Addressing the Prevalence of Biased Algorithms
Algorithms have an immense impact on our lives from everyday tasks to major life-altering decisions. They determine simple things like which movies and shows are recommended to us on our favorite streaming services, whose resume gets rejected for a job position, and even prison sentences. Using algorithms to make suggestions or even decisions for us in itself is not a problem, especially when they determine a fair or good outcome. However, research shows that many algorithms are biased. This means that consequential algorithms like the algorithms being used to suggest prison sentences are often making unfair recommendations. The commonness of biased algorithms showcases the need for bias training among technologists, including training on how to recognize biased data. Thus far, this study suggests that the prevalence of biased algorithms is not only the result of a lack of bias training among technologists but a reflection of society’s acceptance of historical biases. Once addressed, programmers will be more informed and, therefore, more likely to create fair algorithms.