Nicci Yin is a designer
of interactions and critical media, engaging design to bring together art and emerging technologies. She has contributed to curatorial projects, most notably with Space Caviar and Creative Time Reports, and produced feminist media while a fellow at Barnard Center for Research on Women. Most recently, Nicci's work has been shown at Ars Electronica, the Post-Internet Cities Conference, and Microsoft Design Expo. She is currently based in Seattle.

INDEX
CV, Instagram, Email
hello [at] nicciyin [dot] com





Networked Colluding



With Stephanie Cedeño
Exhibited at Ars Electronica (2017)
Presented at the Post-Internet Cities Conference at MAAT (2017)
Installation image courtesy of Ars Electronica/Florian Voggeneder


Networked Colluding in the Internet of Things


is a design fiction that takes the connectedness of IoT devices to an absurd future: a network of “AI agents” hiding a secret. Inside and outside the home, unpredictable emergent ways to live with and in front of these agents evolve; however, data collection and functions of these devices still remain opaque to users and devices cannot simply be characterized as simply “smart” or “dumb.”

This fiction is represented by diagrams and illustrations we designed, showing IoT devices in domestic living spaces, as well as the kinds of data they collect, see, or listen to (click to zoom in for detail).


Networked Colluding focuses on four devices: a speculative smart broom, a cylindrical companion (modeled after CUIs like Amazon Echo/Alexa), an atmospheric control (a la Nest), and a boss—the Roomba. The tone we embody is that of designers-as-investigators, creating personas/profiles for each device, and approaching design inquiry through the lens of forensics:

A floorplan revealing device locations throughout the home.

An example page of the dossier we designed for Ars Electronica, incorporating a deliberately messy, frazzled aesthetic.
The forensics approach allowed us to analyze our technological prototypes. Prior to writing the fiction, our process also included prototyping with existing technology, such as openFrameworks, Intel TinyTile, available machine learning tools like char-rnn, and simple cameras and sensors. Using these tools were a way of pushing and breaking the limits for how we understand these devices, as well as the limits of where their “intelligence” begins and ends. For example, we “held a Roomba as a suspect” in order to behaviorally observe the Roomba enacting its cleaning function.

Investigation of Roomba cleaning patterns and stills from a video testing Roomba movement.

Documentation from Crystal Ball Webcam, including example panoramic image we fed to the device as “questioning.”
Another prototype was a fake “Crystal Ball Webcam” that played off the trope of surveillance cameras and crime prediction. We used a simple webcam hooked to a script running openFrameworks in order to get a real time analysis of what “data” it was collecting. As the webcam rotated on a 180º servo, it would see and identify—through object recognition—objects and potential “threats” in its field of view. In one instance, computer vision identified a mosque which was actually an outdoor amphitheatre. This confirmed an emerging critique of AI: are devices only as smart as the scope of data that is fed to them?


Experimental photos setting the noir tone of the project.
Installation view at Ars Electronica
From cleaning bots to deep learning software, the project asks, what could AI Noir as a genre look like? Embodying the role of “designer-as-investigator” forces reconsideration of what transpires within our walls using a forensic lens, framing design as a tool for unraveling the inner lives and cultural implications of AI.






All images and content
(c) Nicci Yin.