The epistemic ethical concerns involving algorithms in intelligent communication

Abstract

With the development and application of algorithms as catalysts, the changing modes of information production, dissemination, and consumption have also given rise to a myriad of serious ethical challenges. This study employs a multi-case approach and semistructured in-depth interviews to examine three prominent international information technology companies, namely, Meta, Sina, and Byte Dance. By investigating the utilization of algorithms in content creation and distribution and adopting an epistemic, ethical framework, this paper analyzes the phenomenon of information cocooning resulting from inconclusive algorithmic evidence, the presence of algorithmic black boxes stemming from inscrutable evidence, and the issue of algorithmic bias caused by misguided evidence. Consequently, this paper proposes three fundamental ethical principles for algorithmic systems: certainty, interpretability, and reliability.

Keywords

• algorithm bias
• algorithmic black box
• algorithmic ethics
• information cocoon
• interpretability