I have quickly became fascinated with this idea. If developing the actual learning algorithms was an interesting challenge, the usefulness and simplicity of the outcome was pushing me to go for it.

A discovery engine for music, movies and books is something that the Internet, despite its complexity, hasn’t been covering yet, and there are so many people always willing to find some new cool band, to watch an interesting movie or to read a good book.

But “cool”, “interesting” and “good”, when it comes to art, are very subjective appreciations, largely based on one’s current taste. And if we think about how vast the offer is out there, I find that a discovery tool can be very helpful to so many people; a discovery tool that helps you “explore your taste” (this later-on became Taste Kid’s slogan).

Note: This post has been written in retrospect and posted on September 10th, 2008.

Although Emmy’s birthday is in January 2008, like with any other child, she was conceived some months earlier. I don’t recall exactly when and how the idea of this project emerged in my head, but I can tell it was sometime in the fall of 2007 and it wasn’t premeditated, meaning I haven’t been planning to begin working on a personal project.

I won’t go in too much technical details here, but the basic idea was to develop a recommendation system that learns about human tastes from the web. This potential hasn’t been explored before, at least not to my knowledge or not in the way I was about to do it.

Note: This post has been written in retrospect and posted on September 10th, 2008.