In this months issue of Wired Magazine (March 2008, P 114) there is an article about “The Nitflix Challenge.” Basically you can win one million dollars of you can write a recommendation engine the is ten points better the their current “Cinimatech.” As is usual when it comes to technology, people jumped at the chance and made huge leaps for the first while but their innovation has slowed to a crawl. The current leader is at about 8.5% which is good but not worth any money according to the contest rules! All the systems use the movies you have in your queue and the ratings you have given movies you’ve watched to recommend new ones to you..Anyways none of this has anything to do with the issue I’m thinking of. There was no mention of anyone taking into account households that have more then one person rating movies. We had a Zip.ca account for a while, the Canadian Netflix alternative, and both Michelle and I added movies to the queue and rated them. We both have different taste in movies, some overlap but we also compromise and watch movies one of us like and the other isn’t into. We can’t be the odd ones out here!.Wouldn’t this look schizophrenic to the system? Can an engine take care of this in a simple way? Or do they have to add multiple managing users to an account and have the system recommend different stuff to each user? What do you think?.