Ruminations

Ruminations

This project explores data privacy in the context of Amazon's ecosystem, questioning how we might subvert browser fingerprinting and challenge pervasive consumer tracking.

We began with a provocative question: Could we degrade the value of collected data not by avoiding tracking, but by actively engaging with it? Rather than trying to hide from surveillance, could we overwhelm it with meaningful yet unpredictable patterns?

Initially, we considered implementing a random clickbot to introduce noise into the data collection. However, given the sophistication of modern data cleanup algorithms and the sheer volume of data Amazon processes, such an approach would have been ineffective. They would simply filter out the random noise and continue their analysis.

This led us to a more interesting question: How can we create coherent, non-random data that remains fundamentally unpredictable? Our solution was to introduce patterns that exist beyond the predictive capabilities of current algorithms – similar to trying to predict the behavior of someone whose thought patterns follow their own unique logic.

The Concept

We developed a Chrome browser extension that overlays Amazon's web pages with a dynamic entity tracking user behavior. The system employs an image classifier algorithm to analyze the storefront and formulate product queries. After processing, it presents a "perfectly matched" product – a subtle commentary on algorithmic product recommendations.

The Analog Watchdog

The project's physical component consists of a low-tech installation using a smartphone camera running computer vision algorithms to track minute movements. We positioned this camera to monitor the browser console of a laptop running our extension. The camera feed is displayed on a screen, and the system generates robotic sounds based on the type and volume of detected movement. In practice, it serves as an audible alert system for data exchanges between Amazon and the browser.

Implementation

Try It Yourself

Want to explore or contribute to the project? Check out our code repository: