In his lecture, Jeremy Armstrong made an interesting connection between the Mediterranean sea and the internet. As someone spending a lot of time on the internet recently I thought it would be interesting to break down this comparison and ultimately see why this unfortunately is not fully accurate. Armstrong describes the Mediterranean as varied and diverse and as a constant medium for culture and information. On the surface this is a very appealing understanding of the internet; however, if we look deeper we are able to recognise that gatekeeping and echo-chambers are virtually inescapable online. Armstrong told us that you could not stop people from moving across the Mediterranean and that this was what gave it it’s unique power; however, I believe that information is being constantly manipulated and directed through the internet, making it ultimately very different. 

The most fundamental difference between the internet and the Mediterranean is the targeted nature of the content we are exposed to on the internet. In ancient Greece, it was highly unlikely the sea would know you wanted a specific type of vase and direct this right to your port, yet everyday we see products advertised to us which seem to be tailored to our interests. This practice of monitoring of our internet behaviour for advertising is known as ‘Online Behavioural Advertising,’ and is becoming so frequent on the internet that it’s unlikely we will be able to escape it going into the future¹. Similarly to advertising, the content and resultant ideas which we are exposed to on the internet are almost always a reflection of our own ideas, due to both our ability to choose the forums we enter, and also the nature of website algorithms to target us with information they think we will want to engage with. I’ll use the example of the youtube algorithm although I’m sure we can all think of a range of relevant examples from our daily life.

A visual representation of the youtube algorithm. (Paul Covington, Jay Adams, Emre Sargin, “Deep Neural Networks for YouTube Recommendations.”)

Briefly, the youtube algorithm works through the system pictured; by narrowing down millions of videos into just dozens which are determined by information gathered about the user². This is ultimately very concerning for the free flow of information which we would hope to get from our explorations of the internet. However, if we express any partisanship online we will automatically only be exposed to this set of ideas, as this is what the algorithm thinks we want. This results in our viewpoints becoming narrower and less open to alternative ideas. Armstrong made a comment which I see to be relevant to this: “the more positively you feel about the people inside your group, the more negatively you tend to feel about those outside your group.” With our websites becoming more like echo-chambers, showing us only related content, it is frighteningly easy to fall into a group mentality and demonise those of different groups and ideas, a process known as the “radicalisation pipeline” in more extreme situations³Perhaps if the internet were more like the Mediterranean and provided a constant exposure to new ideas and cultures, we could prevent the formation of such disconnected groups. However, unfortunately, the internet is less like the Mediterranean and more like a very specific waterfall.

 

1: Sophie C. Boerman, et al. “Online Behavioral Advertising: A Literature Review and Research Agenda”

2: Paul Covington, Jay Adams, Emre Sargin, “Deep Neural Networks for YouTube Recommendations.”

3: Manoel Horta Ribeiro, et al. “Auditing Radicalization Pathways on YouTube.”

Featured image: https://www.ancientworldmagazine.com/podcast/networks-ancient-mediterranean/