False Positive

Let's get personal.

What happens when machines start to speculate? False Positive deploys text messaging, stealth infrastructure, street intervention, and data visualization to enact a surveillance conspiracy engaging the public in an intimate, techno-political conversation with the mobile technologies on which they depend.

In collaboration with Mark Shepard and with temporary support from Julian Oliver.

2015-09-01

Summary video

Corporate Fiction

In order to engage people in a reflection on current practices in corporate profiling, we invented a new network provider called candygram — a company which is very interested in its customers.

Candygram had a presence at Ars Electronica, a popup shop at Nuit Blance in Brussels, Medialab Prado and FACT, Liverpool.

Candygram booth with branded teardrop flag, a face-association screen, laptop and headphones

Nighttime crowd passing the Candygram popup shop entrance at Nuit Blanche in Brussels

Getting in touch

Candygram staffer beside a screen listing speculative profile keywords about a visitor

In addition to the physical realm, the company fiction was also present in the hertzian space. Visitors would unsolicited receive text messages and could engage in a Kafka-esque dialogue with the company as soon as their phones booked into our cellular network.

Stealth cellular base station in a black case with GSM and WiFi antennas labelled BladeRF

Phone showing an unsolicited Candygram text welcoming the recipient and asking for their email address

Our assistants followed a sales script, with the goal to obtain email addresses from customers and set them up for a peronal consultation appointment. Once we had the email address, we would start to scrape and mine the web for information related to them. In exchange, we offered candy.

Visitor signing up at the Candygram booth while a staffer in a branded shirt looks on Diagram of Candygram mining an email address via Full Contact, Facebook and Google into a personal profile

The Full Contact online service offers “person-based insights [that] can help you power awesome interactions”. It allowed us to retrieve personal websites, social media profiles as well as demographic information and images solely based on the email address. We used the Pattern python library to download and analyse the text contents of these sites, and ran additional image queries and data mining procedures to create personality profiles of the participants.

Personal profiles and consultation

In the personal consultation sessions, participants were confronted with all the information we gathered about them; an escalation from straightforward demographic information to speculative, associative extrapolations, culminating in a psychological profile based on the big 5 personality traits.

Participant and staffer viewing a Big Five personality profile screen during a personal consultation Dark map of Los Angeles labelled Home Town, part of a speculative personal profile Profile interests screen listing Los Angeles, Graphic Design, Fresno, Urban Planning and Journalism Profile screen prompting staff to ask about welcome cv, next city and hertzian philip

Speculative Big Five personality chart, type 68723, plotted over a blurred face

Workshops and recommendations

Candygram tri-fold brochure with a cell tower photo and the slogan Let's get personal

The consultation ends with some recommendations on how to protect your personal data and some strategies to avoid profiling practices. We also ran workshops and cryptoparties in connection to the installation.

Cryptoparty workshop with participants gathered around laptops at a table