The problem of fairness in tools for algorithmic fairness

The biases introduced by intelligent algorithms have had profound social implications, making the pursuit of fairness in algorithms a priority for various sectors. In order to achieve algorithmic fairness, numerous studies have focused on developing and utilizing tools for algorithmic fairness. However, from the perspective of philosopher of technology Andrew Feenberg, these tools for algorithmic fairness overlook the issue of formal biases in algorithms and only concentrate on the narrow concept of algorithmic biases. Additionally, even if these tools were able to achieve fairness in the narrow sense, they still fail to address the problem of formal biases. Furthermore, the concept of fairness is inherently controversial, which implies that these tools not only fall short in attaining algorithmic fairness, but also fuel further controversy, exacerbating algorithmic biases and creating a paradox of algorithmic fairness. On one hand, this paradox amplifies the social impact of algorithmic biases, while on the other hand, it may render current research on algorithmic fairness meaningless and valueless. Consequently, a nihilistic view of algorithmic fairness and social justice could emerge, potentially jeopardizing the harmonious and stable development of society.