Exploring Perceptions of Algorithmic Bias Among Software Engineers: a Case Study of Software Engineers in İzmir, Türkiye

Abstract

This study investigates how software engineers perceive artificial intelligence (AI) and algorithmic bias. The study explores whether the human-like characteristics of AI influence their engineering practices, which traditionally hold a dualistic view of technology and society. Based on semi-structured interviews with software engineers in & Idot;zmir, T & uuml;rkiye, the findings reveal both similarities and differences between classical engineering and software engineering. Classical engineering views technology and society as separate entities, while software engineers adopt an ambivalent sociotechnical stance, acknowledging but neglecting their interconnectedness. Software engineers prioritize technical definitions and efficiency in assessing algorithms, often considering social dimensions secondary. However, they view algorithms not just as tools, but as codes shaping everyday life with social and cultural attributes. This departure from conventional understanding highlights the sociotechnical context in which software engineers operate. Moreover, the study shows that software engineers tend to interpret algorithmic bias through a technical lens, overlooking broader social and human contexts. These findings emphasize the urgent need to reassess the relationship between technology and society within the sociology of artificial intelligence, fostering a deeper understanding of sociality in software engineering.