Media sites are incredibly good at keeping us online by recommending more articles, videos or songs for us to consume based on our online history. Netflix uses as percentage to show me how likely I am to like a certain show or film. YouTube is constantly showing me other videos and channels I might be interested in. Spotify even makes me a new playlist ever week of songs they think I might enjoy.

This has occasionally been the way I’ve found a band, film or tv show that I’ve grown to love all thanks to tried and tested algorithms. But I’m often left wondering how they determine this recommendations, as many are things I’m simply not interested in.

For example, Netflix has a simple rating system of either a thumbs up or down. This allows me to quickly show whether or not I enjoy the film or show, but it doesn’t tell them why I liked it. While I like the occasional romantic comedy with solid jokes and likeable characters, I don’t love everything the genre has to offer. So when Netflix keeps recommending shows like Gossip Girl and The Vampire Diaries, I can kinda see how they got there, but it gets a little irritating.

This problem could be cause by a number of things. Maybe the Australian Netflix library is still growing, meaning they’re struggling to make brilliant recommendations. Perhaps I’m too picky with my taste in film, and it’s too complicated for a computer to understand. Whatever this issue is, the technology is interesting to consider, because when they make the right recommendations, they really get it right.