AIAI/ Georgia Tech
On March 29, we held a public workshop attended by 50 participants in the Spark Center at Georgia Tech. Participants ranged from members of Google’s Trust and Safety team, planners from the City of Atlanta, and community organizers.
About
The workshop was developed from ongoing dissertation research of Anhton Tran who led the design team. He is a PhD candidate in Human Centered Computing at Georgia Tech’s School of Interactive Computing. His work investigates classification systems of eviction and the data generated from these endeavors. The workshop was crafted by a design team of Master’s students in Human Computer Interaction: Jingxuan Wang, Jacob Xu, Cassia Tang, Tao Lu, and Binta Moncur.
Development for the workshop began September 2023, where students participated in a studio seminar. Together, students read studies of eviction data from a variety of fields: sociology, critical geography, computing, and Urban Planning. Coupled with reading were a series of design charrettes that trained students on design-led research practices and speculative design.
The workshop utilized design metaphors as a container to unpack the issues of court generated eviction data which are public records. From reading and prototyping speculative exercises, the design team connected the creation of eviction records to service industry food production. Participants were led through 3 activities. First was an orienting activity where participants had to recreate a food production assembly line to fill orders for burgers. All of the components of the burgers tied to necessary fields and parts of an initial dispossessory affidavit, the initial document that triggers the creation of an eviction data point. Afterwards, participants reflected on the stress of filling orders before walking through a short presentation on the political economy of eviction data that describes the mode of production of eviction data: high volume and low quality.
This led to the second activity: a world café. Participants brainstormed and ideated on alternative modes of production on eviction data or a data set of their choosing. Did they want to design a data production line that prioritizes low volume data but higher quality? Or higher volume data that was also high quality? If so what would that look like and what would be the critical considerations. Participants were provided with restaurant metaphors to help them develop their ideas. A high volume low quality restaurant may be a fast food establishment, whereas a low volume high quality restaurant may be a farm-to-table bistro. This activity consisted of three rounds of brainstorming focused on different parts of the data production line: sourcing, compiling, and using. All of the ideas were mapped onto large printed posters and participants had the opportunity to switch tables during the second round, gather other insights, and bring it back to their group.
The final workshop was a speculative design activity. Participants were given materials to design their own consumer packaged goods, imagining that data products were ones that you could pick off a shelf at a grocery store. These materials ranged from curated magazine pages for collage, stickers, glue, blank labels, and more. Participants designed a variety of ideas, from a subscription box style data company that curates data sets for the consumer to imagining new data standardizations that would be advertised on the box, like “75% less bias” or “ethically source data.” We concluded the workshop with a brief reflection and plan to document and send artifacts back to participants this summer.