Author: James Arvanitakis, Andrew Francis, Oliver Obst
Publisher: The Conversation
Publication Year: 2018
Summary: The following article discusses how, as analysts, the main 2 ethical questions that are being asked are 1). What data should be collected and 2). How should the data be used? However, we should also ask and see who gets to make the decision in the first place. As it stands right now, the people who hold all the power to make the decision in how we use data are generally rich, educated, white males. This leads to the unconscious biases of this group being hiddenly baked into the technological services that we use every day. Bias in algorithms is a problem that is unfortunately widely prevalent due to the lack of diversity among those that make the decisions of how data is being used. It can lead to issues such as being unable to recognize certain faces based on race, bias in criminal sentencing, and gendered language translations. We collect data that is representative, but if the ones making the high-level decisions about the algorithm are biased themselves, then the tool can be biased without our knowledge. In the future, the American education system needs to push for more diversity in these technical areas. Have more women in these spaces, have more people of color in these spaces. Having a diverse workforce allows for a diverse way of thinking and can potentially prevent biases from going under the radar.