What is the difference between bias and fairness in AI?

There are mountains of academic literature on both. But I wanted to see what a diverse group thinks about them.

The following are insights from the Mission Control fireside chat I led.

We came up with three models of the difference between bias and fairness:

➤ Model 1: Fairness is the opposite of bias. They are two ends on the same spectrum.

➤ Model 2: Fairness encompasses non-bias. Combating bias is one of the things we need to do to increase fairness, but not the only one.

➤ Model 3: Fairness and bias are related, but they are conceptually distinct.

Two examples:

--> Bias is about preferences, e.g., preferring one individual over another. Fairness is about ensuring equity and equality. (h/t Tasha Jackson,MBA, MS)

--> Bias is about how social (and other) kinds shape beliefs. Fairness is about how people are treated. (h/t Borhane Blili-Hamelin, PhD)

➤ Join the discussion about this distinction on my LinkedIn page here