Wondering whether to throw in your job, buy a new car or propose to that girl in accounts? Well now there’s a web site that can help you navigate such tricky decisions.

Called hunch.com, it’s the brainchild of Flickr co-founder Caterina Fake and a group of MIT computer students. It uses a clever combination of user input and algorithmic calculations to determine the most appropriate answer to any question you enter.

The process starts with a series of multi-choice questions designed to let the site learn something about you, your tastes and outlook on a range of topics. These vary from obvious ones such as “where do you live?” to the more intriguing “do you sing in the shower?”

The theory is that, using your responses, the site determines which other users you are most like. Then, when you ask a question, responses from like-minded people will be presented as recommended answers? Make sense? Sort of...

It seems the site will get better over time, thanks to the input of users and the clever way in which they are linked together to provide answers to new questions.

The business angle behind the site comes from the fact that, if you ask “what notebook PC should I buy”, it will suggest a suitable model and then display ads from retailers offering that particular item.

Hunch.com may not be able to give you a definitive answer on whether you should confront your boss or sell your share portfolio, but it’s probably at least as accurate as asking a bunch of mates.

In any case, it gives a glimpse of what might be possible in the future as collective knowledge is harnessed and enhanced with some smart mathematics.

Give it a try and let me know if it works for you.