I recently launched a website,

RECOBOT is an artificial intelligence that recommends things to you using a "recommendation algorithm." It uses the sets of items that its users provide to discover statistically co-occurring items. In this way, RECOBOT learns to give better suggestions over time.

Right now RECOBOT doesn't have many submissions, so it is not very good at giving recommendations. I want to find a way to make this site interesting before there are many submissions, and one way to do that is to "seed" the algorithm with data from sources such as news articles or blog posts. But I haven't done that yet. So, consider this an unfinished experiment.