Why we can’t rely on algorithms to make cultural discoveries

I recently tweeted an article on the Guardian by Jonathan Jones justifying why we need art critics. Looking at the buzz around this article there seems to be mixed feeling about his views, but I think he touches on a key point about the wider aspects of our society.

The internet may well have bypassed the need for gatekeepers but without a human content filter, how can we find the diamonds in the rough. While I agree it is an exciting time that opens up opportunities for those that may easily be dismissed. Especially if they are breaking new ground. Most content of value is lost in the noise.

Take for example YouTube, I am sure there are 100s of videos that have been uploaded today of significant merit, but if I am to rely on the collective intelligence of YouTubes viewers based on the most liked videos this week, the number one is Jessie J. Now, Jessie J has a record deal (and marketing budget) behind her so she is going to be popular (she has got through the traditional gate keepers) and people will seek her out. Number two however, is 20yr old Charlie McDonald, he has a huge number of subscribers (clearly in the millions) but without being disrespectful to Charlie, although I am sure what he contributes enriches the lives of many under 20s, it is hardly of cultural value to the wider community.

A new startup was announced recently on reinventing ratings around a stamp of approval rather than the 5* rating systems we commonly see. This is not a criticism of Stamped, as I don’t know (but am keenly interested in) what they are planning. Based on this concept, however, a simplistic rating system just doesn’t work. And this is where critics, journalists and reviewers come in. We will always need the judgements of trusted experts to sort the wheat from the chaff and introduce us to art that enriches us, that would otherwise be lost in the long tail of content.

I am not sure how easy it is for an algorithm (at this time) to make these subjective decisions on our behalf. To make these judgements, we need a background understanding of the domain/subjects history as well as taste. Many sites may have recommendation systems that try to match other content that is popular and based on what you yourself have liked but I have not found one that works well. Even if we rely on our social connections, I am not convinced we could find the next Coen Brothers, for example, on YouTube or Vimeo.

There is therefore an exciting opportunity that the internet does open up. It is the role of the DJ, programme controller or curator. What may have been traditionally done by large media organisation such as a newspapers, radio or TV stations for example can now be achieved by an individual specialist. The problem then is that we need to find these specialists and it needs to be built on trust.

Certainly if someone can somehow encapsulate (knowledge and taste) into an algorithm, they will find huge success. Hopefully this will also allow those with talent to shine through the noise and share in that fortune, getting the greater exposure they deserve.

[wp_popular_posts number=3]