AditWork / About Me 



INDIVUDAL PROJECT - SPRING 2020 ROLE: Designer, Researcher, Technologist

• PRIMER Emerging Designer Exhibition Second Place
• Exhibited at Abierto Mexicano Diseño 
• Exhibited at DeepCity 2021




Unstable Label

What would it look like to use machine learning as a space to contest meaning rather than simply as an operational tool for efficiency and optimization?

Unstable Label is a participatory data labeling application that facilitates conversations of how we each see and imagine the world around us differently through the collective creation of an object detection algorithm. 

See the functional prototype at unstable-label.glitch.me
See full process case study here.
See user manual for full project writeup.




System Overview


︎︎︎ The left panel of the system allows you to navigate through google street view, finding sites of interest to contribute to the collective machine learning model.

︎︎︎ The right panel represents the model, listing all the “data” that has been contributed to the project, featuring not only labels, but also the stories that put them into context. 






Step 1: Local Navigation


Each user contributes from their own local context from within Google Street View or from a physical data collection device. 




Step 2: Creating Categories


  1. The current machine learning model evaluates the image from your current location, generating labels. 
  2. You create new categories by relabeling those existing categories. The relabeling process is usually done in a small group. You discuss what the original category makes you think of within your local context. 

* This is not about “correcting” the algorithm or making it more “accurate”. It’s about introducing your own locally situated perspective into the dataset





Step 3: Drawing Data


Use your personal categories that you’ve just created to annotate other images, creating data. There are two annotation options: label and imagine. You can choose to annotate the image as it currently exists, or as you wish it existed. 




Step 3: Updating the Machine Learning Model


Once submitted, the model is retrained using your data, adding your categories into the model. Future contributors may relabel your data, recontexualizing it to their local neighborhoods.

* The end result of this system isn’t a “better” model, but the process of contestation and negotiation that it facilitates, some of which doesn’t happen in the app at all.