First Year Seminar: Uses and Misuses of Algorithms

Data scientists have used algorithms for many great things: Netflix’s recommendation system; building teams that go on to win the World Series or the Stanley Cup; proving the existence of the Higgs boson; and early detection of cancer. Other data scientists have used algorithms with malicious intent: targeting of vulnerable people by payday loan companies and for-profit higher education companies that provide little—if any—benefit to their students; using of social network bots to spread misinformation and sow discontent within a country. Sometimes algorithms have unintended negative effects: the firing of skilled teachers in Washington, D.C.; racial discrimination in the lengths of prison sentences and the granting (or not granting) of parole; and discriminating against job applicants (including racism, sexism, and ableism).

In this course, we will investigate both successes and failures of algorithms. We will discover that algorithms are efficient at learning from data, including learning and amplifying biases within the data. We will also get hands-on experience working with data, including investigating data for evidence of bias, using algorithms to make predictions, and using data and algorithms to inform decision-making. No prior experience in statistics or programming is necessary. The only requirements are the desire to solve problems using data and an interest in the ethical use of data.