The internet thinks it knows you. What products you like, what movies, music, what you want to buy next. And it advertises to you accordingly, with all kinds of junk that it thinks you will want. And the internet thinks you've told it all there is to know about you by leaving behind data of where you shop, where you visit, a sort of accidental bread crumb trail.
The website Hunch had a big relaunch this week and it approaches things a bit differently. On Hunch, you create a profile of yourself, either directly or by importing your Twitter or Facebook information, and then you answer questions about yourself. Everything from age and location to what kind of lettuce you like to your thoughts on dolphins. There are tons of questions, you could answer them all day.
Then you ask the site a question, maybe about what car to buy, and it makes recommendations based on who you are.
Hunch was co-founded by Caterina Fake who joins us to talk about what she's trying to get it to do. Fake is also the co-founder of the photo site Flickr, so she knows a thing or two about web collaboration.
We also check in with Andreas Weigend, former lead scientist at Amazon.com, a company known for making recommendations. We ask him about whether data provided by an individual provides a more accurate view of who they are than data gathered surreptitiously.