despite the aura of openness that web companies cultivate.

Basically, it looks like the further improvements are getting harder and harder to find and that the prize cannot be won. The problem is the wacky data. People are very inconsistent with their ratings. It depends on what mood they are in. People will rate movies they have not seen just for the heck of it. What appears to be one person is actually a household.

At the root of this is the major lack of incentives for a user to be consistent and tell the truth. Sure, you might be able to get a recommendation but the fact is that must of us ignore those recommendations in the same way we ignore TV commercials. Thus, we have nothing at stake when making a recommendations. How do we fix this? Well, here are some fun ideas:

1. We could force the user to watch the movie we recommend, that way he will want to tell us what he really likes.
2. More plausibly, give him a discount on the top-recommended movie. So, if we recommend "Isthar" then he can rent that one for 50% off. Hmmm, I see you already thinking of ways to manipulate this.
3. Give up on this whole one-dimensional rating system and start to think about user's moods and needs. Build a model of the user, his family, his interests, and come up with recommendations for when he is having family time, relaxing time, intellectual time, sexy time, etc. "A movie that suits your mood" could be the tagline.
4. Go the social network way and have friends recommend movies. After all, that is how this whole idea started. Somewhere along the line we dropped the friends. In fact, lets pay these friends (in the form of discounts, of course) to make correct recommendations.