Ethical AI: Social Justice in Data Science

2 years ago
15

The public debate to define justice with regard to Individualism vs Collectivism is waging strong, and most Data Scientists seem unaware of its impact on their algorithms. It's an area that my unique background as a Data Scientist who studied Philosophy and Religion compels me to address for those less familiar with a topic that seems so far removed from Data Science. This will be controversial for two reasons: 1) there is no measurable way to test and determine which view of justice is right and there is a lot of disagreement regarding the effects each has had on society, and 2) real world algorithms face much more complex results from adjustments than the simplified examples used in this video. Even the definitions of Individualism and Collectivism have different meanings in different areas of study, and different outcomes along the spectrum between the two.

I've been researching and addressing Algorithm Fairness problems as part of my occupation for over two years, and so have an interest and time invested in these conversations. Despite the fears of addressing this topic head on, I think transparency is important, especially in areas intended for public debate like here. This is definitely an area where I'd love to see your constructive thoughts and comments below so we can sharpen and grow one another and maybe even change our thinking on occasion.

Disclaimer: this video is my own and does not necessarily reflect the positions, strategies, or opinions of 84.51° or our parent company, The Kroger Co.

Loading comments...