Screening Bot

About Screening Bot

"Racism has always been about predicting" (Roberts, 2020). Historical perspectives of racism will show that specific racial groups have been tagged as predisposed to engage in criminal activity, and thus, different control systems have been justified. Today, we see that algorithms have been incorporated to predict who has a higher chance of being involved in future criminal activity, to intervene before a crime takes place, or determine whether someone who has been arrested is likely to re-offend. The increased use of machine learning responds to the belief that they are more objective than humans. However, little is known about the transparency of these systems. According to the US Department of Justice, your possibilities of being arrested if you are Black are twice as much than if you are white. Considering that the data the algorithms feed upon are based on previous records, it is sensible to believe that the algorithm will disproportionately target Black people. The algorithms, while apparently neutral, allow past actions to shape future decisions. Thus, we believe it is necessary to engage in reflections around the need to avoid tech-washing racial mechanisms that perpetuate inequity.

Screening Bot is a problem-solving video game where a robot needs to use training data to create algorithms that will stop potential threat alien races from entering a planet. As the Screening Bot processes the algorithms, the player will find that there are flaws in the training data and many individuals are being unfairly imprisoned. By exposing players to the consequences of predictive surveillance technologies, they will be able to reflect upon these technologies. We believe this is a powerful game to engage young people to reflect on their experiences, identify ways in which data is misused to perpetuate racism, and engage in critical design activities to imagine possible futures for machine learning.

Development Team: Ezequiel Aleman, Larysa Nadolny--Iowa State University. Bruno Gabetti, Alejandro Ferreira, Guillermo Ortíz, Martín Zanoniani--Universidad Technolóca, Uruguay.