Who steals bikes?
Who steals bikes studies issues of racism and stereotyping in Amsterdam, Holland, a widespread social problem that has come into the forefront recently in many European cities. Using technologies and data mapping overlaid onto physical space, the project uses locative media not to catch bike thieves, but rather to challenge prejudices by taking individual thoughts and projecting positive cultural messages onto them as a perceived collective ideology.
Using mixed conditions, where racism and social negativity are coupled with crime, there is an examination to understand why people of color or immigrants in the city are often perceived as the people most likely to commit such a crime. To study this phenomenon, there first was a series of videotaped of one scene of a bike theft occurring and the actors being switched to understand the reactions of the public based on the perpetrator’s ethnicity. Secondly, a series of protest control group studies were developed using mobile technology to create crowd-sourcing data, which in real-time helped to effectively question people’s preconceived stereotyping mindset and to challenge traditional ideologies utilizing data and messages that are relevant to Amsterdam residents.
Using mixed conditions, where racism and social negativity are coupled with crime, there is an examination to understand why people of color or immigrants in the city are often perceived as the people most likely to commit such a crime. To study this phenomenon, there first was a series of videotaped of one scene of a bike theft occurring and the actors being switched to understand the reactions of the public based on the perpetrator’s ethnicity. Secondly, a series of protest control group studies were developed using mobile technology to create crowd-sourcing data, which in real-time helped to effectively question people’s preconceived stereotyping mindset and to challenge traditional ideologies utilizing data and messages that are relevant to Amsterdam residents.
Experiment Prototype
An Experiment Testing the Effect of Data manipulation
Imagine a dark room with an R G B L E D light. Then imagine three persons that each has a device with three sliders to determine a value for R, G, and B.
They are told that a controller will receive all the values from slider bottoms, then calculate the average amount of each color channel and send the three final values to the R G B, L E D.
So, through a collective process, they can adjust the light color in the room.
However, a control algorithm will keep the overall R value always bigger than the two other (G average and B average) even when it’s actually smaller. As a result, although there is not a constant color for the R G B, L E D and it’s changing in response to each person’s slider, the room is always more red.
Each person will think the other two are the ones who like red more and keep holding the R value higher. While in fact higher amount of red is because the controller is doing so.
The app, works in a similar way. The online app was looking at the results of people’s data collection and when necessary was adding data from fake identities to the app. By this way, the app was keeping the overall result more or less the same thing: “The rate for Dutch and people with western origins was higher among the bike theft suspects”.
They are told that a controller will receive all the values from slider bottoms, then calculate the average amount of each color channel and send the three final values to the R G B, L E D.
So, through a collective process, they can adjust the light color in the room.
However, a control algorithm will keep the overall R value always bigger than the two other (G average and B average) even when it’s actually smaller. As a result, although there is not a constant color for the R G B, L E D and it’s changing in response to each person’s slider, the room is always more red.
Each person will think the other two are the ones who like red more and keep holding the R value higher. While in fact higher amount of red is because the controller is doing so.
The app, works in a similar way. The online app was looking at the results of people’s data collection and when necessary was adding data from fake identities to the app. By this way, the app was keeping the overall result more or less the same thing: “The rate for Dutch and people with western origins was higher among the bike theft suspects”.
