Despite the acknowledged importance, there is a literature gap on practically implementing diversity and inclusion in AI. FoschVillaronga and Poulsen define diversity as gender and race representation, and inclusion as users' alignment with relevant options, analyzed at technical, community, and user levels.
The article proposes a normative definition and guidelines for integrating diversity and inclusion into AI development, with feedback from Responsible AI and D&I experts. It emphasizes a socio-technological perspective and Responsible Research and Innovation principles.
[
Collection
]
[
|
...
]