In healthcare, work performance evaluation, and criminal justice, machine learning algorithms are increasingly responsible for sorting human behaviors into meaningful categories. For my master's thesis, I was interested in understanding how this extensive reliance on algorithmic meaning-making changes the nature of human decision-making.
Since I come from a background in developing interfaces, I push back on the idea that algorithms are powerful because they are invisible, and argue instead that algorithms are especially affective, persuasive, and meaningful when they are highly visible. Visualization is one way of rendering algorithms meaningfully to people, which does not account for all that algorithms do, but depicts them in a certain, partial light.
Lifestreams ingests a dataset about student academic performance and behavior and organizes it for interactive analysis. Although the interface could be used to understand the data in practice, Lifestreams was developed as a speculative exploration of the visualization and rationalization of algorithms.