Privacy, personal data, and machine learning have been in the news again and again this past year. Often these topics are discussed in the abstract, making them feel disconnected from our lives as individuals. This lack of a personal perspective on how we as users are being used means that we aren’t given the chance to have a real reaction to this new reality we live in. At least not until it’s too late.

The goal of “Hello Stranger” is to make the viewer uncomfortable by confronting them with their own data. By removing the abstraction and making the viewer uncomfortable my hope is that they may begin to question the data they willingly and unwillingly provide to strangers.

I harvested all the data for “Hello Stranger“ using web crawlers created with the library Puppeteer. The crawlers collected photos and names from social media sites and databases. All data used in this piece is publically available or is accessible by such a large number of people it might as well be public.

The harvested data was formattedinto facial descriptors using the library face-api.js. This library also powers the facial detection, expression detection, and facial identification which takes place in the piece. face-api.js uses a type of machine called neural networks to achieve all these complex tasks.