Quantified Humanists - Final Project - The Infestation of Web Trackers

STEFAN SKRIPAK
QUANT HUMANISTS
SPRING 19
04 22 2019

Final Project - “The Infestation of Web Trackers” - Project Link

Screen Shot 2019-05-01 at 11.28.18 AM.png

As many artistic projects tend to do, my concept for this project has morphed and evolved fairly dramatically since I first started my work on it. I was initially inspired by the concepts of data doubles and this idea of a different version of myself “growing” on the web inspired me to consider using biological or biomedical imagery to represent this concept in some way. As I became more attached to that visual idea, I realized that a more concise concept and data source would ultimately be necessary to convey any sort of clear message.

My research into the ways data doubles are created led me to the discovery of the vast networks of first and third party browser trackers which log user behavior as we browse the web. And while collecting data on these trackers would not necessarily fully illuminate the methods of data double creation, bringing attention to this largely invisible behavior in order to initiate critical thought felt like a positive and meaningful goal.

One week’s worth of browser trackers

One week’s worth of browser trackers

So by this point I had settled on a topic and narrowed my visual ideas down to focus on the aesthetic of growing mold, however I was still unsure about the actual form. My initial concept was to create some sort of physical device which would visualize this growing mold in a series of petri dishes fabricated with LCD screens installed in them. I also considered whether I should simply 3D print static representations of this tracking data somehow, or generate and project a video in the style a digital microscope. But what I ultimately realized is that all of these ideas were at least one step removed from the activity itself that I was visualizing and so my desire to more directly relate this concept to the actual behavior of browsing the web and being tracked while doing it led me to my final idea to create a browser extension.

This extension would have two main functions: to collect the data about the trackers being used on every site a user visits and to visualize that data as mold growing across the screen, obstructing the browser use and forcing the user to acknowledge and consider the information they are giving away. My research into how to achieve these led me to a tool which would prove to be integral to my project - the open-source Firefox extension called Lightbeam. Lightbeam is an extension which collects the same tracker data I was looking for, and them presents it in a separate, fairly simple visualization.

I downloaded the full code for Lightbeam from Github in order to attempt to understand how the data collection worked, but quickly realized that it would actually be simpler and more efficient to just adapt the existing code and build my own project on top of it. This is what I ultimately did, and reveals just how crucial this existing extension is to my final project.

While I inherited most of the backend for my project from Lightbeam, the front-end is my own work. Adapting the code from a series of tutorials by Dan Shiffman on fractals I was able to create a satisfying and unnerving mold growth visualization. While the code is somewhat complicated, I am relatively comfortable working in p5.js and the process of the visual development was a fairly straightforward process of development and debugging.

The aspect of the project which took me by far the longest (and during which I learned the most) was the process of then incorporating the p5.js sketch into the browser extension. Dan Shiffman’s tutorials again proved invaluable, but I also had to do a lot of learning on my own (and soliciting of help from my professor Joey) to get my sketch to both incorporate into the existing extension as well as allow it to access the information about the web trackers being collected by Lightbeam. Considering I knew essentially nothing about developing browser extensions when I began, I am very satisfied with my final product.

The final version has only a few simple functions. Using the back end from the Lightbeam extension, every time the user visits a new website the first and third party trackers are added to a JSON object. Then that JSON is passed to the p5.js sketch which creates the mold visualizations. For every website a user visits, a new mold is grown, and an invisible “food” is created for every third party tracker which the mold then grows towards. Each third party tracker has a unique location, so if multiple websites contain the same tracker, they will grow towards the same spot on the screen. Every time the user visits a new site the molds regenerate, with the addition of the most current website data, quickly obscuring the browser window. And finally there is also a browser toolbar button which simply resets the database. My hope is that in its simplicity this extension as a whole will inspire internet users to consider more critically how closely, and by how many people, our internet behavior is tracked.